2010-10-05T00:00:32 *** Appleman1234 has quit IRC (Ping timeout: 252 seconds) 2010-10-05T00:02:43 *** JensTi has quit IRC (Quit: Leaving.) 2010-10-05T00:07:58 *** delt0r_ has joined #aichallenge 2010-10-05T00:09:14 i love going for a rewrite and then finding an hour later that you're already further along than you were in your old code base 2010-10-05T00:09:28 and it's better, at that 2010-10-05T00:09:40 *** delt0r___ has quit IRC (Ping timeout: 272 seconds) 2010-10-05T00:14:44 *** Arthur has joined #aichallenge 2010-10-05T00:17:54 *** znutar has joined #aichallenge 2010-10-05T00:24:48 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-10-05T00:28:04 *** jesionaj has quit IRC (Quit: ( www.nnscript.com :: NoNameScript 4.22 :: www.esnation.com )) 2010-10-05T00:30:40 *** Utkarsh has joined #aichallenge 2010-10-05T00:39:16 *** yasith_ has quit IRC (Ping timeout: 264 seconds) 2010-10-05T00:39:57 !seen DanielVF 2010-10-05T00:39:57 Janzert: DanielVF was last seen in #aichallenge 2 hours, 58 minutes, and 55 seconds ago: * DanielVF out. :( 2010-10-05T00:41:41 *** rwa has joined #aichallenge 2010-10-05T00:41:57 Janzert: why? 2010-10-05T00:44:35 oh, I have elo confidence bounds displayed in the rankings working locally 2010-10-05T00:44:51 but of course it requires adding two columns to the rankings table 2010-10-05T00:45:25 and I'm unsure how to best handle that so the server would get updated correctly 2010-10-05T00:45:49 it's a simple alter table statement of course 2010-10-05T00:59:16 Top 10 players: ruilov(3892), davidjliu(3824), dmj111(3750), felixcoto(3697), rsergio(3683), Hazard(3669), sequoh(3658), EBraun(3630), adumlah(3624), DocBot(3621) 2010-10-05T00:59:30 *** hellman has joined #aichallenge 2010-10-05T01:05:58 *** LazyHK has joined #aichallenge 2010-10-05T01:06:59 *** LazyHK has quit IRC (Client Quit) 2010-10-05T01:10:22 *** yasith_ has joined #aichallenge 2010-10-05T01:13:05 !seen contestbot 2010-10-05T01:13:05 Frontier: I have not seen contestbot. 2010-10-05T01:13:14 The cake is a lie -.- 2010-10-05T01:16:18 I don't feel so bad now 2010-10-05T01:16:21 lol 2010-10-05T01:16:22 contestbot: whoami 2010-10-05T01:16:22 sigh: I don't recognize you. 2010-10-05T01:16:29 ^^ about this 2010-10-05T01:16:38 contestbot: whoami 2010-10-05T01:16:38 amstan: I don't recognize you. 2010-10-05T01:16:54 contestbot: whoami 2010-10-05T01:16:54 amstan: amstan 2010-10-05T01:17:02 it's a login thing 2010-10-05T01:17:06 it's how you administer the bot 2010-10-05T01:17:07 yeah, I figured :P 2010-10-05T01:17:12 it'll say that to everyone 2010-10-05T01:17:14 except me 2010-10-05T01:17:20 :D 2010-10-05T01:17:51 contestbot: echo I predict that $randomNick will win! 2010-10-05T01:17:51 I predict that greghaynes will win! 2010-10-05T01:17:56 *** McLeopold has quit IRC (Quit: Leaving) 2010-10-05T01:18:08 contestbot: unload anonymus\ 2010-10-05T01:18:08 amstan: Error: There was no plugin anonymus\. 2010-10-05T01:18:10 contestbot: unload anonymous 2010-10-05T01:18:10 amstan: Aye, aye, sir 2010-10-05T01:18:20 greghaynes: nice try 2010-10-05T01:18:20 what's that? 2010-10-05T01:18:30 sigh: the say command 2010-10-05T01:18:44 sigh: you can tell it @say #aichallenge something 2010-10-05T01:18:50 but... meh.. it can be abused 2010-10-05T01:19:08 you realise that wasn't because of say... 2010-10-05T01:19:20 oh! 2010-10-05T01:19:22 lol 2010-10-05T01:19:23 wow 2010-10-05T01:19:48 :D 2010-10-05T01:20:03 contestbot: alias add predict "echo I predict that $randomNick will win!" 2010-10-05T01:20:03 amstan: Ready to serve. 2010-10-05T01:20:08 contestbot: predict 2010-10-05T01:20:08 I predict that Mathnerd314 will win! 2010-10-05T01:20:10 contestbot: predict 2010-10-05T01:20:11 I predict that Gorovkhaia2 will win! 2010-10-05T01:20:16 uh oh 2010-10-05T01:20:37 yay, I totally knew I would win :p 2010-10-05T01:20:41 contestbot: predict 2010-10-05T01:20:41 I predict that tty2 will win! 2010-10-05T01:21:18 ok, we've found a way to keep amstan amused for a while 2010-10-05T01:21:34 contestbot: predict 2010-10-05T01:21:34 I predict that rogue780|server will win! 2010-10-05T01:21:36 LOL 2010-10-05T01:21:41 contestbot, im not even in it! 2010-10-05T01:21:42 tty2: User error, it's not my fault. 2010-10-05T01:21:46 hahah 2010-10-05T01:22:01 these error messages are hilarious 2010-10-05T01:22:06 !eightball will tty2 win? 2010-10-05T01:22:06 sigh: You know the answer better than I. 2010-10-05T01:22:11 im here to sponsor other groups, not participate :) 2010-10-05T01:22:12 o_O 2010-10-05T01:24:15 contestbot: synonym sigh 2010-10-05T01:24:16 sigh: snort 2010-10-05T01:24:20 that' 2010-10-05T01:24:30 that's not a very nice synonym :( 2010-10-05T01:25:01 contestbot: dict sigh 2010-10-05T01:25:02 amstan: wn, gcide, and moby-thes responded: wn: sigh n 1: an utterance made by exhaling audibly [syn: {suspiration}] 2: a sound like a person sighing; "she heard the sigh of the wind in the trees" v 1: heave or utter a sigh; breathe deeply and heavily; "She sighed sadly" [syn: {suspire}] 2: utter with a sigh; gcide: Sigh \Sigh\, n. [OE. sigh; cf. OE. sik. See {Sigh}, v. i.] 1. A deep and (9 more messages) 2010-10-05T01:25:19 contestbot: google sigh 2010-10-05T01:25:20 sigh: Sigh - Wikipedia, the free encyclopedia: ; Sigh (band) - Wikipedia, the free encyclopedia: ; Sigh (Aug 10th Venue Changed to The Underworld!!) on MySpace Music ...: ; Sigh - Definition and More from the Free Merriam-Webster Dictionary: I'm a band? o_O 2010-10-05T01:25:45 apparently so 2010-10-05T01:25:46 lol 2010-10-05T01:27:04 *** yasith_ has quit IRC (Ping timeout: 240 seconds) 2010-10-05T01:30:31 i wish i could play against my own bot ;) 2010-10-05T01:30:35 ala galcon 2010-10-05T01:31:02 would be fun 2010-10-05T01:31:11 I heard mention of people writing a human interface 2010-10-05T01:31:15 no! philhassey is gone 2010-10-05T01:31:24 gone? 2010-10-05T01:31:32 like.. not in this channel 2010-10-05T01:31:36 contestbot: seen philhassey 2010-10-05T01:31:36 amstan: philhassey was last seen in #aichallenge 3 weeks, 3 days, 22 hours, 54 minutes, and 51 seconds ago: http://www.galcon.com/news/2010/09/10/google-ai-challenge-featuring-galcon/ - wheee 2010-10-05T01:32:01 who is philhassey? 2010-10-05T01:32:06 galcon guy 2010-10-05T01:32:10 ahh, cool 2010-10-05T01:33:14 does he have an entry, or was that considered unfair? :P 2010-10-05T01:33:30 sigh: why not? lol 2010-10-05T01:33:32 that would be interesting 2010-10-05T01:33:37 yeah, i twould 2010-10-05T01:33:48 sigh: he actually entered his stock galcon ai 2010-10-05T01:33:56 ahh, awesome 2010-10-05T01:33:57 haha the stock ai is not that good ;) 2010-10-05T01:34:38 no, it's terrible 2010-10-05T01:34:48 I haven't played galcon >_> 2010-10-05T01:34:53 http://ai-contest.com/profile.php?user_id=3907 2010-10-05T01:35:29 well, he's getting more wins than me 2010-10-05T01:35:47 win/loss ratio is irellevant with elo scores 2010-10-05T01:35:54 sigh: he has 10 pages of games too though 2010-10-05T01:35:55 and biased matchings 2010-10-05T01:35:58 I know that :P 2010-10-05T01:36:16 with this tcp thing... do i need to register or am i just making up a username / password 2010-10-05T01:36:26 sometimes I worry that people take me seriously 2010-10-05T01:36:28 hexsprite: make up one, then continue using it for later games 2010-10-05T01:36:45 sigh: wait.. no, lol, you're in the top 2010-10-05T01:36:46 *** Tobu has joined #aichallenge 2010-10-05T01:36:55 sigh: why did i fall for that? 2010-10-05T01:37:09 @rankings 2010-10-05T01:37:10 sigh: Top 10 players: ruilov(3921), davidjliu(3809), dmj111(3753), felixcoto(3698), rsergio(3683), Hazard(3676), sequoh(3660), cfaftw(3660), adumlah(3631), EBraun(3630) 2010-10-05T01:37:15 I don't see me anywhere 2010-10-05T01:37:33 @rankings 25 2010-10-05T01:37:34 amstan: Top 25 players: ruilov(3921), davidjliu(3809), dmj111(3753), felixcoto(3698), rsergio(3683), Hazard(3676), sequoh(3660), cfaftw(3660), adumlah(3631), EBraun(3630), DocBot(3619), goldman(3615), _iouri_(3604), Arekku(3602), youwillfail(3593), Mistmanovx(3566), fglider(3563), sighbot(3562), bix0r4ever(3561), ofrias(3560), McLeopold(3540), savchenko(3512), Polsky(3496), Killer(3490), (1 more message) 2010-10-05T01:37:38 #moar 2010-10-05T01:37:38 amstan: Manwe56(3489) 2010-10-05T01:37:51 aye, i make top 25 still at least aha 2010-10-05T01:38:54 I think it will be a week or more before I have a better bot 2010-10-05T01:39:06 need to make major changes that might not work 2010-10-05T01:39:09 yeah, i'm in no state to be working on a bot 2010-10-05T01:39:10 will ruilov reach 4k today? 2010-10-05T01:39:26 contestbot: less 2010-10-05T01:39:26 i have like 10 assignments due in the next two weeks and a wedding in sydney this weekend :| 2010-10-05T01:39:26 amstan: I'm sorry Dave, I cannot less. 2010-10-05T01:39:37 rabidus: he already did previously 2010-10-05T01:39:44 antimatroid2: you can code during the wedding :P 2010-10-05T01:39:54 antimatroid2: yeah, get a laptop 2010-10-05T01:40:08 oh :O 2010-10-05T01:40:09 laptops are counterproductive to me doing actual work 2010-10-05T01:40:11 I wonder when someone will get a negative elo 2010-10-05T01:40:18 :D 2010-10-05T01:40:28 sigh: jcamero translated them so they're all postive, at my request 2010-10-05T01:40:38 lowest 167 atm 2010-10-05T01:40:39 oh 2010-10-05T01:40:46 so they are going to keep adjusting/ 2010-10-05T01:40:48 ? 2010-10-05T01:41:09 that explains the jump in scores at the top then 2010-10-05T01:41:10 i thought it would be nicer for the lower ranked people to at least have a positive score, even if any reaction based off that is somewhat irrational 2010-10-05T01:41:18 i expect they're just + 2000 or something 2010-10-05T01:41:24 antimatroid2: yeah, I agree 2010-10-05T01:41:39 antimatroid2: nah, my point is that even with the 2000 offset it is getting close to zero 2010-10-05T01:42:07 yeah, perhaps some will drop below 0 then 2010-10-05T01:42:13 given 4000+ was breached 2010-10-05T01:42:22 yup 2010-10-05T01:43:04 i decided the best strategy in the second my of my examples this morning was to send all of your ships for whatever moves into the future to your closest neutral 2010-10-05T01:43:15 but i think you can increase the ship size of the neutrals so that no longer works 2010-10-05T01:43:23 giving the kind of example i was originally aiming ofr 2010-10-05T01:43:26 if that makes any sense? 2010-10-05T01:43:33 antimatroid2: my current bot does something similar 2010-10-05T01:43:41 the bot on the site 2010-10-05T01:44:02 it makes it look very aggressive, and it's not *that* much worse than before 2010-10-05T01:44:23 i want lots of human picked maps in the final comp, people don't want that do they? 2010-10-05T01:44:35 i'd even be happy for lots of well picked 3-4 planet maps being in the final tournament 2010-10-05T01:44:50 *** tapwater has quit IRC (Quit: tapwater) 2010-10-05T01:44:55 no, 100000 planet maps 2010-10-05T01:45:18 you can't have that and expect reasonably intelligent strategies 2010-10-05T01:45:23 sure you can 2010-10-05T01:45:38 first pass is to go through and eliminate 99% of planets from consideration 2010-10-05T01:45:45 at least that's what I would do 2010-10-05T01:46:05 yeah, but then when it gets busy? 2010-10-05T01:46:16 what kind of ai is that what can't handle unlimited amount of planets ? 2010-10-05T01:46:22 you're going to have to ignore large chunks (stupid) or not do much (stupid) 2010-10-05T01:46:24 as there is nearly unlimited amount of planets 2010-10-05T01:46:35 infinite planet map! 2010-10-05T01:46:40 :) 2010-10-05T01:46:45 huge star wars 2010-10-05T01:46:48 implicitly defined, expands as requires 2010-10-05T01:46:49 rabidus: would you be able to work out your moves for a game of galcon with 2 billion planets fairly normal? 2010-10-05T01:46:52 there are limits to things 2010-10-05T01:47:02 me? no. someone? yes. 2010-10-05T01:47:14 von neumann maybe, but he's dead 2010-10-05T01:47:27 but still there would be a winner 2010-10-05T01:47:31 some sort 2010-10-05T01:47:31 whatever, 23 planets is a pretty poor galaxy 2010-10-05T01:47:35 :P 2010-10-05T01:47:40 idd 2010-10-05T01:47:41 yeah, i want more than 23 2010-10-05T01:47:54 but 100000 is a LOT 2010-10-05T01:48:00 08:36 < antimatroid2> i'd even be happy for lots of well picked 3-4 planet maps being in the final tournament 2010-10-05T01:48:04 lol 2010-10-05T01:48:20 that's because he wants to solve with his game theory 2010-10-05T01:48:28 aah 2010-10-05T01:48:41 and by "well-picked" means "solved by antimatroid2" 2010-10-05T01:48:44 *where 2010-10-05T01:48:49 :p 2010-10-05T01:49:29 but everyone else would be expected to do the same, i just consider it interesting to see what the best strategies are 2010-10-05T01:49:37 not what sort of did well on randomly generated maps against people 2010-10-05T01:49:39 why not both? 2010-10-05T01:50:05 actually 100 games with each randomly generated maps are something like 23000 planet map :) 2010-10-05T01:50:22 wh? 2010-10-05T01:50:24 *eh? 2010-10-05T01:50:24 rabidus: you aren't accounting for all those planets at once 2010-10-05T01:51:58 yes, not. but. if you use imagination you can count them as... 'partial win'.. or something. When you win a map, and new map begins, you can think that new planet in new map as old galaxy 2010-10-05T01:52:09 :p 2010-10-05T01:55:22 *** Mathnerd314 has quit IRC (Ping timeout: 255 seconds) 2010-10-05T01:55:23 *** JCS^ has quit IRC (Quit: \x00) 2010-10-05T01:59:17 Top 10 players: ruilov(3916), davidjliu(3801), dmj111(3746), felixcoto(3691), rsergio(3673), Hazard(3671), cfaftw(3665), sequoh(3657), adumlah(3629), EBraun(3616) 2010-10-05T02:03:25 *** bhasker has quit IRC (Ping timeout: 276 seconds) 2010-10-05T02:10:03 *** bhasker has joined #aichallenge 2010-10-05T02:11:58 *** Palmik has joined #aichallenge 2010-10-05T02:20:58 *** wvdschel has joined #aichallenge 2010-10-05T02:29:04 *** hellman has quit IRC (Ping timeout: 265 seconds) 2010-10-05T02:34:50 heh, I just put up a code review for issue 88 and it ended up being issue 188 :P 2010-10-05T02:41:04 *** HakanD has joined #aichallenge 2010-10-05T02:41:12 morning 2010-10-05T02:58:40 *** mega1 has joined #aichallenge 2010-10-05T02:59:17 Top 10 players: ruilov(3961), davidjliu(3840), dmj111(3769), felixcoto(3741), cfaftw(3728), Hazard(3723), rsergio(3696), sequoh(3680), _iouri_(3645), adumlah(3644) 2010-10-05T03:04:35 getting close... 2010-10-05T03:10:14 *** aerique has joined #aichallenge 2010-10-05T03:11:52 *** Tobu has quit IRC (Ping timeout: 252 seconds) 2010-10-05T03:16:42 *** wvdschel has quit IRC (Ping timeout: 264 seconds) 2010-10-05T03:18:54 *** hellman has joined #aichallenge 2010-10-05T03:19:08 *** jaspervdj has joined #aichallenge 2010-10-05T03:28:30 *** Mathnerd314 has joined #aichallenge 2010-10-05T03:29:06 despair setting in 2010-10-05T03:29:54 *** wvdschel has joined #aichallenge 2010-10-05T03:31:58 *** Arthur has quit IRC (Ping timeout: 276 seconds) 2010-10-05T03:32:15 *** wvdschel has quit IRC (Client Quit) 2010-10-05T03:33:44 *** perror has joined #aichallenge 2010-10-05T03:33:55 *** wvdschel has joined #aichallenge 2010-10-05T03:34:23 rwa: got nothing new? make a game tree :P 2010-10-05T03:34:59 tried that a bit 2010-10-05T03:36:18 didn't work well? 2010-10-05T03:36:28 i started, but don't really have time to work on ai for at least a week 2010-10-05T03:36:43 was tough to find a way to model things in a way that was tractable but gave useful results 2010-10-05T03:37:02 yeah, it's going to be messy, but i don't see how else to look ahead 2010-10-05T03:37:21 i think it's better to do lots of moves per layer and less layers rather than try to penetrate very far 2010-10-05T03:37:31 agree 100% 2010-10-05T03:37:37 have you considered even letting current terminal nodes running for a certain number of games themselves? 2010-10-05T03:37:48 possibly even trying to somewhat predict characteristics of their bot? 2010-10-05T03:38:15 it's tricky business working out how to evaluate the payoff for a given gamestate 2010-10-05T03:38:22 antimatroid2: pondered those ideas some, yes, but no impl. yet 2010-10-05T03:38:58 if you're going to try to be predictive, what you should do against good vs bad bots varies wildly. 2010-10-05T03:39:19 although, you can just do well against good bots, and assume that will beat a bad bot 2010-10-05T03:39:23 that is my general approach 2010-10-05T03:40:28 antimatroid2: generally, yes 2010-10-05T03:40:57 it hasn't really failed me on this yet either 2010-10-05T03:40:59 *** JensTi has joined #aichallenge 2010-10-05T03:42:12 *** bartwe has quit IRC (Ping timeout: 252 seconds) 2010-10-05T03:48:14 omg... just found huge bug 2010-10-05T03:48:14 *** hellman has quit IRC (Read error: Connection reset by peer) 2010-10-05T03:48:34 *** hellman has joined #aichallenge 2010-10-05T03:49:21 *** ulope has quit IRC (Ping timeout: 252 seconds) 2010-10-05T03:50:05 *** ulope has joined #aichallenge 2010-10-05T03:50:21 *** JensTi has quit IRC (Quit: Leaving.) 2010-10-05T03:59:18 Top 10 players: ruilov(3917), davidjliu(3844), cfaftw(3825), dmj111(3757), felixcoto(3742), Hazard(3723), rsergio(3686), sequoh(3662), EBraun(3641), adumlah(3639) 2010-10-05T04:04:04 *** HakanD has quit IRC (Ping timeout: 255 seconds) 2010-10-05T04:07:47 *** sigh_ has joined #aichallenge 2010-10-05T04:07:48 *** sigh has quit IRC (Read error: Connection reset by peer) 2010-10-05T04:08:05 *** sinan has joined #aichallenge 2010-10-05T04:08:46 *** Florian has joined #aichallenge 2010-10-05T04:09:43 *** Appleman1234 has joined #aichallenge 2010-10-05T04:19:39 *** keynes has joined #aichallenge 2010-10-05T04:20:00 *** Itkovian has joined #aichallenge 2010-10-05T04:24:03 *** HakanD has joined #aichallenge 2010-10-05T04:26:21 *** Meatkat has joined #aichallenge 2010-10-05T04:32:00 *** oPless has joined #aichallenge 2010-10-05T04:38:16 *** askhader has quit IRC (Remote host closed the connection) 2010-10-05T04:38:16 *** Zepp has quit IRC (Remote host closed the connection) 2010-10-05T04:41:05 *** pgpaskar_ has quit IRC (Ping timeout: 264 seconds) 2010-10-05T04:41:46 quit 2010-10-05T04:41:52 *** Meatkat has quit IRC (Quit: Page closed) 2010-10-05T04:45:26 *** hellman has quit IRC (Ping timeout: 276 seconds) 2010-10-05T04:56:29 *** Utkarsh has quit IRC (Ping timeout: 265 seconds) 2010-10-05T04:58:22 *** Tobu has joined #aichallenge 2010-10-05T04:59:19 Top 10 players: cfaftw(3892), ruilov(3880), davidjliu(3850), Hazard(3772), dmj111(3758), felixcoto(3745), rsergio(3682), sequoh(3655), adumlah(3650), EBraun(3648) 2010-10-05T05:00:02 new leader 2010-10-05T05:02:01 *** sigh has joined #aichallenge 2010-10-05T05:02:24 *** Tobu has quit IRC (Ping timeout: 240 seconds) 2010-10-05T05:04:53 changing quite often (: 2010-10-05T05:05:27 *** Utkarsh has joined #aichallenge 2010-10-05T05:09:00 *** Florian has quit IRC (Disconnected by services) 2010-10-05T05:09:45 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-10-05T05:11:33 *** Florian95 has joined #aichallenge 2010-10-05T05:26:48 *** Naktibalda has joined #aichallenge 2010-10-05T05:32:20 *** Utkarsh has joined #aichallenge 2010-10-05T05:32:28 *** otherAntimatroid has joined #aichallenge 2010-10-05T05:33:38 *** Blkt has joined #aichallenge 2010-10-05T05:34:54 *** MrCeeJ has quit IRC (Quit: Leaving.) 2010-10-05T05:35:43 *** tty2 is now known as tty1 2010-10-05T05:45:42 *** Tobu has joined #aichallenge 2010-10-05T05:48:09 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-10-05T05:54:20 *** Zeiris_ has quit IRC (Ping timeout: 255 seconds) 2010-10-05T05:54:57 *** Utkarsh has joined #aichallenge 2010-10-05T05:59:19 Top 10 players: cfaftw(3872), ruilov(3869), davidjliu(3843), Hazard(3795), dmj111(3756), felixcoto(3742), rsergio(3681), EBraun(3649), sequoh(3644), adumlah(3643) 2010-10-05T06:07:34 ya know what would be cool... if this game had humans and bots playing eachother with each knowing if the other is a player or bot, this way the bot could learn from the humans, and the best ranked bot wins. Also would be interesting to see how the bots compare to the humans overall. 2010-10-05T06:07:51 games of human-vs-human and bot-vs-human and bot-vs-bot all getting posted 2010-10-05T06:11:44 humans have no chance :) 2010-10-05T06:14:20 you can still play thousands of games with your own bot to make it learn 2010-10-05T06:14:38 *** holmar has joined #aichallenge 2010-10-05T06:14:51 is anyone *actually* doing any machine learning stuff... I keep hearing talk but no action 2010-10-05T06:20:18 I don't... 2010-10-05T06:20:39 I may start with it as soon as my university courses start 2010-10-05T06:20:55 for the first semester I have Machine Learning course too 2010-10-05T06:21:04 this may be an interesting playground 2010-10-05T06:24:30 *** ichti has joined #aichallenge 2010-10-05T06:25:00 my attempts at machine learning will be running lots of different parameters against each other and having it try to update them based off what does better/worse 2010-10-05T06:25:10 but i see no point in that untill a lot more has been set up 2010-10-05T06:25:30 *** Appleman1234 has quit IRC (Ping timeout: 245 seconds) 2010-10-05T06:32:31 that is so annoying when enemy crashes and gets the point 2010-10-05T06:33:26 yeah, i just reupload when that happens 2010-10-05T06:33:44 once your bot moves up the rankings it's not an issue 2010-10-05T06:35:44 *** ichti has quit IRC (Ping timeout: 255 seconds) 2010-10-05T06:37:23 *** Appleman1234 has joined #aichallenge 2010-10-05T06:40:41 does it move up? 2010-10-05T06:42:08 *** dmj111 has joined #aichallenge 2010-10-05T06:48:02 *** boegel has joined #aichallenge 2010-10-05T06:49:33 *** drodil has joined #aichallenge 2010-10-05T06:49:36 hi 2010-10-05T06:49:39 hi 2010-10-05T06:52:30 *** hellman has joined #aichallenge 2010-10-05T06:55:30 *** boegel has quit IRC (Quit: Leaving) 2010-10-05T06:55:40 @contestbot time 2010-10-05T06:55:41 rabidus: You have no gotten any error messages recently, so here's a random one just to let you know that we care. 2010-10-05T06:55:55 ... how did i check that? 2010-10-05T06:56:16 ? 2010-10-05T06:56:30 servers time 2010-10-05T06:57:05 it did work for me 2010-10-05T06:57:09 contestbot: time 2010-10-05T06:57:10 Blkt: 06:57 AM, October 05, 2010 2010-10-05T06:57:14 lol 2010-10-05T06:57:55 is there going to be bug fixes for incorrect result of the match any time soon ? 2010-10-05T06:58:09 I doubt it 2010-10-05T06:58:13 :( 2010-10-05T06:58:29 I think they'll fix it in time for the final contest 2010-10-05T06:58:42 hopefully :) 2010-10-05T06:59:20 Top 10 players: ruilov(3895), davidjliu(3870), cfaftw(3843), Hazard(3821), dmj111(3784), felixcoto(3761), rsergio(3704), Arekku(3692), EBraun(3672), adumlah(3671) 2010-10-05T07:00:05 *** dmj111 has quit IRC (Ping timeout: 245 seconds) 2010-10-05T07:10:26 there is still quite a lot of variation in the top 10 2010-10-05T07:13:05 yup 2010-10-05T07:13:12 !predict 2010-10-05T07:13:13 I predict that mietek will win! 2010-10-05T07:13:30 !predict 2010-10-05T07:13:30 I predict that AlliedEnvy will win! 2010-10-05T07:13:38 lol 2010-10-05T07:23:05 *** wvdschel has quit IRC (Quit: Leaving) 2010-10-05T07:23:28 *** wvdschel has joined #aichallenge 2010-10-05T07:28:05 *** wvdschel has quit IRC (Client Quit) 2010-10-05T07:28:24 *** wvdschel has joined #aichallenge 2010-10-05T07:29:29 *** wvdschel has joined #aichallenge 2010-10-05T07:30:25 *** wvdschel has joined #aichallenge 2010-10-05T07:33:29 *** otherAntimatroid has left #aichallenge 2010-10-05T07:34:07 *** BtbN has joined #aichallenge 2010-10-05T07:36:34 http://ai-contest.com/visualizer.php?game_id=5471996 2010-10-05T07:37:29 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-10-05T07:44:35 *** Utkarsh has joined #aichallenge 2010-10-05T07:46:09 *** jhawthorn has quit IRC (Ping timeout: 240 seconds) 2010-10-05T07:48:03 sigh: we're letting the country down 2010-10-05T07:48:38 what country 2010-10-05T07:48:59 australia 2010-10-05T07:49:05 ae 2010-10-05T07:49:11 i'm "youwillfail" on the actual site 2010-10-05T07:49:23 in tribute of robot unicorn attack naturally 2010-10-05T07:49:37 :D 2010-10-05T07:49:38 best served during swot vac 2010-10-05T07:50:09 sigh was in top10 yesterday? 2010-10-05T07:50:16 do i remember correctly 2010-10-05T07:50:37 it was earlier than yesterday, more like last week 2010-10-05T07:51:06 yeah, people keep getting better though 2010-10-05T07:51:21 oh.. 2010-10-05T07:51:36 !predict 2010-10-05T07:51:36 I predict that farbe will win! 2010-10-05T07:51:38 yeah, now I'm not sure if my current bot is worse than my old bot, or if everyone got better 2010-10-05T07:51:48 i was 95% sure that it will me 2010-10-05T07:52:35 non heuristic bots will be cool 2010-10-05T07:52:54 wish i could write in to file... 2010-10-05T07:52:58 all bots must have heuristics of some sort 2010-10-05T07:52:59 so bot would actually learn 2010-10-05T07:53:22 rabidus: play on tcp? 2010-10-05T07:53:42 yeah, but it would be nice if that could be don in real contest too 2010-10-05T07:53:48 *done 2010-10-05T07:54:51 *** jhawthorn has joined #aichallenge 2010-10-05T07:56:09 yes, but as far as i know all are still strictly heuristic based 2010-10-05T07:56:47 rabidus: what can you get from that you can't do from testing on tcp before hand? 2010-10-05T07:57:00 you can update for individual bots as you play them if you want 2010-10-05T07:58:07 *** HakanD has quit IRC (Ping timeout: 264 seconds) 2010-10-05T07:59:22 Top 10 players: ruilov(3884), cfaftw(3875), davidjliu(3846), Hazard(3804), dmj111(3773), felixcoto(3750), rsergio(3694), Arekku(3689), EBraun(3664), sequoh(3661) 2010-10-05T07:59:53 *** Utkarsh has quit IRC (Read error: Connection reset by peer) 2010-10-05T07:59:54 i just mean that a learning bot would be nice. it takes matches against other and trys to learn from mistakes 2010-10-05T08:00:11 that is afaik AI 2010-10-05T08:07:55 *** Utkarsh has joined #aichallenge 2010-10-05T08:08:02 *** delt0r___ has joined #aichallenge 2010-10-05T08:08:52 *** Appleman1234 has quit IRC (Read error: Connection reset by peer) 2010-10-05T08:09:20 *** delt0r_ has quit IRC (Ping timeout: 255 seconds) 2010-10-05T08:12:00 rabidus: plenty of time to learn before the close of submissions 2010-10-05T08:12:31 but AI as a category is pretty board, doesn't have to include learning 2010-10-05T08:13:45 e.g. AI taught at uni would often be more accurately described as "Advanced Search Algorithms" :P 2010-10-05T08:13:49 hehe, i mean that BOT would learn. I would teath that bot to learn :) 2010-10-05T08:14:13 teath a bot? 2010-10-05T08:14:17 lol 2010-10-05T08:14:19 teach 2010-10-05T08:14:19 :D 2010-10-05T08:14:26 yeah I know, I meant there was plenty of time for your bot to learn before the close of submissions 2010-10-05T08:14:33 and you can learn from real games on tcp 2010-10-05T08:14:34 * oPless imagines a bot that bites 2010-10-05T08:14:38 :) 2010-10-05T08:14:55 sigh: but that isn't the same 2010-10-05T08:15:07 why not? 2010-10-05T08:15:08 nah, forget it. 2010-10-05T08:15:13 must get back to work 2010-10-05T08:15:14 :) 2010-10-05T08:15:18 haha 2010-10-05T08:22:03 *** Utkarsh has quit IRC (Read error: Connection reset by peer) 2010-10-05T08:23:02 the ai unit at my uni was terrible 2010-10-05T08:23:18 but so are the rest of the cs units at my uni 2010-10-05T08:23:41 it's not a very strong university 2010-10-05T08:23:44 *** krokokrusa has joined #aichallenge 2010-10-05T08:23:55 my uni had something about Prolog, but I forgot everything long time ago (graduated in 2003) 2010-10-05T08:24:08 yeah, we did some stuff in prolog, but nothing substantial 2010-10-05T08:24:23 they are very math averse in that school, which is retarded given the subject matter 2010-10-05T08:25:43 Naktibalda: haha, same... my uni did its AI subject in Prolog 2010-10-05T08:29:16 *** Haggis has joined #aichallenge 2010-10-05T08:30:12 *** Utkarsh has joined #aichallenge 2010-10-05T08:32:35 I don't think prolog was well suited though... 2010-10-05T08:33:57 *** keynes has quit IRC (Quit: Bye) 2010-10-05T08:35:23 *** HakanD has joined #aichallenge 2010-10-05T08:36:25 *** ichti has joined #aichallenge 2010-10-05T08:58:28 so quiet... 2010-10-05T08:59:21 Top 10 players: ruilov(3880), davidjliu(3848), cfaftw(3828), Hazard(3805), dmj111(3773), felixcoto(3755), Arekku(3701), rsergio(3686), EBraun(3664), sequoh(3660) 2010-10-05T08:59:47 !echo $randomNick: this channel is too quiet - can you do something about it? 2010-10-05T08:59:47 jmpespxoreax: this channel is too quiet - can you do something about it? 2010-10-05T09:02:25 click 2010-10-05T09:04:54 lol 2010-10-05T09:05:37 !echo $randomNick and sigh - can you guys make something exciting? 2010-10-05T09:05:37 HakanD and sigh - can you guys make something exciting? 2010-10-05T09:06:03 hah 2010-10-05T09:07:10 !eightball Should I make something exciting? 2010-10-05T09:07:10 sigh: It shall be. 2010-10-05T09:07:29 hmmm... that seems to imply no effort is required on my part 2010-10-05T09:14:32 *** hellman has quit IRC (Remote host closed the connection) 2010-10-05T09:17:07 *** Utkarsh has quit IRC (Ping timeout: 252 seconds) 2010-10-05T09:18:42 *** cfaftw has joined #aichallenge 2010-10-05T09:22:35 *** Utkarsh has joined #aichallenge 2010-10-05T09:27:10 *** Utkarsh has quit IRC (Ping timeout: 265 seconds) 2010-10-05T09:31:47 haha 2010-10-05T09:32:02 *** Utkarsh has joined #aichallenge 2010-10-05T09:34:23 it's great coming into a channel, having no chatter for like 10 minutes, then seeing somebody laugh 2010-10-05T09:35:59 *** sinan has quit IRC (Ping timeout: 265 seconds) 2010-10-05T09:37:43 :) 2010-10-05T09:40:10 *** Nerd_Skywalker has joined #aichallenge 2010-10-05T09:40:38 A Strange game...the only winning move is not to play 2010-10-05T09:40:44 :-D 2010-10-05T09:41:09 *** Nerd_Skywalker has quit IRC (Client Quit) 2010-10-05T09:41:40 *** criis has quit IRC (Ping timeout: 265 seconds) 2010-10-05T09:46:13 he left... but it's only true if there's one neutral planet. i think... 2010-10-05T09:49:31 or in Wargames 2010-10-05T09:53:11 he meant not to play altogether 2010-10-05T09:53:21 as opposed to not to make a move, I think 2010-10-05T09:53:39 oh 2010-10-05T09:54:51 *** criis has joined #aichallenge 2010-10-05T09:59:21 Top 10 players: ruilov(3862), davidjliu(3834), cfaftw(3817), Hazard(3803), dmj111(3771), felixcoto(3742), Arekku(3713), rsergio(3679), sequoh(3667), EBraun(3657) 2010-10-05T10:14:56 *** ichti has quit IRC (Ping timeout: 264 seconds) 2010-10-05T10:15:24 *** sinan has joined #aichallenge 2010-10-05T10:16:47 *** hellman has joined #aichallenge 2010-10-05T10:19:14 *** Appleman1234 has joined #aichallenge 2010-10-05T10:20:47 *** sigh has quit IRC (Remote host closed the connection) 2010-10-05T10:31:45 *** Utkarsh has quit IRC (Ping timeout: 276 seconds) 2010-10-05T10:36:28 *** Utkarsh has joined #aichallenge 2010-10-05T10:44:37 *** yasith_ has joined #aichallenge 2010-10-05T10:45:27 last (4 2010-10-05T10:45:29 shit 2010-10-05T10:48:06 ? 2010-10-05T10:58:31 accidentally disclosed part of your lisp code? :) 2010-10-05T10:59:22 Top 10 players: ruilov(3873), davidjliu(3836), cfaftw(3825), Hazard(3809), dmj111(3787), felixcoto(3756), Arekku(3719), rsergio(3688), sequoh(3686), DocBot(3657) 2010-10-05T11:00:50 does anyone know what determines how the rate that each bot gets to play games? 2010-10-05T11:00:59 *** aerique has quit IRC (Quit: ...) 2010-10-05T11:01:07 seems like once a bot gets to the top, the game rate slows way down 2010-10-05T11:04:23 I think a rule is that new bots play more :) 2010-10-05T11:12:51 yeah, it tries to favor new bots a little 2010-10-05T11:14:36 *** Itkovian has quit IRC (Quit: Itkovian) 2010-10-05T11:23:45 *** ichti has joined #aichallenge 2010-10-05T11:25:59 *** Utkarsh has quit IRC (Read error: Connection reset by peer) 2010-10-05T11:28:29 *** ichti has quit IRC (Ping timeout: 265 seconds) 2010-10-05T11:32:32 *** ichti has joined #aichallenge 2010-10-05T11:33:04 *** Utkarsh has joined #aichallenge 2010-10-05T11:35:00 i'm impatiently waiting for my bot to play ruilov :) 2010-10-05T11:37:30 *** ichti has quit IRC (Ping timeout: 272 seconds) 2010-10-05T11:40:40 *** Utkarsh has quit IRC (Ping timeout: 272 seconds) 2010-10-05T11:41:00 *** ichti has joined #aichallenge 2010-10-05T11:41:34 *** ichti has quit IRC (Read error: Connection reset by peer) 2010-10-05T11:41:52 *** ichti has joined #aichallenge 2010-10-05T11:42:29 *** rwa has quit IRC (Quit: rwa) 2010-10-05T11:42:34 *** HakanD has quit IRC (Quit: Leaving) 2010-10-05T11:42:58 *** rwa has joined #aichallenge 2010-10-05T11:43:32 *** Meatkat has joined #aichallenge 2010-10-05T11:45:56 *** Utkarsh has joined #aichallenge 2010-10-05T11:46:14 *** ichti has quit IRC (Ping timeout: 252 seconds) 2010-10-05T11:47:27 *** rwa has quit IRC (Ping timeout: 265 seconds) 2010-10-05T11:50:40 *** ichti has joined #aichallenge 2010-10-05T11:53:26 *** Meatkat has quit IRC (Quit: Page closed) 2010-10-05T11:54:19 *** dmj111 has joined #aichallenge 2010-10-05T11:54:27 *** yasith__ has joined #aichallenge 2010-10-05T11:54:47 *** irchs has joined #aichallenge 2010-10-05T11:54:52 Hello... 2010-10-05T11:54:57 Anyway of challenging a player directly? 2010-10-05T11:54:58 :) 2010-10-05T11:55:47 !echo "$randomNick, you are drunk again!" -> " Shut up, its burden enough to see you twice." 2010-10-05T11:55:47 dwchandler, you are drunk again! -> Shut up, its burden enough to see you twice. 2010-10-05T11:56:29 irchs: Play on the TCP server. If you both play at the same time chances are good. 2010-10-05T11:56:43 TCP Server? :) 2010-10-05T11:56:54 contestbot, where tcp 2010-10-05T11:56:54 amstan: tcp = http://www.benzedrine.cx/planetwars/ 2010-10-05T11:57:00 Thank you! 2010-10-05T11:57:29 constesbot, how comes you know so much? 2010-10-05T11:57:32 *** yasith_ has quit IRC (Ping timeout: 264 seconds) 2010-10-05T11:58:20 Do you use the username and pass for the main site? 2010-10-05T11:58:20 !echo I am the matrix. 2010-10-05T11:58:20 I am the matrix. 2010-10-05T11:58:32 You can use any login 2010-10-05T11:58:52 irchs: most ppl choose username-version as their names 2010-10-05T11:59:01 there's also a password that can be anything 2010-10-05T11:59:03 The password just makes sure you dont end up messing with someone elses stats ^^ 2010-10-05T11:59:21 ah I see 2010-10-05T11:59:23 Top 10 players: ruilov(3879), cfaftw(3843), davidjliu(3839), Hazard(3823), dmj111(3784), felixcoto(3755), Arekku(3729), rsergio(3697), sequoh(3678), EBraun(3662) 2010-10-05T11:59:41 One guy has 39 logins there :D 2010-10-05T12:00:29 The results are deleted after 24 hours. So if you want to keep a copy of a game, save the html file. 2010-10-05T12:00:34 amstan: could I get ratings.txt from the server backend directory? 2010-10-05T12:01:01 *** Migi32 has joined #aichallenge 2010-10-05T12:01:08 Mainly I'm wondering if all the confidence intervals given are symmetric 2010-10-05T12:01:52 !echo $randomNick and $randomNick are wondering, too... 2010-10-05T12:01:52 dwchandler and Ari2 are wondering, too... 2010-10-05T12:02:33 somewhat random, anyway 2010-10-05T12:02:45 guys. stop it with those.. 2010-10-05T12:03:32 Is it normal for the tcp client to exit without saying win or lose? 2010-10-05T12:04:44 irchs: yes, it doesn't handle Invalid instructions :) 2010-10-05T12:04:59 From my end or the other end? 2010-10-05T12:05:07 your end 2010-10-05T12:05:10 balls 2010-10-05T12:05:18 Meaning I am spitting out some mess that it doesn't recognise? 2010-10-05T12:05:19 you should see 'I' status in your game list 2010-10-05T12:06:15 http://72.44.46.68/getplayer?player=janstaunton 2010-10-05T12:06:16 :( 2010-10-05T12:06:17 me cries 2010-10-05T12:06:21 * irchs cries 2010-10-05T12:06:22 rather 2010-10-05T12:06:34 log your orders 2010-10-05T12:07:04 aye prolly a good idea 2010-10-05T12:07:06 cheers 2010-10-05T12:09:11 Janzert: having seen it, they are symmetric 2010-10-05T12:09:50 ahh, ok 2010-10-05T12:10:07 *** Frontier has left #aichallenge 2010-10-05T12:10:48 *** Frontier has joined #aichallenge 2010-10-05T12:10:51 i've been wanting to make a visualization of the leaderboard. some sort of box and whisker plot maybe 2010-10-05T12:10:59 now I wonder why, since I have seen bayeselo give nonsymmetric ones before 2010-10-05T12:11:05 i'd need more data than is currently shown though 2010-10-05T12:11:12 Janzert: different parameters 2010-10-05T12:11:23 bayeselo has like 50 different ways to compute stuff 2010-10-05T12:11:45 yeah 2010-10-05T12:12:01 * Janzert goes to dig into the specific way bayeselo is being called 2010-10-05T12:12:12 i believe it's a .sh script 2010-10-05T12:13:03 yes, although it is just feeding the output of readgames.py into bayeselo 2010-10-05T12:14:53 http://pastebin.com/v5NUAwvP 2010-10-05T12:15:55 what does + - mean? :) 2010-10-05T12:16:19 confidence interval 2010-10-05T12:16:31 Thanks amstan 2010-10-05T12:16:34 *** bhasker has quit IRC (Quit: bhasker) 2010-10-05T12:21:45 *** irchs has quit IRC (Quit: irchs) 2010-10-05T12:25:03 hmm, bayeselo is being used completely with just the defaults. If I remember right these give a small boost to the second player in a game because of the first mover advantage chess has. 2010-10-05T12:25:35 on the other hand the winner is always fed to bayeselo as the first player 2010-10-05T12:26:30 *** Bobng has joined #aichallenge 2010-10-05T12:28:41 whoever made the TCP server is my hero 2010-10-05T12:28:49 i just now started using it 2010-10-05T12:31:37 Check out the magical GUI 2010-10-05T12:31:40 ;) 2010-10-05T12:32:51 what magical gui? 2010-10-05T12:34:35 what's so magical about it? 2010-10-05T12:34:38 http://github.com/greghaynes/QTcpPlanetWars 2010-10-05T12:34:44 It has buttons 2010-10-05T12:34:50 Which make it magical 2010-10-05T12:34:58 oh really? I missed them when I tried it :) 2010-10-05T12:35:29 Well, one button 2010-10-05T12:35:33 Its a good one though 2010-10-05T12:36:08 what does it do? 2010-10-05T12:36:29 greghaynes: does it include a windows exe? 2010-10-05T12:37:06 No windows...I should make one though 2010-10-05T12:37:17 Naktibalda: Real time shows you your game on tcp server 2010-10-05T12:37:19 greghaynes: don't expect others to use it then 2010-10-05T12:37:49 Ah, I didnt realize theres a lot of win'ers around 2010-10-05T12:37:55 probably are.. 2010-10-05T12:38:00 you really need more buttons 2010-10-05T12:38:06 Ok, class, back in 1:30 2010-10-05T12:38:08 moar buttons! 2010-10-05T12:38:29 greghaynes: i'm fine with no windows exe :) 2010-10-05T12:38:46 i'm just saying.. some ppl might expect it 2010-10-05T12:39:00 meh 2010-10-05T12:41:28 Janzert: you mean you think the server is not hurting the rank of the loser as much as it should? 2010-10-05T12:41:34 s/rank/score/ 2010-10-05T12:42:53 That's what I was thinking, but now I'm pretty sure it is in fact ok. There is no first mover advantage by default, it's just something that can be set. 2010-10-05T12:43:05 in bayeselo that is 2010-10-05T12:44:41 I'm now trying to find an old post by the bayeselo author that explained many of the options and what he considered the best way to use it 2010-10-05T12:45:19 *** bhasker has joined #aichallenge 2010-10-05T12:47:26 Janzert: what do you think of my suggestion on the ticket by the way? to show the mean and range rather than the mean and distances from the mean? 2010-10-05T12:47:26 hmm, the default confidence intervals are not terribly accurate 2010-10-05T12:47:28 http://talkchess.com/forum/viewtopic.php?topic_view=threads&p=306627&t=30677 2010-10-05T12:48:35 that makes sense to me, the only reservation I have is it doesn't seem to be the "standard" way to present the intervals 2010-10-05T12:49:13 but I'm leaning towards it given the usefullness of it 2010-10-05T12:49:41 "In general, it is not a good idea to infer anything from confidence intervals when trying to compare two programs. Use LOS is better. " 2010-10-05T12:49:45 i'm not sure what to take out of that 2010-10-05T12:49:55 not sure what LOS is, either way. 2010-10-05T12:50:15 this was in a discussion testing changes to chess programs 2010-10-05T12:50:15 oh likelihood of superiority 2010-10-05T12:50:19 yeah 2010-10-05T12:50:42 so it's better to decide if your change is an improvement by using the LOS 2010-10-05T12:50:53 i see 2010-10-05T12:50:56 *** sinan has quit IRC (Ping timeout: 245 seconds) 2010-10-05T12:51:10 things like this are why i really really want tarballs from the server 2010-10-05T12:51:41 stuff that it may be impractical for the site to show, but would still be occasionally useful or interesting 2010-10-05T12:51:49 yeah 2010-10-05T12:52:17 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]) 2010-10-05T12:52:19 I started to work on that then decided displaying CI's was more immediately useful... and probably easier to do as well :) 2010-10-05T12:52:46 yeah 2010-10-05T12:53:12 so I'm thinking that we should feed bayeselo "exactdist" as well before we start displaying the intervals on the site 2010-10-05T12:53:27 that may be a good idea 2010-10-05T12:53:54 I have a feeling the covariance method will be to expensive to use 2010-10-05T12:54:46 yeah, n^3 2010-10-05T12:55:06 actually, n is only in number of players, not in number of games 2010-10-05T12:55:11 i'm not sure how expensive that would be 2010-10-05T12:55:19 hmm, maybe 2010-10-05T12:55:45 maybe a local test can crack this 2010-10-05T12:56:50 I think I'll add a note in the source and maybe danielvf can test it on the server to see what the actual cost is 2010-10-05T12:57:50 so in the rankings showing is probably best 2010-10-05T12:58:18 i do think that would be the most immediately informative 2010-10-05T12:58:19 what do you think about the profile page? show +/- or upper/lower? 2010-10-05T12:58:32 does the profile page even show elo score? 2010-10-05T12:58:42 I added that yes 2010-10-05T12:58:44 ah 2010-10-05T12:58:45 nice 2010-10-05T12:59:10 on the profile page i would probably show it the same way 2010-10-05T12:59:20 also called it estimated elo so it would hopefully help give people the idea that it isn't an exact score 2010-10-05T12:59:23 Top 10 players: ruilov(3878), cfaftw(3842), davidjliu(3838), Hazard(3821), dmj111(3783), felixcoto(3754), Arekku(3728), rsergio(3693), sequoh(3678), DocBot(3662) 2010-10-05T13:00:05 Janzert: do you know what the oppo. column in ratings.txt is? 2010-10-05T13:00:18 average elo of opponents 2010-10-05T13:00:22 ah! 2010-10-05T13:00:31 now it makes sense 2010-10-05T13:02:19 heh, now that I look at it "exactdist" used to be fed to bayeselo but is commented out with the note "we don't really need this", which is true as long as the confidence intervals aren't being used 2010-10-05T13:02:19 hmm... i want a way to sort the leaderboard by different criteria now 2010-10-05T13:02:34 ha! 2010-10-05T13:03:20 now that we will have a patch to show the lower confidence bound and stuff, if we could sort the leaderboard by whatever column we want then i could finally have my "rank by lower confidence bound" idea without imposing it on anybody else 2010-10-05T13:03:47 ooh, yeah. I hadn't thought of that 2010-10-05T13:06:02 anyone here know redditalien? the bot? 2010-10-05T13:06:13 they stole my error message! from contestbot, the one with dave! 2010-10-05T13:06:29 well.. i'm not mad, i just wanna know if they have any association with us 2010-10-05T13:07:33 what's the error message 2010-10-05T13:09:39 cschneid: I'm sorry Dave, I'm afraid i can't do that. 2010-10-05T13:09:48 maybe it's just coincidence 2010-10-05T13:09:50 that's... not a very original quote 2010-10-05T13:09:54 ;) 2010-10-05T13:09:58 yeah, probably 2010-10-05T13:10:02 it would of been cool though 2010-10-05T13:10:09 i've seen other bots say that 2010-10-05T13:10:58 definitely coincidence ;) 2010-10-05T13:11:23 *** jfrank_ has joined #aichallenge 2010-10-05T13:21:24 *** Florian95 has quit IRC (Remote host closed the connection) 2010-10-05T13:21:49 *** JCS^ has joined #aichallenge 2010-10-05T13:24:34 *** Meatkat has joined #aichallenge 2010-10-05T13:28:00 *** justin_pdx has joined #aichallenge 2010-10-05T13:29:42 *** Tobu has quit IRC (Ping timeout: 245 seconds) 2010-10-05T13:29:48 hmm, I'm not sure what the best/most intuitive way to display the actual upper and lower bounds on the profile page is. Using the +/- interval is fairly obvious with "Estimated Elo: 1234 (+123/-123)", maybe "Estimated Elo (lower/mean/upper): 1111/1234/1345"? 2010-10-05T13:30:14 but that still looks rather ugly 2010-10-05T13:30:25 Will the +/- always be the same? 2010-10-05T13:30:30 no 2010-10-05T13:30:58 especially as you get to the top and bottom of the rankings 2010-10-05T13:31:10 Right 2010-10-05T13:31:46 "Estimated Elo (l/m/u) : 1111/1234/1345"? 2010-10-05T13:31:56 i wish we could show a little box and whisker plot 2010-10-05T13:32:18 with the mean written to the side or something 2010-10-05T13:32:33 Meatkat: think anyone will really understand l/m/u without having it explained? 2010-10-05T13:32:38 that would be a lot of work though 2010-10-05T13:32:49 i don't think everybody would understand that 2010-10-05T13:32:57 Janzert: nope 2010-10-05T13:33:21 hmm, maybe I should try making it a tooltip instead so they can see (lower/mean/upper) if they hover over it 2010-10-05T13:33:36 hmm 2010-10-05T13:33:41 although nothing else has that 2010-10-05T13:33:43 *** Bobng has quit IRC (Quit: Leaving) 2010-10-05T13:33:49 but then it's more work to compare bounds 2010-10-05T13:33:54 say of the top ten 2010-10-05T13:34:03 hard to remember all those numbers as you hover over one at a time 2010-10-05T13:34:17 just the explanation in the tooltip 2010-10-05T13:34:22 ah 2010-10-05T13:34:24 that could work 2010-10-05T13:34:46 i think i might look into a visualization of the ladder 2010-10-05T13:34:52 might be nice 2010-10-05T13:36:22 also I changed the ranking list to display the upper and lower bounds so that is where most comparisons will happen I imagine 2010-10-05T13:37:59 *** Meatkat has quit IRC (Quit: Page closed) 2010-10-05T13:39:24 oh, i didn't even notice you were talking about the profile page 2010-10-05T13:39:35 :) 2010-10-05T13:39:56 yeah, "Estimated Elo: 1234 (+123/-123) 2010-10-05T13:39:58 "" 2010-10-05T13:40:02 seems alright to me 2010-10-05T13:40:45 although if we ever show opponent scores on the profile page then we would want to think about reconsidering i think 2010-10-05T13:40:57 since that goes back to the realm of comparing against other bots 2010-10-05T13:42:03 Janzert: hey, could you pastebin the html that your patch is generating for the leaderboard? i want to experiment with a javascript visualization 2010-10-05T13:42:07 hmm, I'll play with it a bit 2010-10-05T13:42:17 or are you not really testing? 2010-10-05T13:42:20 sure 2010-10-05T13:42:22 sweet 2010-10-05T13:43:22 http://pastebin.org/30053 2010-10-05T13:43:30 thanks 2010-10-05T13:43:38 *** Bobng has joined #aichallenge 2010-10-05T13:45:30 *** krokokrusa has quit IRC (Ping timeout: 255 seconds) 2010-10-05T13:50:50 *** Tobu has joined #aichallenge 2010-10-05T13:50:50 *** Tobu has joined #aichallenge 2010-10-05T13:51:55 *** Zeiris_ has joined #aichallenge 2010-10-05T13:53:33 just wanted to mention http://www.kongregate.com/games/tjcarlos/civilizations-wars, a cool game with PlanetWars mechanics 2010-10-05T13:53:53 nice thing to play while thinking about what your next strategy will be 2010-10-05T13:54:40 *** BtbN has quit IRC (Quit: Verlassend) 2010-10-05T13:55:19 Is there any way to contact the contest organisers? email? 2010-10-05T13:57:21 *** fawek has joined #aichallenge 2010-10-05T13:57:29 btw, is it really correct to call the "best guess" elo here the mean? Is there a better term to use? 2010-10-05T13:57:55 Utkarsh: best is probably to either catch them on here or post in the forums 2010-10-05T13:59:24 Top 10 players: ruilov(3885), Hazard(3858), davidjliu(3841), cfaftw(3820), dmj111(3767), felixcoto(3753), Arekku(3730), rsergio(3694), sequoh(3675), DocBot(3665) 2010-10-05T14:02:04 *** perror has quit IRC (Quit: Bye all !) 2010-10-05T14:06:29 *** boegel has joined #aichallenge 2010-10-05T14:07:32 Janzert: mean is the usual term for it, but i don't know if that word still applies with non-symmetric bounds 2010-10-05T14:07:48 it is usually true when it's a normal probability distribution 2010-10-05T14:08:04 maximum likelihood? 2010-10-05T14:08:19 that seems accurate, but a bit of a mouthful 2010-10-05T14:08:27 most likely? 2010-10-05T14:08:37 yeah, accurate but a little long for a tooltip 2010-10-05T14:08:41 jmcarthur, is it not a normal probability distribution? 2010-10-05T14:08:54 Migi32: i don't think it is once we are computing non-symmetric confidence bounds 2010-10-05T14:09:10 * Janzert needs something that fits in between Lower/blah/Upper :) 2010-10-05T14:09:51 I seem to remember something like "greatest likelihood estimation" 2010-10-05T14:09:52 sticking with Mean might be best even though it's not quite right 2010-10-05T14:10:00 Lower/Guess/Upper 2010-10-05T14:10:02 but that's too long of a name 2010-10-05T14:10:06 Lower/Estimate/Upper 2010-10-05T14:10:15 yeah 2010-10-05T14:10:24 if you can't call it mean call it estimate 2010-10-05T14:11:17 that might be better although of course all 3 are estimates 2010-10-05T14:12:28 Janzert: http://i.imgur.com/5Vabz.png 2010-10-05T14:12:32 well, if anyone will complain about the inaccuracy of the term he should have a better name 2010-10-05T14:12:58 Migi32: little progress with your bot? :) 2010-10-05T14:13:07 jmcarthur: sweet 2010-10-05T14:13:08 Janzert: of course those are high scores and i set the range of that plot to 0-3000, so they are kind of small 2010-10-05T14:13:22 and i set it to 100 pixels wide 2010-10-05T14:13:30 not sure if that is too wide or twoo narrow 2010-10-05T14:13:33 *too 2010-10-05T14:14:04 not sure either 2010-10-05T14:14:09 boegel, still coding utility functions, but in my mind my bot has made great progress 2010-10-05T14:14:25 migi32, have you submitted yet? what's your bot's name? 2010-10-05T14:14:51 we could always set the range dynamically (via php), so we can be sure not to make it too big or two small in that sense 2010-10-05T14:14:55 Migi32: but, have you tested it! 2010-10-05T14:15:03 but i am not sure about the rendering width 2010-10-05T14:15:07 *** Itkovian has joined #aichallenge 2010-10-05T14:15:12 i keep saying two instead of too 2010-10-05T14:15:18 on another note I found that you can have bayeselo anchor the elo and wonder if we should change to an anchor with the lowest elo being 0 or 100 instead of just adding 2000 to all of them 2010-10-05T14:15:32 Janzert: that would be awesome and i fully support that 2010-10-05T14:15:40 i hate the artificial score inflation 2010-10-05T14:15:46 yeah 2010-10-05T14:15:58 I think I'll leave that as a separate change though 2010-10-05T14:15:59 cftaftw, no I only fight on TCP as "Migi32", but all my games have gone off the server's cache already 2010-10-05T14:16:23 do you like the plots? do you think they are better than the numbers? 2010-10-05T14:16:40 boegel, since I can't use many libraries there are a lot of functions/classes I have to program myself 2010-10-05T14:16:42 I think I would still want to see the numbers 2010-10-05T14:16:50 maybe the numbers could be a tooltip? 2010-10-05T14:16:51 migi32, what language? 2010-10-05T14:16:52 I'm currently working on a simplex algorithm 2010-10-05T14:16:55 c++ 2010-10-05T14:17:07 the plots do add a nice at a glance understanding 2010-10-05T14:17:25 Migi32: that's C++ for you :P 2010-10-05T14:17:27 * Janzert is still hoping to sort by upper and lower bound columns 2010-10-05T14:17:33 yeah that would be nice 2010-10-05T14:17:53 boegel, yea were you the one who was going to convince me of Haskell tomorrow? 2010-10-05T14:18:18 Migi32: Thursday 2010-10-05T14:18:27 Migi32: so yes, I was :P 2010-10-05T14:18:57 hmm, just realised I should change the database to store the upper/lower bound instead of +/- so it can be an easy sort later 2010-10-05T14:20:22 boegel, oh it was thursday? I don't have classes on thursday, and I don't have a "kot" (what's the English for that?) in Ghent... 2010-10-05T14:20:38 Migi32: where are you from? 2010-10-05T14:20:42 yeah that'd be better 2010-10-05T14:20:42 Deinze 2010-10-05T14:20:47 Migi32: "dorm", I think 2010-10-05T14:20:51 Migi32: hmm, right 2010-10-05T14:20:57 *** optimum has quit IRC (Ping timeout: 245 seconds) 2010-10-05T14:21:02 that's not that far though, right? 2010-10-05T14:21:09 *** amstan has quit IRC (Remote host closed the connection) 2010-10-05T14:21:14 no, half an hour bus ride. 2010-10-05T14:21:22 can you link me to a site with more info on when/where/what? 2010-10-05T14:21:43 Janzert: the nice thing about the plot is when you can actually see how good an estimate is. it may even be as useful as sorting by lower bound when it's easy to see visually: http://i.imgur.com/bTGU4.png 2010-10-05T14:22:23 heh i intended for testplayer2 to actually have a lower lower bound than testplayer3 in that, but i didn't set it low enough 2010-10-05T14:23:00 :) 2010-10-05T14:23:24 jmcarthur, I would draw a 1px black border around that, and also a line indicating the mean ELO 2010-10-05T14:23:33 *** optimum has joined #aichallenge 2010-10-05T14:23:35 Migi32: the red line is the mean 2010-10-05T14:23:41 no the mean overall elo 2010-10-05T14:23:48 huh? 2010-10-05T14:23:58 over all bots? 2010-10-05T14:24:07 the 1400 point (or something) 2010-10-05T14:24:14 eh 2010-10-05T14:24:17 *** smellyhippy has quit IRC (Ping timeout: 245 seconds) 2010-10-05T14:24:25 yes, over all bots 2010-10-05T14:24:37 right now it's just 3 lines floating somewhere 2010-10-05T14:24:47 i agree with the border 2010-10-05T14:24:49 it doesn't really tell you how good the bot is 2010-10-05T14:24:54 *** smellyhippy has joined #aichallenge 2010-10-05T14:24:54 *** smellyhippy has joined #aichallenge 2010-10-05T14:25:06 it's not meant to. it's meant give you a basis for comparison against other nearby bots 2010-10-05T14:25:12 which i think it does fine 2010-10-05T14:25:30 ah ok 2010-10-05T14:25:47 even the border i'm iffy about 2010-10-05T14:25:50 Migi32: http://www.haskell.org/haskellwiki/Ghent_Functional_Programming_Group 2010-10-05T14:25:52 *** irchs has joined #aichallenge 2010-10-05T14:26:25 http://www.dropbox.com/gallery/9140666/1/screenshots?h=011b14 latest screenshots showing lower/mean/upper as I have it right now 2010-10-05T14:26:40 gotta go do some real work before I make more changes :/ 2010-10-05T14:27:05 Technicum is like right by my school :P 2010-10-05T14:28:13 jmcarthur: maybe set the range of the plot by highest and lowest values on the page? 2010-10-05T14:28:23 boegel, what's the average age in a FPG meeting? 2010-10-05T14:28:31 that would seem to make best use of the space 2010-10-05T14:29:00 *** smellyhippy has quit IRC (Ping timeout: 240 seconds) 2010-10-05T14:29:36 Migi32: goh, some are students, others PhD students 2010-10-05T14:29:44 Migi32: I'd say 22-23 orso 2010-10-05T14:29:46 *** prowler has joined #aichallenge 2010-10-05T14:29:50 yo all 2010-10-05T14:29:56 hi prowler 2010-10-05T14:30:02 Migi32: just come and see, you'll enjoy it 2010-10-05T14:30:10 wondering if I need to do anything to bug an issue with a match ending early? 2010-10-05T14:30:11 Migi32: this is what it could look like with the overall mean shown: http://i.imgur.com/s2WD4.png 2010-10-05T14:30:12 Migi32: we usually go out for a beer afterwards 2010-10-05T14:30:29 that's without a border of course 2010-10-05T14:30:33 looking into that now 2010-10-05T14:30:39 boegel, ok. And how many people come there? 2010-10-05T14:30:42 on average 2010-10-05T14:30:55 anyone? 2010-10-05T14:31:01 http://ai-contest.com/visualizer.php?game_id=5486971 2010-10-05T14:31:08 Migi32: depends... can be something like 10, but also 30 2010-10-05T14:31:09 its a common topic i guess on the forum 2010-10-05T14:31:14 Migi32: 30 is *a lot* though 2010-10-05T14:32:13 prowler: What do you mean " 2010-10-05T14:32:19 to bug an issue" ? 2010-10-05T14:32:35 well it seems that some times games just randomly stop 2010-10-05T14:32:46 That is usually bad input. 2010-10-05T14:32:46 i dont see any mention of timeout or invalid commands 2010-10-05T14:32:47 boegel, oh so it's more something for people who already are FP experts rather than to convince people into using FP? 2010-10-05T14:33:04 You either: sent an order for zero ships, sent more ships than a planet had in a turn, or sent ships to yourself. 2010-10-05T14:33:04 http://ai-contest.com/visualizer.php?game_id=5486971 2010-10-05T14:33:09 That's usually what does it for me. 2010-10-05T14:33:21 Or sent ships FROM a planet you didn't own. 2010-10-05T14:33:25 humm i have checks for all of that 2010-10-05T14:33:28 but let me make sure.. 2010-10-05T14:33:29 Migi32: it's intended for everyone who has interest in FP, regardless of their level 2010-10-05T14:33:34 Migi32: I'm by no means an expert 2010-10-05T14:34:02 does it say anywhere that my bot sent an invalid command? 2010-10-05T14:34:06 Migi32: it might be intimidating though, I think this meeting's talks are pretty hard-core 2010-10-05T14:34:08 prowler: If you test your bot on benzedrine it will tell you if you lost due to an invalid command. 2010-10-05T14:34:16 anyway, I'm off 2010-10-05T14:34:20 *** smellyhippy has joined #aichallenge 2010-10-05T14:34:23 It wont tell you what the condition was, but it's slightly helpful. 2010-10-05T14:34:27 ttyl y'all 2010-10-05T14:34:35 *** boegel has quit IRC (Quit: Leaving) 2010-10-05T14:35:02 benzedrine? 2010-10-05T14:35:02 Benzedrine is more strict than your local parser, too. For example, the playgame in the starter package allows you to send ships to yourself and issue orders for zero ships. 2010-10-05T14:35:07 the ftp server? 2010-10-05T14:35:11 tcp. 2010-10-05T14:35:17 sorry ftp 2010-10-05T14:35:22 http://72.44.46.68/ 2010-10-05T14:35:24 right? 2010-10-05T14:35:28 Yeah. 2010-10-05T14:35:32 if so I use that.. never had a disconnect 2010-10-05T14:36:02 Cool, well if you lose it will have an I or D next to your loss to tell you if it was a disconnect or invalid input. 2010-10-05T14:36:26 But a game stop like that is DEFINITELY invalid input. 2010-10-05T14:36:53 if (sPlanet == dPlanet) isValid = false; 2010-10-05T14:36:57 if (sPlanet.Owner != 1) isValid = false; 2010-10-05T14:37:00 if (numShips <= 0) isValid = false; 2010-10-05T14:37:04 if (sPlanet.NumShips < numShips) isValid = false; 2010-10-05T14:37:10 are the checks i have 2010-10-05T14:37:20 if any of them are false it wont send the command 2010-10-05T14:37:39 Would your numShips check be invalid if it was sending multiple orders in a turn? 2010-10-05T14:37:43 *** boegel has joined #aichallenge 2010-10-05T14:37:48 Does it sum the orders? 2010-10-05T14:38:31 thats a good point.. i create the fleets but not sure if I modify the planet ships.. let me check that 2010-10-05T14:39:59 Migi32, Janzert: with border and overall mean: http://i.imgur.com/NPM7j.png 2010-10-05T14:41:39 jmcarthur: hmm, I think I like the border but not the mean 2010-10-05T14:41:44 me too 2010-10-05T14:42:01 i don't think the mean is very useful information 2010-10-05T14:42:10 yeah 2010-10-05T14:42:52 would it be hard to scale it by the highest and lowest values on the page? 2010-10-05T14:43:39 if they can't be derived in the script easily the php should be able to supply them 2010-10-05T14:43:47 that would just need to be added to the php, yeah 2010-10-05T14:44:01 i personally think i'd rather it be scaled by all bot scores though 2010-10-05T14:44:23 hhmmmm.... does it bother anyone else that you can just crash out whilst you are ahead and win? 2010-10-05T14:44:32 http://ai-contest.com/visualizer.php?game_id=5486403 2010-10-05T14:44:55 Migi32, Janzert: border with no mean: http://i.imgur.com/GKyHH.png 2010-10-05T14:45:07 *** optimum has quit IRC (Ping timeout: 245 seconds) 2010-10-05T14:46:11 * Janzert likes that best 2010-10-05T14:46:20 irchs: Who was winning? 2010-10-05T14:46:36 I think ben was 2010-10-05T14:46:43 So ben issued an invalid order and won the game 2010-10-05T14:47:02 Does it go on ships total or on ships on a planet? 2010-10-05T14:47:07 That's weird. 2010-10-05T14:47:12 irchs: the person that crashes loses 2010-10-05T14:47:22 No, he's right though, Janzert. 2010-10-05T14:47:28 The crasher won. 2010-10-05T14:47:29 but the visualizer shows the wrong name for invalid orders 2010-10-05T14:47:32 Oh. 2010-10-05T14:47:36 Useful. 2010-10-05T14:47:38 ..... 2010-10-05T14:47:38 known bug at the moment 2010-10-05T14:47:41 :( 2010-10-05T14:47:41 oh 2010-10-05T14:47:41 -_- 2010-10-05T14:47:50 It is highly likely that I crashed out 2010-10-05T14:47:55 I am doing some strange things atm 2010-10-05T14:48:09 (with my code, not irl :P) 2010-10-05T14:48:17 :) 2010-10-05T14:48:41 How are you guys debugging your work? Running within the visualiser is annoying for debugging 2010-10-05T14:49:00 (I am using Java) 2010-10-05T14:49:25 *** optimum has joined #aichallenge 2010-10-05T14:53:12 files 2010-10-05T14:53:55 is there no way to see why your bot failed to compile? 2010-10-05T14:55:55 jmcarthur, I'd definitely do the border. Mean is up to you (though I'd vote yes as well) 2010-10-05T14:58:02 *** bartwe has joined #aichallenge 2010-10-05T14:58:03 *** emtles has joined #aichallenge 2010-10-05T14:58:18 OK something is seriously broken 2010-10-05T14:58:26 I'm uploading old zips that worked 2010-10-05T14:58:31 and they aren't compiling either 2010-10-05T14:59:24 Top 10 players: ruilov(3894), Hazard(3857), davidjliu(3851), cfaftw(3802), dmj111(3767), felixcoto(3767), Arekku(3766), rsergio(3708), sequoh(3684), EBraun(3684) 2010-10-05T15:02:24 irchs: use the c++ visualizer 2010-10-05T15:02:33 its much faster and lighter 2010-10-05T15:02:57 man my bot sucks at choosing planets to attack -__ 2010-10-05T15:02:59 -_- 2010-10-05T15:03:00 and i rewrote the basic bots in python so that i don't have to deal with Java 2010-10-05T15:03:08 welcome to the club cfat 2010-10-05T15:03:15 i need to fix my turn 1 logic 2010-10-05T15:03:22 my current heuristic is horrible 2010-10-05T15:03:29 going to change it into a 0-1 knapsack 2010-10-05T15:03:44 midgame it still does a decent job of selecting planets 2010-10-05T15:03:49 what's a 0-1 knapsack? 2010-10-05T15:04:05 someone writing in perl? i have a problem with getting fleet's destination 2010-10-05T15:04:18 my bot is actually worse with end-game decisions... 2010-10-05T15:04:23 http://en.wikipedia.org/wiki/Knapsack_problem 2010-10-05T15:04:35 meaning, i'll ahve more ships and more planets 2010-10-05T15:04:38 then it does something dumb 2010-10-05T15:04:59 happens sometimes to my bot too, basically would send ships to acquire neutral ^ 2010-10-05T15:05:01 and die 2010-10-05T15:05:16 so i have added a few checks to not do stuff like that 2010-10-05T15:05:30 and sometimes my heuristics will overestimate the number of ships i can send out 2010-10-05T15:05:42 and bot will send out a few more than i should and lose a critical planet 2010-10-05T15:06:01 definitely the most difficult heuristic is "how many ships do i have available" 2010-10-05T15:06:13 well i do a few things to calculat ethat 2010-10-05T15:06:20 ditto 2010-10-05T15:06:23 but it's still hard to balance 2010-10-05T15:06:26 a) is the planet under attack, if no then its much simple 2010-10-05T15:06:30 you need to get the .. send out what I can send based upon what's attacking me in the future.. working 2010-10-05T15:06:50 well you need to take care also of how many enemy ships are parked close to you 2010-10-05T15:07:01 bhasker, yeah that's the hard part 2010-10-05T15:07:01 and how many can you expect to get from your planets in the time the enemy sends his ships at you 2010-10-05T15:07:02 you should be able to be under constant attack and keep the planet at 0 without losing it 2010-10-05T15:07:19 well prowler that is easy when enemy is streaming ships at you 2010-10-05T15:07:28 but some bots don't do that 2010-10-05T15:07:35 they send a huge attack all of a sudden 2010-10-05T15:07:49 right.. i take into account the 'planet future' 2010-10-05T15:07:51 look at this loss to hazard 2010-10-05T15:07:53 and if I'll lose it.. 2010-10-05T15:07:53 that is a hard one to counter 2010-10-05T15:07:53 http://www.ai-contest.com/visualizer.php?game_id=5486684 2010-10-05T15:08:01 then it becomes another planet that needs to be attacked etc 2010-10-05T15:08:05 i had the advantage 2010-10-05T15:08:19 but i underestimate the threat of his two largest planets 2010-10-05T15:08:31 your mistake was to go after a middle planet 2010-10-05T15:08:37 what i do to prevent that now is 2010-10-05T15:08:41 i queue up orders 2010-10-05T15:08:48 and if there are closer planets than the one i am attacking 2010-10-05T15:08:50 i cancel the order 2010-10-05T15:08:58 is it alwasy a mistake, though? 2010-10-05T15:09:04 in most cases yes 2010-10-05T15:09:06 going for a middle planet has actually won me games 2010-10-05T15:09:14 well not in the middle of 7 of his 2010-10-05T15:09:17 and yours so far 2010-10-05T15:09:17 (not that i specifically target middle planets... it just happens sometimes) 2010-10-05T15:10:05 he can reinforce it much faster than you can 2010-10-05T15:10:14 so unless you send a really really big force 2010-10-05T15:10:18 and capture 2-3 at the same time 2010-10-05T15:10:24 or within next 2-3 turns 2010-10-05T15:10:29 you won't be able to hold it 2010-10-05T15:10:33 true 2010-10-05T15:10:57 but the realy problem there was i overestimated how many ships i could send and still have that planet be safe 2010-10-05T15:11:01 real* 2010-10-05T15:11:05 yes thats true 2010-10-05T15:11:51 http://72.44.46.68/canvas?game_id=324181 2010-10-05T15:11:55 my bot doing dumb stuff 2010-10-05T15:12:06 at turn 34 we are roughly even 2010-10-05T15:12:27 yeah 2010-10-05T15:12:28 at rather at turn 24 ish 2010-10-05T15:12:35 my bot decided to take 3 neutrals at the same time 2010-10-05T15:12:38 and burnt 200 ships 2010-10-05T15:12:40 ... 2010-10-05T15:12:52 setting it at a disadvantage of 50% in ships 2010-10-05T15:13:09 need to add some checks so that it doesn't commit more than a certain fraction at any time to acquiring neutrals midgame 2010-10-05T15:13:28 right now its greedy 2010-10-05T15:13:33 it just takes it if it can 2010-10-05T15:13:56 actually, the main mistake there looks like it has to do with the first planet of yours that he takes 2010-10-05T15:14:30 looks like it was abandoned 2010-10-05T15:15:44 actually, haha 2010-10-05T15:15:48 cfaftw: Just looked at that match. Yeah it was interesting to see. I really wonder what my bot would have done there. How can I check that? 2010-10-05T15:15:48 you're right about the neutral planet thing 2010-10-05T15:16:00 you went for the center planet... that was a mistake 2010-10-05T15:16:18 frontier, you mean my match against hazard? 2010-10-05T15:17:37 frontier, and how can you check what? 2010-10-05T15:17:49 exactly. You are rank 4 and still your bot has some funny flaws. 2010-10-05T15:18:16 I'd like to check how my bot would have decided in this situation 2010-10-05T15:18:42 Can I put the replay in a map file or something? 2010-10-05T15:19:30 *** ebrahim has joined #aichallenge 2010-10-05T15:19:32 Janzert: and finally, with tooltip for the numbers: http://i.imgur.com/xlc6R.png 2010-10-05T15:20:40 *** boegel has quit IRC (Quit: Leaving) 2010-10-05T15:21:02 frontier, yeah you can save the map 2010-10-05T15:21:28 frontier, the map data is in the page source... you can easilly convert it to a regular map file 2010-10-05T15:21:28 looks nice 2010-10-05T15:21:41 But how do I get it out of the JavaScript mangling into something readable ^^? 2010-10-05T15:22:31 i copy/paste, then replace all ':' with '\n', all ',' with ' ' and put a 'P 2010-10-05T15:22:34 'P 2010-10-05T15:22:40 'P ' at the beginning of each line 2010-10-05T15:23:00 use a regex replace, replace '^' with 'P ' 2010-10-05T15:23:16 i won't bother writing a proper patch until we actually have the necessary data in the database 2010-10-05T15:23:44 unless you want to just include it with your branch, Janzert... but i have a feeling you'd still rather stick with the raw numbers 2010-10-05T15:23:52 And will that expand it to a complete maps (planets fleets) for each round? 2010-10-05T15:24:26 *** RainCT has joined #aichallenge 2010-10-05T15:24:51 Also there is a pipe symbol | on the end of a line. What does that mean? 2010-10-05T15:26:26 that separates the planet info from the in-flight fleets 2010-10-05T15:26:54 hold on, let me throw up the rudimentary python script i use to turn the fleet data into a mock bot 2010-10-05T15:28:38 a mock bot? You use replays to debug your bot or did I get that wrong? 2010-10-05T15:29:08 Frontier there is a python script 2010-10-05T15:29:12 that can take the url of the page 2010-10-05T15:29:16 and feed the map data to your bot 2010-10-05T15:29:28 so you can see how your bot will react. 2010-10-05T15:29:35 http://pastebin.com/Cw7Nqems 2010-10-05T15:29:43 I just want to see the decisions from your fatal move onwards. 2010-10-05T15:29:44 that's my script... didn't know there was another one 2010-10-05T15:29:53 there is 2010-10-05T15:29:57 its posted on the forum by daniel 2010-10-05T15:30:15 search for something like parse_game_state* 2010-10-05T15:30:20 cool 2010-10-05T15:30:22 that's even better 2010-10-05T15:30:40 my script just spits out a bunch of 'if' statements with orders based on the current turn 2010-10-05T15:31:15 http://pastebin.com/s1iH92Bh 2010-10-05T15:31:32 you can just run it like 2010-10-05T15:31:34 *** hellman has quit IRC (Remote host closed the connection) 2010-10-05T15:31:45 ./parse_game_state_url.py | 2010-10-05T15:31:58 how you will visualize the whole thing i am not sure yet ^ 2010-10-05T15:33:16 hmm 2010-10-05T15:33:41 *** Arekku has joined #aichallenge 2010-10-05T15:33:54 cfaftw: Thank you, I will use that script as reference to parse the string ;) 2010-10-05T15:36:10 my bot's latest iteration is doing decently except when it does something stupid 2010-10-05T15:36:11 http://72.44.46.68/getplayer?player=bhaskerv20.8 2010-10-05T15:36:12 *** Accoun has quit IRC () 2010-10-05T15:36:29 Oh well, it is as I expected... the planet data is only included once and the rest are more or less incremental updates 2010-10-05T15:37:05 its too aggresive to acquire neutrals 2010-10-05T15:37:15 what kind of heuristic are you guys using to decide when to acquire a neutral 2010-10-05T15:38:20 if I have spare ships and the planet is closer to me 2010-10-05T15:38:21 it's a secret 2010-10-05T15:38:31 sorry jm :D 2010-10-05T15:38:39 :p 2010-10-05T15:38:42 heh 2010-10-05T15:38:49 you can blab yours all you want 2010-10-05T15:39:00 cheers 2010-10-05T15:39:04 my heuristic is obviously stupid lol 2010-10-05T15:39:09 yeah i'm not sure how much i can say without compromising my strategy 2010-10-05T15:39:15 *** hexsprite has quit IRC (Read error: Connection reset by peer) 2010-10-05T15:39:19 growth_rate * length of time I plan to hold it - cost to snag it 2010-10-05T15:39:33 hmm 2010-10-05T15:39:58 *** hellman has joined #aichallenge 2010-10-05T15:40:40 bh do you have an example game? 2010-10-05T15:40:58 where it does stupid stuff? 2010-10-05T15:41:05 yep 2010-10-05T15:41:09 the problem with neutrals is that you can often expect others to snipe them 2010-10-05T15:41:16 http://72.44.46.68/canvas?game_id=324411 2010-10-05T15:41:28 i overcommitted ships 2010-10-05T15:41:31 to acquiring neutrals 2010-10-05T15:41:35 which shifted the balance to the enemy 2010-10-05T15:41:43 i was actually leading 2010-10-05T15:41:47 and bam it reversed 2010-10-05T15:42:21 and my planet selection strategy is certainly not optimal 2010-10-05T15:42:24 it went for a 1 growth planet 2010-10-05T15:42:32 with 60 ships 2010-10-05T15:42:45 i generally try to keep my bot from going for neutrals in the demilitarized zone 2010-10-05T15:42:47 You snagged the wrong neutrals, yeah, that growth_rate 1 planet was a poor choice. 2010-10-05T15:42:49 instead of the 58 with 2 growth 4-5 turns away 2010-10-05T15:43:05 or heck even the 71 2010-10-05T15:43:09 as i said my bot is a bit greedy 2010-10-05T15:43:18 if there is a planet that it can acquire it doesn't hold back 2010-10-05T15:43:22 it will send ships to acquire it 2010-10-05T15:43:23 The 77 and the 71 near your home were much better choices. 2010-10-05T15:43:26 yea 2010-10-05T15:43:28 *** hexsprite has joined #aichallenge 2010-10-05T15:43:45 i need to add some check saying hey if there are 5 other planets better than the current target maybe wait a turn to acquire the better one 2010-10-05T15:43:46 or something 2010-10-05T15:44:45 there must be a ton of inlining going on here: Total Physical Source Lines of Code (SLOC) = 235 ... cabal build 10.59s user 0.47s system 99% cpu 11.121 total ... 33M -rwxr-xr-x 1 jake users 33M Oct 5 14:35 dist/build/planet-wars-jmcarthur/planet-wars-jmcarthur 2010-10-05T15:45:11 yeah 2010-10-05T15:45:22 about the same change I tried to implement 2010-10-05T15:45:42 unfortunately it blew WAY above the time limit 2010-10-05T15:46:01 so I had to restore my backup 2010-10-05T15:46:10 arekku, really? 2010-10-05T15:47:08 well I did a really stupid implementation of it with a order 4 or so algorithm 2010-10-05T15:47:20 so far i haven't come anywhere close to the time limit 2010-10-05T15:47:35 and i have a LOT of o(n^2) loops 2010-10-05T15:47:49 some n^3 2010-10-05T15:47:51 either that or extremely buggy code (ie infinitive loops) 2010-10-05T15:48:22 hmm my code is python 2010-10-05T15:48:26 and it take less than 300ms per turn 2010-10-05T15:48:51 I guess it could be reserving 0 ships for a mission 2010-10-05T15:48:51 cfaftw, I have an O(n^5) loop and I'm fine 2010-10-05T15:49:01 fine, but I'm a smeghead :D 2010-10-05T15:49:35 My bot takes > 5 or 6 seconds per turn. I'm working on paring that down. 2010-10-05T15:49:44 5-6 seconds?? 2010-10-05T15:49:55 are you doing like a game tree search or something 2010-10-05T15:49:56 lol :D 2010-10-05T15:49:58 thats a lot of time per turn 2010-10-05T15:49:58 I need to cache common lookups, like planet distances. 2010-10-05T15:50:08 wtf 2010-10-05T15:50:11 that takes like 0.01 seconds 2010-10-05T15:50:15 at most 2010-10-05T15:50:19 that takes like 0.0000000000000001 seconds 2010-10-05T15:50:25 woowwwww 2010-10-05T15:50:51 Yeah but if you do it for every planet for every OTHER planet, 20 times, it adds up. 2010-10-05T15:50:53 ouch 2010-10-05T15:51:01 *** amstan has joined #aichallenge 2010-10-05T15:51:01 *** ChanServ sets mode: +o amstan 2010-10-05T15:51:06 i cache distances 2010-10-05T15:51:07 yeah... 2010-10-05T15:51:09 why recompute 2010-10-05T15:51:13 its static for next 200 turns 2010-10-05T15:51:14 Laziness. 2010-10-05T15:51:21 every planet distance to every planet is still under 0.0000000001 seconds, really 2010-10-05T15:51:24 I don't care about computation time right now. 2010-10-05T15:51:25 depends language tho 2010-10-05T15:51:36 mine's in python 2010-10-05T15:51:45 each turn takes about 15ms 2010-10-05T15:51:47 did you actually profile for this? 2010-10-05T15:51:50 my first turn takes 1.3 seconds or so 2010-10-05T15:51:52 oh, I speak python! hssss hssss sssss 2010-10-05T15:51:58 when i repeat it several million times 2010-10-05T15:52:03 usually my elapsed time is 0ms 2010-10-05T15:52:18 I actually should profile. 2010-10-05T15:52:19 Arekku: del snake.sound 2010-10-05T15:52:24 i will be taking the full 1 sec per turn, hopefully no more and no less 2010-10-05T15:52:30 I wonder where I spend all my time. 2010-10-05T15:52:35 (of course, we're all using very different system configurations) 2010-10-05T15:52:39 of course, i will have to make it less to be safe :\ 2010-10-05T15:52:44 browsing pr0n? 2010-10-05T15:52:51 justin_pdx, maybe your bot is actually smart :) 2010-10-05T15:52:54 5-6 seconds.... sick 2010-10-05T15:52:55 perhaps I will provide my c++ comrades with a c++ starter package that already caches all the common things 2010-10-05T15:52:58 hmm actually i might need to fix a few things 2010-10-05T15:53:04 mine is taking now close to 800ms total 2010-10-05T15:53:07 by mid game 2010-10-05T15:53:20 and doesn't copy-construct every planet whenever you call pw.Planets() 2010-10-05T15:53:32 I think my bot is pretty smart. It can play in under a second if I reduce the maximum look-forward. 2010-10-05T15:53:54 Also, the maps in the example map directory don't match the server maps. Where do I get the server map package? 2010-10-05T15:54:08 justin_pdx: I think my bot is retarded :P 2010-10-05T15:54:25 The maps in the starter packages are bigger, it seems, than the server maps, am I wrong? 2010-10-05T15:54:52 justin_pdx, it varies 2010-10-05T15:55:14 they seem to all be cut from the same mold... 2010-10-05T15:55:14 I code in ruby so I imagine I'm already at an itty bitty runspeed disadvantage over c++ etc. 2010-10-05T15:55:27 itty bitty? :P 2010-10-05T15:55:45 Yeah, itty bitty. 2010-10-05T15:55:48 :p 2010-10-05T15:55:57 java ftw :D 2010-10-05T15:56:05 ruby has one of the slowest interpreters i know of :P 2010-10-05T15:56:20 java has an awesome VM. too bad its warmup time is so long 2010-10-05T15:56:23 java blows 2010-10-05T15:56:29 takes forever to load 2010-10-05T15:56:30 i hate java the language 2010-10-05T15:56:32 annoying as hell 2010-10-05T15:56:43 yeah well the tutorial is for java 2010-10-05T15:56:44 and yes the language is not very nice either 2010-10-05T15:56:54 it is my opinion that java makes for poor interactive programs 2010-10-05T15:57:00 i am writing in python for now, if when i start hitting timeouts i will recode it in c++ 2010-10-05T15:57:01 the JVM rather 2010-10-05T15:57:01 if needed 2010-10-05T15:57:04 *** hellman has quit IRC (Remote host closed the connection) 2010-10-05T15:57:23 yeah, same here. if my haskell bot ends up still being too slow i will rewrite in c 2010-10-05T15:57:53 no you won't :P 2010-10-05T15:58:05 Is there a spec for what kind of hardware the server is running? 2010-10-05T15:58:22 So we can guess at runspeeds? 2010-10-05T15:58:28 but i don't think it will be a big problem this time. in the tron contest i was trying to make a point that haskell could run nearly as fast as c. i think i showed that. this time, i'm also certain to reach the same goal 2010-10-05T15:58:37 Is it more powerful than my laptop or not? 2010-10-05T15:58:45 probably 2010-10-05T15:58:46 justin_pdx, there was. Let me scroll up and see if I can find it back. 2010-10-05T15:58:47 i worked hard in the tron contest to do that. i think it won't be as much work this time if i can use GHC HEAD 2010-10-05T15:58:53 at my job, our server code-base is all python and C, but we will be refactoring soon and my supervisor really wants to do it all in java 2010-10-05T15:58:57 i might have to find a new job 2010-10-05T15:58:59 -_- 2010-10-05T15:59:25 Top 10 players: davidjliu(3905), ruilov(3884), Hazard(3842), cfaftw(3789), Arekku(3784), dmj111(3772), rsergio(3732), EBraun(3683), adumlah(3670), DocBot(3666) 2010-10-05T15:59:30 haha 2010-10-05T15:59:32 why would you take perfectly good python and C and turn it into java? 2010-10-05T15:59:43 because that's the language he likes 2010-10-05T15:59:48 doh! 2010-10-05T16:00:03 can't find it back. But it was an average 32-bit cpu 2010-10-05T16:00:05 oh well 2010-10-05T16:00:52 Migi32: i misparsed that sentence as "average-32-bit cpu" as though the cpu was variable bits 2010-10-05T16:00:57 i was confused 2010-10-05T16:01:01 lol :D 2010-10-05T16:01:28 *** Palmik has quit IRC (Remote host closed the connection) 2010-10-05T16:02:02 cftawtw, I share your dislike for Java. Imho, any programming language that doesn't offer multiple inheritance is not worth looking at :P 2010-10-05T16:02:11 o_O 2010-10-05T16:02:45 you can o_O me all you want until you find a case where you'll actually need multiple inheritance. 2010-10-05T16:02:51 i think multiple inheritance is a bad solution to a real problem. language without a solution to that problem claim it does not exist, and C++ just mutilates it :P 2010-10-05T16:02:58 haha, that's... not the reason i don't like java 2010-10-05T16:03:36 Migi32: i can tell you all kinds of such cases. i just don't like multiple inheritance as a solution 2010-10-05T16:03:45 java is too much like a straitjacket 2010-10-05T16:03:51 it's only needed because the language is retarded in other ways 2010-10-05T16:04:11 other than that it's a fine language, just don't make me use it :P 2010-10-05T16:04:15 I dislike java cause it's very clunky with things that should be easy, like sorting an array. 2010-10-05T16:04:35 yeah, nothing quite like [a,b,c].sort or sort [a,b,c] 2010-10-05T16:04:48 Iterators and parsers and extra objects, and collections, and buffers and blah blah blah. 2010-10-05T16:04:57 Janzert: Agreed 2010-10-05T16:05:01 jmcarthur: Exactly. 2010-10-05T16:05:50 i have no problem with the existence of iterators and parsers and extra objects, and collections, and buffers 2010-10-05T16:05:53 *** Mihaime has joined #aichallenge 2010-10-05T16:05:56 I have to std::binary_sort(my_array.begin(), my_array.end()); but having to type much is not a reason to dislike a language 2010-10-05T16:06:04 i just don't think java provides good interfaces for them 2010-10-05T16:06:16 and i don't like that java requires you to write them out fully for every single little thing 2010-10-05T16:06:35 Migi32: yeah, having to read much is a better reason ;) 2010-10-05T16:07:14 what really was a deal-breaker for Java was the fact that an interface has to know all the types of exceptions any of its implementations can throw 2010-10-05T16:07:25 that's not true 2010-10-05T16:07:30 it's not? 2010-10-05T16:07:34 that's only true for checked exceptions 2010-10-05T16:07:40 it also has unchecked exceptions 2010-10-05T16:08:31 *** delt0r_ has joined #aichallenge 2010-10-05T16:08:40 but i agree that the exception situation still sucks 2010-10-05T16:08:57 my bot is begging for a rewrite 2010-10-05T16:09:10 sometimes i'm afraid somebody will see my code... and judge me 2010-10-05T16:09:47 *** delt0r___ has quit IRC (Ping timeout: 240 seconds) 2010-10-05T16:09:57 my code literally contains this comment: 2010-10-05T16:10:00 If your bot puts out the right moves then they can't judge you. 2010-10-05T16:10:03 / _ 2010-10-05T16:10:03 / / \ 2010-10-05T16:10:03 / / ! \ WARNING! 2010-10-05T16:10:03 / /_____\ HIGHLY EXPLOSIVE. USE WITH CARE. 2010-10-05T16:10:08 *** mceier has joined #aichallenge 2010-10-05T16:10:16 haha, same 2010-10-05T16:10:18 i think i write nice code. i'm definitely going to be making it public once the contest is over 2010-10-05T16:10:18 ha, nice 2010-10-05T16:10:36 I hope the contest goes on past it's end date. 2010-10-05T16:10:39 Just for fun 2010-10-05T16:10:48 jmcarthur, i'm actually going to do the same... i'm really looking forward to seeing what everyone's code was actually doing 2010-10-05T16:10:48 did the Tron contest do that? 2010-10-05T16:10:54 it did for a little while on the tcp server for the tron one 2010-10-05T16:11:02 you can figure out what other bots are doing just by watching, but only to a limited extent 2010-10-05T16:11:28 jmcarthur: I think the confidence bounds branch is ready to go again unless you see something else that should change before it goes in 2010-10-05T16:11:34 I know if I'd win I'd keep my bot running on tcp to see how long it can stay unbeaten 2010-10-05T16:11:45 or well, how long it can stay strong 2010-10-05T16:11:48 once it goes in I'd like to see about adding the whisker plots 2010-10-05T16:11:57 Janzert: sweet 2010-10-05T16:11:58 migi32, i'm looking forward to seeing your bot 2010-10-05T16:12:03 Janzert: i'll take a look in a bit 2010-10-05T16:12:11 and sorting by lower bound too 2010-10-05T16:12:24 i expect nothing short of total domination 2010-10-05T16:12:32 i don't know how we would go about this if we also show the whisker plots :\ 2010-10-05T16:12:37 *** Appleman1234 has quit IRC (Ping timeout: 245 seconds) 2010-10-05T16:12:47 cfaftw, disappoint you, I will not. 2010-10-05T16:13:06 I like Java. It is an easy to learn and read language. That's why supervisors love it. Practically everyone uses the same code style. 2010-10-05T16:13:16 easy to learn. not, IMO, easy to read 2010-10-05T16:13:27 my thought right now is if we don't show columns for upper and lower bound then clicking on the elo column would toggle through elo, lower bound and upper bound 2010-10-05T16:13:58 hmm 2010-10-05T16:14:35 i suppose that could work 2010-10-05T16:15:22 jmcarthur: Compared to C code with lots of macros and conditional defines it is easy to read. But they introduced templates and annotations a while back so it becomes more and more obscure like any other language. ^^ 2010-10-05T16:15:49 Frontier: C and Java are for very different domains though 2010-10-05T16:16:03 Frontier: and doesn't Java have generics, not templates? 2010-10-05T16:16:04 first though I really want to get started recording and displaying the bounds the rest is :) 2010-10-05T16:16:13 Janzert: agreed 2010-10-05T16:16:41 i really want to see the whiskers on the leaderboard though. i've been secretly wanting that for a long time now 2010-10-05T16:16:56 jmcarthur: Right, Java cannot do as much as C++ with templates. I just tend to call them so. 2010-10-05T16:17:23 yeah, they will be cool 2010-10-05T16:17:27 *** Accoun has joined #aichallenge 2010-10-05T16:17:32 generics and templates aren't really so bad semantically. they just suck syntactically in Java and C++ 2010-10-05T16:17:47 arekku, your bot has been gaining a lot of traction 2010-10-05T16:18:14 btw, feeding the javascript the highest and lowest values on the page in a hidden div after the table should be really simple. 2010-10-05T16:18:15 Frontier, no language can do as much as c++. Just look at the c++0x standard. I bet there are less than 100 people on this planet who know every detail of c++ 2010-10-05T16:18:28 if that works for the javascript to get the values 2010-10-05T16:18:32 I also like how everything is an object in Java with no exception. That makes for some awesome features like arrays of classes, having methods as objects or constructors so you can pass them around 2010-10-05T16:18:44 my current bot is going to slip to 5 or 6, most likely 2010-10-05T16:18:46 Migi32: the complexity of the spec says very little about the power or expressiveness of the language 2010-10-05T16:18:48 until i resubmit 2010-10-05T16:18:52 actually I should just put them in a script tag 2010-10-05T16:19:07 Frontier: there are tons of exceptions. none of the primitive types in java are objects 2010-10-05T16:19:27 lol, as i say that i go to 6 2010-10-05T16:19:48 Migi32: I just tried to get a function pointer out of a share library... did you know you cannot do that with C++? You cant just cast the returned void* into a function pointer. It is rediculous ^^ 2010-10-05T16:19:49 cfaftw: good call :) 2010-10-05T16:20:17 *** jaspervdj has quit IRC (Quit: NEVER GONNA GIVE YOU UP NEVER GONNA LET YOU DOWN) 2010-10-05T16:20:25 Frontier: you also can't call main recursively in C++ o_O 2010-10-05T16:20:45 *** ebrahim has quit IRC (Remote host closed the connection) 2010-10-05T16:20:47 jmcarthur: They all have wrappers if you need them. If you don't need them - fine - your program runs faster. 2010-10-05T16:21:05 Frontier: i was just refuting your "no exceptions" statement 2010-10-05T16:21:46 cfaftw: pure luck with the rng 2010-10-05T16:22:06 rng? 2010-10-05T16:22:10 *** Sluf has joined #aichallenge 2010-10-05T16:22:13 hiya 2010-10-05T16:22:16 *** u1tr0n has joined #aichallenge 2010-10-05T16:22:19 Anyone else having problem with the site upload script? 2010-10-05T16:22:20 hi all 2010-10-05T16:22:24 yep 2010-10-05T16:22:29 again? 2010-10-05T16:22:32 jmcarthur: You failed, because Java implicitly converts primitive types to objects where needed! For example when you use generics. A list of Integer objects for example 2010-10-05T16:22:32 cfaftw: random number generator, as I'n I've had favourable matchups 2010-10-05T16:22:34 amstan? ^^ 2010-10-05T16:22:35 Warning: mkdir() [function.mkdir]: Too many links in /var/www/vhosts/ai-contest.com/www/check_submit.php on line 24 2010-10-05T16:22:39 j3camero: ^^ 2010-10-05T16:22:42 ugh 2010-10-05T16:22:58 it'll drop down eventually 2010-10-05T16:23:07 the admins really should just make things go into subdirs 2010-10-05T16:23:31 instead of submissions/botid, it should be submissions/b/o/t/i/d 2010-10-05T16:23:33 for last 2 day found some free time to make bot and blah... uploading error ^( 2010-10-05T16:23:43 damn ^^ i have been working all day on my script, i wanted to test it 2010-10-05T16:23:57 i believe there is a cron job to clean things up now 2010-10-05T16:24:01 lol ultron, it feels like we're cloned 2010-10-05T16:24:01 it's just apparently not running often enough 2010-10-05T16:24:11 yep 2010-10-05T16:24:35 oh 2010-10-05T16:24:46 why isn't the upload working ? 2010-10-05T16:24:47 :< 2010-10-05T16:24:55 *** Naktibalda has joined #aichallenge 2010-10-05T16:25:03 drodil: because somebody did the server wrong! 2010-10-05T16:25:11 crap :< 2010-10-05T16:25:32 oh sh... the server cannot take any more submissions because of the directory structure? 2010-10-05T16:25:33 and now i don't have bot since i tried to upload new one 2010-10-05T16:25:56 amstan: Your call! 2010-10-05T16:26:00 people have uploaded so many uncompiled bots, the server has given up 2010-10-05T16:26:01 yes... sssh 2010-10-05T16:26:08 i'm trying to find jeff's command 2010-10-05T16:26:31 shame on you, who won't compile your bots first >:'( 2010-10-05T16:27:36 *** Appleman1234 has joined #aichallenge 2010-10-05T16:28:17 the server won't use precompiled bots, Sluf 2010-10-05T16:28:29 can someone help me? 2010-10-05T16:28:36 look though the repo for a cleanup script 2010-10-05T16:28:49 it sounded to me like something that wasn't in the repo, last night 2010-10-05T16:28:56 amstan: I'm pretty sure there it sin't in the repo 2010-10-05T16:29:27 well.. there's nothing in his home folder 2010-10-05T16:29:32 err, I'm pretty sure there isn't anything like that in the repo 2010-10-05T16:29:36 check the crontab and see if he added it 2010-10-05T16:30:24 if this thing didn't clear away enough to even last for a day then a better solution needs to be found anyway 2010-10-05T16:30:29 *** wvdschel has quit IRC (Quit: Leaving) 2010-10-05T16:30:55 it's not in the global crontab 2010-10-05T16:31:00 and if it were.. it's not fast enough 2010-10-05T16:31:05 *** Lithosphere has joined #aichallenge 2010-10-05T16:32:00 well, i would have said to run it every hour or so 2010-10-05T16:32:06 because clearly once a day isn't enough 2010-10-05T16:32:12 and then in the meantime find something better 2010-10-05T16:32:22 cron.hourly is empty 2010-10-05T16:32:29 like many of us have said already, submissions/b/o/t/i/d would be great 2010-10-05T16:32:36 modify the upload script to nuke the user's previous entries 2010-10-05T16:32:47 that would be a good fix 2010-10-05T16:32:52 easy too 2010-10-05T16:33:09 well, easier than my suggestion at least, since i assume it would change less code 2010-10-05T16:33:10 *** Appleman1234 has quit IRC (Ping timeout: 276 seconds) 2010-10-05T16:33:16 jmcarthur: this is not a thing that i can do now 2010-10-05T16:33:21 go ask about it on the mailing list 2010-10-05T16:33:27 i need to find the cleanup script 2010-10-05T16:33:27 *** optimum has quit IRC (Ping timeout: 245 seconds) 2010-10-05T16:33:41 *** Umsturz has joined #aichallenge 2010-10-05T16:35:01 *** cfaftw has quit IRC (Quit: Leaving) 2010-10-05T16:35:05 k 2010-10-05T16:35:06 found it 2010-10-05T16:35:13 python delete_some_old_submissions.py 100 2010-10-05T16:35:30 k.. 2010-10-05T16:35:32 cleared 1000 2010-10-05T16:35:41 alright, it's on the mailing list 2010-10-05T16:35:53 that script name sounds.... not good 2010-10-05T16:35:59 delete *some* old submissions? 2010-10-05T16:36:08 only cleared 1000?! 2010-10-05T16:36:11 and a parameter 2010-10-05T16:36:21 i am confused by this 2010-10-05T16:36:25 mv delete_some_old_submissions.py try_delete_some_maybe_old_submissions_maybe.py 2010-10-05T16:36:29 lol 2010-10-05T16:36:33 *** pmachado has joined #aichallenge 2010-10-05T16:36:52 lovely how the parameter doesn't even specify the number to delete. what is it even for? 2010-10-05T16:37:12 i actually ran the 100 2 times, then the 1000 2010-10-05T16:37:13 i assume that it does 2010-10-05T16:37:15 hehe, would a script that deletes all but say any users last 3 submissions be wanted? 2010-10-05T16:37:19 ah 2010-10-05T16:37:33 that would be pretty reasonable, Janzert 2010-10-05T16:38:04 wouldn't it better to have 1 directory per user, and keep only the latest submission here? 2010-10-05T16:38:33 even having a directory per user and keeping all of that user's submissions would be fine 2010-10-05T16:38:42 i doubt a user would upload 32000 submissions, after all 2010-10-05T16:38:51 although that could be a DOS maybe 2010-10-05T16:38:54 yes, but would require a lot of changes in the current code base 2010-10-05T16:39:04 you can't keep all of a user's submissions in one directory 2010-10-05T16:39:20 that either overwrites, or you have subdirectories 2010-10-05T16:39:27 Zannick: i meant with subdirectories 2010-10-05T16:40:00 ah 2010-10-05T16:40:36 actually I've gotta run for about 6 hours so probably won't be able to write that script now 2010-10-05T16:43:59 *** optimum has joined #aichallenge 2010-10-05T16:45:08 *** eager has joined #aichallenge 2010-10-05T16:46:11 *** eager has quit IRC (Client Quit) 2010-10-05T16:47:24 *** pmachado has quit IRC (Quit: Page closed) 2010-10-05T16:48:03 *** emtles has quit IRC (Quit: Page closed) 2010-10-05T16:48:09 *** Appleman1234 has joined #aichallenge 2010-10-05T16:50:44 *** Mathnerd314 has quit IRC (Ping timeout: 276 seconds) 2010-10-05T16:51:10 *** serialx has joined #aichallenge 2010-10-05T16:51:57 *** serialx has quit IRC (Client Quit) 2010-10-05T16:52:14 k.. do submissions work now? 2010-10-05T16:52:55 *** Appleman1234 has quit IRC (Ping timeout: 272 seconds) 2010-10-05T16:57:50 *** Mathnerd314 has joined #aichallenge 2010-10-05T16:57:53 *** amstan has quit IRC (Ping timeout: 240 seconds) 2010-10-05T16:58:37 *** JensTi has joined #aichallenge 2010-10-05T16:59:26 Top 10 players: davidjliu(3904), ruilov(3881), Hazard(3823), dmj111(3779), Arekku(3755), cfaftw(3750), rsergio(3733), EBraun(3684), adumlah(3676), DocBot(3669) 2010-10-05T17:01:21 Yes, Amstan, it's perfect :) 2010-10-05T17:01:28 thank you 2010-10-05T17:01:43 *** wh1teside has quit IRC (Ping timeout: 255 seconds) 2010-10-05T17:03:36 *** wh1teside has joined #aichallenge 2010-10-05T17:18:47 so, invalid input, does sending too many ships from a planet when it doesn't have enough count as invalid, or does it just deal with it? 2010-10-05T17:19:45 So if a planet has 10 ships, and I request 20 to move from it, will it crap out? 2010-10-05T17:20:55 yes 2010-10-05T17:21:23 damn it 2010-10-05T17:21:27 That explains a lot... 2010-10-05T17:21:31 * irchs needs sanity checking 2010-10-05T17:28:32 thanks btw 2010-10-05T17:29:15 *** sinan has joined #aichallenge 2010-10-05T17:33:32 *** Blkt has quit IRC (Ping timeout: 265 seconds) 2010-10-05T17:37:00 *** Blkt has joined #aichallenge 2010-10-05T17:39:27 *** justin_pdx has quit IRC (Quit: justin_pdx) 2010-10-05T17:40:24 *** BtbN has joined #aichallenge 2010-10-05T17:47:19 * jmcarthur has found a function that causes his compilation time to increase by a half second for every time he uses the function 2010-10-05T17:52:31 *** mega1 has quit IRC (Ping timeout: 252 seconds) 2010-10-05T17:52:49 haha 2010-10-05T17:53:48 i think it's because it's creating a fusion point in my code 2010-10-05T17:54:10 so all the intermediate loops between uses of that function are being fused into one 2010-10-05T17:55:01 I started on my loop, of a loop, of a loop of a loop, and thought to myself.. I should rethink this. 2010-10-05T17:55:15 heh, yeah, that happens 2010-10-05T17:55:35 ghc is handling this nested stuff for me. that's where the compilation overhead is coming from 2010-10-05T17:56:07 i'm saying "loop, then loop, then loop, then loop" and ghc is translating that into "loop of loop of loop of loop" 2010-10-05T17:56:56 or actually, i'm saying "loop a, then loop b, then loop c, then loop d" and ghc is tranlating is into "loop of a b c and d" 2010-10-05T17:57:13 and there are nested loops where needed 2010-10-05T17:57:23 when i was taught sml, i remember a series of functions that would take it exponentially long to determine the types of 2010-10-05T17:57:33 yeah, type inference is exponential 2010-10-05T17:57:38 *** optimum has quit IRC (Ping timeout: 245 seconds) 2010-10-05T17:57:43 well, HM type inference is, i mean 2010-10-05T17:58:04 but it's normally not a problem 2010-10-05T17:59:25 Top 10 players: ruilov(3905), davidjliu(3888), Hazard(3848), dmj111(3797), Arekku(3755), cfaftw(3752), DocBot(3691), adumlah(3687), EBraun(3665), sequoh(3643) 2010-10-05T18:00:48 *** amstan has joined #aichallenge 2010-10-05T18:00:48 *** ChanServ sets mode: +o amstan 2010-10-05T18:01:54 damn the top bots keep getting better 2010-10-05T18:02:09 i do not see that as a bad thing 2010-10-05T18:02:22 *** sigh_ is now known as sigh 2010-10-05T18:02:22 well it is:p when my bot isnt improving by that much ^ 2010-10-05T18:02:35 jmcarthur: how is your bot coming along? 2010-10-05T18:02:39 *** pgpaskar_ has joined #aichallenge 2010-10-05T18:02:48 have you submitted anything yet? 2010-10-05T18:02:52 *** sigh is now known as Guest7586 2010-10-05T18:02:58 *** Umsturz has quit IRC (Quit: Umsturz) 2010-10-05T18:03:00 sigh: i'm about to have to write out a simplex function :\ 2010-10-05T18:03:04 *** RainCT has quit IRC (Remote host closed the connection) 2010-10-05T18:03:06 *** optimum has joined #aichallenge 2010-10-05T18:03:14 haha, I'll have to do that too sometime soon 2010-10-05T18:03:23 errr 2010-10-05T18:03:24 i think a lot of us are 2010-10-05T18:03:41 *** Guest7586 is now known as sigh_ 2010-10-05T18:04:30 *** sigh_ is now known as sigh 2010-10-05T18:04:51 ok, got my nick back :P 2010-10-05T18:05:00 hmm... i think i'm at that point now, actually. time to read up on simplex 2010-10-05T18:05:50 :) 2010-10-05T18:06:21 ew, wikipedia is not a good reference for this 2010-10-05T18:06:29 haha 2010-10-05T18:06:43 there are plenty of code implementations around on the net 2010-10-05T18:09:10 * jmcarthur considers just binding to an existing C implementation or something 2010-10-05T18:09:40 yeah, that would be best... assuming that you can get it compile on the contest servers 2010-10-05T18:09:51 that's the main thing with libraries 2010-10-05T18:09:57 i'm still holding out hope for binary submissions 2010-10-05T18:10:11 :) 2010-10-05T18:10:20 i could be really screwed if that falls through 2010-10-05T18:10:28 :( 2010-10-05T18:10:55 at least i will have most of it already speced out for a rewrite in C, if i have to do that 2010-10-05T18:11:27 heh 2010-10-05T18:12:38 *** sinan has quit IRC (Ping timeout: 245 seconds) 2010-10-05T18:15:54 wtf, my bot climbed 10 ranks from yesterday... this is wierd 2010-10-05T18:16:16 @rankings 20 2010-10-05T18:16:17 amstan: Top 20 players: ruilov(3911), davidjliu(3889), Hazard(3833), dmj111(3803), Arekku(3756), cfaftw(3735), DocBot(3695), adumlah(3688), EBraun(3666), sequoh(3642), sighbot(3638), goldman(3638), chotting(3636), Rexxar(3633), _iouri_(3625), bix0r4ever(3623), savchenko(3612), youwillfail(3605), Mistmanovx(3598), fglider(3589) 2010-10-05T18:16:36 amstan: I might get back into the top10 without doing anything :P 2010-10-05T18:16:39 lol 2010-10-05T18:20:16 fun fact: this current bot has a bug that makes it incapable of stealing neutrals... I don't think it makes a difference at this level though 2010-10-05T18:28:13 yeah screw this. i'm using existing code 2010-10-05T18:28:33 that's the way :) 2010-10-05T18:28:42 good night everyone 2010-10-05T18:28:48 bye 2010-10-05T18:31:14 jmcarthur, you're going to use simplex as well? What for? 2010-10-05T18:31:21 *** Blkt has quit IRC (Quit: Error: do not makunbound t please) 2010-10-05T18:31:57 i'm actually going to be using it for two things 2010-10-05T18:33:40 but sorry for not answering your question :) 2010-10-05T18:33:50 *** Meatkat has joined #aichallenge 2010-10-05T18:33:53 *** optimum has quit IRC (Ping timeout: 245 seconds) 2010-10-05T18:34:05 jmcarthur, is the tree search node evaluation one of the two? 2010-10-05T18:34:19 i'm going to use duplex 2010-10-05T18:34:21 saves paper 2010-10-05T18:34:29 well, i'm doing search, but not tree search 2010-10-05T18:34:43 you could say that one of my uses of simplex will have to do with evaluation though 2010-10-05T18:34:47 *** Mathnerd314 has quit IRC (Disconnected by services) 2010-10-05T18:35:17 i am not intending to hint that i'm solving the game or something, btw. i wish i was, but i'm not 2010-10-05T18:35:37 *** Mathnerd314_ has joined #aichallenge 2010-10-05T18:35:40 *** Mathnerd314_ is now known as Mathnerd314 2010-10-05T18:35:48 Zannick, what is duplex? Can't find much about it on google 2010-10-05T18:36:04 it's a method of printing on both sides of a piece of paper 2010-10-05T18:36:08 simplex uses only one side 2010-10-05T18:36:10 ;) 2010-10-05T18:36:26 haha :D 2010-10-05T18:36:30 so simplex is half duplex? 2010-10-05T18:36:37 that joke went totally over my head :P 2010-10-05T18:37:03 I was already reading the wikipedia on "Vedic duplex method for extracting a square root" 2010-10-05T18:37:09 lol 2010-10-05T18:37:12 simplex is duplex with every other page blank 2010-10-05T18:38:14 i guess i should be more honest. my search algorithm is tree search if you squint really really hard 2010-10-05T18:38:26 * sigh squints 2010-10-05T18:38:44 * sigh squints harder 2010-10-05T18:39:06 I can't see jmcarthur's algorithm :( 2010-10-05T18:39:20 haha :) 2010-10-05T18:44:57 *** cfaftw has joined #aichallenge 2010-10-05T18:45:19 *** Utkarsh has quit IRC (Ping timeout: 252 seconds) 2010-10-05T18:45:21 @amstan i compiled the org changes in a parseable format http://pastebin.com/raw.php?i=87hKTtpZ 2010-10-05T18:45:22 jfrank_: You have no gotten any error messages recently, so here's a random one just to let you know that we care. 2010-10-05T18:45:27 *** Arekku has quit IRC (Quit: Page closed) 2010-10-05T18:46:06 *** Utkarsh has joined #aichallenge 2010-10-05T18:46:23 @amstan hey contestbot can I have some? 2010-10-05T18:46:24 Migi32: User error, it's not my fault. 2010-10-05T18:46:31 :( 2010-10-05T18:46:44 @error 2010-10-05T18:46:45 amstan: Run as fast as you can and don't look back. 2010-10-05T18:47:31 I was hoping for the "I'm sorry Dave. I can't error." message 2010-10-05T18:48:57 @echo [error] [error] 2010-10-05T18:48:58 You have no gotten any error messages recently, so here's a random one just to let you know that we care. You have no gotten any error messages recently, so here's a random one just to let you know that we care. 2010-10-05T18:49:05 *** davidjliu has joined #aichallenge 2010-10-05T18:49:07 @echo [error] [error] [error] [error] [error] 2010-10-05T18:49:08 Run as fast as you can and don't look back. No! User error, it's not my fault. I'm sorry Dave, I cannot error. You have no gotten any error messages recently, so here's a random one just to let you know that we care. 2010-10-05T18:49:15 sigh: there you go 2010-10-05T18:49:18 *** Mihaime has quit IRC (Read error: Connection reset by peer) 2010-10-05T18:49:21 heh 2010-10-05T18:49:33 @echo [rankings] 2010-10-05T18:49:35 Top 10 players: ruilov(3912), davidjliu(3880), Hazard(3822), dmj111(3805), Arekku(3761), bix0r4ever(3738), cfaftw(3736), DocBot(3709), adumlah(3689), EBraun(3674) 2010-10-05T18:49:46 amstan: "You have no gotten" 2010-10-05T18:49:47 *** FlagCapper has quit IRC (Quit: Page closed) 2010-10-05T18:49:55 *** Mihaime has joined #aichallenge 2010-10-05T18:50:07 *** jfrank_ has quit IRC (Ping timeout: 265 seconds) 2010-10-05T18:50:28 @echo [rankings], $randomNick(3651) 2010-10-05T18:50:29 Top 10 players: ruilov(3912), davidjliu(3880), Hazard(3822), dmj111(3805), Arekku(3761), bix0r4ever(3738), cfaftw(3736), DocBot(3709), adumlah(3689), EBraun(3674) , jmcarthur(3651) 2010-10-05T18:50:56 lol 2010-10-05T18:50:59 :D 2010-10-05T18:51:06 well done, jmcarthur :P 2010-10-05T18:51:17 *** optimum has joined #aichallenge 2010-10-05T18:52:12 lol 2010-10-05T18:52:50 got to go 2010-10-05T18:52:51 bye 2010-10-05T18:52:51 at first i was like "wtf" 2010-10-05T18:53:02 Migi32: cya 2010-10-05T18:53:12 the space before the comma is a little telling 2010-10-05T18:53:25 yeah :( 2010-10-05T18:53:31 aside from the "11 players in the top ten" thing 2010-10-05T18:53:36 @echo [rankings], $randomNick(3651) 2010-10-05T18:53:38 Top 10 players: ruilov(3912), davidjliu(3880), Hazard(3822), dmj111(3805), Arekku(3761), bix0r4ever(3738), cfaftw(3736), DocBot(3709), adumlah(3689), EBraun(3674) , wh1teside(3651) 2010-10-05T18:53:45 no one is going to count :P 2010-10-05T18:53:47 ah thats quite useful 2010-10-05T18:53:49 contestbot: apropos string 2010-10-05T18:53:49 amstan: No appropriate commands were found. 2010-10-05T18:53:53 Zannick, aside from those two I find the "@echo [rankings], $randomNick(3651)" also a little telling 2010-10-05T18:53:55 contestbot: load string 2010-10-05T18:53:55 amstan: Error: You don't have the owner capability. If you think that you should have this capability, be sure that you are identified before trying again. The 'whoami' command can tell you if you're identified. 2010-10-05T18:54:00 Migi32: :D 2010-10-05T18:54:03 @echo [rankings],$randomNick(3651) 2010-10-05T18:54:05 Top 10 players: ruilov(3914), davidjliu(3881), Hazard(3824), dmj111(3806), Arekku(3763), bix0r4ever(3739), cfaftw(3738), DocBot(3711), adumlah(3690), EBraun(3675) ,bartwe(3651) 2010-10-05T18:54:13 oh duh wrong space 2010-10-05T18:54:14 contestbot: botdoc 2010-10-05T18:54:15 amstan: botdoc = (#1) http://ubottu.com/stdin/supydocs/plugins/, or (#2) http://supybook.fealdia.org/devel/, or (#3) http://sourceforge.net/apps/mediawiki/gribble/index.php?title=Supybot_Resources 2010-10-05T18:54:15 oh well 2010-10-05T18:54:23 *** Migi32 has quit IRC (Quit: http://boinc.berkeley.edu) 2010-10-05T18:54:43 yeah, i can't find it 2010-10-05T18:54:51 there's string functions you can use.. 2010-10-05T18:55:28 @error 2010-10-05T18:55:29 bhasker: An error occured while trying to show the previous error. 2010-10-05T18:55:35 thats a good one 2010-10-05T18:55:37 *** Mihaime has quit IRC (Quit: Leaving) 2010-10-05T18:55:37 @error 2010-10-05T18:55:38 bhasker: An error occured while trying to show the previous error. 2010-10-05T18:55:42 !predict 2010-10-05T18:55:42 I predict that a1k0n will win! 2010-10-05T18:55:46 !predict 2010-10-05T18:55:46 I predict that BtbN will win! 2010-10-05T18:55:55 what other commands can contestbot take? 2010-10-05T18:55:56 @help 2010-10-05T18:55:56 bhasker: (help [] []) -- This command gives a useful description of what does. is only necessary if the command is in more than one plugin. 2010-10-05T18:56:08 !eightball Will bhasker find any more commands? 2010-10-05T18:56:08 sigh: It is so. 2010-10-05T18:56:14 lol 2010-10-05T18:56:14 wtf? 2010-10-05T18:56:27 you guys should try roulette 2010-10-05T18:56:29 !eightball will BtbN win 2010-10-05T18:56:29 bhasker: About as likely as pigs flying. 2010-10-05T18:56:32 lol 2010-10-05T18:56:36 !rainbow rainbow's for everyone! 2010-10-05T18:56:36 sigh: rainbow's for everyone! 2010-10-05T18:56:44 s/'// 2010-10-05T18:56:47 *rainbows 2010-10-05T18:56:51 yes.... me fail 2010-10-05T18:56:53 :) 2010-10-05T18:57:01 !roulette 2010-10-05T18:57:01 amstan: *click* 2010-10-05T18:57:14 contestbot: sigh is awesome 2010-10-05T18:57:14 i know... 2010-10-05T18:57:31 !rainbow rainbows! 2010-10-05T18:57:31 Zannick: rainbows! 2010-10-05T18:57:37 @yay 2010-10-05T18:57:37 Double rainbow all the way! 2010-10-05T18:57:37 !roulette 2010-10-05T18:57:38 bhasker: *click* 2010-10-05T18:57:41 aw 2010-10-05T18:57:43 gah! 2010-10-05T18:57:47 bhasker: try it more.. 2010-10-05T18:57:48 !rainbow 2010-10-05T18:57:48 bhasker: (rainbow ) -- Returns colorized like a rainbow. 2010-10-05T18:57:53 !roulette 2010-10-05T18:57:53 bhasker: *click* 2010-10-05T18:57:55 wait..did it work? 2010-10-05T18:58:00 !roulette 2010-10-05T18:58:00 bhasker: *click* 2010-10-05T18:58:05 what is it supposed to do? 2010-10-05T18:58:07 Zannick: did what work? 2010-10-05T18:58:09 bhasker: try it more.. 2010-10-05T18:58:18 !roulette 2010-10-05T18:58:18 bhasker: *click* 2010-10-05T18:58:20 bah 2010-10-05T18:58:22 you didn't win yet.. 2010-10-05T18:58:24 it will ban you 2010-10-05T18:58:26 I bet 2010-10-05T18:58:33 no, it won't ban you 2010-10-05T18:58:38 contestbot: We are all bored, tell us something! 2010-10-05T18:58:39 Frontier: I'm sorry Dave, I cannot We. 2010-10-05T18:58:46 sigh: hm, 2010-10-05T18:58:48 contestbot: fortune 2010-10-05T18:58:48 amstan: Error: The fortune command is not configured. If fortune is installed on this system, reconfigure the supybot.plugins.Unix.fortune.command configuration variable appropriately. 2010-10-05T18:58:53 notice how the r isn't red? 2010-10-05T18:58:55 nice fortune :) 2010-10-05T18:58:55 meh.. 2010-10-05T18:58:57 it kind of worked... 2010-10-05T18:59:17 contestbot: zen 2010-10-05T18:59:18 amstan: Run as fast as you can and don't look back. 2010-10-05T18:59:22 aww! 2010-10-05T18:59:26 Top 10 players: ruilov(3920), davidjliu(3887), Hazard(3830), dmj111(3812), Arekku(3768), cfaftw(3744), DocBot(3717), adumlah(3697), EBraun(3682), Rexxar(3677) 2010-10-05T18:59:30 oh.. i disabled it 2010-10-05T18:59:37 must just be my terminal 2010-10-05T18:59:53 contestbot: wtf is a terminal 2010-10-05T18:59:53 amstan: (wtf [is] ) -- Returns wtf is. 'wtf' is a *nix command that first appeared in NetBSD 1.5. In most *nices, it's available in some sort of 'bsdgames' package. 2010-10-05T19:00:06 contestbot: wtf is terminal 2010-10-05T19:00:06 amstan: Error: The wtf command is not configured. If it is installed on this system, reconfigure the supybot.plugins.Unix.wtf.command configuration variable appropriately. 2010-10-05T19:00:09 aww! 2010-10-05T19:00:09 !colorize is this better than rainbow? 2010-10-05T19:00:09 sigh: is this better than rainbow? 2010-10-05T19:00:18 nah.. 2010-10-05T19:00:26 k.. let's leave the bot alone 2010-10-05T19:00:27 lol 2010-10-05T19:00:42 I sometimes type 'lol' in my terminals when something doesn't work 2010-10-05T19:00:53 !roulette 2010-10-05T19:00:53 *** jmcarthur was kicked by contestbot (BANG!) 2010-10-05T19:00:54 * contestbot reloads and spins the chambers. 2010-10-05T19:00:59 *** jmcarthur has joined #aichallenge 2010-10-05T19:01:02 I wait for the day I don't read "Command not found: lol" 2010-10-05T19:01:03 i lose 2010-10-05T19:01:23 *** optimum has quit IRC (Ping timeout: 245 seconds) 2010-10-05T19:01:33 !craps 2010-10-05T19:01:34 Zannick: An error occured while trying to show the previous error. 2010-10-05T19:01:47 !texasholdem 2010-10-05T19:01:48 Zannick: I have no idea what you mean. 2010-10-05T19:02:03 bah 2010-10-05T19:05:07 Frontier: alias lol=sl 2010-10-05T19:05:17 *** optimum has joined #aichallenge 2010-10-05T19:05:34 sudo apt-get install sl 2010-10-05T19:05:35 do you guys just use a bash script to run TCP games? 2010-10-05T19:05:45 *** znutar has quit IRC (Read error: Connection reset by peer) 2010-10-05T19:05:50 amstan: let's not have sl in the channel... 2010-10-05T19:05:51 cfaftw: check out greghayes' qt interface for the tcp 2010-10-05T19:05:51 \/should probably check the forum 2010-10-05T19:06:04 ok 2010-10-05T19:06:30 what is s1? 2010-10-05T19:06:41 sl 2010-10-05T19:06:43 cfaftw: http://github.com/greghaynes/QTcpPlanetWars 2010-10-05T19:06:46 sl = common typo of ls 2010-10-05T19:06:58 ah 2010-10-05T19:07:02 also Steam Locomotive 2010-10-05T19:07:23 *** Utkarsh has quit IRC (Ping timeout: 276 seconds) 2010-10-05T19:07:25 some people have a program 'sl' which will run an uninterruptible steam locomotive across your screen 2010-10-05T19:07:27 *** BtbN has quit IRC (Quit: Verlassend) 2010-10-05T19:07:31 *** Meatkat has quit IRC (Ping timeout: 265 seconds) 2010-10-05T19:07:35 haha 2010-10-05T19:07:40 just install it 2010-10-05T19:07:43 it's a fun thing to have 2010-10-05T19:07:45 thanks 2010-10-05T19:07:50 "fun" :) 2010-10-05T19:08:00 it's good to take breaks, though 2010-10-05T19:09:44 omg that is awesome 2010-10-05T19:10:29 it supports -little and -Floating 2010-10-05T19:11:01 *** Utkarsh has joined #aichallenge 2010-10-05T19:12:19 *** jfrank has joined #aichallenge 2010-10-05T19:13:40 i think i'm going to buy minecraft tonight 2010-10-05T19:13:46 that'll be my break 2010-10-05T19:14:12 haha 2010-10-05T19:14:52 @amstan i compiled the org changes in a parseable format http://pastebin.com/raw.php?i=87hKTtpZ with python parser http://pastebin.com/raw.php?i=HLbZTsFY 2010-10-05T19:14:52 jfrank: I'm sorry Dave, I cannot amstan. 2010-10-05T19:15:02 amstan i compiled the org changes in a parseable format http://pastebin.com/raw.php?i=87hKTtpZ with python parser http://pastebin.com/raw.php?i=HLbZTsFY 2010-10-05T19:15:05 I support anyone in the top 10 taking up minecraft 2010-10-05T19:15:34 lol 2010-10-05T19:15:49 jfrank: decent.. decent 2010-10-05T19:15:57 jfrank: so.. if i run that... it'll work? 2010-10-05T19:16:17 as you can see in the script there are TODO lines, no mysql context. but it will parse yes. 2010-10-05T19:16:34 jfrank: yeah.. i already have those sql parts, in python 2010-10-05T19:16:42 maybe i can make planetwars in minecraft 2010-10-05T19:16:46 *** ichti has quit IRC (Ping timeout: 264 seconds) 2010-10-05T19:17:11 *** Sluf has quit IRC (Ping timeout: 265 seconds) 2010-10-05T19:17:20 jfrank: i have this: http://pastebin.com/JPa7R70g 2010-10-05T19:17:22 amstan: you'll have to not include the 'other issues' that i put at the bottom 2010-10-05T19:23:40 *** McLeopold has joined #aichallenge 2010-10-05T19:25:26 jfrank: PlayerLookup(name) returns id? 2010-10-05T19:25:32 or PlayerLookup(id) returns name? 2010-10-05T19:26:00 amstan: i had it return id, i should have used more variable names 2010-10-05T19:26:08 *** Utkarsh has quit IRC (Ping timeout: 264 seconds) 2010-10-05T19:26:55 *** Appleman1234 has joined #aichallenge 2010-10-05T19:27:01 *** fawek has quit IRC (Read error: Connection reset by peer) 2010-10-05T19:27:25 amstan: looking at it again the 'player not found' doesn't help much since it overwrites the id with None. the ids come in as names or numbers, so it is to convert to player id (numeric) 2010-10-05T19:27:25 jfrank: can you help me? 2010-10-05T19:27:32 *** Utkarsh has joined #aichallenge 2010-10-05T19:27:45 jfrank: i want to test these utility funcitons, i need your username and your ord id 2010-10-05T19:28:34 amstan: my username is jfrank, ord id? 2010-10-05T19:28:59 orgid 2010-10-05T19:30:06 amstan: oh im in this list of updates. currently i'm other, so its Mentor Graphics Inc, a line near the bottom. my id is 9463 2010-10-05T19:30:27 ok.. so.. you're other, aka 1 2010-10-05T19:30:30 or was it 0? 2010-10-05T19:30:31 hmm 2010-10-05T19:30:33 darn 2010-10-05T19:30:43 0 2010-10-05T19:30:46 0, yeah 2010-10-05T19:30:59 *** sigh has quit IRC (Remote host closed the connection) 2010-10-05T19:31:13 *** sigh has joined #aichallenge 2010-10-05T19:32:44 *** Utkarsh has quit IRC (Ping timeout: 276 seconds) 2010-10-05T19:33:23 *** McLeopold has quit IRC (Quit: Leaving) 2010-10-05T19:34:47 jfrank: you still in other? 2010-10-05T19:35:01 amstan: yes 2010-10-05T19:35:03 jfrank: now you should be in waterloo 2010-10-05T19:35:15 amstan: yes 2010-10-05T19:35:19 cool 2010-10-05T19:35:22 let's see the none case 2010-10-05T19:38:58 ok.. 2010-10-05T19:39:00 not good 2010-10-05T19:39:02 something froze.. 2010-10-05T19:39:24 amstan: the site looks fine :-) 2010-10-05T19:40:35 *** Utkarsh has joined #aichallenge 2010-10-05T19:42:18 i don't understand how the contest rankings work... currently it appears that my bot is ranked higher than any of the ones that beat it, but it didn't beat anyone of high rank 2010-10-05T19:43:01 jfrank: how does this look: http://pastebin.com/mTWZaxhm 2010-10-05T19:43:24 amstan: looking... 2010-10-05T19:43:33 so.. the functions work, they were tested.. 2010-10-05T19:44:25 amstan: the first qry in addupdateorg is not an org query 2010-10-05T19:44:26 how do you use that qtcp client with a python bot 2010-10-05T19:44:38 jfrank: what? 2010-10-05T19:44:38 i made a simple wrapper bash script that just calls my script 2010-10-05T19:44:46 but it made me timeout against another bot 2010-10-05T19:44:55 jfrank: oh.. i see 2010-10-05T19:44:58 amstan: shouldnt the first of the two queries be select orgname.... 2010-10-05T19:45:38 jfrank: http://pastebin.com/mTWZaxhm 2010-10-05T19:46:06 copy and paste wrong.. anyway, now it should be fine 2010-10-05T19:46:09 i'll be running that test 2010-10-05T19:46:28 jfrank: k.. orgname searching works.. adding org was not tested.. 2010-10-05T19:46:32 but like.. it should kinda work.. 2010-10-05T19:46:40 well.. let's try it 2010-10-05T19:46:45 jfrank: what's your org name? 2010-10-05T19:47:00 my new one is mentor graphics 2010-10-05T19:47:06 currently its waterloo 2010-10-05T19:48:03 cool 2010-10-05T19:48:06 "Mentor Graphics Inc" 2010-10-05T19:48:07 let's try.. 2010-10-05T19:48:41 yep 2010-10-05T19:48:43 worked 2010-10-05T19:48:45 id 224 2010-10-05T19:48:48 244* 2010-10-05T19:49:23 jfrank: k.. so.. can i run it then, was the rest of the script tested? 2010-10-05T19:50:06 i think the rest of the data should be good. there is the case that the username can't be found, which prints stuff, you'll have to not include that 'other issues' 2010-10-05T19:51:07 there was an issue about special chars.. 2010-10-05T19:51:09 in failure of lookup playerLookup returns None, which causes line 55 to not be that helpful 2010-10-05T19:51:27 Mälardalens 2010-10-05T19:51:29 like that.. 2010-10-05T19:52:17 *** hornairs has joined #aichallenge 2010-10-05T19:52:30 eww. yeah that is nasty 2010-10-05T19:52:57 yeah.. 2010-10-05T19:53:00 idk how it handles it.. 2010-10-05T19:53:02 oh well 2010-10-05T19:53:03 let's try 2010-10-05T19:53:15 well.. i can ascii it real quick 2010-10-05T19:55:03 http://pastebin.com/raw.php?i=yVbdsgUi 2010-10-05T19:55:40 none of the usernames had issues, so its just saner versions of org names. can't hurt, may make someone mad who doesn't get an special character 2010-10-05T19:55:46 jfrank: something's wrong: http://pastebin.com/715RGKUu 2010-10-05T19:56:25 *** JensTi has quit IRC (Ping timeout: 240 seconds) 2010-10-05T19:57:28 amstan: well we know Obeleh exists, so it must be the PlayerLookup function 2010-10-05T19:57:55 also.. 2010-10-05T19:57:59 some organizations were added weird.. 2010-10-05T19:58:10 *** Utkarsh has quit IRC (Ping timeout: 264 seconds) 2010-10-05T19:58:45 jfrank: Shippensburg University of Pennsylvania:aNonnyMouse,davidkelly999 2010-10-05T19:59:26 Top 10 players: ruilov(3938), davidjliu(3879), dmj111(3816), Hazard(3814), cfaftw(3767), Arekku(3755), DocBot(3716), bix0r4ever(3710), adumlah(3690), EBraun(3683) 2010-10-05T19:59:32 *** JensTi has joined #aichallenge 2010-10-05T19:59:34 amstan: eww my bad 2010-10-05T19:59:37 yeah.. 2010-10-05T19:59:38 i saw 2010-10-05T19:59:39 so.. 2010-10-05T19:59:41 hmm 2010-10-05T19:59:43 ok 2010-10-05T20:00:48 amstan: but you got all the ones with ids out of the way. the script is also idempotent which is cool 2010-10-05T20:01:07 jfrank: nope.. i cannot find jwijst 2010-10-05T20:01:08 anywhere.. 2010-10-05T20:01:26 jfrank: did you go by the forum username? 2010-10-05T20:01:30 they might not match 2010-10-05T20:01:41 no, i only included username when they explicitly asked for it 2010-10-05T20:02:01 i realized that, they can have typos. Obeleh is in that list though 2010-10-05T20:02:09 and we know he exists, he is on page 1 of the rankings 2010-10-05T20:02:37 so some of those errors are real, but it looks like that function is not finding ids for real users, which is a problem 2010-10-05T20:03:22 *** Utkarsh has joined #aichallenge 2010-10-05T20:03:33 so im imagining there should be 4 or 5 name errors, not 40ish? 2010-10-05T20:05:13 idk.. something's weird.. 2010-10-05T20:08:23 *** irchs has quit IRC (Quit: irchs) 2010-10-05T20:08:24 for some reason this fails: ('select user_id from users where username like "%s";', ('jwijst',)) 2010-10-05T20:10:33 *** Utkarsh has quit IRC (Ping timeout: 245 seconds) 2010-10-05T20:11:01 *** cfaftw has quit IRC (Quit: Leaving) 2010-10-05T20:11:04 oh! that's why! 2010-10-05T20:12:34 wait why 2010-10-05T20:12:42 semi-colon? 2010-10-05T20:13:11 double quotes...? 2010-10-05T20:13:48 is it a prepared statement or a printf? 2010-10-05T20:14:33 jfrank: perfect.. 2010-10-05T20:14:37 yeah.. finally fixed it 2010-10-05T20:14:41 a few problems with it that it had 2010-10-05T20:14:43 player jwijst not found 2010-10-05T20:14:46 that's the only error 2010-10-05T20:15:05 *** u1tr0n has quit IRC () 2010-10-05T20:15:10 *** JensTi has quit IRC (Quit: Leaving.) 2010-10-05T20:15:28 yeah, errors aren't handled properly 2010-10-05T20:16:27 jfrank: but yeah.. cool 2010-10-05T20:16:29 it works 2010-10-05T20:16:43 now we need to find a way to aquire more files like orgchanges.txt 2010-10-05T20:18:30 jfrank: thanks for the help 2010-10-05T20:19:09 i have to go home.. big day tomorrow 2010-10-05T20:19:13 ttyl 2010-10-05T20:19:16 *** amstan has quit IRC (Remote host closed the connection) 2010-10-05T20:20:26 *** Utkarsh has joined #aichallenge 2010-10-05T20:22:13 *** mceier has quit IRC (Quit: leaving) 2010-10-05T20:29:25 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-10-05T20:30:08 *** optimum has quit IRC (Ping timeout: 245 seconds) 2010-10-05T20:31:51 *** optimum has joined #aichallenge 2010-10-05T20:34:22 *** Utkarsh has joined #aichallenge 2010-10-05T20:34:38 *** hornairs has quit IRC (Quit: hornairs) 2010-10-05T20:41:49 *** Bobng has quit IRC (Quit: Leaving) 2010-10-05T20:58:43 real 0m0.936s 2010-10-05T20:58:43 user 0m0.920s 2010-10-05T20:58:43 sys 0m0.012s 2010-10-05T20:59:28 Top 10 players: ruilov(3939), davidjliu(3871), dmj111(3818), Hazard(3805), cfaftw(3783), bix0r4ever(3755), Arekku(3753), DocBot(3724), EBraun(3690), adumlah(3682) 2010-10-05T21:00:09 *** krokokrusa has joined #aichallenge 2010-10-05T21:00:14 Is anyone here running tests on all 100 maps agains the 5 starter bots? 2010-10-05T21:00:51 Frontier: don't bother with anything otehr than rage and dual 2010-10-05T21:01:18 I used to, but now I beat them all on all maps, so it is pointless 2010-10-05T21:01:30 How long did it take for one run? 2010-10-05T21:02:21 I can run about 200 games a minute, that's parallel across 4 cores though and my bot uses hardly any time 2010-10-05T21:03:03 the games don't run the full 200 turns either, so that helps :) 2010-10-05T21:03:27 200 a minute with the java local server? 2010-10-05T21:03:34 yup 2010-10-05T21:03:45 on 2.83 GHz quad core 2010-10-05T21:05:24 crazy speed. for me it took around 15 minutes with all 5 bots i think 2010-10-05T21:06:02 as I said, my bot doesn't use much time... it's getting slower though 2010-10-05T21:07:25 if I run my bots against each other the games are often longer so they take more time 2010-10-05T21:07:33 but still only a few seconds per game 2010-10-05T21:08:24 *** justin_pdx has joined #aichallenge 2010-10-05T21:09:06 Well I have a notebook with a 2.0 Ghz dual core and just ran RageBot against itself and the other 4 on all maps in 1,4 seconds :) 2010-10-05T21:09:59 thats about 3ms per game ^^ 2010-10-05T21:12:11 I rewrote the server and the bots in C++ and let them all run in one process and communicate directly through C++-objects without using stdout and stdin. That will be fun when I start to test different versions of my bot on all maps. 2010-10-05T21:18:02 *** bhasker has quit IRC (Ping timeout: 276 seconds) 2010-10-05T21:20:58 *** Mathnerd314 has quit IRC (Ping timeout: 245 seconds) 2010-10-05T21:22:31 Frontier: nice :) 2010-10-05T21:24:13 The PlayGame.jar takes - I just checked - ~19 CPU minutes on my computer! Now you know why I had to do something ^^ 2010-10-05T21:25:08 *** justin_pdx has quit IRC (Quit: justin_pdx) 2010-10-05T21:26:35 *** Arthur_ has joined #aichallenge 2010-10-05T21:28:08 *** krokokrusa has quit IRC (Ping timeout: 255 seconds) 2010-10-05T21:29:04 yes, PlayGame.jar is terrible 2010-10-05T21:29:10 i use albert's C++ engine thing 2010-10-05T21:29:15 sigh: You contradict yourself ^^ - "I can run about 200 games a minute" - "only a few seconds per game" 2010-10-05T21:29:37 200 games is across 4 cores 2010-10-05T21:29:43 a1k0n: is it open source? 2010-10-05T21:29:49 I mentioned that :P 2010-10-05T21:29:49 yep, it's on the forum 2010-10-05T21:30:13 sigh: shame on me. I never get used to this parallel running stuff 2010-10-05T21:30:35 haha... 200 games in series would be nice though :P 2010-10-05T21:30:40 When I look at the output of 'top' I never know if it is 40% of BOTH cpus or just one 2010-10-05T21:30:57 parallel was easy, I'm just using a forking version of map I found on the net 2010-10-05T21:31:07 (python) 2010-10-05T21:32:05 Ok, you can fork any shell script as well I guess and run 25 maps on each core or so. 2010-10-05T21:34:11 a1k0n: There is only one good thing about PlayGame.jar - it should run out of the box on most setups. 2010-10-05T21:38:16 *** McLeopold has joined #aichallenge 2010-10-05T21:54:34 *** jimmerton has joined #aichallenge 2010-10-05T21:58:47 *** yasith__ has quit IRC (Ping timeout: 265 seconds) 2010-10-05T21:59:00 *** jimmerton_ has joined #aichallenge 2010-10-05T21:59:27 Top 10 players: ruilov(3945), davidjliu(3852), dmj111(3831), Hazard(3808), cfaftw(3794), bix0r4ever(3753), Arekku(3730), DocBot(3705), Rexxar(3703), EBraun(3702) 2010-10-05T22:00:49 *** jimmerton has quit IRC (Ping timeout: 240 seconds) 2010-10-05T22:00:49 im lookin but i cant find this c++ engine .... pointers? would love to get a bit of a speed boost ;) 2010-10-05T22:00:49 *** jimmerton_ is now known as jimmerton 2010-10-05T22:01:04 c++ version of playgame that is 2010-10-05T22:03:01 *** cfaftw has joined #aichallenge 2010-10-05T22:03:53 sup 2010-10-05T22:07:57 *** Accoun has quit IRC (Ping timeout: 276 seconds) 2010-10-05T22:10:13 *** yasith__ has joined #aichallenge 2010-10-05T22:12:41 *** Accoun has joined #aichallenge 2010-10-05T22:15:16 *** cfaftw has quit IRC (Quit: Leaving) 2010-10-05T22:19:04 *** Naktibalda has quit IRC (Remote host closed the connection) 2010-10-05T22:20:25 *** amstan has joined #aichallenge 2010-10-05T22:20:25 *** ChanServ sets mode: +o amstan 2010-10-05T22:26:13 jfrank: you deserve a medal of honour, do you have a forums account? 2010-10-05T22:28:46 *** amstan_ has joined #aichallenge 2010-10-05T22:28:46 *** ChanServ sets mode: +o amstan_ 2010-10-05T22:29:12 *** Cyndre has joined #aichallenge 2010-10-05T22:31:49 *** amstan has quit IRC (Ping timeout: 245 seconds) 2010-10-05T22:32:40 *** Mathnerd314 has joined #aichallenge 2010-10-05T22:35:11 *** jimmerton_ has joined #aichallenge 2010-10-05T22:37:15 *** jimmerton has quit IRC (Ping timeout: 240 seconds) 2010-10-05T22:37:15 *** jimmerton_ is now known as jimmerton 2010-10-05T22:40:36 *** McLeopold has quit IRC (Ping timeout: 276 seconds) 2010-10-05T22:44:33 *** Appleman1234 has quit IRC (Quit: Leaving) 2010-10-05T22:47:43 *** sneilan has joined #aichallenge 2010-10-05T22:48:48 hey, i was wondering if it's possible to use galcon AI's in the flash or executable version of galcon from galcon.com. it would look really cool if that was possible. thank you for your time 2010-10-05T22:52:49 *** jimmerton has quit IRC (Quit: jimmerton) 2010-10-05T22:59:28 Top 10 players: ruilov(3916), davidjliu(3846), dmj111(3836), Hazard(3791), bix0r4ever(3751), DocBot(3713), Arekku(3710), EBraun(3703), Rexxar(3698), adumlah(3674) 2010-10-05T23:05:00 *** Zeiris_ has quit IRC (Ping timeout: 265 seconds) 2010-10-05T23:24:53 !predict 2010-10-05T23:24:53 I predict that jhawthorn will win! 2010-10-05T23:32:27 *** genericbob has quit IRC (Ping timeout: 252 seconds) 2010-10-05T23:32:32 *** genericbob has joined #aichallenge 2010-10-05T23:41:08 *** bhasker has joined #aichallenge 2010-10-05T23:44:30 sneilan: the bot called "galcon" is that AI 2010-10-05T23:44:41 sneilan: made by the galcon creator himself 2010-10-05T23:44:54 bleh, can't remember the passwords I used for the testaccounts on my test server :P 2010-10-05T23:45:44 sneilan: http://ai-contest.com/profile.php?user_id=3907 2010-10-05T23:46:31 yes, but, what i meant to say is if it's possible to replay matches in the galcon fusion game itself (not the java simulator). it looks cooler with all the graphics and all that 2010-10-05T23:48:03 ah i see 2010-10-05T23:48:09 i agree. i'm not sure 2010-10-05T23:48:28 sneilan: the canvas visualizer is pretty nice though 2010-10-05T23:50:02 it is and i'm fine with it. i was just curious if I could use galcon fusion as the visualizer instead, but, it looks like i can't 2010-10-05T23:52:31 thank you for your time tho! 2010-10-05T23:52:35 I've never heard of it being possible so I would assume it's probably not 2010-10-05T23:52:40 would be cool if it was though 2010-10-05T23:52:45 definitely 2010-10-05T23:53:16 *** hexsprite has quit IRC (Quit: hexsprite) 2010-10-05T23:57:12 *** sneilan has quit IRC (Quit: Page closed) 2010-10-05T23:59:28 Top 10 players: ruilov(3931), dmj111(3867), davidjliu(3853), Hazard(3808), Arekku(3726), DocBot(3718), bix0r4ever(3713), EBraun(3710), Rexxar(3709), sequoh(3678)