2010-11-18T00:00:53 *** Utkarsh has quit IRC (Ping timeout: 265 seconds) 2010-11-18T00:01:02 if I make my AI do really well, but assume that the opponent does well also.. will I pretty much for sure end up in the top hundred? or can my assumption that the opponent isn't lousy going to possibly hurt me while climbing the ranks? 2010-11-18T00:01:41 burny: hard to say 2010-11-18T00:01:54 i don't think it's that hard to get in the top 100 still 2010-11-18T00:02:30 like for instance.. if I assume the opponent doesn't send 100 at my base, or do nothing atall, on the first turn.. 2010-11-18T00:03:09 i'm not making any such assumption, but i'll consider them making such a move in my game tree if i leave it vulnerable 2010-11-18T00:04:41 cause aside from very specific maps.. doing nothing, or attacking, on the irst turn.. is a horrible move, and will never win 2010-11-18T00:05:06 "never"? 2010-11-18T00:05:23 never against a bot that's any good 2010-11-18T00:05:59 depends on the map 2010-11-18T00:06:10 it's not going to do as well on average no 2010-11-18T00:06:17 *** Utkarsh has joined #aichallenge 2010-11-18T00:06:47 there's almost no map, that doing nothing on turn 1, is a viable move 2010-11-18T00:06:49 but if you can pick out when it's best to do that, then you do even better 2010-11-18T00:06:56 i disagree 2010-11-18T00:07:04 there was one with other day with 6 in the centre 2010-11-18T00:07:22 on could take a neutral near their planet on the first turn, but if the other player waited they could then take the better center planet 2010-11-18T00:07:35 notice use of the term 'almost' 2010-11-18T00:07:59 i think it will still happen 2010-11-18T00:10:59 *** Utkarsh has quit IRC (Ping timeout: 255 seconds) 2010-11-18T00:15:39 eburnette: 7030 games and counting 2010-11-18T00:15:51 burny: surely a majority of contestants are already coding their bots as though their opponents do well. they can't all be in the top 100 2010-11-18T00:16:09 New map set in play and available for download in the forum 2010-11-18T00:16:32 jmcarthur.. not all of those people do well themselves though :P 2010-11-18T00:16:34 janzert: is it maps running on the official server? 2010-11-18T00:16:40 yes 2010-11-18T00:16:51 Can you make a downloadable zip of all the games for data mining? 2010-11-18T00:17:40 *** davidjliu has quit IRC (Quit: Page closed) 2010-11-18T00:17:49 Top 10 players: bocsimacko(3644), bix0r4ever(3513), protocolocon(3439), dmj111(3428), shangas(3417), george(3400), _Astek_(3397), davidjliu(3389), luksian(3371), Xrillo(3368) 2010-11-18T00:18:22 *** Utkarsh has joined #aichallenge 2010-11-18T00:18:38 *** delt0r___ has joined #aichallenge 2010-11-18T00:19:10 new maps live on my tcp :) 2010-11-18T00:19:23 * eburnette is turning in 2010-11-18T00:19:25 *** Appleman1234 has joined #aichallenge 2010-11-18T00:20:24 i hope to at least be testing on mcleopolds server by the end of today 2010-11-18T00:20:40 i may put up an official submission too 2010-11-18T00:20:51 i need to make sure that works before the day before aha 2010-11-18T00:20:53 *** delt0r has quit IRC (Ping timeout: 276 seconds) 2010-11-18T00:21:47 burny: that was my point 2010-11-18T00:23:42 first new map game: http://zeroviz.us:8080/canvas?game_id=7047 2010-11-18T00:24:56 *** Utkarsh has quit IRC (Ping timeout: 255 seconds) 2010-11-18T00:28:18 my game data is 74MB zipped. O_o 2010-11-18T00:28:52 McLeopold: the new tcp is awesome :) 2010-11-18T00:29:12 thx 2010-11-18T00:29:35 is it possible to get a time added column with the map rankings? 2010-11-18T00:29:47 count seems to still give some old ones? 2010-11-18T00:30:03 so, you want a filter on the table? 2010-11-18T00:30:18 well when you say you put new maps up, i want to be able to go and watch the games on them :P 2010-11-18T00:30:30 it would be an easy way to sort for them 2010-11-18T00:30:33 sort by name is the best way so far 2010-11-18T00:30:45 what are the new maps? 2010-11-18T00:31:32 *** Utkarsh has joined #aichallenge 2010-11-18T00:32:05 the ones janzert generated for official 2010-11-18T00:34:32 *** Appleman1234 has quit IRC (Ping timeout: 260 seconds) 2010-11-18T00:36:11 *** Utkarsh has quit IRC (Ping timeout: 255 seconds) 2010-11-18T00:36:38 *** hellman has joined #aichallenge 2010-11-18T00:37:55 yeah, but what names? 2010-11-18T00:38:26 map_r538* 2010-11-18T00:38:46 thanks 2010-11-18T00:41:25 what about zeroviz_538_*.txt? 2010-11-18T00:42:05 that's just a different sample that's not even on mctcp by now? 2010-11-18T00:42:49 mega1: i think it's meant to be the new maps that are on the official server 2010-11-18T00:42:55 *** Utkarsh has joined #aichallenge 2010-11-18T00:43:40 antimatroid: I'm talking about themaps downloaded from McLeopold_'s tcp server a couple of hours ago. 2010-11-18T00:43:43 are you too? 2010-11-18T00:43:51 don't really know what they are, but from the name I would guess they're ones McLeopold_ generated for his server using the latest generator 2010-11-18T00:44:11 mega1: see the front page description 2010-11-18T00:45:17 oh, thanks 2010-11-18T00:45:23 "the are running now" 2010-11-18T00:45:39 yes 2010-11-18T00:45:44 http://zeroviz.us:8080/by_Map.html 2010-11-18T00:45:53 there is a latest time and view game link now 2010-11-18T00:46:00 antimatroid: there you go ^^ 2010-11-18T00:46:31 McLeopold_: that was a typo 2010-11-18T00:46:42 on the web page 2010-11-18T00:46:50 oh... 2010-11-18T00:48:07 the latest also works in the by player page, so you can keep tabs on mega1 2010-11-18T00:48:40 ooh, html fail... 2010-11-18T00:48:56 :) 2010-11-18T00:49:42 *** Utkarsh has quit IRC (Ping timeout: 260 seconds) 2010-11-18T00:50:38 it's not the most robust web site. 2010-11-18T00:50:46 in fact, it's a pile of crap 2010-11-18T00:52:18 i wouldn't have the first clue about websites :P 2010-11-18T00:53:14 *** greghaynes has quit IRC (Read error: Operation timed out) 2010-11-18T00:53:21 what the hell is pichu? 2010-11-18T00:53:48 eburnette 2010-11-18T00:53:59 *** Appleman1234 has joined #aichallenge 2010-11-18T00:54:00 that was sarcasm :P 2010-11-18T00:54:10 ah 2010-11-18T00:54:24 isn't it another onomatopoeia that represents sneezing? 2010-11-18T00:54:39 lol 2010-11-18T00:55:09 *** Utkarsh has joined #aichallenge 2010-11-18T00:57:18 *** greghaynes has joined #aichallenge 2010-11-18T00:59:28 Zannick: just for my information, were you able to test the check_submit change in a full setup locally? 2010-11-18T00:59:29 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-11-18T01:00:51 useless stats - maps_r538_server: #planets mean 21.7 median 21.5 mode 17; (even:odd 9:41) Symmetry radial:bilat 28:22 2010-11-18T01:01:40 hmm, odd number of planets pretty high 2010-11-18T01:01:55 that's probably because radial needs an odd number to work 2010-11-18T01:02:05 because of the stupid center planet 2010-11-18T01:02:10 rebel just started up 0.23 2010-11-18T01:02:13 if you are still around 2010-11-18T01:02:18 can't load zeroviz:8080 2010-11-18T01:02:20 except the UI is down again 2010-11-18T01:02:22 :-\ 2010-11-18T01:02:40 janzert: i don't have a php install 2010-11-18T01:02:43 McLeopold_: one could remove the very centre planet 2010-11-18T01:02:53 but some people seemed to want players home planets as 1 and 2 2010-11-18T01:02:55 i don't know why 2010-11-18T01:02:58 ok, working on testing it here now 2010-11-18T01:03:05 thanks 2010-11-18T01:03:19 what does one need to run the tcp server and the associated UI code? 2010-11-18T01:03:33 ui back up 2010-11-18T01:04:03 it's running on thttpd 2010-11-18T01:04:10 but I don't think that's a requirement 2010-11-18T01:04:16 php? 2010-11-18T01:04:19 + mysql? 2010-11-18T01:04:21 python 2010-11-18T01:04:25 hmm 2010-11-18T01:04:30 no database, all text files 2010-11-18T01:04:32 what about the tcp server itself? 2010-11-18T01:04:43 cpp, written by dhartmei 2010-11-18T01:04:56 hmm and how much network traffic does this generate 2010-11-18T01:05:00 i could host it on my website 2010-11-18T01:05:05 if required 2010-11-18T01:05:18 not too much 2010-11-18T01:05:27 I don't notice internet slowdown 2010-11-18T01:05:47 *** Utkarsh has joined #aichallenge 2010-11-18T01:05:55 hmm if you can mail me like a single tar.gz i can see if i can host it on my website 2010-11-18T01:06:00 with all the files 2010-11-18T01:06:05 I think it's my lack of understanding of thttpd 2010-11-18T01:06:30 well my website runs on apache so should be fine 2010-11-18T01:06:45 if you want to set it up w/ apache it would probably be more stable 2010-11-18T01:06:50 9 even planets is pretty close, in the perfect word it would be 12.5 2010-11-18T01:07:58 I've got apache running as well. Now that I've replaced the cgi stuff, I could probably switch over. 2010-11-18T01:09:04 proud of this game http://ai-contest.com/visualizer.php?game_id=7775856 2010-11-18T01:09:05 bhasker: I'll try and get it up on bitbucket later 2010-11-18T01:09:09 kk 2010-11-18T01:10:37 rebelxt you and me are 1 rank apart on official 2010-11-18T01:10:43 nice 2010-11-18T01:11:03 bhasker: the only process on the server really doing much work is my bot :) 2010-11-18T01:11:17 hehe 2010-11-18T01:11:39 i wonder if i should abandon my new version and just update 0.23 2010-11-18T01:11:47 i have put so much work into the new one 2010-11-18T01:11:50 and it doesn't do squat 2010-11-18T01:12:30 *** Utkarsh has quit IRC (Ping timeout: 250 seconds) 2010-11-18T01:13:30 what's the minimum planet distance? 2? 2010-11-18T01:13:49 it's 1.05 here: http://zeroviz.us:8080/canvas?game_id=7225 2010-11-18T01:14:44 planet 7 and 10 2010-11-18T01:15:17 McLeopold_: ^^^ 2010-11-18T01:15:41 uhoh, better tell janzert 2010-11-18T01:16:21 um, 1.05 should round up to 2? 2010-11-18T01:16:45 yep 2010-11-18T01:17:05 the actual distance should maybe be minimum 2, though... 2010-11-18T01:17:08 *** Utkarsh has joined #aichallenge 2010-11-18T01:17:20 ah, right 2010-11-18T01:17:33 that could lessen the overlapping problems 2010-11-18T01:17:35 no, it's fine then 2010-11-18T01:17:35 nah, I think it's fine 2010-11-18T01:17:48 Zannick: that's just a display issue 2010-11-18T01:17:49 Top 10 players: bocsimacko(3645), bix0r4ever(3530), protocolocon(3435), dmj111(3421), Raschi(3413), shangas(3411), george(3395), luksian(3374), _Astek_(3370), Xrillo(3366) 2010-11-18T01:18:01 players still have 1 turn to see a fleet incoming 2010-11-18T01:18:09 not pretty, but it's better than further biasing the density towards the edges 2010-11-18T01:18:12 yeah, although i always place planets from 2 outwards with radius 2010-11-18T01:18:19 should that be from 1.02 maybe? 2010-11-18T01:18:32 actually, it probably should 2010-11-18T01:18:41 mega1? 2010-11-18T01:19:00 i think that should maybe be changed now that i think about it 2010-11-18T01:19:03 1.0100000000009 to be precise 2010-11-18T01:19:10 yeah 2010-11-18T01:19:15 you could set it at 1 and it wouldn't matter 2010-11-18T01:19:17 basing the minimum distance check on the actual distance would make the center even more sparse 2010-11-18T01:19:23 because it would reject the integer distance from the centre planet 2010-11-18T01:19:57 or would that matter? 2010-11-18T01:20:20 overall, it's better to generate any number, then throw out the result, than to generate a number in a specific range 2010-11-18T01:20:33 mega1 is agreeing 2010-11-18T01:20:46 pickRadius should be from 1 to maxRadius then 2010-11-18T01:20:54 random_radius i think it is 2010-11-18T01:20:55 I'm not sure we are talking about the same thing. 2010-11-18T01:21:30 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-11-18T01:22:16 i'm not even sure what the rand_radius function is doing anymore :\ 2010-11-18T01:22:26 this is were my merge idea is better 2010-11-18T01:22:45 it might be better just to randomly generate x,y then translate to r,theta? 2010-11-18T01:22:50 (because it's been changed, not cause i forgot what i did ) 2010-11-18T01:23:21 translate? :P convert 2010-11-18T01:23:24 possibly 2010-11-18T01:23:34 translate = convert :P 2010-11-18T01:23:42 Zannick: two things so far, the upload_errors function doesn't pick up the global errors variable so does nothing currently easy fix is pass in the outer variable and then pass it back out as the return 2010-11-18T01:24:24 also upload_errors picks up invalid file type on a direct link to check_submit.php 2010-11-18T01:24:45 not really a problem since it will stop the submission from happening anyway 2010-11-18T01:24:55 just gives a sort of misleading message 2010-11-18T01:25:18 hm 2010-11-18T01:25:31 i can just return the error string 2010-11-18T01:25:44 *** jmcarthur has quit IRC (Quit: WeeChat 0.3.0) 2010-11-18T01:25:44 wait, no, your way is better 2010-11-18T01:26:06 *** Mathnerd314 has quit IRC (Quit: ChatZilla 0.9.86-rdmsoft [XULRunner 1.9.2.12/20101026210630]) 2010-11-18T01:26:12 and as for the other thing 2010-11-18T01:26:42 direct link isn't apparently giving an error 2010-11-18T01:27:02 *** Utkarsh has joined #aichallenge 2010-11-18T01:27:05 so it's dropping to testing against the name 2010-11-18T01:28:10 I have a feeling $_FILES['uploadedfiles'] simply isn't set 2010-11-18T01:28:16 at least it shouldn't be 2010-11-18T01:28:18 checking now 2010-11-18T01:28:20 *** jmcarthur has joined #aichallenge 2010-11-18T01:28:21 antimatroid: isn't there always a planet right in the center of the map? 2010-11-18T01:28:29 yes 2010-11-18T01:28:46 rand_radius is at 1, so it is correct 2010-11-18T01:28:46 although, the "centre" isn't so clear for line symmetric maps 2010-11-18T01:29:12 I though it was still the center of the map area 2010-11-18T01:29:21 it is 2010-11-18T01:30:02 *** Utkarsh has quit IRC (Read error: Connection reset by peer) 2010-11-18T01:30:05 but it could be on the outside point of the convex hull around the map now 2010-11-18T01:30:19 that would have been impossibly before unless all planets were on a line 2010-11-18T01:30:46 well not quite 2010-11-18T01:30:47 you lost me 2010-11-18T01:31:08 so, center planet is center of map area or equidistant between home planets? 2010-11-18T01:31:22 it's at (0,0) 2010-11-18T01:31:32 okay: then everything is fine 2010-11-18T01:31:34 but the line joining the two home planets doesn't necessarily intersect that planet 2010-11-18T01:31:41 yeah, ok 2010-11-18T01:32:17 ah, I get the convex hull thing now 2010-11-18T01:32:29 if all theta were to one side of the map 2010-11-18T01:32:34 yeah 2010-11-18T01:32:47 I haven't seen that yet 2010-11-18T01:33:04 neither, it's just an extreme example as to why it's not such a "central" planet always anymore 2010-11-18T01:34:47 *** sigh has quit IRC (Ping timeout: 272 seconds) 2010-11-18T01:36:33 janzert: another revision is in 2010-11-18T01:37:44 *** Appleman1234 has quit IRC (Ping timeout: 276 seconds) 2010-11-18T01:38:08 *** Utkarsh has joined #aichallenge 2010-11-18T01:39:09 looks like a closing } is missed after line 21? 2010-11-18T01:39:58 oops 2010-11-18T01:40:06 * Zannick embarrassed 2010-11-18T01:40:45 fixed 2010-11-18T01:41:00 looks like it works otherwise 2010-11-18T01:41:40 cool 2010-11-18T01:42:43 *** jmcarthur has quit IRC (Quit: WeeChat 0.3.2) 2010-11-18T01:44:36 i'll push when you give the +1 if it's all good 2010-11-18T01:44:37 *** Utkarsh has quit IRC (Ping timeout: 245 seconds) 2010-11-18T01:45:08 ok, just doing one last functional test and a once over of the code 2010-11-18T01:45:13 http://zeroviz.us:8080/canvas?game_id=7336 2010-11-18T01:45:30 ^^ that's why, ultimately, the ragebot check may not matter 2010-11-18T01:50:04 *** Utkarsh has joined #aichallenge 2010-11-18T01:50:09 Zannick: ok, looks good 2010-11-18T01:52:26 merged, revision 545 2010-11-18T02:01:44 live on the site 2010-11-18T02:04:06 hm, not logged in, it thinks my submission is too recent 2010-11-18T02:05:32 that wouldn't have been something i changed, though 2010-11-18T02:05:40 no 2010-11-18T02:06:06 okay, now it says i forgot to upload a file. 2010-11-18T02:06:07 :D 2010-11-18T02:06:15 *** phxtrack has quit IRC (Read error: Connection reset by peer) 2010-11-18T02:06:22 my profile is intact 2010-11-18T02:06:22 *** phxtrack has joined #aichallenge 2010-11-18T02:06:36 so i'm marking the issue fixed 2010-11-18T02:08:26 *** Utkarsh has quit IRC (Ping timeout: 255 seconds) 2010-11-18T02:10:33 fixed the not logged in problem now 2010-11-18T02:11:04 cool 2010-11-18T02:11:30 *** virdo has quit IRC (Ping timeout: 240 seconds) 2010-11-18T02:13:17 *** lavalamp_ has quit IRC (Ping timeout: 265 seconds) 2010-11-18T02:13:47 *** jmcarthur has joined #aichallenge 2010-11-18T02:17:48 Top 10 players: bocsimacko(3647), bix0r4ever(3532), protocolocon(3429), dmj111(3419), shangas(3414), george(3393), _Astek_(3378), luksian(3371), felixcoto(3360), Xrillo(3360) 2010-11-18T02:22:04 *** jmcarthur has quit IRC (Quit: WeeChat 0.3.2) 2010-11-18T02:31:51 *** Appleman1234 has joined #aichallenge 2010-11-18T02:34:21 *** Palmik has joined #aichallenge 2010-11-18T02:40:07 *** jmcarthur has joined #aichallenge 2010-11-18T02:40:11 hi 2010-11-18T02:42:24 *** Mistmanov has joined #aichallenge 2010-11-18T02:54:39 *** yasith has joined #aichallenge 2010-11-18T03:01:17 *** jaspervdj has joined #aichallenge 2010-11-18T03:03:55 *** tapwater has quit IRC (Quit: tapwater) 2010-11-18T03:10:34 *** Mjothvitnir has quit IRC (Quit: :hi-ho hi-ho its off to bed I go) 2010-11-18T03:12:07 http://zeroviz.us:8080/canvas?game_id=7591 2010-11-18T03:12:30 for anyone who cares, I've updated the visualizer so that the active planet will be drawn on top. 2010-11-18T03:12:54 now you can finally see the planets covered by those gigantic growth 5 planets. 2010-11-18T03:13:21 janzert: ^^ 2010-11-18T03:14:51 nice, I'll try and backport it a little later 2010-11-18T03:15:09 let me know if you need help. 2010-11-18T03:15:16 My changes are off of the official version 2010-11-18T03:15:41 ok, great 2010-11-18T03:16:49 I was trying to debug that game, but I couldn't see JamesMG's fleet count. :) 2010-11-18T03:17:49 Top 10 players: 2010-11-18T03:18:09 contestbot is sleepy 2010-11-18T03:18:10 McLeopold_: Run as fast as you can and don't look back. 2010-11-18T03:20:30 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-11-18T03:32:40 *** lavalamp_ has joined #aichallenge 2010-11-18T03:42:24 McLeopold: your site appears to be down 2010-11-18T03:42:39 wait, no, it appears my net is fd 2010-11-18T03:42:40 brb 2010-11-18T03:45:26 *** antimatroid1 has joined #aichallenge 2010-11-18T03:47:22 *** antimatroid has quit IRC (Ping timeout: 265 seconds) 2010-11-18T03:51:42 *** aerique has joined #aichallenge 2010-11-18T03:57:10 *** yasith has quit IRC (Ping timeout: 250 seconds) 2010-11-18T04:03:02 main site does appear to be down now though aha 2010-11-18T04:09:05 WOW 2010-11-18T04:09:07 *** yasith has joined #aichallenge 2010-11-18T04:09:31 I just got blinded by lightning right over my house. 2010-11-18T04:09:47 hope the server stays up :) 2010-11-18T04:10:46 antimatroid1: check my fix to the visualizer on my site. :) 2010-11-18T04:13:28 *** McLeopold_ has quit IRC (Read error: Connection reset by peer) 2010-11-18T04:14:23 much more helpful 2010-11-18T04:18:00 amstan: Error: There was a problem accessing the interface to ai-contest.com 2010-11-18T04:22:27 i have another idea for a fun ai contest 2010-11-18T04:22:33 have any of you played minesweeper flags? 2010-11-18T04:22:41 hmm> 2010-11-18T04:22:48 where you're actually trying to hit the most bombs, although it's not simultaneous 2010-11-18T04:23:11 isn't it trivial? 2010-11-18T04:23:36 not sure, sometimes you only know the probability of a bomb being somewhere 2010-11-18T04:23:44 large parts of it probably are, and luck 2010-11-18T04:23:55 would be fun to just write such an ai anyway 2010-11-18T04:25:03 you can do that in a day though 2010-11-18T04:26:22 it'd take longer than that 2010-11-18T04:26:35 but it probably wouldn't make for a great contest game, didn't think that through very well :P 2010-11-18T04:27:04 *** Naktibalda has joined #aichallenge 2010-11-18T04:27:20 *** mceier has joined #aichallenge 2010-11-18T04:36:00 *** tim__ has joined #aichallenge 2010-11-18T04:37:08 hey, I just saw this challenge and downloaded the python starter package. however, it does not seem to work well and the forum are currently having sql problems. 2010-11-18T04:37:35 i've got two problems. first, if java does not crash myBot.py times out in the first turn. 2010-11-18T04:38:03 second, often java crashes instantly with Exception in thread "Thread-1" java.lang.ArrayIndexOutOfBoundsException: 1 2010-11-18T04:38:03 at ViewerPanel.run(ViewerPanel.java:85) 2010-11-18T04:38:04 at java.lang.Thread.run(Thread.java:662) 2010-11-18T04:38:32 hm I wonder how old the viewer and stuff is in that package 2010-11-18T04:39:02 is there any other place to get a more recent python version? 2010-11-18T04:39:19 *** phreeza has joined #aichallenge 2010-11-18T04:39:24 tim__ it could "time out" if it fails to actually start, you're sure you've got python set up and the path/command is correct etc? 2010-11-18T04:39:58 lavalamp_, : yes, python should work. i'm using: java -jar tools/PlayGame.jar maps/map7.txt 1000 1000 log.txt "python myBot.py" "java -jar example_bots/RandomBot.jar" | java -jar tools/ShowGame.jar 2010-11-18T04:40:10 just like in the starters guide 2010-11-18T04:40:20 hm looks right 2010-11-18T04:40:24 tim__: just a forewarning, the contest ends in a little over a week and is an almost 3 month contest 2010-11-18T04:40:33 don't let that stop you having some fun, just a warning :) 2010-11-18T04:40:47 yes, i know, but thanks. I'm just in for some fun :) 2010-11-18T04:40:57 :) 2010-11-18T04:40:58 if you "python mybot.py" and type "go", does it say "go" back to you? 2010-11-18T04:41:26 yes it does 2010-11-18T04:41:54 hmm 2010-11-18T04:42:01 http://ai-contest.com/forum/viewtopic.php?f=19&t=1087 2010-11-18T04:42:01 http://ai-contest.com/forum/viewtopic.php?f=17&t=1094 2010-11-18T04:42:06 my guess is something is up with java on your system 2010-11-18T04:42:12 that was for someone else 2010-11-18T04:42:13 :P 2010-11-18T04:43:11 @jbotman 2010-11-18T04:43:12 lavalamp_: jbotman = http://www.mabs.at/ewing/JBotManager.jar 2010-11-18T04:43:30 I recommend getting jbotmanager from that link 2010-11-18T04:43:52 it's better than the included tools ^^ 2010-11-18T04:44:01 jBotManager is what I use 2010-11-18T04:44:51 l, 2010-11-18T04:45:06 where can I find documentation how to use it? 2010-11-18T04:47:01 just open the jar file :P 2010-11-18T04:47:11 it's pretty self explanatory how to use it after that 2010-11-18T04:47:47 ok, I'll try that. need to reboot. thank you guys 2010-11-18T04:47:47 if you're playing in human mode, press a digit for it to send that proportion of the ships on that planet 2010-11-18T04:47:52 0 == 100% 2010-11-18T04:48:07 kk 2010-11-18T04:48:28 *** tim__ has quit IRC (Remote host closed the connection) 2010-11-18T04:50:12 *** perror has joined #aichallenge 2010-11-18T04:58:10 doh the site really is hose right now... :/ 2010-11-18T04:58:15 *hosed 2010-11-18T05:00:18 *** Ice_Harley has joined #aichallenge 2010-11-18T05:01:06 *** tim__ has joined #aichallenge 2010-11-18T05:02:30 hey guys :-) first of all this JBotManager seems pretty nice :) I just let this default bot play against itself. is it normal that after eg bot 2 won over bot 1 it says "Player 1/2 Message: you have timed out"? 2010-11-18T05:07:12 tim__: no 2010-11-18T05:07:23 It should only do that if you really time out 2010-11-18T05:08:26 however there is a few outstanding bugs --but thats not listed... mostly it will give you a lose when you have no fleets buts still own a planet 2010-11-18T05:08:56 http://ai-contest.com/forum/viewtopic.php?f=18&t=751 2010-11-18T05:09:31 another thing I just saw is that letting the given bots play each other it always ends up in "illegal fleet requests" 2010-11-18T05:09:32 also i just don't have time right now to fix anything. 2010-11-18T05:09:55 well u need to set up your own bot command lines 2010-11-18T05:10:06 they ones that are there by default won't work 2010-11-18T05:10:17 oh, you're the guy who did it :-) thanks! 2010-11-18T05:10:37 why is that so? did the rules change? are the bots just crappy? 2010-11-18T05:10:57 ?? 2010-11-18T05:11:06 your paths won't be right 2010-11-18T05:11:29 it wont find anything untill you set up the bot command lines properly 2010-11-18T05:11:32 the paths seem to work 2010-11-18T05:11:35 just timeouts 2010-11-18T05:11:41 yes it does find the bots 2010-11-18T05:11:57 but they are playing some turns end then it ends up in an illegal fleet request 2010-11-18T05:12:27 Well its never done this for me. The protocol has not changed so --i don't know 2010-11-18T05:12:39 the server ignores # 2010-11-18T05:13:13 check the standard error/out thats where most mesages get sent 2010-11-18T05:13:46 sometimes if you get something wrong... the only way to fix it is with a restart... I though i got all those bugs --but you never know 2010-11-18T05:13:55 It was always a little rought 2010-11-18T05:13:58 rough 2010-11-18T05:14:14 ok 2010-11-18T05:17:51 Top 10 players: bocsimacko(3650), smiley1983(3459), protocolocon(3430), dmj111(3417), george(3402), shangas(3390), luksian(3371), davidjliu(3371), felixcoto(3358), Xrillo(3354) 2010-11-18T05:23:06 *** tim__ has quit IRC (Remote host closed the connection) 2010-11-18T05:50:37 *** bhasker has quit IRC (Read error: Connection reset by peer) 2010-11-18T05:50:42 *** hellman has quit IRC (Remote host closed the connection) 2010-11-18T05:57:40 *** bhasker has joined #aichallenge 2010-11-18T06:00:35 *** mceier has quit IRC (Quit: leaving) 2010-11-18T06:17:52 Top 10 players: bocsimacko(3646), smiley1983(3457), protocolocon(3435), dmj111(3412), asavis(3410), george(3395), shangas(3378), Xrillo(3354), felixcoto(3353), davidjliu(3350) 2010-11-18T06:20:19 god damnit 2010-11-18T06:20:24 there's an australian in second place 2010-11-18T06:24:40 janzert: shouldn't the main page of the aicontest tell sth about the mapchange? 2010-11-18T06:44:23 *** Appleman1234 has quit IRC (Ping timeout: 264 seconds) 2010-11-18T06:44:31 *** Appleman1234 has joined #aichallenge 2010-11-18T07:03:18 @tcp 2010-11-18T07:03:19 medrimonia1: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/, or (#3) http://zeroviz.us:8080, or (#4) 98.247.248.39:995 (zeroviz.us) 2010-11-18T07:06:53 Am I the only one who can't access to zeroviz.us:8080? 2010-11-18T07:06:55 *** VegaForces has joined #aichallenge 2010-11-18T07:07:21 I can "play" through the tcp server, but I can't see my games 2010-11-18T07:11:00 website also doesn't load for me 2010-11-18T07:11:26 thanks for the answer Mistmanov 2010-11-18T07:17:53 Top 10 players: bocsimacko(3665), bix0r4ever(3615), Raschi(3560), smiley1983(3453), luksian(3435), protocolocon(3428), dmj111(3416), davidjliu(3412), george(3409), _Astek_(3386) 2010-11-18T07:19:32 *** yasith has quit IRC (Ping timeout: 255 seconds) 2010-11-18T07:20:41 *** Cold-Phoenix has joined #aichallenge 2010-11-18T07:20:43 *** smellyhippy has quit IRC (Ping timeout: 265 seconds) 2010-11-18T07:21:43 *** Error323 has quit IRC (Read error: Connection reset by peer) 2010-11-18T07:23:18 *** Error323_ has joined #aichallenge 2010-11-18T07:23:25 *** Error323_ is now known as Error323 2010-11-18T07:25:32 *** smellyhippy has joined #aichallenge 2010-11-18T07:32:50 *** iamluck|2 has joined #aichallenge 2010-11-18T07:32:50 *** iamluck has quit IRC (Read error: Connection reset by peer) 2010-11-18T07:39:00 *** yasith has joined #aichallenge 2010-11-18T07:39:59 * lavalamp_ stumbles upon fount of all wisdom 2010-11-18T07:41:16 lol how is that 2010-11-18T07:41:17 *** flag has quit IRC (Quit: Lost terminal) 2010-11-18T07:42:22 realized why my anti-snipe/fortification code sucks... 2010-11-18T07:42:49 i see :P 2010-11-18T07:44:04 Also found a map on which my 2-ply lookahead made up for my sucky move gen, which made me happy 2010-11-18T07:44:46 McLeopold: 70-100MB or more is no problem, can you make it available as a download link and update it every so often? It would save people from scraping games one by one 2010-11-18T07:45:07 he ain't here, and his server is down 2010-11-18T07:45:08 :/ 2010-11-18T07:45:44 now how to implement this, hmmm 2010-11-18T07:46:17 http://zeroviz.us:8080/ down 2010-11-18T07:46:24 keyboard=] 2010-11-18T07:46:56 Accoun: yeah, down for everyone it seems 2010-11-18T07:54:04 @rankings 2010-11-18T07:54:05 antimatroid1: Top 10 players: Raschi(3704), bocsimacko(3667), bix0r4ever(3617), luksian(3488), smiley1983(3471), protocolocon(3422), dmj111(3415), george(3405), davidjliu(3398), shangas(3384) 2010-11-18T07:54:22 that wont last long 2010-11-18T07:57:10 hehe 2010-11-18T07:59:08 *** yasith has quit IRC (Ping timeout: 276 seconds) 2010-11-18T08:00:54 *** mceier has joined #aichallenge 2010-11-18T08:11:31 *** yasith has joined #aichallenge 2010-11-18T08:13:58 hmmm 2010-11-18T08:16:28 hey bhasker 2010-11-18T08:16:33 how is it going 2010-11-18T08:17:38 i just noticed an ancient bot of mine is performing just as good as my current on the new mapset 2010-11-18T08:17:44 which is kind of scary 2010-11-18T08:17:51 haha 2010-11-18T08:17:53 Top 10 players: Raschi(3749), bocsimacko(3678), bix0r4ever(3602), luksian(3503), smiley1983(3486), protocolocon(3478), davidjliu(3419), george(3418), shangas(3412), asavis(3388) 2010-11-18T08:17:54 because on the old the new version raped it 2010-11-18T08:17:54 welcome to the club:) 2010-11-18T08:18:20 yea.. 2010-11-18T08:18:21 i am going to see if i can get my new version to beat my old version on most maps 2010-11-18T08:18:53 me too 2010-11-18T08:19:03 and that'll be the end of it 2010-11-18T08:19:39 *** delt0r has joined #aichallenge 2010-11-18T08:19:52 *** delt0r___ has quit IRC (Ping timeout: 240 seconds) 2010-11-18T08:21:00 but first i need to really cleanup in here 2010-11-18T08:22:11 *** Mistmanov has quit IRC (Quit: Page closed) 2010-11-18T08:24:52 *** hellman has joined #aichallenge 2010-11-18T08:25:30 hmm so on say map1 2010-11-18T08:25:37 what do you guys think is a good start 2010-11-18T08:25:39 *** Appleman1234 has quit IRC (Quit: Leaving) 2010-11-18T08:26:20 map1? 2010-11-18T08:27:29 like the first map on the original set 2010-11-18T08:29:33 bhasker: which original set? 2010-11-18T08:29:47 oh the first 100 maps that come with the starter packs 2010-11-18T08:29:55 ah 2010-11-18T08:30:00 i don't have those anymore :P 2010-11-18T08:32:16 wher is hidden TCP server? 2010-11-18T08:37:37 what do you mean by hidden tcp server? 2010-11-18T08:43:24 The global level of benzendrine is lower than the level of zeroviz, right? 2010-11-18T08:45:00 i would expect so 2010-11-18T08:46:06 I hope that zeroviz will be fixed soon (I love loosing and the visualizer is better) 2010-11-18T08:47:53 *** medrimonia1 is now known as medrimonia 2010-11-18T08:53:24 too good for the 1? :P 2010-11-18T08:54:04 raschi still hasn't lost a game yet 2010-11-18T08:55:14 confused how did he unseat bockismacko 2010-11-18T08:55:18 without playing him 2010-11-18T08:55:22 with just 7 games? 2010-11-18T08:55:30 that doesn't seem right 2010-11-18T08:55:36 100% win rate? 2010-11-18T08:58:36 *** p4p4 has joined #aichallenge 2010-11-18T09:00:00 yes 2010-11-18T09:00:16 its just a sideeffect from bayesian elo 2010-11-18T09:09:44 yeah, people always overshoot at the start 2010-11-18T09:09:49 it wont matter at the end 2010-11-18T09:10:02 not always 2010-11-18T09:10:06 but if they get a couple of decent wins 2010-11-18T09:10:48 dmj111: new submission isn't doing very well for you :P: 2010-11-18T09:13:37 Hmm it seems that my new bot is not so bad after all 2010-11-18T09:13:54 I was planning to destroy it, but on second sight... 2010-11-18T09:13:57 *** Naktibalda has left #aichallenge 2010-11-18T09:15:37 Does anyone know if it's a problem if My bot "write" a file in the official server (I assume that the bot hasn't the rights to do that but I'm not sure) 2010-11-18T09:16:24 against the rules 2010-11-18T09:16:28 you cannot write to file 2010-11-18T09:16:31 you can read from a file 2010-11-18T09:17:17 yes I know that it's against the rules, but the problem is just that when my bot run, it writes a file just to help me for debugging 2010-11-18T09:17:52 something like info.write('send '+str(num_ships)+' from planet a to planet b') 2010-11-18T09:17:55 Top 10 players: Raschi(3785), bocsimacko(3689), bix0r4ever(3665), smiley1983(3547), protocolocon(3469), davidjliu(3415), george(3409), shangas(3407), luksian(3396), asavis(3373) 2010-11-18T09:18:04 And it overwrite it each time 2010-11-18T09:19:03 I don't care about that file in official server but it's just to avoid the little # at the beginning of each line with info.write 2010-11-18T09:19:10 i'm not sure if it would screw up or not, it'd be heavily frowned upon 2010-11-18T09:19:42 hmm 2010-11-18T09:19:50 i would personally advise against it :) 2010-11-18T09:19:53 raschi is on a scary winning streak 2010-11-18T09:20:18 i dunno, hasn't played bocsimacko yet 2010-11-18T09:20:28 I'll comment it again so^^ 2010-11-18T09:20:30 he usually brings people back down to reality 2010-11-18T09:21:46 http://ai-contest.com/profile.php?user_id=9325 2010-11-18T09:21:49 how is that ranked 98? 2010-11-18T09:21:58 It's funny, because a lot of time ago (he was already first) I was winning almost one game of two against him but... I was loosing all the others 2010-11-18T09:22:06 worst pairing luck ever? 2010-11-18T09:22:37 *** andrei92 has joined #aichallenge 2010-11-18T09:22:48 antimatroid1: I don't know he's usually ranked in the top 50 2010-11-18T09:22:50 hello 2010-11-18T09:22:53 hello 2010-11-18T09:23:21 can someone tell me why i'm getting this error andrei92 issued an invalid order ? 2010-11-18T09:23:31 because you sent more ships than a planet had 2010-11-18T09:23:48 hm...this is strange 2010-11-18T09:23:58 because i check if i have that number of ships... 2010-11-18T09:24:02 i will check again , thanks 2010-11-18T09:24:17 i 2010-11-18T09:24:25 Maybe because you sent a negative number of ship 2010-11-18T09:24:28 i've taken all checks out of my code so i know it isn't buggy 2010-11-18T09:24:30 or if you're on tcp 2010-11-18T09:24:39 maybe you've sent 0 ships 2010-11-18T09:24:56 i'll add tehm back in before the final contest 2010-11-18T09:25:03 but i don't expect they'll be needed 2010-11-18T09:27:26 woohoo , thanks guys 2010-11-18T09:30:36 quit 2010-11-18T09:30:40 *** andrei92 has quit IRC (Quit: Page closed) 2010-11-18T09:31:58 how do you guys decide when a neutral is safe to acquire with +1 ships 2010-11-18T09:32:02 and when to send extra ships? 2010-11-18T09:34:01 hmmm that's a big problem to me too 2010-11-18T09:34:20 @tcp 2010-11-18T09:34:21 Error323: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/, or (#3) http://zeroviz.us:8080, or (#4) 98.247.248.39:995 (zeroviz.us) 2010-11-18T09:34:34 but usually if minlen_friend(p) hmm 2010-11-18T09:40:39 how does that work out for you 2010-11-18T09:40:46 my heuristic is kinda awful 2010-11-18T09:50:27 guys the tcp server of dhartmei is only hosting old maps right? 2010-11-18T09:52:00 *** rebelxt_ has joined #aichallenge 2010-11-18T09:52:22 bhasker: http://72.44.46.68/canvas?game_id=1078869 - nice game! 2010-11-18T09:54:08 *** ColdPhoenix has joined #aichallenge 2010-11-18T09:54:13 Error323: right 2010-11-18T09:54:21 that sux 2010-11-18T09:54:23 >_< 2010-11-18T09:55:00 no one has seen dhartmei 2010-11-18T09:55:13 does someone know him personally? 2010-11-18T09:55:19 it should be the case that there will be an official tcp next time 2010-11-18T09:55:23 no idea 2010-11-18T09:55:31 i know nobody from this contest personally 2010-11-18T09:55:32 because for a guy like us being away from the pcc that loong is unnatural :P 2010-11-18T09:55:49 i suspect he just lost interest in this contest 2010-11-18T09:56:09 well possible but unlikely imo 2010-11-18T09:57:35 *** Cold-Phoenix has quit IRC (Ping timeout: 276 seconds) 2010-11-18T09:57:48 it has happened with quite a few people 2010-11-18T09:58:44 oh man i just found a massive problem with my bot 2010-11-18T09:59:01 time to celebrate? :P 2010-11-18T10:03:04 *** mceier has quit IRC (Quit: leaving) 2010-11-18T10:15:05 *** JamesMG has joined #aichallenge 2010-11-18T10:15:18 @tcp 2010-11-18T10:15:19 JamesMG: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/, or (#3) http://zeroviz.us:8080, or (#4) 98.247.248.39:995 (zeroviz.us) 2010-11-18T10:17:56 Top 10 players: bocsimacko(3702), bix0r4ever(3624), Raschi(3537), smiley1983(3515), SUA.khb(3486), protocolocon(3473), shangas(3422), george(3416), davidjliu(3411), asavis(3398) 2010-11-18T10:19:09 Hmmmm It's frustrating, I've got to go to my university if I want to test my bot properly on the tcp server... 2010-11-18T10:19:15 *** p4p4 has quit IRC (Ping timeout: 245 seconds) 2010-11-18T10:19:16 too much time outs at home 2010-11-18T10:22:02 that would be one way for me to do lots of testing at once 2010-11-18T10:22:09 *** aerique has quit IRC (Quit: ...) 2010-11-18T10:22:14 my uni is pretty much empty atm too, we just finished exams 2010-11-18T10:22:40 We will soon begin the exams 2010-11-18T10:23:02 That's why the contest has to stop the 27^^ 2010-11-18T10:23:19 Because if it don't I'll screw my semester 2010-11-18T10:23:23 *** rebelxt_ has quit IRC (Ping timeout: 265 seconds) 2010-11-18T10:24:05 haha, my semester was terrible 2010-11-18T10:24:13 i studied for each of my exams the day before 2010-11-18T10:24:16 one of them on the day 2010-11-18T10:25:28 In fact the biggest problem is that I don't even go to classes 2010-11-18T10:25:55 I've something like two months of classes to work. 2010-11-18T10:27:01 I'm still in second year of Licence, so it's not so hard, but we've still got to work, at least a little bit. 2010-11-18T10:29:34 *** ColdPhoenix has quit IRC () 2010-11-18T10:29:48 what's the current updatefrequency of the rating at the official server? 2010-11-18T10:32:25 i need to re-prioritize my attack/defend scheme 2010-11-18T10:34:44 *** RainCT has joined #aichallenge 2010-11-18T10:34:59 I need to see where my bot is most broken, I guess zeroviz webserver is down though? 2010-11-18T10:35:10 interesting game rebel though i would have lost if it had continued 2010-11-18T10:35:21 or maybe not 2010-11-18T10:36:45 interesting this version isn't too bad now 2010-11-18T10:39:00 *** lavalamp_ has quit IRC (Quit: Page closed) 2010-11-18T10:40:37 bhasker: my version is waaaaaaaaaaaaaay better with new mapset 2010-11-18T10:40:42 so be carefull :P 2010-11-18T10:41:20 i mean, dhartmei's tcp is not really representative 2010-11-18T10:41:21 :/ 2010-11-18T10:42:17 heh true 2010-11-18T10:42:23 but its doing better there too now 2010-11-18T10:42:26 its still worse than 0.23 2010-11-18T10:42:32 but its abit of an improvement 2010-11-18T10:44:27 http://72.44.46.68/canvas?game_id=1079247 wut, major comeback 2010-11-18T10:45:35 nice 2010-11-18T10:45:53 wow 2010-11-18T10:55:33 wow 2010-11-18T10:55:38 amazing 2010-11-18T10:55:58 don't show that to iouri, he might fix it :) 2010-11-18T10:58:16 i'm a little worried about smiley1983 2010-11-18T10:58:22 i was planning to take first in australia :( 2010-11-18T10:59:49 I'm just shooting for top 5% in country/overall at this point 2010-11-18T11:07:20 which country are you from? 2010-11-18T11:07:28 australia isn't one of the more competitive countries 2010-11-18T11:09:33 *** lavalamp_ has joined #aichallenge 2010-11-18T11:09:39 USA 2010-11-18T11:09:47 ah, well yeah, slightly harder :P 2010-11-18T11:11:00 at least I'm not in hungary :) 2010-11-18T11:11:11 haha :P 2010-11-18T11:11:18 those poor buggers 2010-11-18T11:11:31 there's only 29 of them 2010-11-18T11:11:38 we went over 100 for australia the other day 2010-11-18T11:11:45 maybe if they all got together into a team 2010-11-18T11:11:47 and mine's not included in that atm, as nothings on the official 2010-11-18T11:12:01 I think there are about 1000 frmo the US right now 2010-11-18T11:12:01 it goes 1, 380, 446, ... 2010-11-18T11:12:41 yeah, I'm 51 out of 965 2010-11-18T11:13:16 bah, my hopes are dropping every day 2010-11-18T11:13:22 now i think i'd be happy in the top 50 2010-11-18T11:13:35 i haven't played against people in a while though 2010-11-18T11:13:39 I don't remember what revision of my code is on official though, haha 2010-11-18T11:13:43 so i have no idea where it'll be once i have it finished 2010-11-18T11:13:48 one of my old revisions is by far my best still 2010-11-18T11:14:04 i have a complete rewrite that i haven't tested against people yet 2010-11-18T11:14:09 oh fun 2010-11-18T11:14:18 I stress when I first play new stuff 2010-11-18T11:14:30 i still haven't added in redistribution or redirecting stuff 2010-11-18T11:14:38 i'm trying to get my minimax tree to work as i want 2010-11-18T11:14:39 *** p4p4 has joined #aichallenge 2010-11-18T11:15:02 *** bhasker has quit IRC (Quit: bhasker) 2010-11-18T11:17:50 I've debated combining my minimax bot into my real one, but I think I don't have time to do it right anymore 2010-11-18T11:17:57 Top 10 players: bocsimacko(3706), bix0r4ever(3609), Raschi(3529), protocolocon(3478), Silver_S(3463), smiley1983(3456), davidjliu(3439), george(3426), shangas(3424), asavis(3417) 2010-11-18T11:18:57 *** Palmik has quit IRC (Remote host closed the connection) 2010-11-18T11:19:38 i've spent the last couple of days redoing it 2010-11-18T11:19:52 i think it's getting the information right now 2010-11-18T11:20:00 i just need to get the evaluation working properly 2010-11-18T11:20:27 then i can add the other stuff in and start testing/fixing 2010-11-18T11:20:41 *** rebelxt has joined #aichallenge 2010-11-18T11:21:52 *** Ice_Harley has quit IRC (Ping timeout: 265 seconds) 2010-11-18T11:29:25 *** superflit has joined #aichallenge 2010-11-18T11:34:38 *** VegaForces has quit IRC (Quit: Page closed) 2010-11-18T11:34:39 *** McLeopold has joined #aichallenge 2010-11-18T11:36:38 McLeopold: site's down :P 2010-11-18T11:37:00 ? 2010-11-18T11:37:07 zeroviz 2010-11-18T11:37:14 it's been down for a few hours 2010-11-18T11:37:22 *** mega1 has quit IRC (Ping timeout: 250 seconds) 2010-11-18T11:38:01 k, it's back up 2010-11-18T11:38:08 I'm moving to apache now 2010-11-18T11:38:35 cool 2010-11-18T11:38:50 At least port 995 stays up 2010-11-18T11:39:08 and I'm glad flexicoto is playing in it now :) 2010-11-18T11:39:09 yeah people said that still works 2010-11-18T11:39:15 page still wont load for me 2010-11-18T11:40:12 hmm 2010-11-18T11:40:17 hi 2010-11-18T11:40:18 are you sure? 2010-11-18T11:40:40 it's up for me 2010-11-18T11:41:00 hmm, maybe my cache is being annoying or something 2010-11-18T11:41:09 i did try ctrl f5 numerous times 2010-11-18T11:41:13 yeah, it doesn't send out good cache headers 2010-11-18T11:41:43 ctrl+shift+del 2010-11-18T11:42:41 hmmm, that didn't do it, the rest of the internet seems to work, but i'll try the old power d/c method 2010-11-18T11:44:53 if it doesn't send out good cache headers, antimatroid's ISP might be incorrectly locally caching it 2010-11-18T11:44:54 *** antimatroid has joined #aichallenge 2010-11-18T11:44:55 works for me now, thanks McLeopold 2010-11-18T11:45:47 I kinda wish ai-contest would show a graph on player's profiles of their rank 2010-11-18T11:45:55 .... and now it works :\ 2010-11-18T11:46:04 d0ugal: it did last time 2010-11-18T11:46:08 Ah 2010-11-18T11:46:12 and there was rankings for countries etc. 2010-11-18T11:46:18 i don't know what they got rid of those things 2010-11-18T11:46:22 I just think the number isn't that useful without the context because ranks go up and down so much 2010-11-18T11:46:34 the old one was a graph of your ranking over time 2010-11-18T11:46:50 That's be nice 2010-11-18T11:46:50 That'd* 2010-11-18T11:47:22 bocsimackos would be impressive 2010-11-18T11:47:43 ideally it should do your modal rank over a certain period i think 2010-11-18T11:47:52 otherwise you'd have huge jumps where people resubmit 2010-11-18T11:47:57 *** antimatroid1 has quit IRC (Ping timeout: 276 seconds) 2010-11-18T11:47:58 well, possibly median 2010-11-18T11:48:10 modal would only work well for bocsimacko :P 2010-11-18T11:48:13 guys fyi: e-mailed dhartmei asking to update maps on his tcp server, but e-mails bounced marked as spam :( PM'ed him as well on the forum since he used to respond to PM 2010-11-18T11:48:24 graph should be of final rank for each submission 2010-11-18T11:48:39 rebelxt: already tried emailing him and got the bounce like a week ago :) 2010-11-18T11:49:01 lavalamp_: yep 2010-11-18T11:49:05 that'd work well 2010-11-18T11:49:26 would also show who resubmits all the time ehe 2010-11-18T11:49:38 ...and maybe it should graph percentile, not elo or place # 2010-11-18T11:50:04 since both of those depends on how many other people have entries 2010-11-18T11:51:01 antimatroid: I'd settle for a new graph per submission 2010-11-18T11:51:37 i prefer lavalamps idea 2010-11-18T11:51:38 antimatroid: he is still paying for the server so he is around.. probably just busy with RL stuff 2010-11-18T11:51:53 yeah, i'm pretty sure he has the server for more than just that 2010-11-18T11:52:03 it doesn't require too much traffic to do the tcp thing i don't think 2010-11-18T11:52:12 (not taking away from how awesome he is) 2010-11-18T11:52:30 hmm, maybe - but you'd have to do a shed load of submissions to make interesting graphs 2010-11-18T11:52:47 you don't want interesting graphs 2010-11-18T11:53:00 just a basic idea of how the user progressed through the tournament 2010-11-18T11:53:25 Suppose' 2010-11-18T11:53:32 but I also really want interesting graphs :) 2010-11-18T11:53:36 i thought the graphs were fine last time as they were 2010-11-18T11:53:50 jmcarthur: so what happened to them? :P 2010-11-18T11:53:54 and the country rankings 2010-11-18T11:54:13 they went up and down a lot since they showed your rank continuously even when you resubmitted, but you could still see through the noise when you wanted to only compare your final rank at any point 2010-11-18T11:54:37 and it also worked well for those cases that you didn't submit often. you could see how your bot was falling as others got better 2010-11-18T11:54:49 it was actually kind of motivating, IMO 2010-11-18T11:54:58 "i better get off my ass and make this thing better!" 2010-11-18T11:55:03 heh 2010-11-18T11:55:08 *** rebelxt has left #aichallenge 2010-11-18T11:55:12 So you must have thought we were too motivated and removed em? 2010-11-18T11:55:14 antimatroid: i have no idea why they are gone 2010-11-18T11:55:32 yeah, that's why i would think maybe a kind of average value over a smaller time period say 6-12 hours 2010-11-18T11:55:35 graph of a single submission is not meaningful because the submission plays at a constant level 2010-11-18T11:55:38 d0ugal: if i had been a decision maker i would have left them in, unless there is information i don't know about 2010-11-18T11:55:48 lavalamp_: not true 2010-11-18T11:55:49 Fair 'nuff 2010-11-18T11:55:49 so you could potentially avoid the bumps at resubmissions but capture the changes over longer time periods 2010-11-18T11:55:59 Guess the graphs would be fairly process intensive 2010-11-18T11:56:00 lavalamp_: other bots improve, so your submission drops over time 2010-11-18T11:56:05 but probably not that much more 2010-11-18T11:56:12 and lots of results to store too.. 2010-11-18T11:56:28 jmcarthur, yeah ok but I don't think that's why people want to watch graphs ^^ 2010-11-18T11:56:37 lavalamp_: were you around for the last contest? 2010-11-18T11:56:42 nope 2010-11-18T11:56:50 it worked pretty well as a continuous graph of rank, IMO 2010-11-18T11:57:03 I wish I'd known about this one when it started and not a month later :( 2010-11-18T11:57:26 I can believe that 2010-11-18T11:57:30 i found out about this one a few days before it opened 2010-11-18T11:57:42 someone stumbled across the new page and posted it on reddit 2010-11-18T11:57:53 antimatroid: or maybe the smoothing could be weighted by time since last submission rather than a constant sliding window 2010-11-18T11:58:02 not sure what that would look like, but it might be nice 2010-11-18T11:58:23 yeah, doesn't need to be perfect, just give a decent idea of how one has done/ is doing 2010-11-18T11:58:55 actually, if we were using trueskill we could just show the mean score for the graph. it wouldn't change upon resubmission. it would just become more bouncy for a while 2010-11-18T11:59:12 how does trueskill work? 2010-11-18T11:59:13 or perhaps lower bound 2010-11-18T11:59:23 ah no not lower bound 2010-11-18T11:59:28 * antimatroid is very unknowledgable on ranking methods 2010-11-18T12:00:29 antimatroid: it's complicated, but essentially you have a running mean and a running confidence interval, and both are taken into acount for each game. unlike bayeselo, it's iterative. there's also a well-studied leaderboard-ranking and matching algorithm that comes with it 2010-11-18T12:00:46 s/running confidence interval/running deviation/ 2010-11-18T12:00:56 confidence interval i suppose *is* the combination of both 2010-11-18T12:01:57 antimatroid: i forget who now, but somebody ran some tests with it on this contest and said it gave bad results, but i suspect that it wasn't being used as intended. the matching algorithm is designed to maximize the information gained from each game and stuff 2010-11-18T12:02:36 the matching algorithm is also probabilistic, so you still get variety 2010-11-18T12:03:02 i rather like trueskill on paper, but i admit i have never used it for anything 2010-11-18T12:03:11 seems to work well on xbox live though 2010-11-18T12:03:44 *** mega1 has joined #aichallenge 2010-11-18T12:04:52 yeah fair enough 2010-11-18T12:06:00 trueskill is intended for contestants that vary in strength over time, which only happens here during resubmissions 2010-11-18T12:06:36 if it was used then maybe you wouldn't have to clear history every time people resubmit 2010-11-18T12:07:24 lavalamp_: right that was why i was pushing for it earlier in the contest 2010-11-18T12:07:32 ah 2010-11-18T12:07:53 lavalamp_: a small tweak to the algorithm would even make it assume bots keep the same skill level until they are resubmitted 2010-11-18T12:08:01 conversely maybe it doesn't work well for contestants of constant strength? 2010-11-18T12:08:09 jmcarthur: yeah that would be awesome 2010-11-18T12:08:13 on paper anyway 2010-11-18T12:08:32 unfortunately the only way to test it is to also pick up its matching algorithm and stuff 2010-11-18T12:09:10 would that be too bad? 2010-11-18T12:09:39 the current matching algorithm... leaves room for improvement 2010-11-18T12:09:39 well, i think it would be at this point. we should have done it earlier. we could have always reverted back if it did turn out badly 2010-11-18T12:09:52 oh yeah it's too late now, definitely 2010-11-18T12:09:54 people are just resistant to change :\ 2010-11-18T12:10:20 but I strongly believe that the final should not be determined by a rating algorithm 2010-11-18T12:10:30 because of intransitivities 2010-11-18T12:10:33 IMO one of the nice things about the games not counting until the final tournament is that we can experiment with things like this 2010-11-18T12:10:50 but very few other people seem to share my view 2010-11-18T12:10:54 rating algorithm can determine seed # into a tournament of some kind 2010-11-18T12:11:17 jmcarthur I'll vote for you ^^ 2010-11-18T12:11:20 ha 2010-11-18T12:11:26 i'm not even an organizer 2010-11-18T12:11:27 jmcarthur: i also think trueskill would be a good idea. 2010-11-18T12:11:44 i do agree with lavalamp_ that it would be a poor choice for the tournament 2010-11-18T12:11:58 i just think it would be about ideal for the three month testing period 2010-11-18T12:12:01 I think a series of round-robin tournaments would be best 2010-11-18T12:12:13 i'm not going to guess what would be best for the tournament 2010-11-18T12:12:14 yes, i agree, for the runup. 2010-11-18T12:12:34 i'd rather suggest things that have very well established backgrounds for it, and i don't know the literature well enough 2010-11-18T12:13:42 i keep thinking running swiss tournaments over and over again for the final tournament. 2010-11-18T12:13:50 as many as time permits. 2010-11-18T12:14:16 dmj111: or one looooong one 2010-11-18T12:14:25 you can always just keep adding rounds 2010-11-18T12:14:27 lavalamp_: what about a merge sort like process with the final round being round robbins? 2010-11-18T12:14:32 Hmmmm. Sure, that would work :) 2010-11-18T12:15:13 antimatroid: it would be interesting to try that and see what results it gives! 2010-11-18T12:15:14 antimatroid: i think a reealllly long swiss would end up looking like a round robin at the top, and every bot would still play the same number of games. 2010-11-18T12:15:26 yeah, that's correct 2010-11-18T12:15:43 maybe the last half day for only the top, or something, but I like everyone getting to play. 2010-11-18T12:15:45 the large RL tournaments I've been to are all swiss 2010-11-18T12:15:46 so do completely random pairings inside groups, after certain time intervals, split groups in half, repeat, then in the final round with groupd of like 25, round robbin the shit out of that 2010-11-18T12:16:10 some variation of swiss I should say 2010-11-18T12:16:13 and yeah, i'd give the top group a full day to itself 2010-11-18T12:16:44 i think a bunch of these need to be tried with the submitted bots after this contest is run 2010-11-18T12:16:49 and see what the different results are 2010-11-18T12:16:56 absolutely agree with that 2010-11-18T12:17:16 so that next time they know what to do 2010-11-18T12:17:47 after this is over, I was going to try some experiments like that. I have a two player true skill program written too, so I wanted to see how it did too. 2010-11-18T12:17:48 i'm only willing to contribute so far as i'm still eligible to compete 2010-11-18T12:17:57 Top 10 players: bocsimacko(3812), bix0r4ever(3753), Raschi(3670), protocolocon(3582), luksian(3563), davidjliu(3560), smiley1983(3549), george(3537), shangas(3530), jimrogerz(3530) 2010-11-18T12:18:07 same here ^^ 2010-11-18T12:18:14 i expect a lot of people are in that boat 2010-11-18T12:18:16 oh, antimatroid yeah, my bot is all messed up. 2010-11-18T12:18:27 i noticed :P 2010-11-18T12:18:35 i'm ironing out minimax 2010-11-18T12:18:37 9 rounds of swiss apparently should be enough for a clear winner 2010-11-18T12:18:48 i think i have state calculations right, i just need to get evaluation working properly 2010-11-18T12:19:04 jmcarthur: that's equilent to a single elimination tournament at the top 2010-11-18T12:19:05 according to wikipedia: "Determining a clear winner (and, incidentally, a clear loser) usually requires the same number of rounds as a knockout tournament, that is the binary logarithm of the number of players rounded up." 2010-11-18T12:19:20 how many players are you assuming? 2010-11-18T12:19:27 the number we have in the tournament currently 2010-11-18T12:19:46 it would take 2**9 is only 512, isn't it? 2010-11-18T12:19:49 single elimination is definitely not good for this type of contest 2010-11-18T12:19:49 the last tournament have like 750 bots entered into the final contest, this will presumable have a little over 4000 2010-11-18T12:19:55 if it keeps growing.. 2010-11-18T12:19:59 hmm maybe i used the wrong log base 2010-11-18T12:20:05 had* 2010-11-18T12:20:05 you're right of course 2010-11-18T12:20:06 13 rounds. 2010-11-18T12:20:22 yeah that sounds right 2010-11-18T12:20:31 (too much time using FFT's with power of two sizes...) 2010-11-18T12:21:01 although are there really any people with bots ranked below 1000 that are actually trying? 2010-11-18T12:21:17 i see bots submitted right at the start from people at those ranks 2010-11-18T12:21:31 but if you keep adding rounds to a swiss it will be more and more like a round robin at the top, which would be very very good 2010-11-18T12:22:02 antimatroid: yeah.. but how can we avoid dropping people ranked that low who are trying? I have a coworker ~1500 who likes to keep watching his play, even if he hasn't changed it. 2010-11-18T12:22:16 and, i just contradicted myself. 2010-11-18T12:23:06 yeah, i do advocate for giving those people a decent go 2010-11-18T12:23:27 would he be opposed to having to resubmit every few weeks? 2010-11-18T12:23:42 as a guess 2010-11-18T12:24:11 probably not ideal 2010-11-18T12:24:17 No, in fact I wrote an issue on the issue tracker suggesting something like that. His idea as an arcade analogy... log in and click a button to drop a quarter in the machine, and get 50 more games. 2010-11-18T12:24:33 yeah, that'd be good 2010-11-18T12:24:34 run a tournament with everyone for a day, then drop the bottom half. rinse and repeat. 2010-11-18T12:24:48 lavalamp_: don't give the bottom people good enough rankings 2010-11-18T12:24:55 some of those people may have spent 3 months on their bot 2010-11-18T12:25:07 good point 2010-11-18T12:25:29 oh awesome, there's now 6 submissions in go: http://ai-contest.com/language_profile.php?lang=Go 2010-11-18T12:26:25 yeah, I guess swiss tournament with as many rounds as possible is my ideal final 2010-11-18T12:26:44 maybe with extra rounds for the top 100 2010-11-18T12:26:58 now, i don't know much about tournament systems, but what's wrong with an n-elimination tournament for some large value of n? does n just have to be too large to be practical? 2010-11-18T12:28:03 the advantages of swiss over knockout given on wikipedia seem to mainly be that if you aren't limited to a certain number of games at a time it might converge more quickly... but we *are* limited to a certain number of games at a time 2010-11-18T12:29:05 oh, and swiss doesn't eliminate people of course, so you get to determine a clear loser as well as a clear winner 2010-11-18T12:29:22 "Compared with a round-robin tournament, a Swiss can handle many players without requiring an impractical number of rounds. An elimination tournament is better suited to a situation in which only a limited number of games may be played at once, e.g. tennis. In a Swiss system, all players can be playing a round at the same time." 2010-11-18T12:32:02 name something that isn't limited to number of games at a time? 2010-11-18T12:32:12 chess 2010-11-18T12:32:17 how? 2010-11-18T12:32:23 there's a finite number of chess boards on earth 2010-11-18T12:32:26 every player can play at the same time 2010-11-18T12:32:28 and finite number of people 2010-11-18T12:32:29 *sigh* 2010-11-18T12:32:33 oh, :P 2010-11-18T12:32:37 practically it's possible :P 2010-11-18T12:32:50 okay, i get it now 2010-11-18T12:32:53 i takes things literally 2010-11-18T12:32:59 too much sometimes 2010-11-18T12:33:44 so clearly the solution is to get enough servers such that every bot can play at once :P 2010-11-18T12:34:21 just as the solution in tennis tournaments it to buy enough tennis courts that every player can play at once 2010-11-18T12:34:23 have everyone install something and leave a port open for the 3 days... 2010-11-18T12:34:25 *is to buy 2010-11-18T12:34:30 yes 2010-11-18T12:34:58 JamesMG: can't do it through tcp like ways 2010-11-18T12:35:17 people use whatever computing power they want, and if you have a crap connection, say goodbye to a good rank 2010-11-18T12:35:29 it's impractical for other reasons too 2010-11-18T12:35:43 bots change 2010-11-18T12:35:47 etc. etc. yeah 2010-11-18T12:40:20 hey, in my dream, i solved it ! man, it was soo easy, i just assumed, i'd own all planets and just A*'ed backwards! 2010-11-18T12:40:56 haha 2010-11-18T12:40:57 haha 2010-11-18T12:41:24 that's awesome 2010-11-18T12:41:45 lol 2010-11-18T12:48:29 http://ai-contest.com/visualizer.php?game_id=7834518 2010-11-18T12:48:37 bocsimacko was a bit unlucky there 2010-11-18T12:48:52 i think he had the win if the turn limit was different 2010-11-18T12:49:42 it looks like a clear early mistake to me 2010-11-18T12:50:16 didn't take as many planets as bix0r4ever 2010-11-18T12:50:40 hmm 2010-11-18T12:50:41 yeah, but assuming bix0r didn't manage to take the planet with 5 ships at the end 2010-11-18T12:50:44 but it did waver a bit anyway 2010-11-18T12:50:46 nevermind 2010-11-18T12:50:58 it'd put him behind to catch up on growth 2010-11-18T12:51:14 the point is that the advantage was held almost the entire time 2010-11-18T12:51:31 my point is that excluding time limits, it probably would have switched 2010-11-18T12:51:56 sure it is now the game to optimise for turn 200 2010-11-18T12:52:11 just making a comment on the fact that the outcome out have been different if the turn limit was different 2010-11-18T12:52:17 might* have 2010-11-18T12:52:24 :) 2010-11-18T12:52:30 i am still unconvinced that time limits taint the game 2010-11-18T12:52:37 to me they are simply a part of it 2010-11-18T12:52:43 especially now that it's actually in stone 2010-11-18T12:52:44 i think it's very rare that games like that happen 2010-11-18T12:52:46 but i have had a few 2010-11-18T12:52:59 it was always the game to optimize for turn n, we just didn't decide what n should be until recently 2010-11-18T12:53:01 yeah 2010-11-18T12:53:08 no debate there 2010-11-18T12:53:18 yeah, i don't mind at all, i understand it has to happen 2010-11-18T12:53:23 it was merely a comment :P 2010-11-18T12:53:26 *** bhasker has joined #aichallenge 2010-11-18T12:53:47 but then again, if the turn limit was different then many bots would also be coded differently. we can't really guess what bots are good at what turn limits 2010-11-18T12:53:48 woah, dhartmei posted in the forums 2010-11-18T12:53:56 well, different constants in the code at least 2010-11-18T12:54:04 which can have big effects 2010-11-18T12:54:04 new maps are on benzadrine 2010-11-18T12:54:09 Accoun: ^^ 2010-11-18T12:55:27 nice 2010-11-18T12:57:32 hey so is the dhartmei's server running the new maps now? 2010-11-18T12:57:53 he added janzerts 50 maps and 450 he generated with the current generator 2010-11-18T12:57:55 antimatroid: the benzendrine visualizer has some problems with the new maps 2010-11-18T12:57:57 i expect old maps are stil there 2010-11-18T12:58:05 *** mceier has joined #aichallenge 2010-11-18T12:58:15 mega1: post in the forum? 2010-11-18T12:58:35 *** mceier has quit IRC (Client Quit) 2010-11-18T12:58:54 it does though, hmm 2010-11-18T12:58:57 *** mceier has joined #aichallenge 2010-11-18T13:00:03 i've no idea when dhartmei will next be on the forums, hopefully he will check again or the games 2010-11-18T13:00:43 mega1: http://ai-contest.com/visualizer.php?game_id=7834518 do you think you would have won that given more time? 2010-11-18T13:01:04 not that it really matters, i just find it interesting when results probably would have changed 2010-11-18T13:01:48 i can't tell if you're about to lose the planet with 5 ships or not, i think it depends on that 2010-11-18T13:02:48 antimatroid: possible, but my bot was rather braindead several times. Deserved to lose. 2010-11-18T13:02:53 rather surprising to me it appears draws between top 100 competitors is much lower than overall draw rate on the drawish maps 2010-11-18T13:03:41 janzert: not really 2010-11-18T13:03:57 stupid bots are more likely to not see any possible moves 2010-11-18T13:04:15 whereas smarter bots will see a game that is actually there to be played 2010-11-18T13:04:31 so the stupider bots are drawing on maps that shouldn't necessarily be draws 2010-11-18T13:04:38 at least that'd be what i'd have thought 2010-11-18T13:04:44 that seems to be what is happening anyway 2010-11-18T13:05:05 one of my problems is my stupid bot deciding it'd be best to just sit there 2010-11-18T13:05:06 I thought it was more likely for the smarter bots to see when a map was truly a draw and play for it 2010-11-18T13:05:23 i don't think they happen as often as a stupid bot misses possible moves 2010-11-18T13:05:32 well, clearly :P 2010-11-18T13:05:38 otherwise we wouldn't be discussing this 2010-11-18T13:06:08 also, there are now checks in to try and remove those very trivial draw maps 2010-11-18T13:07:16 yeah, must be effective :) 2010-11-18T13:07:44 is there a map in the current set that is a theoretical draw even though the simple draw is avoided? 2010-11-18T13:07:54 not sure sorry 2010-11-18T13:07:58 ok 2010-11-18T13:08:09 mcleopolds server might be possible to find one 2010-11-18T13:08:29 http://zeroviz.us:8080/by_Map.html 2010-11-18T13:08:32 sort by draw percent 2010-11-18T13:09:12 on the contest server with name, overall games, draws, top 100 games, draws 2010-11-18T13:09:14 map_r528_4 2 16662 402 (2.4%) 127 9 (7.1%) 2010-11-18T13:09:31 is the highest draw percentage among top 100 2010-11-18T13:09:34 what's r528_17 like? 2010-11-18T13:09:49 that has 25% draws on zeroviz, but look at it 2010-11-18T13:09:54 http://zeroviz.us:8080/canvas?game_id=179 2010-11-18T13:10:02 map_r528_27 2 16707 369 (2.2%) 111 (0.0%) 2010-11-18T13:10:07 no draws in top 100 2010-11-18T13:10:16 oops, wrong one 2010-11-18T13:10:31 map_r528_17 2 16335 2133 (13.1%) 104 (0.0%) 2010-11-18T13:12:24 I was going to remove maps that hit a draw percentage of something like 10 or 20 percent among games between top 100... 2010-11-18T13:13:00 while that seems like it's still a good idea it appears fairly unlikely with the current generator 2010-11-18T13:13:17 i would have thought even that kind of number is fairly acceptable 2010-11-18T13:13:21 janzert: and you'd risk removing maps that are not draws 2010-11-18T13:13:51 mega1: yes, there is that risk 2010-11-18T13:13:52 it sucks to get a draw map, but it also sucks to get a map removed that your bot plays well on 2010-11-18T13:14:05 yeah, i saw leave them 2010-11-18T13:14:08 say* 2010-11-18T13:14:20 there are a fair number of the new maps that get ~30% draw rates overall though 2010-11-18T13:14:44 map_r538_15 1 1383 420 (30.4%) 15 (0.0%) 2010-11-18T13:15:24 actually here http://ai-contest.com/map_stats.php but please don't post it in the forums 2010-11-18T13:16:06 the query takes a few seconds to run and I don't want to have it hit too many times 2010-11-18T13:16:36 http://zeroviz.us:8080/canvas?game_id=7047 2010-11-18T13:16:39 that's the map 2010-11-18T13:16:48 i think it might be a draw map 2010-11-18T13:16:49 mega1? 2010-11-18T13:17:12 although even if it's a draw map in perfect play, a lot of top bots might fail on it 2010-11-18T13:17:27 so i think ideally that 30% number is stil more than acceptable 2010-11-18T13:17:37 i'd probably be inclined to keep it if it was up to me 2010-11-18T13:17:47 that's not a trivial draw map in my book at all 2010-11-18T13:17:58 Top 10 players: bocsimacko(3719), bix0r4ever(3623), Raschi(3505), protocolocon(3488), davidjliu(3455), smiley1983(3446), george(3442), shangas(3432), luksian(3418), asavis(3408) 2010-11-18T13:18:13 yeah, I guess unless I see maps hitting really high percentages I'll leave it alone 2010-11-18T13:18:14 i can't tell easily without distances 2010-11-18T13:19:11 I think the only ones we _may_ want to remove are the maps that are proven draws and they are also have high draw % in practice 2010-11-18T13:19:41 and i don't think it's even going to happen 2010-11-18T13:19:57 janzert: any idea of how many maps we'll be looking at having used? 2010-11-18T13:20:17 not really 2010-11-18T13:20:58 > 30 < 100000 :) 2010-11-18T13:21:25 not general enough! /s 2010-11-18T13:21:28 :p 2010-11-18T13:25:30 *** hellman has quit IRC (Remote host closed the connection) 2010-11-18T13:28:04 oh nice my bot broke into the 300's with some lucky wins 2010-11-18T13:30:35 *** tapwater has joined #aichallenge 2010-11-18T13:31:29 *** asbadrobot has joined #aichallenge 2010-11-18T13:32:36 draw map? http://zeroviz.us:8080/zeroviz_538_36.txt.html 2010-11-18T13:34:37 I think so. 2010-11-18T13:35:15 in practice however ... 2010-11-18T13:35:58 yeah, not everyone can see to wait... 2010-11-18T13:36:59 *** tim__ has joined #aichallenge 2010-11-18T13:38:33 yeah, i don't think i'd even remove that map 2010-11-18T13:38:33 the map generator v528 produced a larger percentage of draw maps. 2010-11-18T13:39:13 I'm itching to make an example of one of the maps, but their too good. 2010-11-18T13:39:28 what? 2010-11-18T13:39:58 nvm 2010-11-18T13:40:06 200 turns map http://zeroviz.us:8080/zeroviz_538_10.txt.html 2010-11-18T13:40:22 hum...what's changed in 538 compared to 528? Im crashing in EVERY game now 2010-11-18T13:40:35 no enough to cause that 2010-11-18T13:40:47 maybe max distance? 2010-11-18T13:40:59 or planet count? 2010-11-18T13:41:06 wasnt it 24 in 528? 2010-11-18T13:41:07 yeah, you have upper limit of 30 planets * 34 distance 2010-11-18T13:41:09 yes 2010-11-18T13:41:11 ah 2010-11-18T13:41:13 I see 2010-11-18T13:41:16 but it was 34 before that 2010-11-18T13:41:20 makes perfect sense :) 2010-11-18T13:41:21 i made a mistake 2010-11-18T13:41:25 :) 2010-11-18T13:41:35 30 distance, not 34 2010-11-18T13:41:35 i made many mistakes actually, that was one of them 2010-11-18T13:41:43 yeah I just read its going to change to 24 so I removed the extra prediction and hard coded it to 24 2010-11-18T13:41:57 30? is that for same map area? 2010-11-18T13:42:05 radius 15 2010-11-18T13:42:27 alright. Thanks for this, got to fix it it when Ive got time 2010-11-18T13:42:36 final max distance is going to be 30 just to be totally clear 2010-11-18T13:42:44 yes 2010-11-18T13:42:45 that was a compromise to balance max distance and density 2010-11-18T13:43:00 ...but my bot will still detect it :P 2010-11-18T13:43:07 previous maps were 24x24 in area with 23 planets 2010-11-18T13:43:19 so 30*62 is my current max number of states :| 2010-11-18T13:43:35 i think i can handle it 2010-11-18T13:44:03 although the way i'm currently doing stuff, i think i could shipd that 62 in half 2010-11-18T13:44:11 chop** 2010-11-18T13:44:27 shipd is not even close to chopped, it's 5:44am, night :) 2010-11-18T13:44:30 that was an area/planet ratio of 25.0 2010-11-18T13:44:52 at 23 planets, radius 15, its now 30.7 2010-11-18T13:45:20 at 30 planets, it's now 23.56 2010-11-18T13:45:41 hmm... looks like maps will be less dense 2010-11-18T13:46:14 i think i'm pretty happy with the generalness of maps 2010-11-18T13:46:24 yes 2010-11-18T13:46:27 *** choas has joined #aichallenge 2010-11-18T13:46:59 actually, it's hard to compare densities with different shaps, because distance has a big affect 2010-11-18T13:47:35 density inside the convex hull? 2010-11-18T13:48:01 would total lenght of the min span tree be useful at all? 2010-11-18T13:48:06 length* 2010-11-18T13:48:20 *** tim__ has quit IRC (Ping timeout: 255 seconds) 2010-11-18T13:48:47 i wouldn't like to be the one to work out how to calculate that first one 2010-11-18T13:49:53 yell loudly when I break the visualizer 2010-11-18T13:49:58 if 2010-11-18T13:53:00 one thing that irritates me for no reason at all is that the mouse over planet showing reverses for the "dominant" planet, even when you're still not on the boundary of the undominant one 2010-11-18T13:53:03 if that makes any sense 2010-11-18T13:53:16 it really doesn't matter at all, i'm just making a comment about it annoying me :P 2010-11-18T13:53:25 the fixes are awesome otherwise 2010-11-18T13:54:16 I don't understand 2010-11-18T13:54:42 you know how if i put my mouse over a "behind" planet it comes to the front?" 2010-11-18T13:55:11 if i move the mouse back over to the original dominant planet, without actually going outside the new dominant circle, it switches back over to the old one 2010-11-18T13:55:20 oh, I see 2010-11-18T13:55:21 it's really unimportant 2010-11-18T13:55:34 I just hooked into the active planet stuff 2010-11-18T13:55:58 I think they would need to do a distance check rather than a boundary check 2010-11-18T13:56:24 it really wasn't necessary to pass the distances and growths of planets to us every turn 2010-11-18T13:56:36 although that'd be interesting, a game where new neutrals pop up as the game progresses 2010-11-18T13:56:56 i couldn't handle that atm 2010-11-18T13:57:00 or your planet gets nuked 2010-11-18T13:57:04 ooo dynamic maps.... that would break me totally 2010-11-18T13:57:27 it would be a rather stupid game 2010-11-18T13:57:27 it might break me in a few spots, but not much 2010-11-18T13:57:29 but fun to watch 2010-11-18T13:57:38 no more draws 2010-11-18T13:57:50 I pre-calculate a bunch of distance tables 2010-11-18T13:57:53 if no neutral captures for 20 turns, pop up new planets 2010-11-18T13:57:57 you'd probably want to bound planet maximums too 2010-11-18T13:58:01 so might have to nuke planets 2010-11-18T13:58:28 better than adding a new neutral, just subtract ships from existing ones 2010-11-18T13:58:33 this game deserves much more time devotion than a 3 month contest 2010-11-18T13:58:37 it's rather complicated 2010-11-18T13:58:38 too bad you can't do scorched earth 2010-11-18T13:58:58 take a planet and destroy it? 2010-11-18T13:59:03 if you're going to lose a planet and there is nothing you can do, sacrifice your ships to reduce growth to 0 2010-11-18T13:59:09 hehe that would be awesome 2010-11-18T13:59:16 netrual planets could work in reverse with growth 2010-11-18T13:59:17 defense becomes super important 2010-11-18T13:59:19 till reaching 0 2010-11-18T13:59:43 you could take an enemy planet just for a turn and destroy it 2010-11-18T13:59:49 yeah :) 2010-11-18T13:59:49 or make it a real brain fuck and do proportional growth rates 2010-11-18T13:59:53 *** Bobng has joined #aichallenge 2010-11-18T14:02:48 how about have the planets move along predefined paths? like orbiting the center planet... 2010-11-18T14:03:22 so you can increase growth by consolidating? 2010-11-18T14:03:34 or increase growth by spreading out evenly? 2010-11-18T14:03:35 well, I guess it would just change distance calculations 2010-11-18T14:03:55 move planets would be awesome 2010-11-18T14:04:20 hop a ride on an orbiter and swing into the enemy base and reck havok 2010-11-18T14:05:03 how about you can destroy growthrate*10 ships on a planet to get +1 to the growthrate... 2010-11-18T14:05:13 redirecting fleets would be interesting 2010-11-18T14:05:28 although not sure you could do it with current dist calcs 2010-11-18T14:05:38 actually you could, but it'd be messyish 2010-11-18T14:05:40 redirecting fleets would cause havok, oh man 2010-11-18T14:05:53 the holy grail of mess* 2010-11-18T14:07:10 actually, you'd just pick your optimal ship accumulation points 2010-11-18T14:07:16 i don't think it'd be that much harder 2010-11-18T14:07:34 you'd sort of be making your own dynamic 0 growth planets 2010-11-18T14:08:22 i guess the best way is lots of 0 growth 0 ship neutrals 2010-11-18T14:08:30 and fleets can battle "in mid air" 2010-11-18T14:11:04 *** rebelxt has joined #aichallenge 2010-11-18T14:11:32 when the contest first started, the battle sequence wasn't set 2010-11-18T14:11:48 fyi all: dhartmei has just replied - he replaced the maps on his tcp server with new ones! 2010-11-18T14:11:56 Awesome 2010-11-18T14:12:00 I saw code that did an orbital battle first, then the remaining ships when to the surface to battle the on planet ships 2010-11-18T14:12:15 sweet 2010-11-18T14:12:26 I liked the orbital method better. 2010-11-18T14:12:47 for the cost of 1 ships, you could have prevented the enemy from taking a neutral 2010-11-18T14:12:59 then taken it the next turn 2010-11-18T14:13:10 slightly better than growth + 2 2010-11-18T14:13:23 *** narnach has quit IRC (Quit: Leaving.) 2010-11-18T14:13:57 seems like it'd make for more sophistication in sniping 2010-11-18T14:13:59 yeah, i liked another resolution method better 2010-11-18T14:14:19 how was the current one settled on? 2010-11-18T14:14:23 *** rebelxt has left #aichallenge 2010-11-18T14:14:34 the stoning of dissenters, lavalamp_ 2010-11-18T14:14:39 dhartmei needs to update the visualizer 2010-11-18T14:14:47 planets are being draw out of the box 2010-11-18T14:14:58 lol 2010-11-18T14:15:29 lavalamp_: none were an obvious best choice 2010-11-18T14:15:32 a decision was made 2010-11-18T14:16:18 heh... I guess we're lucky one was made at all xP 2010-11-18T14:17:21 that had to be done 2010-11-18T14:17:31 each option resulted in a completely different game 2010-11-18T14:17:58 Top 10 players: bocsimacko(3717), bix0r4ever(3572), protocolocon(3483), Raschi(3481), davidjliu(3442), luksian(3437), george(3431), shangas(3422), asavis(3407), Xrillo(3384) 2010-11-18T14:18:54 chances of bocsimacko winning this are looking very very good 2010-11-18T14:19:08 wow, expansion failure: http://zeroviz.us:8080/canvas?game_id=9428 2010-11-18T14:20:02 hey, you got three whole planets... 2010-11-18T14:20:11 good ones too 2010-11-18T14:20:16 better believe it 2010-11-18T14:27:10 it's interesting that bocsimacko has a huge point lead over everyone else in the top 10 :) 2010-11-18T14:28:42 by watching his games i get the feeling his strategy is rather different than everyone elses - some kind of advance planning taking place there 2010-11-18T14:39:42 *** perror has quit IRC (Quit: Bye all !) 2010-11-18T14:41:45 hello 2010-11-18T14:43:06 @tcp 2010-11-18T14:43:07 McLeopold: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/, or (#3) http://zeroviz.us:8080, or (#4) 98.247.248.39:995 (zeroviz.us) 2010-11-18T14:43:25 contestbot forget tcp 3 2010-11-18T14:43:25 McLeopold: Ready to serve. 2010-11-18T14:43:54 contestbot learn tcp as http://tcp.zeroviz.us/ 2010-11-18T14:43:54 McLeopold: OK 2010-11-18T14:43:59 @tcp 2010-11-18T14:44:00 McLeopold: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/, or (#3) 98.247.248.39:995 (zeroviz.us), or (#4) http://tcp.zeroviz.us/ 2010-11-18T14:47:08 I'm on apache now. Let me know if you find issues. 2010-11-18T14:48:18 *** Ice_Harley has joined #aichallenge 2010-11-18T14:50:58 I wish they would let us use multiple threads.... my design could run in parallel really easily 2010-11-18T14:54:51 Is anyone experiencing a lot of timeout to my tcp? 2010-11-18T14:55:40 someone on the forum was 2010-11-18T14:55:46 yeah, I saw that 2010-11-18T14:56:06 I do see http://zeroviz.us:8080/TIMEOUT.html 2010-11-18T14:56:08 it was running fine for me last night 2010-11-18T14:56:25 225, I'm not sure if that's my end or the other people's end 2010-11-18T14:58:32 Yeah! First time inside top20! 2010-11-18T15:05:35 *** RainCT has quit IRC (Remote host closed the connection) 2010-11-18T15:07:13 *** davidd has quit IRC (Ping timeout: 260 seconds) 2010-11-18T15:08:07 *** Mistmanov has joined #aichallenge 2010-11-18T15:15:05 *** yasith has quit IRC (Ping timeout: 245 seconds) 2010-11-18T15:16:25 I'm going to drink the mercurial cool-aid 2010-11-18T15:17:59 Top 10 players: bocsimacko(3723), bix0r4ever(3614), protocolocon(3493), luksian(3472), george(3442), davidjliu(3441), shangas(3426), Raschi(3413), jimrogerz(3409), smiley1983(3396) 2010-11-18T15:18:36 * lavalamp_ buys McLeopold a round 2010-11-18T15:18:50 8) 2010-11-18T15:19:07 bitbucket.org is your friend 2010-11-18T15:19:13 :D 2010-11-18T15:20:39 *** Accoun has quit IRC () 2010-11-18T15:21:36 commit vs. push? 2010-11-18T15:21:56 commit is local 2010-11-18T15:22:06 commit before push 2010-11-18T15:23:32 ah 2010-11-18T15:25:00 github is my friend 2010-11-18T15:25:18 github doesn't do free private repos 2010-11-18T15:25:26 i can't hg push to github 2010-11-18T15:25:30 ;) 2010-11-18T15:25:30 True that 2010-11-18T15:25:41 I was paying for them already though 2010-11-18T15:26:07 TortoiseHG is really nice-- though I do recommend using the commandline for a while just so you know it 2010-11-18T15:26:10 Zannick: You can svn commit though 2010-11-18T15:26:28 another strike against it ;) 2010-11-18T15:26:30 that's a poor alternative 2010-11-18T15:26:50 I'm using the eclipse plugin 2010-11-18T15:27:29 ah never tried that 2010-11-18T15:27:52 Indeed, I prefer github just cause its got more of a following I suppose 2010-11-18T15:27:55 and its hardly expensive 2010-11-18T15:28:19 but anyways.. 2010-11-18T15:28:32 So on the tcp server, what are the two numbers on the planet? I feel like I'm missing something? 2010-11-18T15:29:13 id+growth? 2010-11-18T15:29:33 Oh, ID, duh. :) 2010-11-18T15:29:48 yeah... that thing that is really nice to know when debugging 2010-11-18T15:30:20 Indeed 2010-11-18T15:31:43 *** Accoun has joined #aichallenge 2010-11-18T15:37:47 *** Ice_Harley has quit IRC (Quit: Page closed) 2010-11-18T15:38:06 right now I am wishing I had kept more consistant code versions in merc 2010-11-18T15:42:40 *** Rubicon-|-Cross has joined #aichallenge 2010-11-18T15:54:20 *** Mistmanov has quit IRC (Quit: Page closed) 2010-11-18T15:59:22 grrrrrrrrr I hate timeouts 2010-11-18T15:59:26 *** narnach has joined #aichallenge 2010-11-18T16:02:27 *** davidd has joined #aichallenge 2010-11-18T16:04:52 *** narnach has quit IRC (Ping timeout: 252 seconds) 2010-11-18T16:09:45 *** narnach has joined #aichallenge 2010-11-18T16:10:15 *** smellyhippy has quit IRC (Ping timeout: 265 seconds) 2010-11-18T16:13:54 *** smellyhippy has joined #aichallenge 2010-11-18T16:13:54 *** smellyhippy has joined #aichallenge 2010-11-18T16:17:59 Top 10 players: bocsimacko(3733), bix0r4ever(3582), protocolocon(3492), medrimonia(3462), luksian(3445), jimrogerz(3438), george(3437), davidjliu(3436), shangas(3418), asavis(3389) 2010-11-18T16:19:09 *** delt0r___ has joined #aichallenge 2010-11-18T16:20:30 *** delt0r has quit IRC (Ping timeout: 245 seconds) 2010-11-18T16:27:02 *** Rubicon-|-Cross has quit IRC (Ping timeout: 255 seconds) 2010-11-18T16:36:24 *** tim__ has joined #aichallenge 2010-11-18T16:37:04 *** McLeopold has quit IRC (Ping timeout: 240 seconds) 2010-11-18T16:40:19 *** p4p4 has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2010-11-18T16:41:49 *** d0ugal has left #aichallenge ("Leaving...") 2010-11-18T16:50:30 *** Naktibalda has joined #aichallenge 2010-11-18T16:51:04 *** asbadrobot has left #aichallenge 2010-11-18T16:54:43 *** sigh has joined #aichallenge 2010-11-18T17:00:36 *** Appleman1234 has joined #aichallenge 2010-11-18T17:04:22 *** McLeopold has joined #aichallenge 2010-11-18T17:04:59 i have a (probably very basic) question regarding planetwars-python-kit and python: notmyplanets = list(self.universe.not_my_planets()) TypeError: 'Planets' object is not callable --> Whats the problem here? 2010-11-18T17:07:18 so my feature to show the latest game was actually showing the first game 2010-11-18T17:08:50 take off the () 2010-11-18T17:08:53 tim__: guess: it's not a function, lose the () after it? 2010-11-18T17:09:20 wait... 2010-11-18T17:09:32 No it is a function that should return a list. 2010-11-18T17:09:47 I don't know python or the python starter package 2010-11-18T17:09:50 tim__: pastebin your code 2010-11-18T17:09:55 * lavalamp_ goes back to work 2010-11-18T17:11:12 http://pastebin.com/Zw7JAkL4 2010-11-18T17:11:29 does @property mean its actually NOT a function!? 2010-11-18T17:11:32 medrimonia: was the timeout due to your code, or a bad connection to the server? 2010-11-18T17:11:48 its a decorator 2010-11-18T17:12:35 hmm, the starter package changed alot since I downloaded it 2010-11-18T17:12:47 *** Rubicon-|-Cross has joined #aichallenge 2010-11-18T17:13:16 *** Palmik has joined #aichallenge 2010-11-18T17:13:39 which package are you using? 2010-11-18T17:15:11 @property does mean it's not a function anymore 2010-11-18T17:15:12 McLeopold: An error occured while trying to show the previous error. 2010-11-18T17:15:17 this one: https://github.com/ulope/planetwars-python-kit . I lost the brackets and dont have the error any more. Now I tried to let the Bot play but I always get a timeout :( 2010-11-18T17:15:47 looks like you'll need some logging 2010-11-18T17:16:38 this is my code (dont be too hard on my, I started 20min ago ;D): http://pastebin.com/DuHxsUSf 2010-11-18T17:16:53 is there an easy way to automatically do logging? 2010-11-18T17:17:24 yeah, look at the logging bot in the package 2010-11-18T17:17:59 Top 10 players: bocsimacko(3737), bix0r4ever(3574), protocolocon(3494), luksian(3460), jimrogerz(3455), davidjliu(3438), george(3437), shangas(3421), Raschi(3395), Xrillo(3385) 2010-11-18T17:19:04 logging will be great, and you should do it, but also try this: run your bot, and paste contents of the map into it, and type in go... you should see the error your bot makes, I think. 2010-11-18T17:24:57 found some problems. thanks for the logging tip :) 2010-11-18T17:27:24 does 'continue' inside a for loop, make it go back to the start of the loop(and incriment the counter) 2010-11-18T17:27:47 *** iamluck|2 is now known as iamluck 2010-11-18T17:28:07 in Python? yes 2010-11-18T17:28:14 usually 2010-11-18T17:29:08 *** Appleman1234 has quit IRC (Ping timeout: 255 seconds) 2010-11-18T17:30:07 it does in every language I know 2010-11-18T17:31:04 *** Appleman1234 has joined #aichallenge 2010-11-18T17:38:44 *** choas has quit IRC (Quit: leaving) 2010-11-18T18:00:59 *** davidjliu has joined #aichallenge 2010-11-18T18:06:35 *** mceier has quit IRC (Quit: leaving) 2010-11-18T18:08:37 *** Bobng has quit IRC (Ping timeout: 260 seconds) 2010-11-18T18:10:09 does any know if you will lose the game if you only own a zero growth planet? 2010-11-18T18:13:12 *** JamesMG has quit IRC (Quit: Leaving.) 2010-11-18T18:14:44 only ships are counted 2010-11-18T18:14:56 planets growth doesn't enter the picture 2010-11-18T18:15:56 janzert: around? the servers are seemingly coming to a halt 2010-11-18T18:18:00 Top 10 players: bocsimacko(3744), bix0r4ever(3593), protocolocon(3493), davidjliu(3454), cobracom(3448), george(3444), jimrogerz(3434), shangas(3428), luksian(3419), Raschi(3407) 2010-11-18T18:18:55 most likely it's just mostly playing games with bots near the top of the rankings 2010-11-18T18:19:08 it tends to slow way down during those periods 2010-11-18T18:19:55 although the average for the last *is* pretty low 2010-11-18T18:20:02 *** tim__ has quit IRC (Read error: Connection reset by peer) 2010-11-18T18:20:16 last hour 2010-11-18T18:24:19 *** Naktibalda has quit IRC (Ping timeout: 272 seconds) 2010-11-18T18:27:06 *** tapwater_ has joined #aichallenge 2010-11-18T18:27:37 *** tapwater has quit IRC (Read error: Operation timed out) 2010-11-18T18:27:46 *** tapwater has joined #aichallenge 2010-11-18T18:28:43 *** tapwater__ has joined #aichallenge 2010-11-18T18:29:58 *** tapwater___ has joined #aichallenge 2010-11-18T18:30:42 *** tapwater____ has joined #aichallenge 2010-11-18T18:31:13 the top bots use so much more cpu? 2010-11-18T18:31:23 *** tapwater_ has quit IRC (Ping timeout: 255 seconds) 2010-11-18T18:31:27 *** tapwater has quit IRC (Read error: Operation timed out) 2010-11-18T18:31:27 *** tapwater____ is now known as tapwater 2010-11-18T18:31:34 or time, more likely 2010-11-18T18:31:36 I though almost everyone was still pretty quick 2010-11-18T18:31:41 on average at least yes they tend to use more of their time 2010-11-18T18:31:56 *** narnach has quit IRC (Quit: Leaving.) 2010-11-18T18:33:02 *** tapwater__ has quit IRC (Ping timeout: 255 seconds) 2010-11-18T18:33:07 the bottom bots all mostly just crash immediately on the first turn, so it's hard to keep games that quick on the upper end ;) 2010-11-18T18:34:28 I'm not really sure how much of the effect is just due to that, I haven't really looked into it only noticed that it does happen 2010-11-18T18:34:47 *** tapwater___ has quit IRC (Ping timeout: 264 seconds) 2010-11-18T18:34:54 oh i see 2010-11-18T18:35:07 the 1-turn games still count as games in game_stats 2010-11-18T18:35:10 are there known quantities (example bots) on the leaderboard? 2010-11-18T18:35:39 so less of those makes the gpm appear to go down 2010-11-18T18:35:41 *** Palmik has quit IRC (Remote host closed the connection) 2010-11-18T18:35:49 not purposefully that I'm aware of 2010-11-18T18:36:00 no, but i would think that's the effect 2010-11-18T18:36:14 I found one 2010-11-18T18:36:19 sorry, that was in response to mega1 2010-11-18T18:36:24 a1k0n submitted on sep 11 2010-11-18T18:36:38 Zannick: to what you said, yes 2010-11-18T18:36:43 he was on the top for some days 2010-11-18T18:36:52 janzert: ah 2010-11-18T18:37:23 now that's worth 2563 elo and rank 611 2010-11-18T18:38:30 mega1: wow 2010-11-18T18:38:52 i'm pretty sure he has some better ones he hasn't submitted yet 2010-11-18T18:39:23 I'm trying to assess how many contestants take it a bit seriously 2010-11-18T18:39:41 Zannick: I think he has not the time 2010-11-18T18:40:01 between sept 11 and midoctober i thought he was working on it 2010-11-18T18:40:09 there's also kcm1700, who was #1 for a long while 2010-11-18T18:40:20 where is he now? 2010-11-18T18:40:22 i worked through oct 30 and my bot is floundering around 600-700s 2010-11-18T18:40:27 now he's 141 2010-11-18T18:40:47 *** tapwater_ has joined #aichallenge 2010-11-18T18:40:48 he was leading later than a1k0n if I recall correctly 2010-11-18T18:40:52 uploaded september 18 2010-11-18T18:42:43 *** tapwater has quit IRC (Ping timeout: 252 seconds) 2010-11-18T18:42:43 *** tapwater_ is now known as tapwater 2010-11-18T18:43:01 a1k0n was only there at the very start 2010-11-18T18:43:16 then it was McLeopold (who's bot is still the same) 2010-11-18T18:43:29 *official bot 2010-11-18T18:44:18 then people started knocking down McLeopold until dmj came to the top... kcm got #1 in that period 2010-11-18T18:46:02 *** virdo has joined #aichallenge 2010-11-18T18:47:46 yay, I killed the conversation 2010-11-18T18:47:50 @predict 2010-11-18T18:47:50 I predict that yellowseed will win! 2010-11-18T18:48:46 McLeopold is in pos 174 with 2965 elo 2010-11-18T18:49:02 I really need to submit soon 2010-11-18T18:49:15 that will be interesting to watch :) 2010-11-18T18:49:24 iouri.66 looks pretty good on benzedrine 2010-11-18T18:49:28 I will be dissapointing it think. 2010-11-18T18:49:38 McLeopold: because you won't beat meag1? :P 2010-11-18T18:49:47 I'm losing to a lot of people 2010-11-18T18:49:52 ah :( 2010-11-18T18:50:24 *** davidjliu has quit IRC (Ping timeout: 265 seconds) 2010-11-18T18:50:36 I've got more ideas, just not time. 2010-11-18T18:51:12 sigh: I'll still beat you, though :) 2010-11-18T18:51:27 McLeopold: that's not very difficult :P 2010-11-18T18:52:14 but it is flattering that you claim that as an achievement :) 2010-11-18T18:53:36 McLeopold: you would have a good chance at top 10, I think? 2010-11-18T18:53:50 Yeah, I'm hoping for top 10 2010-11-18T19:01:58 *** lavalamp_ has quit IRC (Quit: Page closed) 2010-11-18T19:02:15 *** Mjothvitnir has joined #aichallenge 2010-11-18T19:11:04 woooooooooow 2010-11-18T19:11:11 did dhartmei return from the dead? 2010-11-18T19:11:14 ^_^ 2010-11-18T19:18:00 Top 10 players: bocsimacko(3748), bix0r4ever(3597), protocolocon(3493), davidjliu(3476), george(3445), luksian(3436), shangas(3426), jimrogerz(3422), cobracom(3407), smiley1983(3392) 2010-11-18T19:19:37 E323 are you not running your latest on dhartmei's server? 2010-11-18T19:22:12 *** davidjliu has joined #aichallenge 2010-11-18T19:22:14 *** Error323 has quit IRC (Read error: Connection reset by peer) 2010-11-18T19:23:44 *** Error323_ has joined #aichallenge 2010-11-18T19:28:18 *** Naktibalda has joined #aichallenge 2010-11-18T19:30:10 http://bitbucket.org/McLeopold/planetwars-mcmaps 2010-11-18T19:34:56 not bad 2010-11-18T19:34:57 http://72.44.46.68/getplayer?player=malazan0.23 2010-11-18T19:35:02 holding at 340 elo 2010-11-18T19:35:02 now 2010-11-18T19:40:52 *** tapwater_ has joined #aichallenge 2010-11-18T19:41:49 *** tapwater__ has joined #aichallenge 2010-11-18T19:42:43 *** tapwater___ has joined #aichallenge 2010-11-18T19:42:56 *** tapwater has quit IRC (Ping timeout: 255 seconds) 2010-11-18T19:42:56 *** tapwater___ is now known as tapwater 2010-11-18T19:43:52 *** tapwater_ has quit IRC (Read error: Operation timed out) 2010-11-18T19:45:55 *** tapwater__ has quit IRC (Ping timeout: 245 seconds) 2010-11-18T19:49:14 *** Appleman1234 has quit IRC (Ping timeout: 255 seconds) 2010-11-18T19:51:06 *** Appleman1234 has joined #aichallenge 2010-11-18T20:00:21 *** phreeza has quit IRC (Quit: kthxbai) 2010-11-18T20:05:24 *** cfaftw has joined #aichallenge 2010-11-18T20:11:04 *** tapwater_ has joined #aichallenge 2010-11-18T20:12:29 *** tapwater has quit IRC (Ping timeout: 240 seconds) 2010-11-18T20:12:30 *** tapwater_ is now known as tapwater 2010-11-18T20:17:07 *** cfaftw has quit IRC (Quit: Leaving) 2010-11-18T20:17:35 *** Appleman1234 has quit IRC (Ping timeout: 245 seconds) 2010-11-18T20:18:01 Top 10 players: bocsimacko(3749), bix0r4ever(3562), davidjliu(3488), protocolocon(3486), wagstaff(3485), george(3445), _Astek_(3426), shangas(3426), luksian(3418), cobracom(3409) 2010-11-18T20:19:32 *** tapwater_ has joined #aichallenge 2010-11-18T20:23:22 *** tapwater has quit IRC (Ping timeout: 260 seconds) 2010-11-18T20:23:22 *** tapwater_ is now known as tapwater 2010-11-18T20:34:00 tcp2.zeroviz.us is up 2010-11-18T20:34:10 taking stupid map suggestions 2010-11-18T20:35:45 *** McLeopold has quit IRC (Quit: Leaving) 2010-11-18T20:47:30 *** superflit has quit IRC (Quit: superflit) 2010-11-18T20:47:38 *** superflit has joined #aichallenge 2010-11-18T20:52:11 *** superflit has quit IRC (Ping timeout: 264 seconds) 2010-11-18T20:55:19 *** Naktibalda has quit IRC (Remote host closed the connection) 2010-11-18T20:55:53 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-11-18T21:01:41 *** bhasker has quit IRC (Ping timeout: 255 seconds) 2010-11-18T21:18:02 Top 10 players: bocsimacko(3749), wagstaff(3683), bix0r4ever(3592), davidjliu(3476), protocolocon(3473), george(3449), luismi(3443), shangas(3430), Silver_S(3410), luksian(3390) 2010-11-18T21:25:30 *** rifinio has joined #aichallenge 2010-11-18T21:26:07 *** superflit has joined #aichallenge 2010-11-18T21:26:29 *** rifinio has quit IRC (Client Quit) 2010-11-18T21:36:24 @tcp 2010-11-18T21:36:25 medrimonia: tcp = (#1) http://www.benzedrine.cx/planetwars/, or (#2) http://72.44.46.68/, or (#3) 98.247.248.39:995 (zeroviz.us), or (#4) http://tcp.zeroviz.us/ 2010-11-18T21:38:38 *** tapwater_ has joined #aichallenge 2010-11-18T21:39:23 *** tapwater__ has joined #aichallenge 2010-11-18T21:40:23 *** tapwater___ has joined #aichallenge 2010-11-18T21:41:24 *** tapwater____ has joined #aichallenge 2010-11-18T21:41:28 *** tapwater has quit IRC (Ping timeout: 252 seconds) 2010-11-18T21:41:29 *** tapwater____ is now known as tapwater 2010-11-18T21:42:27 *** tapwater____ has joined #aichallenge 2010-11-18T21:42:40 *** tapwater_ has quit IRC (Ping timeout: 240 seconds) 2010-11-18T21:44:23 *** tapwater__ has quit IRC (Ping timeout: 264 seconds) 2010-11-18T21:44:40 *** tapwater___ has quit IRC (Ping timeout: 240 seconds) 2010-11-18T21:45:33 *** tapwater has quit IRC (Ping timeout: 240 seconds) 2010-11-18T21:45:33 *** tapwater____ is now known as tapwater 2010-11-18T21:49:35 *** tapwater_ has joined #aichallenge 2010-11-18T21:50:25 *** tapwater__ has joined #aichallenge 2010-11-18T21:51:33 *** tapwater___ has joined #aichallenge 2010-11-18T21:52:50 *** tapwater has quit IRC (Ping timeout: 255 seconds) 2010-11-18T21:52:50 *** tapwater___ is now known as tapwater 2010-11-18T21:53:03 *** tapwater__ has quit IRC (Read error: Operation timed out) 2010-11-18T21:54:11 *** tapwater_ has quit IRC (Ping timeout: 255 seconds) 2010-11-18T21:55:07 *** medrimonia1 has joined #aichallenge 2010-11-18T21:56:17 *** medrimonia has quit IRC (Ping timeout: 240 seconds) 2010-11-18T22:06:31 *** Rubicon-|-Cross has joined #aichallenge 2010-11-18T22:09:12 *** rebelxt_ has joined #aichallenge 2010-11-18T22:16:21 is there a way to tell why your bot loses on first turn? 2010-11-18T22:18:01 Top 10 players: bocsimacko(3746), wagstaff(3726), bix0r4ever(3553), davidjliu(3474), protocolocon(3466), george(3444), shangas(3428), Silver_S(3407), jimrogerz(3384), hutabhu(3374) 2010-11-18T22:19:37 *** tapwater_ has joined #aichallenge 2010-11-18T22:20:39 *** davidjliu has quit IRC (Ping timeout: 265 seconds) 2010-11-18T22:21:39 *** tapwater has quit IRC (Ping timeout: 265 seconds) 2010-11-18T22:21:39 *** tapwater_ is now known as tapwater 2010-11-18T22:25:51 wah, 1. lisp, 2. haskell 2010-11-18T22:26:59 the top100 is pretty random in language 2010-11-18T22:27:30 McLeopold: i like the idea of no neutrals for crazy maps 2010-11-18T22:27:39 of course you're not here though 2010-11-18T22:27:40 3 of the top 15, only appear once in teh top100 though 2010-11-18T22:27:55 crazy maps? 2010-11-18T22:28:04 http://tcp2.zeroviz.us/ 2010-11-18T22:28:12 place to play with more general maps if you want 2010-11-18T22:28:17 but they wont be used in the actual contest 2010-11-18T22:28:53 ahh 2010-11-18T22:34:50 *** rebelxt_ has quit IRC (Quit: Page closed) 2010-11-18T22:37:15 *** bhasker has joined #aichallenge 2010-11-18T22:49:32 *** tapwater_ has joined #aichallenge 2010-11-18T22:50:26 *** tapwater__ has joined #aichallenge 2010-11-18T22:51:26 *** tapwater___ has joined #aichallenge 2010-11-18T22:51:47 *** tapwater has quit IRC (Ping timeout: 255 seconds) 2010-11-18T22:51:47 *** tapwater___ is now known as tapwater 2010-11-18T22:52:27 *** tapwater___ has joined #aichallenge 2010-11-18T22:53:39 *** tapwater____ has joined #aichallenge 2010-11-18T22:53:52 *** tapwater_ has quit IRC (Ping timeout: 240 seconds) 2010-11-18T22:54:28 *** tapwater_ has joined #aichallenge 2010-11-18T22:54:45 *** tapwater has quit IRC (Read error: Operation timed out) 2010-11-18T22:54:46 *** tapwater_ is now known as tapwater 2010-11-18T22:55:23 *** tapwater__ has quit IRC (Ping timeout: 272 seconds) 2010-11-18T22:55:27 *** tapwater_ has joined #aichallenge 2010-11-18T22:56:45 *** tapwater___ has quit IRC (Ping timeout: 245 seconds) 2010-11-18T22:58:11 *** tapwater____ has quit IRC (Ping timeout: 264 seconds) 2010-11-18T22:58:51 *** tapwater has quit IRC (Ping timeout: 240 seconds) 2010-11-18T22:58:52 *** tapwater_ is now known as tapwater 2010-11-18T23:02:58 *** Cyndre has joined #aichallenge 2010-11-18T23:05:55 *** Cyndre_ has quit IRC (Ping timeout: 245 seconds) 2010-11-18T23:07:00 *** Cyndre has quit IRC (Read error: Connection reset by peer) 2010-11-18T23:07:29 *** Cyndre has joined #aichallenge 2010-11-18T23:09:20 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-11-18T23:18:02 Top 10 players: bocsimacko(3727), wagstaff(3575), bix0r4ever(3537), davidjliu(3459), protocolocon(3452), george(3437), shangas(3414), Silver_S(3405), hutabhu(3372), DarkEmi(3370) 2010-11-18T23:19:40 *** tapwater_ has joined #aichallenge 2010-11-18T23:20:23 *** tapwater__ has joined #aichallenge 2010-11-18T23:21:17 *** tapwater has quit IRC (Ping timeout: 245 seconds) 2010-11-18T23:22:51 *** tapwater_ has quit IRC (Read error: Operation timed out) 2010-11-18T23:42:41 *** yasith has joined #aichallenge 2010-11-18T23:49:54 *** Appleman1234 has joined #aichallenge