2011-04-06T00:06:57 *** antimatroid has joined #aichallenge 2011-04-06T00:13:42 *** iFire has joined #aichallenge 2011-04-06T00:55:33 *** delt0r___ has joined #aichallenge 2011-04-06T00:57:00 *** delt0r_ has quit IRC (Ping timeout: 246 seconds) 2011-04-06T01:05:21 *** Mathnerd314 has quit IRC (Read error: Connection reset by peer) 2011-04-06T01:20:15 *** davidjliu has quit IRC (Quit: Page closed) 2011-04-06T02:08:46 *** antimatroid has quit IRC (Ping timeout: 246 seconds) 2011-04-06T02:20:58 *** Cerealklr has quit IRC (Ping timeout: 248 seconds) 2011-04-06T02:22:51 *** Cerealklr has joined #aichallenge 2011-04-06T02:23:28 *** antimatroid has joined #aichallenge 2011-04-06T02:26:50 *** janzert has quit IRC (Ping timeout: 248 seconds) 2011-04-06T02:36:33 *** janzert has joined #aichallenge 2011-04-06T02:39:55 *** stocha has joined #aichallenge 2011-04-06T02:40:58 *** stocha has quit IRC (Client Quit) 2011-04-06T03:02:12 rank 2011-04-06T03:02:43 rank? 2011-04-06T03:03:23 *** kwketh has quit IRC () 2011-04-06T03:03:26 :P 2011-04-06T03:03:45 *** Keth has joined #aichallenge 2011-04-06T03:04:32 Keth: strate aka sets of orbits 2011-04-06T03:04:35 strata* 2011-04-06T03:05:18 :3 2011-04-06T03:06:44 if you have a group G and a set X with x in X, the orbit of x is orb_G(x ) = {g.x: g in G} 2011-04-06T03:07:03 then you have a stabiliser set, so a strata is a set of orbits such that the stabiliser groups are the same 2011-04-06T03:07:48 antimatroid: I see. 2011-04-06T03:07:51 at least i think, like i said rank :P trying to work out wtf i'm doing 2011-04-06T03:08:36 antimatroid: What you're doing is being slightly bored 2011-04-06T03:08:44 i have to find the 3 different strata for G = O(1,1) and X = R^2 2011-04-06T03:08:51 it's an assignment question :P 2011-04-06T03:09:01 Is that for your AI? 2011-04-06T03:09:02 or school? 2011-04-06T03:09:06 for school 2011-04-06T03:09:48 I never studied astronomy 2011-04-06T03:10:06 this is group theory (maths) 2011-04-06T03:10:18 it's technically a physics unit, but i haven't done any physics before, it's really all maths 2011-04-06T03:11:00 I can do physics and maths pretty well, except I'm not familiar with the English terms such as stabiliser groups' 2011-04-06T03:11:20 I haven't studied maths in English 2011-04-06T03:11:30 where are you from? 2011-04-06T03:12:17 I speak Polish, then I spent some time in Sweden and currently living in England 2011-04-06T03:12:19 i'm disappointed by the lack of (not really sure what else to say) asians around for these 2011-04-06T03:12:29 i want to get the japanese more interested at least 2011-04-06T03:12:35 Are you Asian? 2011-04-06T03:12:43 no, i'm from Tasmania in Australia 2011-04-06T03:12:53 Ah 2011-04-06T03:13:04 What do you do for Google AI? 2011-04-06T03:13:15 google is very much just a sponsor name 2011-04-06T03:13:28 there are a bunch of us helping set up that are just past competitors 2011-04-06T03:13:35 amstan and j3camero are from waterloo 2011-04-06T03:13:46 j3camero started it i'm pretty sure 2011-04-06T03:14:24 maps mostly 2011-04-06T03:14:31 *** Keth_ has joined #aichallenge 2011-04-06T03:14:32 ergh, i was scrolled up :P 2011-04-06T03:14:39 meh 2011-04-06T03:14:57 My internet is messed up. So what do you do for AI-Contest? 2011-04-06T03:15:19 maps mostly, although i've had input on the game mechanics for ants and wrote the c++ starter bot 2011-04-06T03:15:30 aha, i see 2011-04-06T03:15:31 and i did a lot of the wiki stuff 2011-04-06T03:15:41 were you the one I spoke about the C++ starter bot and how it's lacking the structure? 2011-04-06T03:15:42 i'm just a past competitor helping out 2011-04-06T03:15:48 that was i :P 2011-04-06T03:15:51 I remember you 2011-04-06T03:15:53 =D 2011-04-06T03:16:17 You've seen stocha bot? 2011-04-06T03:16:27 i haven't looked at it, but i know it's up 2011-04-06T03:16:38 i haven't got my own bot, i started working on my own "starter bot" yesterday 2011-04-06T03:16:41 47 wins for me, 46 for him on random maps. He's really good. 2011-04-06T03:16:48 Want to see replays? 2011-04-06T03:16:51 yes please 2011-04-06T03:17:17 *** Keth has quit IRC (Ping timeout: 246 seconds) 2011-04-06T03:23:13 *** aerique has joined #aichallenge 2011-04-06T03:27:14 *** boegel has joined #aichallenge 2011-04-06T03:41:13 *** sigh_ has quit IRC (Read error: Operation timed out) 2011-04-06T03:49:51 *** chris__0076 has joined #aichallenge 2011-04-06T03:51:40 *** Chris_0076 has quit IRC (Ping timeout: 246 seconds) 2011-04-06T04:01:19 *** antimatroid has quit IRC (Read error: Operation timed out) 2011-04-06T04:16:05 *** antimatroid has joined #aichallenge 2011-04-06T04:22:35 *** antimatroid has quit IRC (Ping timeout: 276 seconds) 2011-04-06T04:32:17 *** AlliedEnvy has quit IRC (Ping timeout: 246 seconds) 2011-04-06T04:33:17 *** sigh has joined #aichallenge 2011-04-06T04:34:17 *** AlliedEnvy has joined #aichallenge 2011-04-06T04:39:42 *** AlliedEnvy has quit IRC (Ping timeout: 264 seconds) 2011-04-06T04:40:49 *** antimatroid has joined #aichallenge 2011-04-06T04:43:30 *** AlliedEnvy has joined #aichallenge 2011-04-06T04:47:44 aichallenge: sigh epsilon * r5121d42 / (ants/ants.py worker/engine.py): Implemented new (json) replay format in engine - http://bit.ly/hVTc86 2011-04-06T04:48:07 Frontier: new replay format has been implemented in the engine 2011-04-06T04:48:16 although it is not pretty printed or anything 2011-04-06T04:55:29 Checking it out, had lots of issues with json being outputted correctly 2011-04-06T05:08:29 *** amstan has quit IRC (Remote host closed the connection) 2011-04-06T05:12:30 Keth_: what? it's not working? 2011-04-06T05:13:08 can you pastebin me the error you are getting or output? :) 2011-04-06T05:23:25 Well, I have no clue how python works and 2011-04-06T05:23:35 If you do play_one_game.cmd > log.txt 2011-04-06T05:23:46 You get python stack errors and callback stuff in log 2011-04-06T05:24:00 and replay doesn't get created 2011-04-06T05:24:10 I'm not sure why would batch script impact on python code 2011-04-06T05:24:17 I guess something with std output 2011-04-06T05:25:06 can you paste me the error that you are getting? 2011-04-06T05:25:10 sure 2011-04-06T05:28:15 http://pastebin.com/VYEcyQeJ 2011-04-06T05:28:20 without > output.txt it works 2011-04-06T05:29:05 *** Stocha has joined #aichallenge 2011-04-06T05:29:15 hmm... that error message is odd 2011-04-06T05:29:23 Yeah I know 2011-04-06T05:29:27 I tried python to output the console into a file, same error 2011-04-06T05:30:03 either way, if you want the replay, it is is viewer/0.replay (if you are using play_one_game.cmd)... if you want it on stdout then use the "--stdout replay" option 2011-04-06T05:30:36 --stdout replay - same error 2011-04-06T05:30:44 Oh waiat 2011-04-06T05:30:57 Yeah, same error 2011-04-06T05:31:02 File "../worker\engine.py", line 235, in write 2011-04-06T05:31:02 file.write(data) 2011-04-06T05:31:02 IOError: [Errno 9] Bad file descriptor 2011-04-06T05:31:24 I tried playgame.py [...] > output.txt and same thing 2011-04-06T05:31:39 so workaround is playgame.py [...] and then copy the replay 2011-04-06T05:32:11 Japaneese have a lot on their mind lately 2011-04-06T05:32:18 that error doesn't look like it's from the latest version 2011-04-06T05:32:30 are you sure that you are running the latest checkout 2011-04-06T05:32:38 yeah, it's old svn but I tried on newest 5 minutes ago as well 2011-04-06T05:32:44 Stocha: is that to me? 2011-04-06T05:32:49 i agree to that 2011-04-06T05:32:53 I will try again 2011-04-06T05:32:55 koreans would be good too 2011-04-06T05:33:22 yeah, same error 2011-04-06T05:33:28 Keth_: in particular ensure that worker/engine.py is up to date 2011-04-06T05:33:33 fp.write(chunk) 2011-04-06T05:33:33 File "../worker\engine.py", line 230, in write 2011-04-06T05:33:33 file.write(data) 2011-04-06T05:33:33 IOError: [Errno 9] Bad file descriptor 2011-04-06T05:33:33 close failed in file object destructor: 2011-04-06T05:33:33 Error in sys.excepthook: 2011-04-06T05:33:40 up to date. 2011-04-06T05:33:52 ok, that's a different error 2011-04-06T05:34:01 now the code matches up with the line number :) 2011-04-06T05:34:23 keth: you've sent the replay to antimatroid in private ? :p 2011-04-06T05:34:23 just different line :) 2011-04-06T05:34:32 *** Gorgoroth has joined #aichallenge 2011-04-06T05:34:36 Stocha: Your bot wins 71 times 2011-04-06T05:34:40 Stocha: my wins 72 2011-04-06T05:34:42 Keth: i guess you'll be posting stuff on the forum soon ? 2011-04-06T05:34:59 Anyway, your bot really made happy :) 2011-04-06T05:35:09 It's a lot of fresh air. I love it :p 2011-04-06T05:35:41 I can't wait to see the other 200 top bots :) 2011-04-06T05:36:03 Stocha: check pm 2011-04-06T05:36:26 *** Eruonen has joined #aichallenge 2011-04-06T05:36:28 Keth_: can you paste me the exact command you are running? 2011-04-06T05:36:29 Stocha: I can't wait to see the contest up and running either 2011-04-06T05:36:56 sigh: play_one_game.cmd unchanged, straight from svn (just as every file in whole directory) 2011-04-06T05:37:01 the only thing I'm running is 2011-04-06T05:37:08 Oh wait 2011-04-06T05:37:19 I only added --stdout replay to play_one_game.cmd 2011-04-06T05:37:20 you mean git, right? 2011-04-06T05:37:35 and then: 2011-04-06T05:37:39 play_one_game.cmd > output.txt 2011-04-06T05:37:58 what happens if you run without the output redirection? 2011-04-06T05:38:27 same error 2011-04-06T05:38:39 just tested 2011-04-06T05:39:21 without --stdout and output directory, no error but no output either 2011-04-06T05:39:33 output.txt 0 bytes 2011-04-06T05:40:05 yeah, without --stdout there is no output on stdout, that's supposed to happen 2011-04-06T05:40:47 hmm... this is on windows, yeah? 2011-04-06T05:40:48 with --stdout error if I try with > 2011-04-06T05:41:00 yeah 2011-04-06T05:42:37 I'll try to get it set up in my windows vm :S 2011-04-06T05:42:50 well, it 2011-04-06T05:43:33 well, it's not that important but it would be really helpful to do some loop outputting into different files 2011-04-06T05:44:55 Keth_: can you paste exactly what's in your play_one_game.cmd file, exactly what you are running, and the full output? 2011-04-06T05:45:05 so I can try to replicate it best I can 2011-04-06T05:45:10 sure 2011-04-06T05:46:24 http://pastebin.com/7mFKmE4N 2011-04-06T05:47:47 replay file ends up 0 bytes (with output directory) 2011-04-06T05:48:21 *** Gorgoroth has left #aichallenge 2011-04-06T05:51:57 Keth_: can you change "--stdout replay" to "--stdout stream" and tell me if you get the same error? 2011-04-06T05:52:54 I then end up with the same error instantly, let me paste the logs 2011-04-06T05:53:13 http://pastebin.com/Xa0i8H73 2011-04-06T05:54:53 thanks, well at least that narrows down where the problem could be :S 2011-04-06T05:55:23 I'm guessing something with the two pipeline at once or something 2011-04-06T05:55:40 It doens't happen in linux? 2011-04-06T05:55:56 or whatever system you are on 2011-04-06T05:56:09 I'm using OSX here, and it doesn't happen 2011-04-06T05:56:16 Hey, you know 2011-04-06T05:56:21 I forgot to tel 2011-04-06T05:56:25 It was working before few SVNs 2011-04-06T05:56:38 I had my profiler working and then it stopped and now I get those errors 2011-04-06T05:56:58 so there must've been some change. I am 100% it was working 3 days ago 2011-04-06T05:57:19 100% sure* 2011-04-06T05:57:21 yeah, it's isolated to when I'm trying to write to two places at once 2011-04-06T05:58:47 So you know what was changed, right? 2011-04-06T06:01:16 Keth_: nevermind, I think I found the problem 2011-04-06T06:02:02 =D 2011-04-06T06:02:35 actually, no :( 2011-04-06T06:02:56 D= 2011-04-06T06:03:22 I'm testing the old revisions 2011-04-06T06:03:28 I have one on my hard drive that was working 2011-04-06T06:03:41 *** Stocha has quit IRC (Ping timeout: 252 seconds) 2011-04-06T06:03:54 Keth_: are you running with the --output viewer option, or not? 2011-04-06T06:04:09 I tried both 2011-04-06T06:04:11 they both give error 2011-04-06T06:04:21 (same error to clarify) 2011-04-06T06:04:26 :) 2011-04-06T06:04:46 well, they should give errors in different places :P 2011-04-06T06:04:54 nope, same place. 2011-04-06T06:05:08 really? o_O 2011-04-06T06:05:16 (double checking) 2011-04-06T06:06:23 replay, stream or json? 2011-04-06T06:07:37 *** Naktibalda has joined #aichallenge 2011-04-06T06:07:37 replay gives error and replay is 0 bytes and output.txt is 0 bytes 2011-04-06T06:07:41 json gives no error and replay is OK but the output.txt is 0 bytes 2011-04-06T06:07:50 stream gives instant error 2011-04-06T06:08:11 Keth_: I'm interested in the difference between having the --output_dir viewer option and not having it 2011-04-06T06:08:25 say with --replay stream (just to make it fail quicker) 2011-04-06T06:08:29 okay 2011-04-06T06:08:54 both exactly same errors, instantly 2011-04-06T06:09:02 with and without --output_dir viewer 2011-04-06T06:09:07 wait 2011-04-06T06:09:22 Yes 2011-04-06T06:10:35 it gives the same line number and everything? 2011-04-06T06:11:45 you're right, one line differs, it's missing. let me paste 2011-04-06T06:12:07 http://pastebin.com/kX2QNqaC 2011-04-06T06:12:25 thought so :) 2011-04-06T06:12:25 was hard to spot it 2011-04-06T06:12:46 Keth_: yeah, that's why I emphasised exact :) 2011-04-06T06:12:50 I knew you're looking at the code so I guessed you know better 2011-04-06T06:12:53 yeah 2011-04-06T06:13:41 I'm not AI to spot a difference the first attempt =D 2011-04-06T06:13:50 :) 2011-04-06T06:14:08 So do you know what's wrong? 2011-04-06T06:14:22 seems like your system doesn't like when I do a .write() call to sys.stdout... which is really wierd 2011-04-06T06:14:34 Windows has no such thing as sys.stdout though 2011-04-06T06:14:36 you should consider that 2011-04-06T06:14:51 It might use some wrapper/emulates that under windows 2011-04-06T06:14:55 yeah, but python works around that 2011-04-06T06:14:59 exactly 2011-04-06T06:15:07 I've got it working on my windows vm 2011-04-06T06:15:12 Huh? 2011-04-06T06:15:16 You got it working? 2011-04-06T06:15:22 yeah 2011-04-06T06:15:27 Windows XP? 2011-04-06T06:15:43 yes 2011-04-06T06:15:48 interesting 2011-04-06T06:15:51 hmm... python 2.6.6 though 2011-04-06T06:15:57 Oh yeah 2011-04-06T06:15:57 you were running 2.7, yeah? 2011-04-06T06:15:59 Well 2011-04-06T06:16:01 I'll try 2.7 2011-04-06T06:16:04 I had 3.2, it didnd't work 2011-04-06T06:16:06 at all 2011-04-06T06:16:12 so i downgraded to 2.7 I think and it worked 2011-04-06T06:16:18 let me see 2011-04-06T06:16:18 yeah, 3 and 2 are completely incomptible 2011-04-06T06:16:38 C:\Program Files (x86)\ICW>python 2011-04-06T06:16:38 Python 2.7 (r27:82525, Jul 4 2010, 09:01:59) [MSC v.1500 32 bit (Intel)] on win32 2011-04-06T06:16:51 I could try 2.6.6 if you want me to 2011-04-06T06:17:00 and I have XP machine next to me 2011-04-06T06:17:04 you choose. 2011-04-06T06:17:57 or you could upgrade to 2.7 and see 2011-04-06T06:18:12 I will run 2.7 to try to replicate the problem 2011-04-06T06:18:30 but 2.6.6 is what is going to be on the contest servers from my understanding anyway 2011-04-06T06:19:24 let me know if how 2.7 works out 2011-04-06T06:20:05 yup, I'll check it later, gtg for now 2011-04-06T06:20:10 sure 2011-04-06T06:20:15 Thanks for help 2011-04-06T06:28:58 fyi, downgrading to 2.6.6 and double checking the version doesn't resolve it 2011-04-06T06:32:18 Keth_: you are on xp? 2011-04-06T06:32:38 i was until a few weeks ago, i finally switched to ubuntu 2011-04-06T06:32:52 seems we do a lot of things the same :P abuse operator overloading, c++, windows.. 2011-04-06T06:36:07 *** Keth_ has quit IRC (Ping timeout: 264 seconds) 2011-04-06T06:37:51 *** ltriant has quit IRC (Ping timeout: 246 seconds) 2011-04-06T06:38:43 *** ltriant has joined #aichallenge 2011-04-06T06:42:18 *** Keth has joined #aichallenge 2011-04-06T06:55:28 I'm on win7 2011-04-06T06:55:33 Why? 2011-04-06T07:07:37 Keth: 2.7 didn't help :( 2011-04-06T07:07:43 can you try running this: http://pastebin.com/6cdBPuBG 2011-04-06T07:08:01 it just tests various ways of outputting to stdout 2011-04-06T07:16:07 sigh: Thanks for the json output format. 2011-04-06T07:16:12 Btw have you seen the loading time improvements I achieved with large replays, antimatroid? 2011-04-06T07:20:03 *** Eruonen has quit IRC (Ping timeout: 276 seconds) 2011-04-06T07:21:32 Frontier: no? 2011-04-06T07:22:21 Compare those two links in Opera or even Firefox to see the difference: http://marco-leise.homedns.org/antgui/stable/visualizer.php?game_id=3 // http://marco-leise.homedns.org/antgui/work/visualizer.php?game_id=3 2011-04-06T07:22:50 i don't have either of those browsers installed :P 2011-04-06T07:22:51 *** kaemo has joined #aichallenge 2011-04-06T07:22:56 chrome <3 2011-04-06T07:23:41 it was faster anyway i think :) 2011-04-06T07:29:28 yeah, 2.6.6 didn't help either 2011-04-06T07:29:36 I will run it 2011-04-06T07:30:16 C:\[work]\Google AI\Ants>dsa.py 2011-04-06T07:30:16 hello 1 2011-04-06T07:30:17 hello 2 2011-04-06T07:30:17 hello 3 2011-04-06T07:30:17 hello 4 2011-04-06T07:30:17 hello 5 2011-04-06T07:30:18 *** FireFly has joined #aichallenge 2011-04-06T07:30:36 >dsa.py > 123 2011-04-06T07:30:36 close failed in file object destructor: 2011-04-06T07:30:36 Error in sys.excepthook: 2011-04-06T07:30:36 Original exception was: 2011-04-06T07:31:05 C:\[work]\Google AI\Ants>dsa.py > 123 2011-04-06T07:31:05 close failed in file object destructor: 2011-04-06T07:31:05 Error in sys.excepthook: 2011-04-06T07:31:05 Original exception was: 2011-04-06T07:31:43 not a single hello, not in console, not in empty file 123 2011-04-06T07:32:23 #!/usr/bin/env python 2011-04-06T07:32:23 import sys 2011-04-06T07:32:23 print "hello 1" 2011-04-06T07:32:26 the above 3 lines even make error 2011-04-06T07:33:32 You would be surprised, i tested them all invidually, they all make same error. 2011-04-06T07:34:05 #!/usr/bin/env python 2011-04-06T07:34:05 import sys 2011-04-06T07:34:05 ^ 2 lines above make no error 2011-04-06T07:42:17 *** Keth has quit IRC (Read error: Connection reset by peer) 2011-04-06T07:57:27 *** Keth has joined #aichallenge 2011-04-06T07:59:34 *** needsch has joined #aichallenge 2011-04-06T08:03:17 sigh: How do I log stderr of a bot to console or to a file? 2011-04-06T08:04:52 Frontier: Explain a little further, you mean outputting playgame.py result into a file? 2011-04-06T08:05:44 I mean, I have a bot that outputs debug information. Where can I find that after a game? 2011-04-06T08:06:11 That's up to your bot, depends on your bot code. You should know where it saves if it's your code. 2011-04-06T08:06:20 Which starter package are you using? 2011-04-06T08:06:21 Frontier: can't yet... will have to be implemented 2011-04-06T08:06:31 2011-04-06T08:07:03 But it can be implemented inside the bot on your own, right? 2011-04-06T08:07:17 sure, just write to a log file 2011-04-06T08:07:35 yeah 2011-04-06T08:08:07 sigh: I ran your thing up there, let me know if you need any more testing 2011-04-06T08:10:17 I don't really know what else to do :( that produced another error altogether. If you have time, write up an issue about it, maybe someone else will be able to figure it out 2011-04-06T08:11:53 okay, I will figure it out then and let you know 2011-04-06T08:14:47 sigh: just figured it out 2011-04-06T08:15:01 play_one_game.cmd needs a fix 2011-04-06T08:16:22 should be "python playgame.py" not "playgame.py" 2011-04-06T08:17:01 so the bat always worked but not for output with > 2011-04-06T08:17:03 works perfect now :) 2011-04-06T08:17:08 o_O 2011-04-06T08:17:42 I don't know anything about bat files 2011-04-06T08:17:46 haha 2011-04-06T08:18:01 well, the bat WORKED 2011-04-06T08:18:21 I just didn't realize it could be bat fault when you see the output in the console 2011-04-06T08:18:30 ahhhh 2011-04-06T08:18:44 so with the bat, sys.stdout is not set? 2011-04-06T08:18:49 or something like that 2011-04-06T08:18:56 basically what it tries to do is 2011-04-06T08:19:01 same thing as test1.txt > test2.txt 2011-04-06T08:19:08 instead of 2011-04-06T08:19:15 python x > output.txt 2011-04-06T08:19:56 what does "test1.txt > test2.txt" do? 2011-04-06T08:20:15 copies 2011-04-06T08:20:21 and runs 2011-04-06T08:20:22 WOW 2011-04-06T08:20:24 haha 2011-04-06T08:20:27 that explains it 2011-04-06T08:20:31 O_o 2011-04-06T08:20:41 test1.txt > ewq 2011-04-06T08:20:45 it runs test1.txt in notepad 2011-04-06T08:20:48 and ewq is created empty 2011-04-06T08:21:47 well, it makes sense. 2011-04-06T08:22:01 somebody just needs to fix play_one_game.cmd in the gik 2011-04-06T08:24:29 @later tell McLeopold Keth says that "playgame.py" should be replaced by "python playgame.py" in play_one_game.cmd to avoid errors when redirecting output from it 2011-04-06T08:24:29 sigh: Yes master! 2011-04-06T08:25:37 I'm just wondering how the hell did it work for you on your vmware machine 2011-04-06T08:25:40 =D 2011-04-06T08:26:00 dunno, maybe it's an XP vs win7 issue 2011-04-06T08:26:20 hm, could be 2011-04-06T08:28:43 It is strange, but I cant seem to run Java bots in the engine 2011-04-06T08:29:32 The bots use System.out.println(); System.out.flush(); but in 0.botN.output there is nothing but the '# turn x' lines 2011-04-06T08:30:11 You need to create new file using in your bot code 2011-04-06T08:30:21 open 'debug.log' and then put all your logs into that file 2011-04-06T08:30:27 don't do print 2011-04-06T08:30:43 Keth: I'm talking about the communication with the engine. 2011-04-06T08:30:49 he's printing bot orders 2011-04-06T08:31:05 hm okay. But why would you want to contact the engine? 2011-04-06T08:31:18 to tell it the move to make :P 2011-04-06T08:31:19 The bot just seems to stay at one spot, but I doesn't time out 2011-04-06T08:31:42 ah I understand now 2011-04-06T08:32:41 Frontier: are you on windows? 2011-04-06T08:32:42 hi, when will the new contest start? 2011-04-06T08:32:50 Frontier: that's odd... if it is not timing out it must be outputting "go" 2011-04-06T08:33:10 or is the ants specification already fixed? 2011-04-06T08:34:22 Frontier: is your bot outputting valid moves? 2011-04-06T08:34:38 0.botN.output only shows the valid moves that a bot has made 2011-04-06T08:36:49 sigh: ok... I am using the starter bot for an experiment 2011-04-06T08:37:03 in the next few days I'll make the bot output show the complete stdout and stderr of the bot 2011-04-06T08:38:07 sigh: The bot outputs col/row is that correct? 2011-04-06T08:38:31 remember col/row != x,y - took me ages to get used to 2011-04-06T08:38:35 =D 2011-04-06T08:38:54 Frontier: "o row col direction" eg "0 1 2 n" 2011-04-06T08:39:03 sorry: "o 1 2 n" 2011-04-06T08:39:29 and it is row/col not col/row 2011-04-06T08:40:33 sigh: What could cause a bot to win on the same map and then lose with the same seed in both games? It shouldn't happen right? 2011-04-06T08:41:18 sigh: who is responsible for the java starter? 2011-04-06T08:41:19 Keth: are you running the python starter bots? 2011-04-06T08:41:29 I'm running my own bots 2011-04-06T08:41:36 are you saying my own bots might not be deteministic? 2011-04-06T08:41:42 oh that's true 2011-04-06T08:41:46 the enemy bot has crazy code 2011-04-06T08:42:09 thanks, that explains it 2011-04-06T08:42:10 Keth: a seed option is passed in to the bot if the engine is using a seed 2011-04-06T08:42:25 Yeah, but the bot I'm fighting with has it's own seed every game 2011-04-06T08:42:27 I forgot that 2011-04-06T08:42:31 :) 2011-04-06T08:42:50 http://gyazo.com/eb21947b379dc710664f221cffff5c2f.png 2011-04-06T08:42:58 since i got surprised when one of the games went different 2011-04-06T08:43:50 yeah, it's the bot's responsibility to seed their random number generator if a seed is passed in 2011-04-06T08:44:13 would that be legal if the bot makes up it's own seed? 2011-04-06T08:44:19 in a real tournament? 2011-04-06T08:44:22 it's only for testing in any case... in the contest the bots wont be passed seeds :) 2011-04-06T08:44:39 seed to help make it easier to test only 2011-04-06T08:44:46 yeah I see 2011-04-06T08:44:52 I thought seed might be used as a proof of bot winning few times 2011-04-06T08:45:03 I could say "results are fake, proove it again" 2011-04-06T08:45:03 :) 2011-04-06T08:45:08 :) 2011-04-06T08:45:35 but i doubt that will be the issue 2011-04-06T08:45:53 well, no one has brought that up so far... I only implemented seeds so I could do deterministic regression tests 2011-04-06T08:46:02 yeah, seed is really helpful 2011-04-06T08:46:15 helped me to recompile my code, add changes and go through the same movement again 2011-04-06T08:46:29 :) 2011-04-06T08:46:35 *** kaemo has quit IRC (Ping timeout: 260 seconds) 2011-04-06T08:46:44 @later tell antimatroid: The Java starter and sample bots use col/row in some places and are unusable atm. 2011-04-06T08:46:44 Frontier: I come to serve. 2011-04-06T08:47:03 Frontier: antimatroid hates java :P 2011-04-06T08:47:14 can i haz huge map? like 500x500? 2011-04-06T08:47:14 git blame says that McLeopold wrote it 2011-04-06T08:47:38 but antimatroid is head of starter bot department 2011-04-06T08:47:56 he is? ok :P) 2011-04-06T08:48:43 sigh: last question, in the tournament, what happens if bot0 has 500 ants and bot1 has 10 and they never kill each other 2011-04-06T08:48:55 does the bot with 500 ants win? 2011-04-06T08:49:06 after like 1000 turns? 2011-04-06T08:49:58 seems like it would be a draw 2011-04-06T08:50:31 just because the enemy ants are running, creating a fake draft? 2011-04-06T08:50:34 lol 2011-04-06T08:51:25 the main aim seems to be to kill other ants, in particular so that you can't get ahead in multiplayer by hiding in a corner 2011-04-06T08:51:45 but I haven't been in on the scoring discussions much 2011-04-06T08:51:49 i see 2011-04-06T08:52:23 scoring is still not settled, in any case 2011-04-06T08:52:55 will remember 2011-04-06T08:53:43 if you want to discuss scoring (or anything related to game rules), harass antimatroid when you see him :) 2011-04-06T08:55:06 I might, he's been nice to me 2011-04-06T08:55:39 *** delt0r_ has joined #aichallenge 2011-04-06T08:56:59 *** delt0r___ has quit IRC (Ping timeout: 240 seconds) 2011-04-06T09:03:20 Frontier: I like your Replay.js class 2011-04-06T09:03:22 if (lit.moar()) { 2011-04-06T09:03:22 tl = lit.gimme(); 2011-04-06T09:04:00 I like that except when this happens: Object # has no method 'gimme'at Replay.txtToJson (http://kwketh.com/ants/visualizer/js/Replay.js:442:13) 2011-04-06T09:04:05 Fuck 2011-04-06T09:04:16 Sorry 2011-04-06T09:04:34 aichallenge: sigh epsilon * r82de839 / worker/engine.py : Add logging on invalid moves to bot output - http://bit.ly/hiTsxn 2011-04-06T09:05:19 Frontier: I added logging of invalid moves 2011-04-06T09:05:32 supposed to be a private link 2011-04-06T09:10:29 Keth: Well.. it would have errored out in the next line anyway 2011-04-06T09:11:03 *** Eruonen has joined #aichallenge 2011-04-06T09:11:21 yea I know but still 2011-04-06T09:11:22 replay doesn't load 2011-04-06T09:11:27 well, a magic replay 2011-04-06T09:11:41 http://pastebin.com/xXBrzjc9 2011-04-06T09:13:15 all the rest work fine 2011-04-06T09:13:35 *** needsch1 has joined #aichallenge 2011-04-06T09:14:10 I don't know what to tell you other than what the error message says: There is extra data after the last line (#556) 2011-04-06T09:14:27 *** boegel has quit IRC (Ping timeout: 252 seconds) 2011-04-06T09:15:11 *** needsch has quit IRC (Ping timeout: 258 seconds) 2011-04-06T09:15:18 sigh: I got it! I got the engine stalled like I promised it would happen. 2011-04-06T09:16:27 it's true, extra data from the engine 2011-04-06T09:16:50 I guess engine's fault 2011-04-06T09:17:15 I really have no clue, you should know frontier and decide if it's important or not 2011-04-06T09:17:21 i'm just letting you know it won't load 2011-04-06T09:17:51 aichallenge: Marco Leise epsilon * r438a5fb / (ants/visualizer/js/Replay.js ants/util/EngineKill0r.java): added engine killing bot test case - http://bit.ly/hNnwVW 2011-04-06T09:18:51 Keth: sigh is working on the engine. I can't do anything about that. 2011-04-06T09:19:10 oh wait 2011-04-06T09:19:11 you just fixed that 2011-04-06T09:19:18 mleise did actually 2011-04-06T09:19:21 - tl = lit.gimme(); 2011-04-06T09:19:21 442 2011-04-06T09:19:22 + tl = lit.gimmeNext(); 2011-04-06T09:19:33 Actually I'm always happy when someone finds bugs in the engine through error messages in the visualizer :) 2011-04-06T09:19:47 Keth: I fixed the error message, yes. 2011-04-06T09:20:57 Oh yeah, thanks 2011-04-06T09:21:18 Frontier: how does your EngineKiller work? 2011-04-06T09:21:39 hehe, that you want to know. but the dark side is hideous 2011-04-06T09:21:55 write a whole bunch of stuff? 2011-04-06T09:21:59 Well, compile it and run it on a really large map with 1000 turns 2011-04-06T09:22:09 500x500 should be fine as a map size 2011-04-06T09:22:37 It writes a debug.txt file to keep you udated on its attack progress 2011-04-06T09:22:47 You'll see what it does then. 2011-04-06T09:24:17 It *may* run into a false positive and just be killed by your engine. 2011-04-06T09:24:57 ah ok, you stall on input 2011-04-06T09:25:06 On my Linux setup it runs up to 64kb btw. So you may need a little patience 2011-04-06T09:25:09 I'll run it tomorrow and see what it actuall does 2011-04-06T09:26:07 kk, have fun. McLeopold once said there is nothing he can do about it form within Python. 2011-04-06T09:27:14 oh well, I don't see it as a big issue... in the last contest the rules were pretty clear that you got disqualified if you tried to fuck with the contest 2011-04-06T09:27:27 I don't see how someone could do this by accident :P 2011-04-06T09:28:10 *** boegel has joined #aichallenge 2011-04-06T09:28:15 a bot could simply hang 2011-04-06T09:28:38 *** boegel has quit IRC (Client Quit) 2011-04-06T09:28:48 it is just a matter of not reading enough bytes from stdin after the last turn 2011-04-06T09:28:57 *** boegel has joined #aichallenge 2011-04-06T09:29:33 I have to admit, that I was convinced that the buffer size was 4kb, but maybe it has been increased on all systems now. 2011-04-06T09:30:32 if I get bored I'll see if there is a way around that 2011-04-06T09:45:18 *** smellyhippy has quit IRC (Read error: Connection reset by peer) 2011-04-06T09:46:38 *** smellyhippy has joined #aichallenge 2011-04-06T09:46:38 *** smellyhippy has joined #aichallenge 2011-04-06T09:51:38 *** sigh has quit IRC (Remote host closed the connection) 2011-04-06T10:22:25 *** stocha has joined #aichallenge 2011-04-06T10:22:27 *** p4p4p5 has joined #aichallenge 2011-04-06T10:31:02 *** stocha has quit IRC (Ping timeout: 252 seconds) 2011-04-06T10:35:54 *** needsch1 has quit IRC (Quit: Leaving.) 2011-04-06T10:37:30 *** kaemo has joined #aichallenge 2011-04-06T10:38:21 *** needsch has joined #aichallenge 2011-04-06T10:39:23 *** Stocha has joined #aichallenge 2011-04-06T10:46:22 *** Stocha has quit IRC (Quit: Page closed) 2011-04-06T10:56:44 *** namina has joined #aichallenge 2011-04-06T11:00:54 *** kaemo has quit IRC (Ping timeout: 276 seconds) 2011-04-06T11:28:33 *** aerique has quit IRC (Quit: ...) 2011-04-06T11:32:28 *** boegel has quit IRC (Quit: *poof!*) 2011-04-06T11:36:46 *** kaemo has joined #aichallenge 2011-04-06T12:09:35 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-06T12:10:44 *** p4p4p5 has joined #aichallenge 2011-04-06T12:10:45 *** p4p4p5 is now known as p4p4p4 2011-04-06T12:39:46 *** BtbN_ is now known as BtbN 2011-04-06T12:41:36 aichallenge: Marco Leise epsilon * r46fd366 / (3 files): visualizer: fixed regression that food pop up instantly / food is now a circle - http://bit.ly/go7iyX 2011-04-06T12:41:37 *** p4p4p4 has quit IRC (Read error: Connection timed out) 2011-04-06T12:46:24 *** p4p4p4 has joined #aichallenge 2011-04-06T12:47:16 *** McLeopold1 has joined #aichallenge 2011-04-06T12:47:40 http://ants.zeroviz.us/ixbPX/ 2011-04-06T12:51:18 *** mceier has joined #aichallenge 2011-04-06T13:15:46 *** Eruonen has quit IRC () 2011-04-06T13:15:55 is the ants specification fixed already? is it documented anywhere already? 2011-04-06T13:16:28 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 3.6.16/20110319135224]) 2011-04-06T13:24:53 *** contestbot_ has joined #aichallenge 2011-04-06T13:24:53 *** contestbot has quit IRC (Disconnected by services) 2011-04-06T13:26:49 *** contestbot_ is now known as contestbot 2011-04-06T13:26:49 *** ChanServ sets mode: +o contestbot 2011-04-06T13:36:46 McLeopold: very nice :D 2011-04-06T13:40:55 McLeopold1: I wonder if I should adopt your color scheme. Bright player colors work better on a dark ground. Then again the bright colors make a more friendly impression. 2011-04-06T13:49:39 McLeopold1: :) 2011-04-06T13:50:06 McLeopold1: are you happy for me to try and get jeff to decide on battle and scoring relatively soon? 2011-04-06T13:50:25 i feel like we need to get some decisions made 2011-04-06T13:50:31 *** amstan has joined #aichallenge 2011-04-06T13:50:31 *** ChanServ sets mode: +o amstan 2011-04-06T13:52:52 *** needsch has quit IRC (Quit: Leaving.) 2011-04-06T13:56:11 antimatroid: I'm ambivalent 2011-04-06T13:57:03 *** rebelxt has joined #aichallenge 2011-04-06T13:57:07 *** rebelxt has left #aichallenge 2011-04-06T13:57:10 *** needsch has joined #aichallenge 2011-04-06T13:57:22 antimatroid: jeff want's the game so simple to understand there is not much room for tactics, if any 2011-04-06T13:57:38 is the ants specification fixed already? is it documented anywhere already? 2011-04-06T13:57:46 needsch: not fixed yet 2011-04-06T13:57:54 thank you 2011-04-06T13:57:57 github.com/aichallenge/aichallenge/wiki 2011-04-06T13:58:12 McLeopold1: I am the same, but eventually the decisions need to be made 2011-04-06T13:58:21 i will make a similar scoring page to the battle resolution 1 2011-04-06T13:58:25 and make him read through both 2011-04-06T13:58:32 scoring? 2011-04-06T13:58:45 people are still complaining about scoring 2011-04-06T13:58:56 the scoring is perfect 2011-04-06T13:59:11 i tend to agree, that can stay as is 2011-04-06T13:59:14 I must have missed the complaints? 2011-04-06T13:59:21 yes :P 2011-04-06T13:59:46 is it "why can I win if I hide in the corner, that's unfair" type of complaints? 2011-04-06T13:59:59 *can't* 2011-04-06T14:00:26 no, i'm having trouble remembering them though 2011-04-06T14:00:46 one argument is to enforce the last participating player automatically wins 2011-04-06T14:03:30 *** andy__ has joined #aichallenge 2011-04-06T14:03:35 we do that with the food bonus 2011-04-06T14:03:44 I just read stocha's post 2011-04-06T14:05:06 antimatroid: how about ants don't die, just race for food? 2011-04-06T14:05:24 and collisions? 2011-04-06T14:05:41 who knows 2011-04-06T14:06:00 there is no good attack that jeff will approve 2011-04-06T14:06:42 Looks like you resigned on that front ^^ 2011-04-06T14:07:11 I think the last version was pretty easy to understand. 2011-04-06T14:07:41 The one where you accumulate damage first and then kill all ants that have received >= 1 dmg point in that turn 2011-04-06T14:08:44 So, here is what jeff has told me 2011-04-06T14:08:49 If I were to make up an ai challenge I would probably have multiple leagues with different rule sets. 2011-04-06T14:10:53 Don't require players to have to zoom in to determine what happened, Don't make any hidden state determine an ant's death... 2011-04-06T14:11:20 Don't make a rule system that can't be discovered by watching a game 2011-04-06T14:11:52 I agree on point 1 :). But point 3 is really bad... 2011-04-06T14:12:03 That leaves us with "die if ants are close", and that's it. 2011-04-06T14:12:33 As long as we don't find a way to visualize other rules efficiently 2011-04-06T14:13:17 He really wants the 1:1 ratio kills because he thinks it's the only way people will understand the game 2011-04-06T14:13:22 but what about his proposed kill iterative closest 2011-04-06T14:13:42 that was my proposal, not his 2011-04-06T14:13:46 everyone i have talked to since the introduction of damage has listed that as their preferred option 2011-04-06T14:14:03 That involves "hidden state" 2011-04-06T14:14:08 how? 2011-04-06T14:14:19 I'm just the messenger, I don't agree. 2011-04-06T14:14:29 i don't see how that involved hidden information 2011-04-06T14:14:34 I know 2011-04-06T14:14:52 but, it does allow blocking 2011-04-06T14:15:11 only in small situations that can be prevented from happening 2011-04-06T14:15:20 the only way to prevent blocking is limit an ants attack to 1 square 2011-04-06T14:15:37 or not allow passages of width 1 or 2 2011-04-06T14:16:12 I thought it was still possible? 2011-04-06T14:16:18 don't think so 2011-04-06T14:16:26 what attack radius? 2011-04-06T14:16:29 damage with attack radius 1 is better than all ants in range die 2011-04-06T14:16:34 sqrt 6 2011-04-06T14:18:31 antimatroid: Okay, give me a minute and I show blocking with sqrt 6 2011-04-06T14:19:20 McLeopold1: is the same blocking possible with sqrt4? 2011-04-06T14:22:15 antimatroid: I think a solid line of ants of any length will block with sqrt6 2011-04-06T14:23:23 antimatroid: a diagonal line blocks sqrt4 2011-04-06T14:24:00 basically, if an ant's attack shape has a line on it, you can block any width 2011-04-06T14:26:18 McLeopold1: why can't an enemy line also break it? 2011-04-06T14:26:28 oh corridor 2011-04-06T14:26:29 hmmm 2011-04-06T14:26:54 I suppose you can take some out, but at a less than 1:1 ratio 2011-04-06T14:27:05 yes, less than 1:1 is fine 2011-04-06T14:27:17 if people want to try and set up defensive structures with those qualities, that's fine 2011-04-06T14:27:23 we just don't want full blocking 2011-04-06T14:27:26 and i don't think that's possible 2011-04-06T14:27:44 if we have open enough maps, people should be able to go "around" these things 2011-04-06T14:28:26 http://www.youtube.com/watch?v=80k7Jl1o8Lc 2011-04-06T14:28:33 i want one of those, but they aren't techniclick D: 2011-04-06T14:30:32 *** boegel has joined #aichallenge 2011-04-06T14:31:20 antimatroid: I doesn't work if you write cursive :) 2011-04-06T14:31:58 *** andy__ has quit IRC (Ping timeout: 252 seconds) 2011-04-06T14:32:31 you still have to lift off every word 2011-04-06T14:32:49 and i'm a maths student, a lot of the time i'm not writing sentences of text 2011-04-06T14:38:17 antimatroid: so, take a crack at convincing jeff of the damage rule 2011-04-06T14:40:47 i think people will avoid fighting when possible without something like damage 2011-04-06T14:40:55 i guess that ignores scoring 2011-04-06T14:41:05 but i don't think the scoring makes sense without something like damage either 2011-04-06T14:44:19 McLeopold: what is preferred if not damage? 2011-04-06T14:44:32 colliding ants only or all ants within range 2011-04-06T14:45:51 colliding? 2011-04-06T14:46:19 if your damage contributed to an ants death, you get a percent of that point 2011-04-06T14:47:42 that works for any attack method 2011-04-06T14:50:39 yeah, but in all other methods you don't want to have more than one ant in the fight 2011-04-06T14:50:53 otherwise you're just losing more ants and giving away more points 2011-04-06T14:51:10 so fighting basically resolves down to don't send more than one ant 2011-04-06T14:56:42 yes, but the scoring method stays the same 2011-04-06T14:57:09 *** delt0r___ has joined #aichallenge 2011-04-06T14:57:49 *** delt0r_ has quit IRC (Read error: Operation timed out) 2011-04-06T15:00:37 McLeopold1: but what does the scoring even do then? 2011-04-06T15:00:58 like, you want to be involved in the most separate fights? 2011-04-06T15:01:03 that's such a boring rule :P 2011-04-06T15:01:12 i know you agree with me though, oh well, i shall try to get damage 2011-04-06T15:01:34 i'm tempted to do a forum poll between damage and everything in range dies 2011-04-06T15:07:26 forum polls suck 2011-04-06T15:08:44 they do, but i want to point out to jeff just how much people prefer damage 2011-04-06T15:17:46 *** Accoun has quit IRC () 2011-04-06T15:18:48 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-04-06T15:21:14 *** Stocha has joined #aichallenge 2011-04-06T15:21:57 *** javagamer has quit IRC (Quit: WeeChat 0.3.4) 2011-04-06T15:23:32 hi in there 2011-04-06T15:23:50 *** javagamer has joined #aichallenge 2011-04-06T15:25:39 McLeopold1: how does your checklist look like? 2011-04-06T15:29:05 is the ant die when close with 1:1 ratio implemented ? 2011-04-06T15:29:28 is it equivalent to die on same square (with the possibility of swaping ) ? 2011-04-06T15:29:34 *** Accoun has joined #aichallenge 2011-04-06T15:40:05 *** p4p4p4 has quit IRC (Read error: Connection timed out) 2011-04-06T15:41:54 *** p4p4p5 has joined #aichallenge 2011-04-06T15:48:01 *** kaemo has quit IRC (Ping timeout: 248 seconds) 2011-04-06T16:07:53 *** andy__ has joined #aichallenge 2011-04-06T16:11:51 aichallenge: Marco Leise epsilon * rd3f7fce / (3 files in 2 dirs): visualizer: added quick and dirty zoom for Stocha and antimatroid - http://bit.ly/foVknt 2011-04-06T16:12:01 :D 2011-04-06T16:13:22 how do i zoom? 2011-04-06T16:14:26 Frontier: ^? 2011-04-06T16:14:38 nevermind 2011-04-06T16:17:05 I'll put up a second directory with my current 'stable' release, so you don't hit the work page when I'm in the middle of refactoring :) 2011-04-06T16:18:17 aichallenge: Marco Leise epsilon * r306c31e / ants/visualizer/js/Replay.js : visualizer: fixed a visual bug when using minified js - http://bit.ly/iea5hD 2011-04-06T16:18:21 here it is: http://marco-leise.homedns.org/antgui/stable/visualizer.php?game_id=2 2011-04-06T16:20:48 aichallenge: Marco Leise epsilon * r2a16e03 / ants/visualizer/js/Application.js : visualizer: fixed clicks on the minimap while the animation is stopped - http://bit.ly/fItTbk 2011-04-06T16:21:55 Frontier: what battle resolution is used for that game? 2011-04-06T16:22:07 Don't ask me, it is ancient :p 2011-04-06T16:22:14 i'm guessing iterative closest 2011-04-06T16:22:52 you can click on the minimap btw, but i guess that is considered standard, hehe 2011-04-06T16:23:00 i need a beer, cu 2011-04-06T16:24:47 didn't realise that 2011-04-06T16:24:56 the zoom is cool thanks :) 2011-04-06T16:31:38 *** boegel has quit IRC (Quit: Leaving) 2011-04-06T16:40:09 *** delt0r___ has quit IRC (Remote host closed the connection) 2011-04-06T16:43:15 *** delt0r has joined #aichallenge 2011-04-06T16:46:32 hey, I made an online version of the map generator I uploaded yesterday. could perhaps be useful for testing different maps.. check it ou heret: http://antmapgenerator.appspot.com/ 2011-04-06T16:47:43 Looks really cool 2011-04-06T16:53:34 *** McLeopold1 is now known as McLeopold 2011-04-06T16:57:34 amstan: the list is still all there, unless you've done some work? 2011-04-06T16:57:50 McLeopold: i don't have the link 2011-04-06T16:57:56 McLeopold: could me migrate it to the issue list? 2011-04-06T16:58:38 http://ietherpad.com/sMmBH9htlr 2011-04-06T16:59:09 amstan: it would be nice if we used a real tracking system 2011-04-06T16:59:41 McLeopold: wgat's wrong with the current one? 2011-04-06T17:00:12 it's like a phpbb forum vomited in github 2011-04-06T17:00:58 Oh 2011-04-06T17:00:59 lol 2011-04-06T17:01:04 I get it now 2011-04-06T17:06:20 McLeopold: well.. any suggestions? 2011-04-06T17:06:29 McLeopold: we can try using lighthouse 2011-04-06T17:06:40 it integrates into github 2011-04-06T17:07:02 the zoom do the wrapping :p 2011-04-06T17:07:17 want can't we have wrapping without zoom now ? :p 2011-04-06T17:07:44 oh he is having a bear. 2011-04-06T17:09:03 *** Stocha has left #aichallenge 2011-04-06T17:10:10 wrapping without zoom? sure, just disable the border 2011-04-06T17:10:44 amstan: maybe I just need to learn the github way 2011-04-06T17:11:03 McLeopold: i like it because it's simple, no fuss about submitting stuff in it 2011-04-06T17:12:13 amstan: I'm watching the video right now, I might then move my list over 2011-04-06T17:12:24 the video? lol 2011-04-06T17:12:40 McLeopold: but yeah.. seriously.. if you really don't feel comfortable with it, we can change it 2011-04-06T17:12:49 you're pretty important for the development 2011-04-06T17:12:56 so you should be ok with it 2011-04-06T17:16:25 *** kaemo has joined #aichallenge 2011-04-06T17:17:00 aichallenge: Marco Leise epsilon * rbe45253 / ants/visualizer/js/Replay.js : visualizer: fiexed regression that disabled fog of war when using minified js - http://bit.ly/gHCcDs 2011-04-06T17:17:32 can i haz vip status too? 2011-04-06T17:18:48 *** bss03 has joined #aichallenge 2011-04-06T17:20:25 Sorry I just disappeared on everyone. My main system had hardware issues; I think it is fine now, but it had to be recapped and the file systems had to be repaired. 2011-04-06T17:21:33 You had luck that you haven't lost all files :) 2011-04-06T17:22:10 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-06T17:23:51 *** p4p4p5 has joined #aichallenge 2011-04-06T17:32:15 bss03: nice to see you 2011-04-06T17:32:41 Frontier: I only had 100 files in lost+found, and all of those are now back in the right place. 2011-04-06T17:32:43 amstan: I'm moving issues over to github 2011-04-06T17:33:02 amstan: we need to make sure every launch issue is actually assigned to someone 2011-04-06T17:33:17 amstan: and by "we" I mean "you" 2011-04-06T17:33:21 contestbot: topic 2011-04-06T17:33:21 amstan: Official Google AI Challenge: http://ai-contest.com/ || Channel Logs: http://contestbot.hypertriangle.com/ || Code Repo: http://github.com/aichallenge/aichallenge || Launch Checklist: https://github.com/aichallenge/aichallenge/issues/labels/launch 2011-04-06T17:33:44 McLeopold: you wanna use tags for assignments too? 2011-04-06T17:33:46 McLeopold: Did my task get completed without me, or do we still need SQL for doing matching? 2011-04-06T17:35:08 I think there are a few issues/ideas/nice-to-haves that aren't formalized. For example I have no support in the visualizer for loading maps separately from replays 2011-04-06T17:38:55 *** skynetz has joined #aichallenge 2011-04-06T17:42:05 *** sigh_ has joined #aichallenge 2011-04-06T17:42:58 i have news from j3camero 2011-04-06T17:43:30 McLeopold: once we're close to being finished, he wants to setup a full beta contest 2011-04-06T17:43:33 hidden from most ppl 2011-04-06T17:44:02 so.. aim for that 2011-04-06T17:45:05 McLeopold: did you write the Java starter bot and sample bots? 2011-04-06T17:45:10 yes 2011-04-06T17:45:36 amstan: yes, tags for assignment, as per the github video I just watched 2011-04-06T17:45:52 made one for myself 2011-04-06T17:45:56 I used them to create a bot that stalls the engine and realized that in some locations it uses col/row format. Espeicially in the construction of Aim enums and when issuing orders. 2011-04-06T17:46:35 Frontier: whoops 2011-04-06T17:46:44 that was before we settled on row/col 2011-04-06T17:47:03 bss03: I'm almost done with the sql stuff 2011-04-06T17:49:01 Kk. 2011-04-06T17:51:43 *** mceier has quit IRC (Quit: leaving) 2011-04-06T17:52:33 So it is possible for a bot to stall the engine, but the case is very unlikely (on my linux at least). The engine would have to send > 64k (size may vary among systems) of data in a single turn, that the bot doesn't read out of the buffer. 2011-04-06T17:54:56 *** javagamer is now known as guavajaime 2011-04-06T17:54:56 My test game had >6000 ants to produce that much data. But I think it is still note-worthy if we come across a game idea in the future that can easily fill up the pipe buffer. 2011-04-06T17:56:30 *** FireFly has quit IRC (Quit: swatted to death) 2011-04-06T17:56:53 what if when bot figures out he's losing and will spam the pipeline with something? that's what we should be worried about 2011-04-06T17:57:42 we don't want to draft occur when that happens 2011-04-06T17:57:46 the draft* 2011-04-06T17:57:55 or engine to freeze 2011-04-06T17:58:52 This case is easier to handle, and I'm sure the engine can kill a bot that doesn't send a go in time among all the other data it is sending. 2011-04-06T18:01:48 *** Eruonen has joined #aichallenge 2011-04-06T18:01:48 ah shit, i forgot to implement globalAlpha in the java canvas emulation 2011-04-06T18:01:59 *** p4p4p5 has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2011-04-06T18:05:55 *** guavajaime is now known as promethium 2011-04-06T18:08:06 *** kaemo has quit IRC (Ping timeout: 260 seconds) 2011-04-06T18:09:19 *** promethium is now known as javagamer 2011-04-06T18:09:24 It's not about the time limit but more about engine not to crash 2011-04-06T18:11:53 it could run out of memory if it is spammed with data :p 2011-04-06T18:12:30 hey, who assigned https://github.com/aichallenge/aichallenge/issues/labels/Frontier#issue/47 to me? That is a clear web-site task for ... mm... yeah... amstan :) 2011-04-06T18:13:14 probably McLeopold 2011-04-06T18:13:23 i guess i can look over it 2011-04-06T18:14:44 McLeopold: hey! why did you change my tag colour? i like blue! 2011-04-06T18:17:43 McLeopold: also.. can you reset your branch? those 2 commits from way back there still appear in the network: https://github.com/aichallenge/aichallenge/network 2011-04-06T18:17:51 McLeopold: s/branch/fork 2011-04-06T18:18:32 I just changed my tag color to yellow, you still look blue on my screen 2011-04-06T18:18:42 yeah, i changed myself back 2011-04-06T18:20:24 amstan: I was trying to reduce visual noise by making all user tags an orangy-red 2011-04-06T18:20:40 oh.. that's ok then 2011-04-06T18:20:41 then you can glance and see that an issue is assigned or not 2011-04-06T18:20:49 good idea 2011-04-06T18:22:47 amstan: what's the command to reset a fork? 2011-04-06T18:23:00 McLeopold: just delete your repo, refork it 2011-04-06T18:26:56 amstan: done 2011-04-06T18:32:31 *** amstan has quit IRC (Remote host closed the connection) 2011-04-06T18:36:52 aichallenge: Marco Leise epsilon * recd7102 / (3 files in 2 dirs): visualizer: fixed crashes in java version, when zooming in and when showing fog of war - http://bit.ly/hE6hZC 2011-04-06T18:38:07 McLeopold: funny, I thought the random color generator was broken and that is why everyone is red, so i changed my name to visualizer yellow. 2011-04-06T18:38:53 :) 2011-04-06T18:45:21 Frontier: you still show up as orange for me. It's not user specific is it? 2011-04-06T18:47:09 no, you changed me back to orange 2011-04-06T18:48:24 oh 2011-04-06T18:48:26 Stocha will have a hard time with your shades of red with green. He is color blind :p 2011-04-06T18:48:33 :) 2011-04-06T18:48:42 only zeta is green 2011-04-06T18:49:06 the 'reds' all seem to have a little green in them 2011-04-06T18:49:28 even I can't tell the reds apart, and that's kinda the point 2011-04-06T18:49:32 I'll have to do something about the visualizer, too 2011-04-06T18:49:52 I've got 2 links for you to checkout in regards to ant colors 2011-04-06T18:50:30 http://ethanschoonover.com/solarized 2011-04-06T18:51:37 this was my 'tool' so far: http://marco-leise.homedns.org/antgui/work/visualizer/colors.html 2011-04-06T18:51:47 http://blog.asmartbear.com/color-wheels.html? scroll down to the four primary colors wheel 2011-04-06T18:53:03 I like your color tool :D 2011-04-06T18:58:29 Solarized is actually intersting. The color wheel and pysiology of the eye is not new to me. 2011-04-06T18:59:37 *** amstan has joined #aichallenge 2011-04-06T18:59:37 *** ChanServ sets mode: +o amstan 2011-04-06T19:08:35 actually there are only 5 colors which I can use: http://en.wikipedia.org/wiki/Category:Articles_with_images_not_understandable_by_color_blind_users 2011-04-06T19:11:43 *** Naktibalda has joined #aichallenge 2011-04-06T19:16:20 *** needsch has quit IRC (Quit: Leaving.) 2011-04-06T19:34:47 *** sigh has joined #aichallenge 2011-04-06T19:54:40 *** Keth has quit IRC (Read error: Connection reset by peer) 2011-04-06T19:55:00 *** Keth has joined #aichallenge 2011-04-06T19:57:09 *** skynetz has quit IRC (Quit: .) 2011-04-06T20:05:32 aichallenge: McLeopold epsilon * rb5b6224 / sql/2_generate_matchup.sql : procedure to create matchup and matchup_player rows on demand - http://bit.ly/fPMRqg 2011-04-06T20:05:46 *** McLeopold has left #aichallenge 2011-04-06T20:11:08 *** Cyndre has joined #aichallenge 2011-04-06T20:13:45 *** andy__ has quit IRC (Ping timeout: 252 seconds) 2011-04-06T20:39:15 *** McLeopold has joined #aichallenge 2011-04-06T20:50:55 *** Eruonen has quit IRC () 2011-04-06T21:06:13 *** Kingpin13 has joined #aichallenge 2011-04-06T21:19:09 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 3.6.16/20110323143040]) 2011-04-06T21:28:15 *** Cyndre has quit IRC (Ping timeout: 276 seconds) 2011-04-06T21:37:24 *** namina has quit IRC (Remote host closed the connection) 2011-04-06T21:49:33 *** fgump has joined #aichallenge 2011-04-06T21:58:22 *** dlila has joined #aichallenge 2011-04-06T22:02:47 *** amstan has quit IRC (Ping timeout: 246 seconds) 2011-04-06T22:08:34 *** Zannick has quit IRC (*.net *.split) 2011-04-06T22:08:34 *** pgpaskar1 has quit IRC (*.net *.split) 2011-04-06T22:08:35 *** CIA-90 has quit IRC (*.net *.split) 2011-04-06T22:08:35 *** spacebat has quit IRC (*.net *.split) 2011-04-06T22:09:34 *** sigh has quit IRC (*.net *.split) 2011-04-06T22:09:34 *** janzert has quit IRC (*.net *.split) 2011-04-06T22:09:34 *** jbroman has quit IRC (*.net *.split) 2011-04-06T22:10:09 *** sigh has joined #aichallenge 2011-04-06T22:10:09 *** jbroman has joined #aichallenge 2011-04-06T22:10:13 *** znutar has quit IRC (Excess Flood) 2011-04-06T22:10:17 *** znutar has joined #aichallenge 2011-04-06T22:10:47 *** Zannick has joined #aichallenge 2011-04-06T22:10:47 *** pgpaskar1 has joined #aichallenge 2011-04-06T22:10:47 *** spacebat has joined #aichallenge 2011-04-06T22:11:07 *** janzert has joined #aichallenge 2011-04-06T22:14:13 *** CIA-33 has joined #aichallenge 2011-04-06T22:19:04 hi janzert 2011-04-06T22:22:15 *** amstan has joined #aichallenge 2011-04-06T22:22:15 *** ChanServ sets mode: +o amstan 2011-04-06T22:30:23 *** greghaynes has quit IRC (Quit: WeeChat 0.3.2) 2011-04-06T22:31:19 *** greghaynes has joined #aichallenge 2011-04-06T22:58:59 *** delt0r_ has joined #aichallenge 2011-04-06T23:00:25 *** delt0r has quit IRC (Ping timeout: 240 seconds) 2011-04-06T23:10:16 *** fgump has quit IRC (Quit: Leaving) 2011-04-06T23:48:34 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-04-06T23:50:14 * janzert waves 2011-04-06T23:52:48 * amstan waves back 2011-04-06T23:53:18 contestbot: botdoc 2011-04-06T23:53:19 amstan: botdoc = (#1) http://ubottu.com/stdin/supydocs/plugins/, or (#2) http://supybook.fealdia.org/devel/, or (#3) http://sourceforge.net/apps/mediawiki/gribble/index.php?title=Supybot_Resources 2011-04-06T23:54:07 contestbot: if 2011-04-06T23:54:08 amstan: User error, it's not my fault. 2011-04-06T23:55:42 janzert: matchup generation is taking about 1500 milliseconds for a single match 2011-04-06T23:56:10 I swear I had it faster on my other laptop, but I can't seem to mimic the indexes correctly