2011-05-21T00:00:22 oh duh i figured it out 2011-05-21T00:00:51 *** Nebri has quit IRC (Quit: Page closed) 2011-05-21T00:01:19 *** chris___0076 has quit IRC (Ping timeout: 240 seconds) 2011-05-21T00:09:11 i'm stuck on ready to be compiled? 2011-05-21T00:09:22 oshit, forgot to turn on the worker 2011-05-21T00:09:36 there 2011-05-21T00:09:50 antimatroid: done compiling 2011-05-21T00:11:50 bleh i really don't care enough to debug th cs bot 2011-05-21T00:12:13 would help if i knew for sure it needed to be debugged lol 2011-05-21T00:12:34 demerzel: did you get the local testing going? 2011-05-21T00:12:43 painfully 2011-05-21T00:12:49 it's not helping me much since i can barely get it to work 2011-05-21T00:12:51 *** chris___0076 has joined #aichallenge 2011-05-21T00:12:59 ? 2011-05-21T00:13:00 much less use it to pinpoint problems in the bots 2011-05-21T00:13:11 my bot fails when i have it collecting food on whatever the tests are 2011-05-21T00:13:30 k, fine 2011-05-21T00:13:31 where did you see that 2011-05-21T00:13:36 * amstan looks at what the test thing actually does 2011-05-21T00:15:38 so.. it runs a game against this bot: https://github.com/aichallenge/aichallenge/blob/epsilon/ants/submission_test/TestBot.py 2011-05-21T00:15:40 *** chris__0076 has quit IRC (Ping timeout: 246 seconds) 2011-05-21T00:15:55 https://github.com/aichallenge/aichallenge/blob/epsilon/worker/worker.py#L358 2011-05-21T00:16:06 as long as your bot does not "if result['status'][0] in ('crashed', 'timeout', 'invalid'):" 2011-05-21T00:16:09 you pass 2011-05-21T00:16:43 hmm 2011-05-21T00:17:00 antimatroid: i could try pause the worker, and see what problem you have 2011-05-21T00:17:03 also i pulled replay data for 2938 and it has matchupid:2950 in the json dat 2011-05-21T00:17:10 is that right 2011-05-21T00:17:23 demerzel: matchupid is not the same as gameid 2011-05-21T00:17:31 ok juts checking 2011-05-21T00:18:03 i tried playing with that testbot.py locally and it crashed 2011-05-21T00:18:18 antimatroid: who did? your bot? 2011-05-21T00:18:42 no, TestBot.py crashed 2011-05-21T00:18:47 hmm 2011-05-21T00:18:48 interesting 2011-05-21T00:19:50 haha 2011-05-21T00:19:53 vindicatino 2011-05-21T00:20:16 antimatroid: testbot works fine for me 2011-05-21T00:20:16 http://pastebin.com/MjfNULrY 2011-05-21T00:20:42 antimatroid: do you have latest version? 2011-05-21T00:20:49 not where i ran that 2011-05-21T00:20:53 :P 2011-05-21T00:21:04 it takes me some effort to copy new stuff over to where i test 2011-05-21T00:21:08 so, why do you expect it to work? 2011-05-21T00:21:17 copy? you just do git pull 2011-05-21T00:21:28 no, i have a local copy of everything i can mess around with 2011-05-21T00:21:36 without having git bitch at me for my local copy being different 2011-05-21T00:21:42 ... 2011-05-21T00:21:44 you're doing it wrong 2011-05-21T00:22:04 :p 2011-05-21T00:22:21 this is all natural only if you are a linux python person 2011-05-21T00:22:33 .cmd to run python is so against my normal usage pattern 2011-05-21T00:22:51 demerzel: that .cmd file was made by a windows user as well 2011-05-21T00:23:18 it runs now 2011-05-21T00:23:35 amstan: they only work if .py files are associated with python and the paths are set 2011-05-21T00:23:41 which in my case neither was true 2011-05-21T00:23:56 plus who uses .cmd for batch files sheesh 2011-05-21T00:24:07 linux users is the answer 2011-05-21T00:24:31 normal programs should have paths set 2011-05-21T00:24:41 you know what play_one_game.cmd did 2011-05-21T00:24:48 the fact that 99% of windows programs don't do it... that's just sad 2011-05-21T00:24:52 opened playgame.py in notepad++ lol 2011-05-21T00:25:09 when i installed python in my windows system, i had to do it 2011-05-21T00:25:19 they don't need to they rely on other variables 2011-05-21T00:25:28 %USERENV% %WINDOWS% etc 2011-05-21T00:25:33 and i had to navigate through like 10 menu systems until i got to do it 2011-05-21T00:26:49 woah! this is major: http://rss.slashdot.org/~r/Slashdot/slashdot/~3/e19caWJdUtg/Confirmed-Microsoft-Says-It-Will-Open-Source-VB-6 2011-05-21T00:26:53 ergh, no it didn't, i'm confused 2011-05-21T00:27:20 amstan: bottom line nobody runs command line anything any more in windows 2011-05-21T00:27:33 not even developers most of the time 2011-05-21T00:27:48 demerzel: yes, that's why you need an IDE most of the time in windows 2011-05-21T00:28:00 yes they are quite handy 2011-05-21T00:28:01 or else it's suicide to dev on it 2011-05-21T00:28:10 are you anti-ide lol] 2011-05-21T00:28:44 at some point you just gotta say 'well i /could/ control every aspect of my computer but that's a huge waste of time for minimal gain' 2011-05-21T00:28:49 i didn't say that, i said that windows makes it impossible to code without an ide 2011-05-21T00:28:56 nah its not impossible 2011-05-21T00:29:01 painful* 2011-05-21T00:29:05 i coded C last year in notepad++ 2011-05-21T00:29:24 demerzel: and how did you compile? 2011-05-21T00:29:27 gcc 2011-05-21T00:29:33 just a batch file 2011-05-21T00:29:38 didn't you need to set paths for that? 2011-05-21T00:29:39 doubleclicked it 2011-05-21T00:29:40 OHNO 2011-05-21T00:29:42 haha 2011-05-21T00:29:48 no just used absolute 2011-05-21T00:29:49 so you had the gcc location hardcoded? 2011-05-21T00:29:59 now.. what if you need to give your program to another person? 2011-05-21T00:30:06 do they have to install gcc in the same location as you? 2011-05-21T00:30:10 no 2011-05-21T00:30:13 what if they have a different letter for the drive? 2011-05-21T00:30:14 if they wanted to use the batch file yes 2011-05-21T00:30:23 but hopefully if they are getting my code they know how to compile 2011-05-21T00:30:43 yes, this breaks down if you have more than 3 source files 2011-05-21T00:30:52 your bash script gets more complex 2011-05-21T00:31:02 and nobody will have to go through it to replace all your hardcoded paths 2011-05-21T00:31:03 yea well its more useful in windows to do it that way 2011-05-21T00:31:06 as opposed to trying to command line it 2011-05-21T00:31:18 bash files are command line.. 2011-05-21T00:31:30 bash files are essentially the same thing as batch files 2011-05-21T00:31:37 .cmd or whatever 2011-05-21T00:31:39 batch* 2011-05-21T00:31:45 batch = windows 2011-05-21T00:31:50 oh oh 2011-05-21T00:31:51 batch files are command line.. 2011-05-21T00:31:52 i get the same error with the current copy of everything 2011-05-21T00:32:01 yea but the point is you can double click a batch file 2011-05-21T00:32:02 antimatroid: the testbot works for me just fine 2011-05-21T00:32:11 rather than typing it out 2011-05-21T00:32:14 demerzel: you can double click a .sh file, what's your point? 2011-05-21T00:32:23 i can run the games with hunter and lefty 2011-05-21T00:32:25 i dont even remember what we're talking about any more i am going t back to decoding replays 2011-05-21T00:32:30 but when i add testbot, it crashses :\ 2011-05-21T00:32:53 antimatroid: and by it you mean ... 2011-05-21T00:33:01 TestBot.py 2011-05-21T00:33:05 i'm not even running my bot in the game 2011-05-21T00:33:13 antimatroid: can you send me your command? 2011-05-21T00:33:40 http://pastebin.com/Z9pBErhE 2011-05-21T00:34:14 supposed to be submission_test/TestBot.py 2011-05-21T00:34:30 i just copied the file you linked to, hang on 2011-05-21T00:34:41 antimatroid: i linked the one from submission_test 2011-05-21T00:34:55 ./playgame.py --seed 42 --end_wait=0.25 --verbose --log_dir game_logs --turns 100 --map_file maps/symmetric_maps/symmetric_10.map "$@" "python submission_test/TestBot.py" "python dist/sample_bots/python/LeftyBot.py" "python dist/sample_bots/python/HunterBot.py" "python dist/sample_bots/python/LeftyBot.py" 2011-05-21T00:35:17 okay, that runs 2011-05-21T00:35:50 antimatroid: there is no dist/sample_bots/python/TestBot.py 2011-05-21T00:35:50 wait food is generated dynamically 2011-05-21T00:35:54 demerzel: yes 2011-05-21T00:35:58 and has a dynamic shelf-life 2011-05-21T00:35:59 there is when i make it 2011-05-21T00:36:01 man this is gonna be rough 2011-05-21T00:36:03 demerzel: part of it is made from start too 2011-05-21T00:37:48 my bot plays fine against it locally :\ 2011-05-21T00:38:44 antimatroid: ugh.. can you send me your bot? 2011-05-21T00:38:56 antimatroid: make sure it includes a makefile or something 2011-05-21T00:38:59 so i can easily compile it 2011-05-21T00:40:53 what email do you want it to? :P 2011-05-21T00:41:00 amstan@ai-contest is good 2011-05-21T00:41:02 this is my food collection algo, i'm not just uploading the code 2011-05-21T00:41:10 i'll keep it safe 2011-05-21T00:41:13 don't worry 2011-05-21T00:43:08 send it to me 2011-05-21T00:43:13 oh you already did ty 2011-05-21T00:43:20 sent 2011-05-21T00:43:24 lol 2011-05-21T00:43:25 demerzel: you wish 2011-05-21T00:43:27 like i can read haskell 2011-05-21T00:43:33 it's c++ 2011-05-21T00:43:38 oh ok yea then send it 2011-05-21T00:43:39 demerzel: fail 2011-05-21T00:44:51 bbiab, gotta finish cooking lunch 2011-05-21T00:45:19 antimatroid: is it supposed to go for food, then stop moving? 2011-05-21T00:45:21 it does that 2011-05-21T00:45:29 against testbot too 2011-05-21T00:45:33 me is confused... 2011-05-21T00:45:42 ugh you gave away the secrets 2011-05-21T00:46:33 * amstan advises for everyone to wait until https://github.com/aichallenge/aichallenge/issues/114 is fixed 2011-05-21T00:46:49 * amstan has a feeling that the test games and the actual games are mixed up 2011-05-21T00:46:51 amstan: yeah 2011-05-21T00:47:09 is dhartmei gonna host again this year 2011-05-21T00:47:09 antimatroid: oh yeah, what happens to your bot with no food 2011-05-21T00:47:16 * amstan is testing 2011-05-21T00:47:21 it should just do nothing 2011-05-21T00:47:22 *** amriedle has quit IRC (Quit: Lost terminal) 2011-05-21T00:47:51 energy conservation, brilliant 2011-05-21T00:47:58 like a coiled snake 2011-05-21T00:48:01 status survived survived survived survived 2011-05-21T00:48:03 all good 2011-05-21T00:50:37 *** sigh has joined #aichallenge 2011-05-21T00:51:04 antimatroid: oh.. got something interesting 2011-05-21T00:52:00 antimatroid: your bot timed out on this 2011-05-21T00:52:14 ./playgame.py --seed 42 --food none --end_wait=0.25 --verbose --log_dir game_logs --turns 30 --debug --map_file submission_test/test.map "$yourbot" "python submission_test/TestBot.py" 2011-05-21T00:52:20 doesn't work, it times out 2011-05-21T00:54:54 try seeing if it works with 2 players 2011-05-21T00:54:57 instead of 4 2011-05-21T01:01:14 hmmm okay, shouldn't make a difference but i'll have a play later 2011-05-21T01:01:19 eating atm then i need to go out for a little while 2011-05-21T01:01:27 *** amstan_ has joined #aichallenge 2011-05-21T01:01:27 *** ChanServ sets mode: +o amstan_ 2011-05-21T01:05:44 *** choas has quit IRC (Quit: leaving) 2011-05-21T01:32:38 *** Que__ has quit IRC (Ping timeout: 252 seconds) 2011-05-21T01:35:49 *** amstan_ has quit IRC (Ping timeout: 246 seconds) 2011-05-21T01:40:31 *** smiley1983 has joined #aichallenge 2011-05-21T02:02:05 *** sigh has quit IRC (Remote host closed the connection) 2011-05-21T02:36:19 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-05-21T02:38:28 *** Mathnerd314 has quit IRC (Ping timeout: 246 seconds) 2011-05-21T02:50:38 *** mcstar has quit IRC (*.net *.split) 2011-05-21T02:53:36 *** chris__0076 has joined #aichallenge 2011-05-21T02:56:19 *** chris___0076 has quit IRC (Ping timeout: 246 seconds) 2011-05-21T03:22:47 *** berak has joined #aichallenge 2011-05-21T03:22:58 *** sigh has joined #aichallenge 2011-05-21T03:46:06 *** demerzel has quit IRC (Ping timeout: 252 seconds) 2011-05-21T03:55:24 *** delt0r___ has joined #aichallenge 2011-05-21T03:56:37 *** delt0r_ has quit IRC (Ping timeout: 252 seconds) 2011-05-21T03:59:21 *** sigh has quit IRC (Remote host closed the connection) 2011-05-21T04:04:51 *** sigh has joined #aichallenge 2011-05-21T04:41:17 *** sigh has quit IRC (Remote host closed the connection) 2011-05-21T04:42:57 When installing Ubuntu server as a base for setting up the contest code, should I allow tasksel to install LAMP server software, or will server_setup do this? 2011-05-21T04:58:43 *** amstan has quit IRC (Remote host closed the connection) 2011-05-21T05:00:31 *** mcstar has joined #aichallenge 2011-05-21T06:04:23 *** Redgis has joined #aichallenge 2011-05-21T06:25:42 *** yasith_ has quit IRC (Read error: Operation timed out) 2011-05-21T06:25:56 *** FireFly has joined #aichallenge 2011-05-21T06:43:18 *** yasith_ has joined #aichallenge 2011-05-21T06:57:29 *** olexs has joined #aichallenge 2011-05-21T07:07:41 *** olexs has quit IRC (Quit: Leaving.) 2011-05-21T07:07:57 *** Naktibalda has joined #aichallenge 2011-05-21T07:17:38 *** Kingpin13 has quit IRC (Ping timeout: 276 seconds) 2011-05-21T07:17:46 *** berak has quit IRC (Ping timeout: 246 seconds) 2011-05-21T07:59:43 *** janzert has joined #aichallenge 2011-05-21T08:06:56 contestbot: later tell amstan no, compilation has not been sandboxed yet. 2011-05-21T08:06:56 janzert: Aye, aye, sir 2011-05-21T08:09:32 contestbot: later tell amstan also I'm working out of town until Thursday so so probably won't be coding anything until after that 2011-05-21T08:09:32 janzert: As you wish. 2011-05-21T08:10:06 *** janzert has left #aichallenge 2011-05-21T08:11:18 contestbot: self-destruct! 2011-05-21T08:11:19 mcstar: User error, it's not my fault. 2011-05-21T08:11:35 *** sigh has joined #aichallenge 2011-05-21T08:20:40 *** AlliedEnvy_ has joined #aichallenge 2011-05-21T08:22:02 *** AlliedEnvy has quit IRC (Read error: Connection reset by peer) 2011-05-21T08:24:15 I've set up a VM with Ubuntu and the contest code, completed the install instructions - I can access the website, create users, upload code and so on. Can anyone tell me how to set the contest code running so it plays matches, updates rankings and so on? 2011-05-21T08:37:52 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 4.0.1/20110504060726]) 2011-05-21T08:38:13 *** olexs has joined #aichallenge 2011-05-21T08:39:33 *** ahmedhelal has quit IRC (Read error: Connection reset by peer) 2011-05-21T09:04:01 *** mcstar has quit IRC (Ping timeout: 246 seconds) 2011-05-21T09:08:40 *** mcstar has joined #aichallenge 2011-05-21T09:17:07 *** Accoun has quit IRC (Ping timeout: 246 seconds) 2011-05-21T09:18:14 *** Nebri has joined #aichallenge 2011-05-21T09:20:09 *** Boom has joined #aichallenge 2011-05-21T09:21:06 *yawns and waves* 2011-05-21T09:21:43 *** Accoun has joined #aichallenge 2011-05-21T09:22:30 *** Frontier has quit IRC (Quit: Leaving.) 2011-05-21T09:24:36 I'm running setup_worker.py on my virtual machine, hoping it will make it run the contest code, but I'm beginning to think I've made a mistake, that this is actually for doing the entire OS install and other setup... somewhere 2011-05-21T09:26:31 *** Nebri has quit IRC (Quit: Page closed) 2011-05-21T09:26:42 *** Nebri has joined #aichallenge 2011-05-21T09:28:45 *** choas has joined #aichallenge 2011-05-21T09:29:01 *** smiley1983 has quit IRC (Ping timeout: 246 seconds) 2011-05-21T09:30:25 *** smiley1983 has joined #aichallenge 2011-05-21T10:04:02 *** sigh has quit IRC (Remote host closed the connection) 2011-05-21T10:36:01 *** berak has joined #aichallenge 2011-05-21T10:53:20 *** McLeopold has joined #aichallenge 2011-05-21T10:55:55 antimatroid: how goes it? 2011-05-21T10:57:13 McLeopold: when setting up the contest code inside a virtual machine for testing, do you know if I should run worker_setup.py? 2011-05-21T10:58:18 To clarify: when I say "for testing", I mean including use of website, trueskill and so on, not just for running local matches. 2011-05-21T10:59:35 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-05-21T10:59:55 *** amriedle has joined #aichallenge 2011-05-21T11:07:34 smiley1983: you need both server_setup and worker_setup. I would suggest without the jails. 2011-05-21T11:08:29 If I've already set it running with jails, I would probably need to redo most of it in order to run without, am I right? 2011-05-21T11:11:07 *** Cyndre has quit IRC (Quit: Leaving) 2011-05-21T11:19:50 smiley1983: no, just turn them off 2011-05-21T11:20:03 you can run everything without running any of the setup stuff 2011-05-21T11:20:20 you just need to setup the server_info.py files yourself and create directories 2011-05-21T11:20:54 dmd just failed compiling... what directories do I need to create? Is it documented somewhere? 2011-05-21T11:21:09 put this key in worker/server_info.py, "secure_jails": False 2011-05-21T11:21:21 dmd is hard, as janzert about htat 2011-05-21T11:22:07 I don't need it myself if I can get everything running without completing worker_setup 2011-05-21T11:22:39 what language are you going to write in? 2011-05-21T11:22:43 Once secure_jail is set, is there some command I should run to set the contest code running? 2011-05-21T11:22:46 OCaml 2011-05-21T11:23:04 probably, unless I decide to learn haskell in time :) 2011-05-21T11:23:07 OCaml is working good on the test server 2011-05-21T11:23:35 Yes, I was happy to see the starter pack compiling and playing games :) 2011-05-21T11:33:07 *** Eruonen has joined #aichallenge 2011-05-21T11:33:07 *** Eruonen has left #aichallenge 2011-05-21T11:49:16 *** mcstar has quit IRC (Remote host closed the connection) 2011-05-21T11:52:15 I've seen mention of a start_worker.sh script, but it doesn't seem to exist in any of my clones - is it generated during worker_setup.py? 2011-05-21T11:53:24 *** amriedle has quit IRC (Quit: Lost terminal) 2011-05-21T11:55:52 *** delt0r_ has joined #aichallenge 2011-05-21T11:57:11 *** delt0r___ has quit IRC (Ping timeout: 240 seconds) 2011-05-21T12:00:03 *** stride is now known as dominictarr 2011-05-21T12:00:10 *** dominictarr is now known as stride 2011-05-21T12:05:25 I've now persuaded worker_setup.py to get up to the point where it complains about the lack of start_worker.sh - if anyone knows what the script is meant to do, or why it isn't there (in my clone or on github) I'd like to know :) 2011-05-21T12:07:44 for context, if it's relevant: running python worker.py -t -n 0 gives endless "Get task error: [Errno http error] no host given" 2011-05-21T12:08:04 *** demerzel has joined #aichallenge 2011-05-21T12:10:42 *** bbitmaster has quit IRC () 2011-05-21T12:17:57 smiley1983: you need an http server runing for worker to pull tasks 2011-05-21T12:21:30 I have apache2 installed - is there any special setup, and is it automated somewhere? 2011-05-21T12:22:51 it is 2011-05-21T12:23:07 look at apache.template 2011-05-21T12:25:34 Hm. I just restarted, and now my virtual machine won't boot. I'm going to have to do some more setup before I can proceed. Thankyou for your help, and sorry to have drawn you away from your other tasks. 2011-05-21T12:28:08 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-05-21T12:37:07 *** olexs has quit IRC (Quit: Leaving.) 2011-05-21T12:40:45 *** amstan has joined #aichallenge 2011-05-21T12:40:45 *** ChanServ sets mode: +o amstan 2011-05-21T12:41:14 amstan: what happened to our test server? 2011-05-21T12:41:20 McLeopold: it's off, hold on 2011-05-21T12:41:48 where is the ants i need the ants 2011-05-21T12:42:02 McLeopold: have you seen my recent bug reports? 2011-05-21T12:42:05 in your pants 2011-05-21T12:42:09 ty 2011-05-21T12:42:11 lol 2011-05-21T12:42:12 no, on github? 2011-05-21T12:42:14 yes 2011-05-21T12:42:26 most of the games are with 30 turns and no food 2011-05-21T12:42:27 idk why 2011-05-21T12:42:38 https://github.com/aichallenge/aichallenge/issues?sort=created&direction=desc&_pjax=true&state=open 2011-05-21T12:42:40 top 3 2011-05-21T12:42:44 that's the functional test settings... 2011-05-21T12:42:50 yes, i know 2011-05-21T12:42:57 also, some bots fail functional testing 2011-05-21T12:42:59 I'll take a look 2011-05-21T12:44:18 *** Naktibalda has joined #aichallenge 2011-05-21T12:44:47 there, it's up 2011-05-21T12:44:51 mine does 2011-05-21T12:44:52 k 2011-05-21T12:44:55 index.php page is annoying though 2011-05-21T12:44:59 since memcached is empty 2011-05-21T12:44:59 but it might be because i rewrote it 2011-05-21T12:45:08 i still can't fidn anything wrong with it tho 2011-05-21T12:45:08 let me bring the daemons up 2011-05-21T12:45:15 is the server set up for C# 2011-05-21T12:45:23 i assume so since it says compiled but who knows 2011-05-21T12:45:32 demerzel: it is 2011-05-21T12:45:55 http://pastebin.com/N7LBTc8D 2011-05-21T12:46:09 i dont even wanna make my botto in c# i thought it could be quick to make one for starter lol 2011-05-21T12:47:09 McLeopold: http://hypertriangle.com:13080/ they're starving man! 2011-05-21T12:47:39 they have food 2011-05-21T12:47:48 I see food being created 2011-05-21T12:47:57 oh, wait then for other games 2011-05-21T12:48:31 McLeopold: well, still: https://github.com/aichallenge/aichallenge/issues/114 2011-05-21T12:48:39 it gives you an example and the debug from the worker 2011-05-21T12:49:17 *** Frontier has joined #aichallenge 2011-05-21T12:49:49 amstan: I bet it is a global variable that was set after a functional test, and didn't get set back 2011-05-21T12:50:08 we shouldn't be using global variables 2011-05-21T12:50:15 no, we shouldn't 2011-05-21T12:50:19 there 2011-05-21T12:50:23 wait.. nvm 2011-05-21T12:50:24 it has food 2011-05-21T12:50:46 also, why is it that sometimes the worker outputs all the orders, and sometimes it just outputs the turn lines? 2011-05-21T12:51:03 probably same issue 2011-05-21T12:51:18 in the same game though 2011-05-21T12:51:20 the worker code got a little complicated, it's kind of messy 2011-05-21T12:51:23 it just switches randomly 2011-05-21T13:01:25 *** Cyndre has joined #aichallenge 2011-05-21T13:07:01 *** yasith__ has joined #aichallenge 2011-05-21T13:09:43 *** yasith_ has quit IRC (Ping timeout: 244 seconds) 2011-05-21T13:10:47 the replays are odd - so basically all food/ants are 'b' but then when the game starts it just randomly picks some of them to be real ants and others to be food? 2011-05-21T13:36:09 *** olexs has joined #aichallenge 2011-05-21T13:45:50 *** Kingpin13 has joined #aichallenge 2011-05-21T13:50:42 *** yasith__ has quit IRC (Read error: Operation timed out) 2011-05-21T14:01:51 *** smiley1983 has quit IRC (Quit: leaving) 2011-05-21T14:03:39 *** yasith__ has joined #aichallenge 2011-05-21T14:13:12 *** smiley1983 has joined #aichallenge 2011-05-21T14:15:11 Can anyone suggest sane values for api_base_url and api_key in server_info.py? This is on a VM where the server will only be accessed locally (browser localhost) 2011-05-21T14:15:58 I have apache2 running, website is accessible. Running worker.py fails, and I think it's because I left those two values mentioned above blank 2011-05-21T14:18:08 smiley1983: are you using setup/setup_worker.py? 2011-05-21T14:18:54 api_base_url should be where the worker can access the apache stuff, like http://localhost/api_something_somethin.php 2011-05-21T14:19:18 and the key you get when you access api_server_setup.php 2011-05-21T14:19:21 yes, I've run setup_worker.py successfully 2011-05-21T14:19:21 from the worker 2011-05-21T14:19:37 ignore everything except the key in there, the other stuff that the php outputs is out of date 2011-05-21T14:27:33 okay... I think that fixed the problem, so now I've encountered the next problem: "file not returned by server" - but I'll have to look into that tomorrow after work, because it's time for sleep now. Thank you for your help in getting this far. 2011-05-21T14:27:48 *** smiley1983 has quit IRC (Quit: leaving) 2011-05-21T14:32:30 amstan: how many hours have you put into this so far? 2011-05-21T14:32:38 Nebri: lost count 2011-05-21T14:32:45 the logs might be an indication 2011-05-21T14:35:54 well now I know where you guys got your practice from for Mohawk. PathFinder did a great job on the mazes it did solve :) 2011-05-21T14:36:14 real shame that bug affected ya during the competition :( 2011-05-21T14:36:21 complete heartbreaker 2011-05-21T14:36:21 i could have done better if it wasn't java 2011-05-21T14:36:32 it was a pain to debug 2011-05-21T14:36:40 i think i'll use jython next 2011-05-21T14:37:05 haven't played around with python yet 2011-05-21T14:37:21 procedural or oop for python? 2011-05-21T14:37:38 opp probably 2011-05-21T14:37:53 and i'll implement all thoes objects for compass and turn properly 2011-05-21T14:37:58 so you can add them, and compare them 2011-05-21T14:38:07 since java doesn't have operator overloading 2011-05-21T14:38:21 yea that is definitely a short coming of java :( 2011-05-21T14:38:30 if it was up to me, that stuff should have been numbers, not objects 2011-05-21T14:38:42 0 for north, 1 for east up to 3 2011-05-21T14:38:50 -1 for left, +1 for right 2011-05-21T14:38:53 you could have done that if you wanted to in your own class :) 2011-05-21T14:39:14 yeah, but i already had some code dependent on it 2011-05-21T14:39:21 ah I see 2011-05-21T14:39:33 i spent like 10 hours in total on it, i should have done more 2011-05-21T14:39:42 so how did you get involved with the google AI challegne that waterloo is running? 2011-05-21T14:39:54 well, i used to be a student at waterloo 2011-05-21T14:40:08 comp science or other program? 2011-05-21T14:40:11 cs yeah 2011-05-21T14:40:15 i won the RPS competition 2011-05-21T14:40:22 RPS? 2011-05-21T14:40:23 and i became an organizer for tron 2011-05-21T14:40:26 rock paper scissors 2011-05-21T14:40:36 ah, didn't hear about that one. :) 2011-05-21T14:40:47 i had like 100 ppl, mostly waterloo students 2011-05-21T14:40:56 only in tron we became more popular 2011-05-21T14:41:11 the tron one looks really fun. I'm considering make a tron bot since all the materials needed are still available, just for the fun of it 2011-05-21T14:41:30 we might run that again 2011-05-21T14:41:36 maybe with 3vs3 2011-05-21T14:41:57 one AI controlling 3 lightcycles? 2011-05-21T14:42:08 yes 2011-05-21T14:42:20 or teams? that could work too 2011-05-21T14:42:33 teams would be harder to cordinate like that though' 2011-05-21T14:42:47 I think your better off with one AI controlling all three bots. 2011-05-21T14:42:53 i guess 2011-05-21T14:43:09 we were contemplating making future contests with communication in between bots 2011-05-21T14:43:17 let them decide on a protocol, see what happens 2011-05-21T14:43:20 if nothing else, it'd go a long way in debugging cause three programmers and theres a bug, that'd be a nightmare to find and lots of finger pointing :P 2011-05-21T14:43:34 by teams i mean random pairs 2011-05-21T14:43:52 kk 2011-05-21T14:43:52 so you have to play nice with the others, even though you don't know who they are 2011-05-21T14:44:01 yea that could be interesting. 2011-05-21T14:44:53 Nebri: how did you hear about the contest? my forum post? 2011-05-21T14:45:12 amstan: nah, good ol Google :) 2011-05-21T14:45:31 I stumbled upon it in the last couple weeks post amazebot. Started searching for other AI competitions 2011-05-21T14:49:03 love stuff like this, and currently unemployed at the moment. So looking for something fun to work on. 2011-05-21T14:49:14 Nebri: yeah.. i know the feeling 2011-05-21T14:49:18 Nebri: what program are you in? 2011-05-21T14:50:09 amstan: Software Development, and I just graduated from the Network Engineering & Security Analyst program at the end of 2009 2011-05-21T14:50:21 cool 2011-05-21T14:50:24 i'm doing comp eng 2011-05-21T14:50:39 is that more hardware related? 2011-05-21T14:50:43 yeah 2011-05-21T14:50:47 we have 1 programming class 2011-05-21T14:51:08 i picked it because i wanted to stay away from c# 2011-05-21T14:51:56 heh :P. The new AI contest that me and aravin are discussing we're choosing C# as our language for development, to get students more exposure to whats out there in software development 2011-05-21T14:53:16 what's it about? 2011-05-21T14:53:58 *** olexs1 has joined #aichallenge 2011-05-21T14:53:59 no idea, he only approached me to ask if I would be interested in venturing into this with him and we tossed a couple game ideas for the AI's to play. As for right now I'm waiting for him to toss up a repo and get coding on a platform that we could use. 2011-05-21T14:54:37 absolutely nothing has been decided except for that I volunteered to help out the engine development. 2011-05-21T14:54:47 cool 2011-05-21T14:54:47 amstan: You left CS? 2011-05-21T14:54:58 jbroman: yeah, i was doing rather poor 2011-05-21T14:55:11 might come back in a few years 2011-05-21T14:55:16 amstan: CompEng at ___? 2011-05-21T14:55:24 mohawk college, hamilton 2011-05-21T14:55:33 I see. How's Hamilton? 2011-05-21T14:55:38 that's where I'm attending as well :) 2011-05-21T14:55:38 good, home 2011-05-21T14:56:37 *** olexs has quit IRC (Ping timeout: 246 seconds) 2011-05-21T14:58:02 * amstan wonders where jeff is 2011-05-21T14:58:42 any particular reason why you guys decided to go with ANTS for the next AI Challenge? 2011-05-21T14:58:57 we like changing the contest every term 2011-05-21T14:59:07 or else it wouldn't be fun, the barrier of entry would go higher and higher 2011-05-21T14:59:26 obviously :), but ants seems to be considerably more complex to me 2011-05-21T14:59:29 there were lots of suggestions(http://ai-contest.com/forums/viewforum.php?f=21), we eventually went with ants 2011-05-21T14:59:37 you should have seen the first proposal 2011-05-21T14:59:59 one process per ant, and they had to communicate using scents 2011-05-21T15:00:05 but that's not a lot of bandwidth 2011-05-21T15:00:28 we eventually ditched it because of the computational power to run such a game 2011-05-21T15:01:00 I think ants is straining the boundaries of what the complexity should be. But on the other hand you don't want something for which excellent algorithms exist and can easily be found (e.g. chess). 2011-05-21T15:01:08 yea no kidding, one thread per ant? HA! :P 2011-05-21T15:01:36 yeah, we even have docs for picking a contest subject: https://github.com/aichallenge/aichallenge/wiki/Game-Criteria 2011-05-21T15:03:09 very nice :) 2011-05-21T15:03:35 I'm very impressed with how you guys are running this contest so far. What kind of back end server do you have running this? 2011-05-21T15:03:53 for tron we used the csclub resources at uwaterloo 2011-05-21T15:04:03 we moved to amazon ec2 for planet wars 2011-05-21T15:04:20 we'll probably stay with ec2, they're cheap 2011-05-21T15:04:54 yea and cloud computing is awesome no matter how you look at it :) 2011-05-21T15:04:57 but pretty much any comp(with default ubuntu natty install) can become a worker 2011-05-21T15:05:09 Nice thing about EC2 is uniformity (= fairness). 2011-05-21T15:05:40 oh? so you guys are using a distrubted computing software to create a cloud on user/volunteer resources? 2011-05-21T15:05:55 Nebri: we rolled our own 2011-05-21T15:06:04 oh.. wait, no, nvm 2011-05-21T15:06:14 lol, I was gunna say! :P 2011-05-21T15:06:17 we don't have volunteers for workers, security concerns 2011-05-21T15:06:36 you don't want contestants to run their own bots on their hardware then send back the results 2011-05-21T15:07:14 the rule is that anyone with system access on the live contest cannot compete 2011-05-21T15:07:22 which makes sense 2011-05-21T15:07:41 but you guys did roll out some software to create distrubuted computing cloud, but the contestants can't set up a worker pc? 2011-05-21T15:08:07 only organizers/volunteers who don't wish to compete? 2011-05-21T15:08:15 Nebri: The worker software runs on multiple machines. If you want to call that a cloud, sure. 2011-05-21T15:08:39 Nebri: so far we only used ec2 for workers for uniformity, as jbroman said 2011-05-21T15:09:27 ah gotcha, so no cloud, just multiple machines running multiple instances 2011-05-21T15:09:40 with all the results being collected and displayed online? 2011-05-21T15:09:40 yes 2011-05-21T15:09:50 makes much more sense now :) 2011-05-21T15:10:06 Nebri: this might help: https://github.com/aichallenge/aichallenge/wiki/Beginner%27s-guide 2011-05-21T15:10:19 i wrote that for a friend wanting to roll his own contest 2011-05-21T15:12:26 it's just such a good idea, and the turn out you guys got for tron was amazing. I'm sure planetwars was just as well (didn't look too much at planet wars) 2011-05-21T15:12:42 we had 7 times more ppl for pw 2011-05-21T15:13:02 WHAT!? 2011-05-21T15:13:16 that is absolutely incredible :) 2011-05-21T15:13:36 for tron we had 708 contestants 2011-05-21T15:13:39 funding must be a pain to get something of thise scope up and running though 2011-05-21T15:13:47 and for planet wars we had 4600 2011-05-21T15:13:58 not really, i think we spent like 500 bucks 2011-05-21T15:14:13 amstan: Only because of the immense amount of free labour. :P 2011-05-21T15:14:17 yes 2011-05-21T15:15:07 yea, volunteers are awesome 2011-05-21T15:16:49 Nebri: were you the bot that was going backwards the whole time? 2011-05-21T15:16:56 are all the games selected purely at random? or do you guys have some match making solution in the background? 2011-05-21T15:17:16 some match making, the bots that didn't play against eachother 2011-05-21T15:17:19 so we can rank them better 2011-05-21T15:17:25 because we're not playing all vs all 2011-05-21T15:17:28 that would be impossible 2011-05-21T15:17:56 right due to CPU power needed to make that happen. 2011-05-21T15:18:02 how many games then to establish elo? 2011-05-21T15:18:30 the more you have the more accurate the results are 2011-05-21T15:18:50 i don't remember how many games we had for planetwars at the end(for each submission) 2011-05-21T15:30:59 *** yasith__ has quit IRC (Ping timeout: 276 seconds) 2011-05-21T15:36:48 McLeopold: pong :P 2011-05-21T15:37:02 antimatroid: dong 2011-05-21T15:37:57 My son is holding a remote to his ear and babbling into it... 2011-05-21T15:38:10 *** demerzel has quit IRC (Ping timeout: 252 seconds) 2011-05-21T15:39:04 antimatroid: either you forgot you pinged me or you are typing a novel... 2011-05-21T15:40:51 (00:55:30) McLeopold: antimatroid: how goes it? 2011-05-21T15:40:52 :p 2011-05-21T15:41:03 you also said yesterday you wanted my maths :P 2011-05-21T15:41:10 oh yeah 2011-05-21T15:41:30 I implemented a food rate function, but wanted to make it cooler 2011-05-21T15:42:04 is that why there is no food? :p; 2011-05-21T15:42:18 that would make a lot of sense :P 2011-05-21T15:42:41 also, i think the server should show people the "test game" if they fail it 2011-05-21T15:42:47 or at least be told why they failed it 2011-05-21T15:42:50 given 2 numbers (#players, food per turn) can you make a function to return whole numbers of food for each player? food_amount_per_player( turn#) 2011-05-21T15:44:12 so, for a rate of 3 with 4 players, it should return 0,1,1,1,0,1,1,1,... for increasing values of turn# 2011-05-21T15:46:06 i don't quite get that? 2011-05-21T15:46:20 does it work for symmetric spawning? 2011-05-21T15:46:36 i really don't care for asymmetric, i thought they'd suck, and any games i've seen on them suck :P 2011-05-21T15:47:13 so if the rate is 3, then there should be 3 new foods per turn, but you only want to spawn food when you have enough for everyone 2011-05-21T15:47:38 so at a rate of 3, the total amount spawned would be : 3,6,9,... 2011-05-21T15:47:39 oh okay, yeah that's sick :) 2011-05-21T15:48:10 the total amount handed out each turn would be: 0,4,4,4,0,4,4,4,... 2011-05-21T15:48:32 so you just bank the food until you have enough to hand out evenly to all players 2011-05-21T15:48:38 yep 2011-05-21T15:48:47 I wanted that as a function 2011-05-21T15:49:03 i didn't know it, but so did i, i was affraid of the amount of ants 2011-05-21T15:49:49 I want that as a function. I wrote it with an extra variable to keep track of left over food, but a function would be nicer, since you can calculate it. 2011-05-21T15:51:05 btw, I turned the rate down to 1 food every 2 turns, so for a 5 player game, food spawns for everyone once every 10 turns. 2011-05-21T15:51:24 there were a lot of 30 turn games yesterday with no food at all? 2011-05-21T15:51:34 I've got that fixed. 2011-05-21T15:51:40 It was bad code. 2011-05-21T15:51:58 A case of oop done poorly. 2011-05-21T15:52:01 my bot fails the test for some reason, amstan managed to break it yesterday but then i had to go out 2011-05-21T15:52:06 so i need to work out what's causing that later 2011-05-21T15:52:11 score one for the functional programmers. 2011-05-21T15:52:32 I got your bot running on my system just fine. 2011-05-21T15:52:47 McLeopold: it works on 4 players, but breaks on 2 players 2011-05-21T15:52:54 oh 2011-05-21T15:53:35 i don't get that still, i will play later though 2011-05-21T15:53:40 the bot doesn't know how many players there are :\ 2011-05-21T15:54:12 antimatroid: so, can you solve my math question? 2011-05-21T15:54:42 McLeopold: test_bot.sh is broken btw 2011-05-21T15:54:59 feel free to fix it 2011-05-21T15:56:10 I have a feeling it will include the lcm 2011-05-21T15:56:41 aichallenge: Alexandru Stan fix_test_bot * rf7e9323 / ants/test_bot.sh : copied the options from play_one_game.sh - http://bit.ly/kOC4K4 2011-05-21T15:57:37 oh! 2011-05-21T15:57:39 it does work 2011-05-21T15:57:44 i forgot python in this test 2011-05-21T16:04:56 mcleopold: i can't make sense of what you want :P 2011-05-21T16:05:09 "whole numbers of food for each player? food_amount_per_player( turn#)"? 2011-05-21T16:05:33 the food functions take a whole number, and multiply it by the number of players 2011-05-21T16:05:57 *** Accoun has quit IRC () 2011-05-21T16:06:10 if I want a rate of 1 food spawned per turn, I should only call the food function once every x turns 2011-05-21T16:06:23 where x in number of players 2011-05-21T16:06:39 oh, i think i get it 2011-05-21T16:07:05 *** demerzel has joined #aichallenge 2011-05-21T16:07:18 you want to call "food_amount_per_player( turn#)" and get how many squares to spawn that turn for each player? 2011-05-21T16:07:35 so for a rate of 1 with 4 players, the function would be: f(1)=0, f(2)=0, f(3)=0, f(4)=1, ... 2011-05-21T16:08:02 so the function would take turn#, #players and rate 2011-05-21T16:08:05 do you want the engine to store an array of these values already calcualted? or to recalculate every turn? 2011-05-21T16:08:23 I was hoping for a math answer, not a code answer 2011-05-21T16:08:33 i could probably do the maths 2011-05-21T16:08:45 but i don't think it'll be as fast as just storing the info in one sweep 2011-05-21T16:09:03 aichallenge: Alexandru Stan epsilon * rf7e9323 / ants/test_bot.sh : copied the options from play_one_game.sh - http://bit.ly/kOC4K4 2011-05-21T16:09:03 aichallenge: Alexandru Stan epsilon * rb204515 / ants/test_bot.cmd : fixed the windows one too - http://bit.ly/l1qvgu 2011-05-21T16:09:14 yeah, probably 2011-05-21T16:09:30 I could find the lcm, generate a schedule and loop through it 2011-05-21T16:09:30 McLeopold: can you test if test_bot.cmd works on windows? i don't ahve access to one 2011-05-21T16:09:31 anyway, i'll work it out 2011-05-21T16:09:43 antimatroid: then use it on your bot, it'll time out 2011-05-21T16:09:45 I'd have to use my win machine then... 2011-05-21T16:09:48 antimatroid: it works for other bots 2011-05-21T16:10:22 amstan: yes, I'll test, give me a few 2011-05-21T16:10:32 or antimatroid could 2011-05-21T16:11:25 i have windows installed, but it'd mean closing everything and restarting 2011-05-21T16:11:47 *** mcstar has joined #aichallenge 2011-05-21T16:13:35 antimatroid: then do it in linux 2011-05-21T16:13:38 i think you might be able to do it somewhat cheaply 2011-05-21T16:13:44 antimatroid: it'll still demonstrate how your bot fails 2011-05-21T16:15:48 i got it to time my bot out 2011-05-21T16:16:18 hmm, okay cheers :) 2011-05-21T16:16:25 at least i have something to test locally now 2011-05-21T16:18:02 McLeopold: +1 for that thing btw, people can functionally test their bot now 2011-05-21T16:18:06 *** Accoun has joined #aichallenge 2011-05-21T16:18:09 McLeopold: would this work... where N = players, R = rate, t = turn, food = floor(t.R/N) - floor((t-1)R/N)? 2011-05-21T16:18:25 let me try... 2011-05-21T16:19:11 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 4.0.1/20110504060726]) 2011-05-21T16:20:04 it requires floor(4(3)/4) == 4, not 3 2011-05-21T16:20:10 ie. rounding could break it? 2011-05-21T16:24:42 whats teh average row/col 2011-05-21T16:24:52 for a visualizer just trying to get a feel for it 2011-05-21T16:24:55 2938 is 80x104 2011-05-21T16:24:55 what do you mean? 2011-05-21T16:25:07 i mean what are the bounds of map row/col 2011-05-21T16:25:13 and/or average/set amount 2011-05-21T16:25:18 oh, that's completely undecided yet 2011-05-21T16:25:23 the max is 256x256 2011-05-21T16:25:27 ah cool ty 2011-05-21T16:25:34 but we wont every be able to feasibly use that 2011-05-21T16:25:42 ever* (it's 6 30am :P) 2011-05-21T16:25:52 just trying to figure out if this 80x104 is an average size map or not 2011-05-21T16:26:04 leaves about 15pixels per at fullscreen 2011-05-21T16:26:34 also why the hell did we call them row/col i knwo i'm gonna be debugging that later 2011-05-21T16:26:38 already have to mentally switch x/y 2011-05-21T16:26:54 don't use x/y 2011-05-21T16:27:11 i think that way i do call them row/col in code 2011-05-21T16:27:11 then you have to mentally remember when you need to transpose the map for reading/writing information 2011-05-21T16:27:24 if you use row/col, you never have to think about that again 2011-05-21T16:27:31 it's how we index writing on pages 2011-05-21T16:27:37 and it's how mathematicians index matrices 2011-05-21T16:27:43 hmm 2011-05-21T16:27:46 programmers are silly for using x,y :P 2011-05-21T16:27:49 i am so used to plot points i guess 2011-05-21T16:27:54 i will buy that 2011-05-21T16:28:04 hey while you're talkative have you looked at replays 2011-05-21T16:28:06 it was x,y 2011-05-21T16:28:11 it looks like maps are just 'b' for every possible food/ant 2011-05-21T16:28:17 but i eventually convinced people it actually is annoying to do it that way :P 2011-05-21T16:28:18 and you have to decode the ants object to determine which turn which 2011-05-21T16:28:21 on turn 1 that is 2011-05-21T16:28:29 unless my map decoder is wrong 2011-05-21T16:28:36 'b'? 2011-05-21T16:28:56 yea looking at this 2011-05-21T16:28:57 http://hypertriangle.com:13080/game/2938 2011-05-21T16:29:24 i have no idea about the replay format :P 2011-05-21T16:29:28 oh poo 2011-05-21T16:29:29 ok then 2011-05-21T16:29:50 McLeopold? 2011-05-21T16:29:52 Frontier? 2011-05-21T16:30:06 mcleopold: playercolors = playerinfo? 2011-05-21T16:30:21 since playerinfo doesn't exist on the wiki and playercolors doesn't exist in the json 2011-05-21T16:31:01 if playerinfo wasn't a blank array i could probably answer that myself 2011-05-21T16:34:22 what? 2011-05-21T16:37:47 *** berak has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2011-05-21T16:38:10 demerzel: we haven't fully decided what to stuff in player_info yet 2011-05-21T16:38:27 the wiki will always be the description that is out of date 2011-05-21T16:39:20 I don't know where the b's came from, it looks like an issue 2011-05-21T16:39:28 ending map state? 2011-05-21T16:42:25 i didn't look yet i was decodingt he initial map 2011-05-21T16:43:21 the weird thing is the initial map state has a billion ants 'b' and also food '*' 2011-05-21T16:43:25 it's just odd iand i'm not sure why 2011-05-21T16:43:48 it would seem it's an end game state where player b won 2011-05-21T16:43:52 like if the ants turned into food on turn 1 that would make sense 2011-05-21T16:44:07 ohhhh so the map->data is endstate map? that is news to me 2011-05-21T16:44:24 I don't think it should be, but it wouldn't matter 2011-05-21T16:44:28 weird so basically to get the intiial state of a map you ignore anything but water 2011-05-21T16:44:31 that is odd 2011-05-21T16:44:45 or food i guess but that's covered in ants[] 2011-05-21T16:44:50 does that sound right? 2011-05-21T16:44:53 it should be starting state, otherwise you don't know the start positions 2011-05-21T16:44:55 McLeopold: did that food calc thing work? 2011-05-21T16:45:04 wouldn't you know from ants[] 2011-05-21T16:45:12 antimatroid: I was hoping for something more cleaver 2011-05-21T16:45:15 clever 2011-05-21T16:45:25 that's cheap and easy :P 2011-05-21T16:46:24 i'll try to work out something cooler 2011-05-21T16:46:31 but i don't think it'll be as cheap 2011-05-21T16:46:58 mcleopold: look at this http://hypertriangle.com:13080/game/2938 2011-05-21T16:47:07 there are 628 'b' in the map->data 2011-05-21T16:47:26 on last turn http://hypertriangle.com:13080/visualizer.php?game=2938 there are 318 living ants 2011-05-21T16:48:09 i am not sure what that corresponds to 2011-05-21T16:49:46 there are 797 ants in the ants[] object 2011-05-21T16:50:06 only thing that it /might/ be is unique starting positions for those ants since i assume food can appear in teh same spot as an ant or other food has previously 2011-05-21T16:50:19 otherwise i am completely stumped as to why there are 628 'b' in teh initial map data 2011-05-21T16:52:43 McLeopold: hmmmm :P 2011-05-21T16:52:53 ? 2011-05-21T16:53:01 if you do n = 5, r = 3, then you get food like 0 1 0 1 1 0 1 0 1 1 .. 2011-05-21T16:53:11 i'm not sure how to calculate that nicer 2011-05-21T16:53:30 where is the thing that makes the replay data JSON 2011-05-21T16:53:31 or r = 5, n = 3, 1,2,2,1,2,2 2011-05-21T16:53:35 i can just go look at that 2011-05-21T16:53:57 demerzel: ants/ants.py look in get_state or get_replay 2011-05-21T16:54:02 ty 2011-05-21T17:00:43 McLeopold: meh, just use floor :P 2011-05-21T17:08:51 ""oh, but you don't have enough stat."Bitch. I can derive fucking stats right here right now." ahaha 2011-05-21T17:09:21 is this python for power of 2? d_col**2 2011-05-21T17:10:45 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-05-21T17:10:51 oh god its not that simple 2011-05-21T17:11:08 leftover keywords jesus i hate python 2011-05-21T17:11:36 *** McLeopold has left #aichallenge 2011-05-21T17:27:50 *** Kingpin13 has quit IRC (Quit: this used to be the life, but I don't need another one) 2011-05-21T17:31:46 lol, rapture's at 6 pm, not sure why you guys are even bothering with coding :P. XD XD 2011-05-21T17:37:05 dude, it's 7 36am in australia, 22nd :P 2011-05-21T17:37:32 I know but it's 5:42 pm 21st here where we live :P 2011-05-21T17:37:36 i was out riding at 6pm, no earthquakes, sorry :P 2011-05-21T17:37:36 the jokes still runnin 2011-05-21T17:38:10 i wish i could have been there for the people who actually expected it 2011-05-21T17:38:14 i am hoping for a video 2011-05-21T17:38:52 me too! :D 2011-05-21T17:41:25 *** AlliedEnvy_ has quit IRC (Remote host closed the connection) 2011-05-21T17:42:36 *** ahmedhelal has joined #aichallenge 2011-05-21T17:44:14 *** AlliedEnvy has joined #aichallenge 2011-05-21T17:48:24 oh shit rapture already i'm still translating python 2011-05-21T17:48:29 hell of a way to go out 2011-05-21T17:49:27 gosh yea python is so readable this.revealed = [[[False for col in range(this.width)] 2011-05-21T17:49:37 oh mcleopold isn't here 2011-05-21T17:49:42 lol 2011-05-21T17:53:00 tbf i didn't have to do a line by line translation but i want my engine to be as close to the original as possible 2011-05-21T17:58:17 *** akranis has joined #aichallenge 2011-05-21T17:59:09 does anybody know what this actually does 2011-05-21T17:59:09 score = [Fraction(0,1)]*this.num_players 2011-05-21T17:59:36 i really don't understand, it does an (0/1) = 0 * num_players but i'm not sure wtf the brackets do 2011-05-21T18:02:38 its a list builder: makes a list of size this filled with Fraction instances 2011-05-21T18:02:51 wait is that a really convoluted way of declaring an int array with all 0s 2011-05-21T18:03:09 granted there is no such thing as int in python afaik but for my mind thats how i think of it 2011-05-21T18:03:11 its not convoluted, if you know what it means 2011-05-21T18:03:34 so i will have an array with a length of num_players all 0'd right? 2011-05-21T18:03:53 assuming fraction(0, 1) isn't some kind of magic number in python 2011-05-21T18:11:02 i just hand-coded 586 coordinate indices 2011-05-21T18:12:40 and vim has trouble matching parantheses 2011-05-21T18:12:46 and indenting too 2011-05-21T18:14:18 coding in vim 2011-05-21T18:14:19 damn 2011-05-21T18:15:07 what? 2011-05-21T18:15:19 u swear emacs? 2011-05-21T18:15:24 (by) 2011-05-21T18:17:57 why did you hand code it? 2011-05-21T18:19:45 its fun to see a dense half page of parantheses 2011-05-21T18:20:08 i dont want to generate them at runtime 2011-05-21T18:20:29 its the indices of the radii that an ant can see 2011-05-21T18:20:49 and the spawn and the attack radii 2011-05-21T18:21:57 btw ty okayzed i would have had no idea what that was 2011-05-21T18:23:58 isnt it like multiplying a string? 2011-05-21T18:24:05 "something"*10000000 2011-05-21T18:24:20 then u get a lot of something which is almost anything 2011-05-21T18:26:07 talk to me people im bored :( 2011-05-21T18:26:30 im at the phase when programming is a chore 2011-05-21T18:28:50 havent gotten that far 2011-05-21T18:29:27 how far have u gotten? what is ur logic based on? 2011-05-21T18:31:00 im not asking you to reveal state secrets 2011-05-21T18:31:15 heuristics? genetic something? 2011-05-21T18:32:18 come on ppl, install a voice recognition software so that you can use irc while you code 2011-05-21T18:45:35 i am dissecting the engine atm 2011-05-21T18:45:42 only known about it for like 2 days 2011-05-21T18:54:32 really? why? 2011-05-21T18:55:01 i mean why do you want to see whats behind the engine? 2011-05-21T18:58:48 *** choas has quit IRC (Quit: leaving) 2011-05-21T19:02:26 *** AlliedEnvy has quit IRC (Remote host closed the connection) 2011-05-21T19:05:04 *** AlliedEnvy has joined #aichallenge 2011-05-21T19:06:21 *** sigh has joined #aichallenge 2011-05-21T19:07:54 why wouldn't i? 2011-05-21T19:08:12 specifically though i'm making a debugger so it's vital i can replicate the engine 2011-05-21T19:14:52 a debugger for what? 2011-05-21T19:15:02 for the official game engine? 2011-05-21T19:15:53 but to develop your own bot i dont see why would you need to know how the engine works 2011-05-21T19:16:15 if it interests you thats ok, im just saying that it consumes your time 2011-05-21T19:18:48 so does writing an ai bot 2011-05-21T19:28:28 ill be outside if anyone needs me just shout 2011-05-21T19:31:36 *** Frontier has quit IRC (Read error: Connection reset by peer) 2011-05-21T19:37:01 *** Frontier has joined #aichallenge 2011-05-21T19:44:04 *** akranis has quit IRC (Remote host closed the connection) 2011-05-21T19:49:31 *** Frontier has quit IRC (Quit: Leaving.) 2011-05-21T19:55:50 *** delt0r___ has joined #aichallenge 2011-05-21T19:58:08 *** delt0r_ has quit IRC (Ping timeout: 276 seconds) 2011-05-21T20:04:27 are the coordinates received absolute or relative? 2011-05-21T20:04:42 ie if i get an ant coord at 8, 9 then is that absolute 8,9 or relative to what i can see 2011-05-21T20:04:47 or have seen 2011-05-21T20:04:50 absolute 2011-05-21T20:04:56 awesome ty 2011-05-21T20:05:09 at least I handle them as such and it seems to work :) 2011-05-21T20:05:30 i cant imagine relative coordinates in this context 2011-05-21T20:05:31 just setting up search trees and if it was relative it juts got a lot more complex 2011-05-21T20:05:42 you would be surprised at the shit in the engine 2011-05-21T20:05:53 i don't know if it uses it but it's got bfs, tracking per player of explored etc 2011-05-21T20:06:01 did you mean shift? 2011-05-21T20:06:01 i haven't fully translated it yet so i wasn't sure 2011-05-21T20:06:04 (joke) 2011-05-21T20:21:51 *** FireFly has quit IRC (Quit: swatted to death) 2011-05-21T20:26:29 *** amstan has joined #aichallenge 2011-05-21T20:26:29 *** ChanServ sets mode: +o amstan 2011-05-21T20:29:53 antimatroid: how's your bot? 2011-05-21T20:32:02 *** mcstar has quit IRC (Quit: WeeChat 0.3.5) 2011-05-21T20:44:02 *** Nebri has quit IRC (Read error: Connection reset by peer) 2011-05-21T20:44:25 *** Nebri has joined #aichallenge 2011-05-21T21:04:04 amstan: do you guys know the problem behind the 30 turn no food games? 2011-05-21T21:08:54 *** No has quit IRC (Ping timeout: 252 seconds) 2011-05-21T21:09:18 *** sigh has quit IRC (Remote host closed the connection) 2011-05-21T21:13:01 *** Redgis has quit IRC (Quit: ... mains libres) 2011-05-21T21:13:11 demerzel: mcleopold might 2011-05-21T21:13:27 ah just noticed it was still there and i know you talked about it last night 2011-05-21T21:52:36 *** olexs1 has quit IRC (Quit: Leaving.) 2011-05-21T22:00:05 *** amstan_ has joined #aichallenge 2011-05-21T22:00:05 *** ChanServ sets mode: +o amstan_ 2011-05-21T22:30:11 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-05-21T22:33:01 amstan: is worker down? 2011-05-21T22:33:26 was putting in a glorified go-bot just to see if it would take it 2011-05-21T22:35:54 and it did not 2011-05-21T22:36:13 i have to officially say i think the test is not right because that bot worked fine using ants.py locally 2011-05-21T22:42:25 demerzel: should be up 2011-05-21T22:42:31 worker is up yes 2011-05-21T22:42:39 but bot failed despite working fine locally 2011-05-21T22:42:44 how did it fail? 2011-05-21T22:43:07 you should try testing it with ants/test_bot.sh 2011-05-21T22:43:10 git pull first 2011-05-21T22:43:19 i just fixed that script today 2011-05-21T22:43:30 antimatroid had it fail with it 2011-05-21T22:43:32 so.. 2011-05-21T22:43:45 *** amstan_ has quit IRC (Remote host closed the connection) 2011-05-21T22:43:54 yea i get an error with test_bot.bat 2011-05-21T22:44:06 it works when doing play_one_game 2011-05-21T22:44:12 i mean it doesn't do jack but it doesn't break either 2011-05-21T22:44:20 just my test case to see if i could get anything to upload period 2011-05-21T22:45:22 amstan: i get this error (sorry windows) running it http://pastebin.com/ABMCXd5L 2011-05-21T22:45:56 i had to take out the -o viewer part too otherwise i got 'got 3 bots expected 2' error 2011-05-21T22:46:22 demerzel: as i said, git pull first 2011-05-21T22:46:32 i am not working with a clone atm i have too many repos already 2011-05-21T22:46:36 but i will pull the latest 2011-05-21T22:46:46 you only need one.. 2011-05-21T22:46:52 this is yesterday 2011-05-21T22:47:02 have you changed play_game or test_bot.sh since then? 2011-05-21T22:47:05 i have one, and i merge like daily, maintain a test server and so on 2011-05-21T22:47:38 ah yea you have 2011-05-21T22:47:39 ok 2011-05-21T22:48:13 i recommend adding a pause to the end of the .cmd 2011-05-21T22:48:25 in windows it doesn't keep the command window open after an error if you do that 2011-05-21T22:48:38 probably less problems with people asking why it broke down the ilne if you add pause to the end lol 2011-05-21T22:48:39 isn't there like a right click option for that? 2011-05-21T22:48:54 if there is i've missed it for the last ~10 years 2011-05-21T22:50:26 actually when i get this done i'll give you a windows version 2011-05-21T22:50:36 moving the %1 to a variable to set 2011-05-21T22:50:53 it's much less of a pain in the ass to just edit that variable than open a command window and navigate wherever the hell 2011-05-21T22:51:10 as i said before.. you're doing it wrong 2011-05-21T22:52:59 no you just refuse to consider that people don't use unix-based :P 2011-05-21T22:53:17 some* people 2011-05-21T22:53:23 most afaik 2011-05-21T22:54:06 hmm now it times out 2011-05-21T22:54:27 is the --end-wait 0.25 giving it 0.25 before a timeout? does that include initial? it might take more than 250ms for it to spin up 2011-05-21T22:56:59 *** delt0r___ has quit IRC (Ping timeout: 240 seconds) 2011-05-21T23:04:09 amstan i really have no damn idea - i yanked the parsing code straight out of my pw bot and it's still saying timed out 2011-05-21T23:05:34 oh well guess it's verbose logging to file time 2011-05-21T23:20:50 omfg i think i figured it out haha 2011-05-21T23:21:02 i missed that we have to send go prior to turn 1 2011-05-21T23:31:39 ok what am i doing wrong here 2011-05-21T23:31:41 http://pastebin.com/Sa76Yqsi 2011-05-21T23:31:44 i am now getting to turn 1 2011-05-21T23:31:52 2 ms processing inital settings 2011-05-21T23:31:57 then 14 ms processing turn 1 data 2011-05-21T23:32:02 at 16 ms i send 'go\n' 2011-05-21T23:32:07 and play_game times me out 2011-05-21T23:32:40 amstan: i would greatly appreciate iti you could look at that log, it logs everything i send/receive and has a turnelapsed ms timer 2011-05-21T23:32:53 i have no idea why its timing out on turn 1 unless i am missing something 2011-05-21T23:33:17 demerzel: i don't actually know anything more about the bots 2011-05-21T23:33:22 idk anything about the protocol 2011-05-21T23:33:37 crap 2011-05-21T23:33:52 i have it logging everything and i see zero problem with it (after i figured out i was suppsoed to send a go on turn 0) 2011-05-21T23:34:11 demerzel: did you flush the outputs and inputs after every read? 2011-05-21T23:34:36 i've done both ways but it accepted the first 'go\n' 2011-05-21T23:34:45 last yeari didn't need to but i will ptu it back in to see if it helps since i am further along now 2011-05-21T23:35:39 nope same thing :( 2011-05-21T23:36:47 tried making it wait 100ms before sending go 2011-05-21T23:36:50 didn't work either 2011-05-21T23:37:13 demerzel: so you did flush the outputs? 2011-05-21T23:37:28 i took it out but put it back in when you said something jic 2011-05-21T23:37:32 didn't make a diff 2011-05-21T23:37:46 manual flushing is back in there 2011-05-21T23:38:05 it takes the first 'go\n' becuase it was stopping there before 2011-05-21T23:38:20 maybe i have to move or it kills it 2011-05-21T23:38:31 i'll put in a 'move north' every turn lol 2011-05-21T23:38:48 demerzel: try doing the same type of debugging on another bot 2011-05-21T23:38:49 that works 2011-05-21T23:38:51 see what they output 2011-05-21T23:39:11 well problem is other bots are other languages 2011-05-21T23:39:24 i'm almost positive it's a .NET stdio problem or i'm just missing something 2011-05-21T23:39:25 so? 2011-05-21T23:39:47 well, i have a feeling you're not compliant to the protocol 2011-05-21T23:39:51 see what the other bots do 2011-05-21T23:40:05 you mgiht be right i'll look if this doesn't work 2011-05-21T23:42:06 oh thank god i found it 2011-05-21T23:42:07 extra newline 2011-05-21T23:42:17 floating out in the ether 2011-05-21T23:44:23 May 21st 23:43:59 successfully entered into contest C# 2011-05-21T23:44:24 woop 2011-05-21T23:44:37 i've got your starter package haha just needed 1 damn line 2011-05-21T23:47:45 You can upload your code as often as you want. 2011-05-21T23:47:46 Sorry, you have to wait at least 10 minutes between submissions.