2011-04-25T00:00:51 *** kaemo0 has quit IRC (Read error: Operation timed out) 2011-04-25T00:02:59 *** kaemo has quit IRC (Ping timeout: 258 seconds) 2011-04-25T00:15:59 *** kaemo0 has joined #aichallenge 2011-04-25T00:16:27 *** kaemo has joined #aichallenge 2011-04-25T00:34:02 *** jaycobcoleman has quit IRC (Quit: Page closed) 2011-04-25T00:46:35 aichallenge: sigh epsilon * r969fa82 / (ants/playgame.py worker/engine.py worker/sandbox.py): Implement logging of stderr for bots - http://bit.ly/frk4qt 2011-04-25T00:46:36 aichallenge: sigh epsilon * ra316d79 / (3 files in 2 dirs): Make --log_stderr the default - http://bit.ly/flO9wD 2011-04-25T00:49:18 sigh: is there a reason not to just leave stderr alone in the default case so it simply continues to print to the console? 2011-04-25T00:49:49 although I suppose that would mix multiple bots output together if more than one are printing to it 2011-04-25T00:50:05 yeah, I wasn't sure 2011-04-25T00:50:24 currently there are two cases, log to file and ignore altogether 2011-04-25T00:51:11 before it was just printing stderr when the bot died, but that resulted in blocking 2011-04-25T00:51:13 yeah, I see your new code. certainly better than just eating it (without really eating it) like it was before 2011-04-25T00:51:36 maybe three options? file, stdout and ignore? 2011-04-25T00:51:39 ahh, I didn't realize it even printed it at the end 2011-04-25T00:52:03 I'd go for file, stderr and ignore 2011-04-25T00:52:14 and make stdout the default so it sort of acts like before? 2011-04-25T00:52:26 makes it easy to use a redirect if they want to that way 2011-04-25T00:52:45 what do you mean? 2011-04-25T00:53:06 keep the bots stderr the engines stderr 2011-04-25T00:53:20 don't redirect it to the engines stdout 2011-04-25T00:53:25 ah yes... of course 2011-04-25T00:53:47 and yeah, I think that is probably the least surprising default 2011-04-25T00:54:06 yeah, that'll be easy... I'll do that 2011-04-25T00:54:51 first... I need to go to the supermarket because I have no food 2011-04-25T00:55:01 yesterday it was closed all day :( 2011-04-25T00:55:05 that's important :) 2011-04-25T00:59:10 aichallenge: janzert epsilon * r5276a95 / (7 files in 3 dirs): Add individual jailuser chroots to worker setup - http://bit.ly/dQCZlt 2011-04-25T01:33:09 yay... aquired food 2011-04-25T01:33:17 I didn't spawn though :( 2011-04-25T01:37:17 *** siddharth has quit IRC (Ping timeout: 258 seconds) 2011-04-25T01:37:50 *** Ice_Harley has joined #aichallenge 2011-04-25T01:45:25 aichallenge: sigh epsilon * r4142601 / (5 files in 2 dirs): Let log_stderr take file, stderr or ignore arguments - http://bit.ly/eYhPqD 2011-04-25T01:58:49 *** ahmedhelal has quit IRC (Read error: No route to host) 2011-04-25T01:59:24 *** ahmedhelal has joined #aichallenge 2011-04-25T02:40:05 sigh: do you want to send people first seen land squares add some point too? 2011-04-25T02:40:11 keep stocha happy 2011-04-25T02:40:30 and then starter bots can strip vision code too, and just mark a square as seen once it gets water/land information about it 2011-04-25T02:41:33 is there some kind of consensus on that? 2011-04-25T02:46:55 is there any opposition? 2011-04-25T02:47:05 i've been waiting to get mcleopolds okay, but i miss him 2011-04-25T02:47:23 i don't think there should be any problem with it 2011-04-25T02:50:03 ergh, so many pictures on reddit lately are so small 2011-04-25T02:51:49 so small? 2011-04-25T02:54:26 that it's hard to read any text 2011-04-25T02:57:57 you can embiggen the images 2011-04-25T02:59:49 not nicely 2011-04-25T04:09:29 antimatroid: did you delete enemies list from starter package? 2011-04-25T04:16:19 yeah :P 2011-04-25T04:16:28 because it was hard to give a vector of food squares 2011-04-25T04:16:38 i could have made it a set, but went meh and deleted them 2011-04-25T04:20:13 *** _0x47 <_0x47!~slash@91.37.227.86> has joined #aichallenge 2011-04-25T04:22:01 *** berak has joined #aichallenge 2011-04-25T04:23:16 *** Frontier has joined #aichallenge 2011-04-25T04:26:55 where is information about enemies now? :) 2011-04-25T04:27:41 in the grid, i see 2011-04-25T04:38:57 Ice_Harley: yeah same with food information 2011-04-25T04:39:06 and i took out lastSeen information and made is isVisible 2011-04-25T04:40:57 *** Ice_Harley has quit IRC (Ping timeout: 252 seconds) 2011-04-25T04:47:00 *** siddharth has joined #aichallenge 2011-04-25T04:48:37 *** mceier has joined #aichallenge 2011-04-25T05:18:12 *** Naktibalda has joined #aichallenge 2011-04-25T05:39:41 *** kaemo0 has quit IRC (Ping timeout: 252 seconds) 2011-04-25T05:39:41 *** kaemo has quit IRC (Ping timeout: 240 seconds) 2011-04-25T05:46:13 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T05:47:28 *** berak has joined #aichallenge 2011-04-25T05:58:01 *** boegel has joined #aichallenge 2011-04-25T06:06:40 *** siddharth has quit IRC (Quit: Leaving) 2011-04-25T06:06:48 *** siddharth has joined #aichallenge 2011-04-25T06:07:49 *** sids_aquarius has joined #aichallenge 2011-04-25T06:12:20 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T06:15:08 *** sids_aquarius has quit IRC (Ping timeout: 264 seconds) 2011-04-25T06:15:09 *** berak has joined #aichallenge 2011-04-25T06:15:25 *** sids_aquarius has joined #aichallenge 2011-04-25T06:20:22 *** Stocha has joined #aichallenge 2011-04-25T06:20:35 what is the current official rad2 for fighting range ? 2011-04-25T06:20:46 i made some test with 5 and 4 lately. 2011-04-25T06:21:07 with a bot that tend to cluster ants, and another one that tend to spread them. 2011-04-25T06:21:36 5 definetly favors the clustering one. 4 is far less obvious. (although it still favor clustering) 2011-04-25T06:21:58 They do not make special actions for optimizing the combat. 2011-04-25T06:22:48 *** Stocha has quit IRC (Client Quit) 2011-04-25T06:25:45 current parameters are spawn2 = 1, attack2 = 4, view2 = 55(i think on this last one) 2011-04-25T06:28:52 sigh: http://pastebin.com/8ZAnR4jU :P 2011-04-25T06:29:00 all translates for vision :\ 2011-04-25T06:29:45 ok...? 2011-04-25T06:32:14 antimatroid? what was that about? 2011-04-25T06:32:31 that's the translate amount for visible squares of an ant 2011-04-25T06:32:41 so one doesn't actually need to search for them from each ant 2011-04-25T06:32:58 why would you search for them? 2011-04-25T06:32:58 i can't be bothered doing vision past naive, it's fairly cheap anyway 2011-04-25T06:33:10 because that was easy :P 2011-04-25T06:33:19 *** Naktibalda has quit IRC (Ping timeout: 250 seconds) 2011-04-25T06:33:24 I still don't get your point :S 2011-04-25T06:33:45 that list is what the vision algorithm in the engine calculates and caches 2011-04-25T06:35:45 *** Stocha has joined #aichallenge 2011-04-25T06:36:02 vision used to be 96. why change this value ? 2011-04-25T06:36:15 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T06:37:04 i agree that food spawn should be 1. But i was happy with the fight=5 and view=96 2011-04-25T06:38:35 *** berak has joined #aichallenge 2011-04-25T06:39:19 i didn't mind with fight and mcleopold wanted 4 2011-04-25T06:39:35 greentea suggested lowering the view radius a while ago and i tended to agree along with mcleopold 2011-04-25T06:39:39 so we decreased it a bit 2011-04-25T06:39:54 sigh: yeah, i just generated them for myself 2011-04-25T06:40:13 i just have it as a constant int array now 2011-04-25T06:40:36 i see 2011-04-25T06:46:05 *** needsch has joined #aichallenge 2011-04-25T06:51:11 why does mcloepold want 4 ? 2011-04-25T06:51:45 arguably it is worth trying reducing the sight radius. Although i find that it is low enough as it stand. you don't have much time to plan already. 2011-04-25T06:53:05 Once the beta start, i hope every change will be justified and submitted to the community for review. 2011-04-25T06:54:02 why algorithm do a fly use ? 2011-04-25T06:54:10 what algorithm do a fly use. i wonder. 2011-04-25T06:54:42 *** delt0r_ has joined #aichallenge 2011-04-25T06:56:12 *** Redgis has joined #aichallenge 2011-04-25T06:56:38 *** delt0r___ has quit IRC (Ping timeout: 276 seconds) 2011-04-25T07:09:56 hmmm, i think i just decided to not use game trees :D 2011-04-25T07:17:57 *** sids_aquarius has quit IRC (Ping timeout: 250 seconds) 2011-04-25T07:30:11 *** sids_aquarius has joined #aichallenge 2011-04-25T07:30:33 .... are we going to want to take up as much memory as we can at the start and keep it? 2011-04-25T07:31:05 sigh: that might be a problem 2011-04-25T07:33:42 ants/memory wars 2011-04-25T07:33:57 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T07:34:27 *** sids_aquarius has quit IRC (Ping timeout: 252 seconds) 2011-04-25T07:39:37 *** berak has joined #aichallenge 2011-04-25T07:46:47 *** FireFly has joined #aichallenge 2011-04-25T07:51:11 *** sids_aquarius has joined #aichallenge 2011-04-25T08:06:11 *** Eruonen has joined #aichallenge 2011-04-25T08:41:53 *** Mathnerd314 has quit IRC (Ping timeout: 260 seconds) 2011-04-25T08:50:19 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T08:51:21 *** berak has joined #aichallenge 2011-04-25T09:03:23 *** helo_ has joined #aichallenge 2011-04-25T09:10:26 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T09:11:20 *** berak has joined #aichallenge 2011-04-25T09:31:55 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T09:34:13 *** berak has joined #aichallenge 2011-04-25T09:44:20 how did the tree went ? 2011-04-25T09:44:32 antimatroid: did you had trouble with trees ? 2011-04-25T09:45:50 *** Kingpin13 has joined #aichallenge 2011-04-25T09:48:11 *** boegel has quit IRC (Quit: This computer has gone to sleep) 2011-04-25T09:51:08 if you can't limit the memory a bot take, you can always decide beforehand how much is the maximum. Then launch the bots, and then kill a bot that take too much. (auto forfeit) then relaunch the game with this bot forfeit. 2011-04-25T09:51:30 it's not ideal though. Then you have the virtual machine and other stuff to take into account ideally. 2011-04-25T09:52:02 *** Stocha has quit IRC (Quit: Page closed) 2011-04-25T09:52:15 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T09:54:13 *** berak has joined #aichallenge 2011-04-25T09:59:41 *** helo_ has quit IRC (Quit: Page closed) 2011-04-25T10:14:58 *** bhasker has joined #aichallenge 2011-04-25T10:20:52 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T10:22:01 *** berak has joined #aichallenge 2011-04-25T10:24:12 *** smiley1983 has joined #aichallenge 2011-04-25T10:48:17 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T10:49:50 *** berak has joined #aichallenge 2011-04-25T10:54:33 *** sigh has quit IRC (Remote host closed the connection) 2011-04-25T10:56:27 *** Naktibalda has joined #aichallenge 2011-04-25T11:06:44 *** bhasker has quit IRC (Quit: bhasker) 2011-04-25T11:10:20 *** bhasker has joined #aichallenge 2011-04-25T11:18:05 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T11:24:04 *** Blkt has joined #aichallenge 2011-04-25T11:25:26 *** berak has joined #aichallenge 2011-04-25T11:27:11 *** Apophis_ has joined #aichallenge 2011-04-25T11:30:16 *** Apophis has quit IRC (Ping timeout: 246 seconds) 2011-04-25T11:47:55 *** amstan has joined #aichallenge 2011-04-25T11:47:55 *** ChanServ sets mode: +o amstan 2011-04-25T11:47:58 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T11:50:16 *** berak has joined #aichallenge 2011-04-25T11:51:49 *** Redgis has quit IRC (Remote host closed the connection) 2011-04-25T11:57:39 *** andy_ has joined #aichallenge 2011-04-25T11:58:04 *** andy_ is now known as Guest53055 2011-04-25T12:01:22 *** Guest53055 has quit IRC (Client Quit) 2011-04-25T12:08:08 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T12:09:35 aichallenge: Alexandru Stan epsilon * re688f01 / website/css/typo.css : added css for code and pre, closes #77 - http://bit.ly/gUZe8V 2011-04-25T12:09:39 aichallenge: Alexandru Stan epsilon * r30cb899 / (11 files in 6 dirs): Merge branch 'epsilon' of github.com:aichallenge/aichallenge into epsilon - http://bit.ly/gIIMLT 2011-04-25T12:10:25 *** berak has joined #aichallenge 2011-04-25T12:21:11 *** znutar_ has quit IRC (Remote host closed the connection) 2011-04-25T12:22:09 *** znutar has joined #aichallenge 2011-04-25T12:25:21 aichallenge: Alexandru Stan epsilon * r594ecc4 / website/css/layout.css : added box shadow around the wrapper, fixed opera support - http://bit.ly/hNcKk7 2011-04-25T12:31:41 *** stride has quit IRC (Ping timeout: 250 seconds) 2011-04-25T12:31:41 *** sids_aquarius has quit IRC (Read error: Connection reset by peer) 2011-04-25T12:36:51 *** Redgis has joined #aichallenge 2011-04-25T12:38:21 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T12:40:10 *** berak has joined #aichallenge 2011-04-25T12:43:37 *** stride has joined #aichallenge 2011-04-25T12:51:11 *** smiley1983 has quit IRC (Quit: leaving) 2011-04-25T12:52:09 *** kaemo0 has joined #aichallenge 2011-04-25T12:52:58 *** kaemo has joined #aichallenge 2011-04-25T12:53:02 *** sids_aquarius has joined #aichallenge 2011-04-25T12:56:03 *** boegel has joined #aichallenge 2011-04-25T12:58:08 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T13:00:08 *** berak has joined #aichallenge 2011-04-25T13:00:37 *** ahmedhelal has quit IRC (Remote host closed the connection) 2011-04-25T13:04:10 *** janzert has left #aichallenge 2011-04-25T13:07:27 aichallenge: Alexandru Stan epsilon * rf2b3caa / (website/css/layout.css website/menu.php website/rules.php): fixed random chars at the end of rules, fixed min-height, removed commented css, more html compliant - http://bit.ly/fw6mjQ 2011-04-25T13:08:12 aichallenge: Alexandru Stan epsilon * r19e4705 / (2 files): Merge remote branch 'ruby/epsilon' into epsilon - http://bit.ly/feWW4o 2011-04-25T13:09:00 aichallenge: Alexandru Stan epsilon * rb008df2 / (3 files): Merge remote branch 'hbhasker/epsilon' into epsilon - http://bit.ly/gp2IWK 2011-04-25T13:18:37 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T13:18:41 *** boegel has quit IRC (Quit: This computer has gone to sleep) 2011-04-25T13:19:56 *** berak has joined #aichallenge 2011-04-25T13:29:25 amstan: IMO it should merge nicely 2011-04-25T13:29:37 Naktibalda: ok, i'll try 2011-04-25T13:31:15 Naktibalda: nope 2011-04-25T13:31:24 big conflict? 2011-04-25T13:31:26 http://pastebin.com/q5CT5DDw 2011-04-25T13:31:27 yes 2011-04-25T13:31:47 mcleopold introduced a new way to do queries 2011-04-25T13:34:43 aichallenge: Gintautas Miselis neworg * r606871b / (9 files in 2 dirs): User can create a new organization during registration - http://bit.ly/e0YxGz 2011-04-25T13:34:44 Naktibalda: see that ^ 2011-04-25T13:34:45 aichallenge: Alexandru Stan neworg * rddc2d32 / (10 files in 2 dirs): (log message trimmed) 2011-04-25T13:34:45 aichallenge: Merge remote branch 'Naktibalda/epsilon' into neworg 2011-04-25T13:34:45 aichallenge: Conflicts: 2011-04-25T13:34:45 aichallenge: website/account_confirmation.php 2011-04-25T13:34:46 aichallenge: website/header.php 2011-04-25T13:34:46 aichallenge: website/process_registration.php 2011-04-25T13:34:47 aichallenge: website/register.php 2011-04-25T13:34:47 aichallenge: Gintautas Miselis neworg * r41b44ff / website/faq.php : Discarded answer about creating new organization - http://bit.ly/gdjre4 2011-04-25T13:34:49 aichallenge: Gintautas Miselis neworg * r3ca2808 / (website/process_registration.php website/register.php): Don't create empty organizations, organization name is trimmed before creation - http://bit.ly/g4Safd 2011-04-25T13:35:15 link: https://github.com/aichallenge/aichallenge/commit/ddc2d32cd77aa76cd44f2a81deb6740b1d2839de 2011-04-25T13:35:22 you could continue that.. or just fix your branch 2011-04-25T13:35:36 i just pushed that for showing you 2011-04-25T13:38:03 *** jaycobcoleman has joined #aichallenge 2011-04-25T13:38:12 Frontier: ping 2011-04-25T13:40:14 animatroid: any particular theories you think should be covered in strategy guides? 2011-04-25T13:45:15 amstan: do I have to for again? 2011-04-25T13:45:22 amstan: do I have to fork again? 2011-04-25T13:45:26 Naktibalda: no 2011-04-25T13:45:40 just pull from our repo 2011-04-25T13:45:46 resolve the merge conflicts 2011-04-25T13:45:53 and then ping me 2011-04-25T13:46:25 Naktibalda: see http://help.github.com/fork-a-repo/ 2011-04-25T13:46:30 Naktibalda: Pull in upstream changes 2011-04-25T13:46:50 amstan: pong 2011-04-25T13:46:52 From https://github.com/aichallenge/aichallenge 2011-04-25T13:46:54 * [new branch] neworg -> upstream/neworg 2011-04-25T13:47:04 Naktibalda: ignore neworg 2011-04-25T13:47:07 Naktibalda: just pull epsilon 2011-04-25T13:47:09 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T13:47:20 Frontier: hey, i'm ready to integrate your visualizer 2011-04-25T13:48:31 *** Chris_0076 has joined #aichallenge 2011-04-25T13:48:51 ok, tell me how you plan to deploy everything. so far I have an ant task that will copy everything over to a user defined destination folder 2011-04-25T13:48:55 *** berak has joined #aichallenge 2011-04-25T13:49:37 Frontier: well, i'll probably want to put that into janzert's script 2011-04-25T13:49:42 (including a zip file for the offline html visualizer) 2011-04-25T13:49:46 *** mcstar has joined #aichallenge 2011-04-25T13:50:13 Frontier: you already took care of compiling and minimizing, correct? 2011-04-25T13:50:21 sure 2011-04-25T13:50:38 so, what do i do to use your ant task? 2011-04-25T13:53:56 just run "ant deploy" with the variable 'deploy.path' set to where you want the website to be. A subdirectory 'visualizer' will be created in there. 2011-04-25T13:57:33 "ant -Ddeploy.path=/my/www/target/directory" should work 2011-04-25T13:58:42 hello guys 2011-04-25T13:58:51 i have a quick question 2011-04-25T13:59:04 answer is: when it's done 2011-04-25T13:59:21 how long does it take for an experienced programmer to preapare for the challenge? 2011-04-25T14:00:24 how long does it take for you to prepare for a new project? 2011-04-25T14:00:27 mcstar: what do you mean by prepare. You get a starter package with a working bot, so there is literally no delay to getting into the game 2011-04-25T14:00:55 true 2011-04-25T14:01:04 but u cant compete with that bot really 2011-04-25T14:01:19 You'll have to read one or two wiki pages to understand the rules, then it is all about your available time and creativity 2011-04-25T14:01:36 let me rephrase it 2011-04-25T14:01:44 how much time will u invest? 2011-04-25T14:01:53 approximately 2011-04-25T14:02:17 before i start doing anything i must have a timeframe 2011-04-25T14:04:50 let's say the contest will last two months and I have no idea how long I will be motivated. if you need a timeframe, you should set it for yourself. 2011-04-25T14:04:53 *** jaycobcoleman has quit IRC (Ping timeout: 252 seconds) 2011-04-25T14:05:39 *** jaycobcoleman has joined #aichallenge 2011-04-25T14:05:47 you say two months, does it mean that the bots compete for two months? 2011-04-25T14:05:53 So I blatantly mis-typed that captcha 2011-04-25T14:05:59 amstan: committed 2011-04-25T14:06:26 i suppose once my bot is uploaded, theres no room for any modifications 2011-04-25T14:06:40 you can upload new versions as many times as you want 2011-04-25T14:06:46 i see 2011-04-25T14:06:47 ty 2011-04-25T14:06:51 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T14:07:16 mcstar: When the contest starts you have ~2 months to code your bot and when you upload something you will get results almost instantly because the bots compete against each other from hour 0. 2011-04-25T14:07:45 ty 2011-04-25T14:07:49 Naktibalda: did you test it? 2011-04-25T14:08:03 yes, I did 2011-04-25T14:08:17 after the 2 months no more modifications are allowed and the contest server will run a final round of games to produce the final results 2011-04-25T14:08:18 Naktibalda: do you have a link? i wanna see how it works 2011-04-25T14:08:19 registration,activation, login,edit works 2011-04-25T14:08:34 no, it's local 2011-04-25T14:08:55 *** berak has joined #aichallenge 2011-04-25T14:09:06 aww 2011-04-25T14:09:08 ok 2011-04-25T14:09:27 Naktibalda: i'm gonna have to resync the db or something 2011-04-25T14:09:30 right? 2011-04-25T14:10:48 I added 2 unique keys where they should exist, removed some unnecessary indexes and deleted 2 duplicated organizations 2011-04-25T14:11:03 so will it work with the current schema? 2011-04-25T14:11:35 yes, it will 2011-04-25T14:12:20 k 2011-04-25T14:12:21 http://hypertriangle.com:13080/ 2011-04-25T14:12:23 let's see 2011-04-25T14:12:53 wait, you can only make an org when you register? 2011-04-25T14:13:09 what about the user profile? 2011-04-25T14:17:58 Where did we end up on the background discussion? Because I imagine a dirt image would look terribly nice 2011-04-25T14:18:33 jaycobcoleman: huh? 2011-04-25T14:18:54 some people didn't like the blue background 2011-04-25T14:19:58 jaycobcoleman: show me a better version then we'll talk 2011-04-25T14:21:41 Alright :) personally it doesn't bother me, but I know some people didn't like it and I think dirt might look good. I'll try to create an image from scratch when I'm not busy and see how that looks 2011-04-25T14:22:36 personally i like it, and i'm not going to work on replacing something that i'm already ok with 2011-04-25T14:26:04 Guess I'm just trying to please everybody :P not always the best decision, but...it's not so much work 2011-04-25T14:32:45 *** jaycobcoleman has quit IRC (Ping timeout: 252 seconds) 2011-04-25T14:34:20 *** janzert has joined #aichallenge 2011-04-25T14:36:36 aichallenge: Gintautas Miselis epsilon * r6ed1dd6 / (96 files in 22 dirs): (log message trimmed) 2011-04-25T14:36:36 aichallenge: Merge branch 'epsilon' of https://github.com/aichallenge/aichallenge into epsilon 2011-04-25T14:36:36 aichallenge: Conflicts: 2011-04-25T14:36:36 aichallenge: website/account_confirmation.php 2011-04-25T14:36:36 aichallenge: website/header.php 2011-04-25T14:36:36 aichallenge: website/process_registration.php 2011-04-25T14:36:37 aichallenge: website/register.php 2011-04-25T14:36:37 aichallenge: Alexandru Stan epsilon * r129bfea / website/css/typo.css : implemented lists css, added strong - http://bit.ly/hcE8MV 2011-04-25T14:39:03 *** Frontier1 has joined #aichallenge 2011-04-25T14:40:14 Frontier: does it automatically move the required files in deploy.path? 2011-04-25T14:40:23 Frontier1: same question ^ 2011-04-25T14:40:26 *** Frontier has quit IRC (Ping timeout: 258 seconds) 2011-04-25T14:41:47 *** chris__0076 has joined #aichallenge 2011-04-25T14:44:43 *** Chris_0076 has quit IRC (Ping timeout: 250 seconds) 2011-04-25T14:47:35 amstan: The files are first compiled and minimized into a directory inside the source tree, then moved over to /visualizer. 2011-04-25T14:47:44 it didn't work for me 2011-04-25T14:47:53 my deploy.path is still empty 2011-04-25T14:48:10 what was the error message? 2011-04-25T14:48:29 Frontier1: no error: http://pastebin.com/FJpNidFX 2011-04-25T14:51:02 you did not specify any task, use "deploy" 2011-04-25T14:51:42 Frontier1: ok.. now it did more things, but my other folder is still empty 2011-04-25T14:52:15 oh.. wait.. error 2011-04-25T14:52:24 it seems like it wants to use cvs? 2011-04-25T14:52:35 /home/alex/Projects/aichallenge/ants/visualizer/build.xml:17: java.io.IOException: Cannot run program "cvs" (in directory "/home/alex/Projects/aichallenge/ants/visualizer/3rdparty"): java.io.IOException: error=2, No such file or directory 2011-04-25T14:53:13 Yes, it loads the Rhino JavaScript engine from Mozilla's CVS 2011-04-25T14:53:46 http://pastebin.com/JaBhrqQj 2011-04-25T14:53:49 *** chris__0076 has quit IRC (Ping timeout: 250 seconds) 2011-04-25T14:55:02 It seems, like ant is not working around a missing JAVA_HOME variable. 2011-04-25T14:55:11 *** delt0r___ has joined #aichallenge 2011-04-25T14:55:43 Frontier1: i can give you root on the test server, you can play around with it in there 2011-04-25T14:57:06 amstan: Does it work if JAVA_HOME is defined? Since I don't know what to do about it access to the test server wont help me much there. 2011-04-25T14:57:07 *** delt0r_ has quit IRC (Ping timeout: 260 seconds) 2011-04-25T14:57:26 idk, what do i define it as? 2011-04-25T14:59:23 Set it to your Java installation directory which contains a bin and lib directory 2011-04-25T15:00:31 Frontier1: i have /usr, where there's a bin and a lib, both with javas in it 2011-04-25T15:01:37 ${java.home}/lib/plugin.jar must exist 2011-04-25T15:01:58 probably do a search for plugin.jar :) 2011-04-25T15:02:36 /usr/lib/jvm/java-6-openjdk/jre/lib/plugin.jar 2011-04-25T15:03:35 http://pastebin.com/jEEhhxsn 2011-04-25T15:04:39 I realize that Ubuntu Java packages are all supposed to work withouth JAVA_HOME defined, but I could not find a solution for the plugin.jar problem. And ant has a predefined java.home directory dereived from the system's JAVA_HOME. Maybe in 10 years the OS developers will have agreed on a new standard. I like the env variable though, since it works with Windows as well. 2011-04-25T15:05:24 don't you have to do 'export JAVA_HOME= ...' if you run it like this? 2011-04-25T15:06:19 adding export doesn't fix it 2011-04-25T15:08:42 try omitting the tailing slash from the path. other than that I don't really know what the problem could be. 2011-04-25T15:09:30 what about sun.plugin.javascript 2011-04-25T15:09:38 am i supposed to have that installed separately or something? 2011-04-25T15:10:18 It is a Java package (namespace) which resides in plugin.jar 2011-04-25T15:10:25 ok 2011-04-25T15:10:46 Frontier1: try #java or something 2011-04-25T15:10:50 If you open plugin.jar you'll see the folder structure 'sun/plungin/javascript/... 2011-04-25T15:10:56 maybe they'll guide you 2011-04-25T15:11:04 or #ant 2011-04-25T15:11:20 nope.. not #ant, lol 2011-04-25T15:12:06 maybe #unix, how do i set env variables ^^ 2011-04-25T15:12:30 i don't think this should be an env variables thing 2011-04-25T15:12:36 most of the time you don't have to touch them 2011-04-25T15:13:02 but most of the time you have to set them up once 2011-04-25T15:13:34 (or they are set up by installed packages) 2011-04-25T15:14:55 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T15:16:25 I guess Ubuntu is a bit isolated there. I have yet to find another OS where the java package does not define JAVA_HOME 2011-04-25T15:25:21 (ok seems to be a Debian thing as well) 2011-04-25T15:26:05 *** berak has joined #aichallenge 2011-04-25T15:29:27 amstan: Ok, give me access to the test server. At least I can try to find out why it wont work even IF java_home is defined. 2011-04-25T15:29:56 I don't really need root access though :p 2011-04-25T15:30:11 Frontier1: some packages are not installed, you'll have to install 2011-04-25T15:30:21 quit 2011-04-25T15:30:25 *** mcstar has quit IRC (Quit: WeeChat 0.3.4) 2011-04-25T15:31:11 *** boegel has joined #aichallenge 2011-04-25T15:32:15 Frontier1: if you do have to install something, please document it, as we'll have to add it to the server-setup.py 2011-04-25T15:39:07 amstan: Ok, first of all I need ant on the test server ^^ and I am not root. 2011-04-25T15:39:15 sudo apt-get ant 2011-04-25T15:39:55 oh, no password? cool 2011-04-25T15:40:15 omg 102 mb of additinal disk space 2011-04-25T15:40:20 it's because you're in the admin group 2011-04-25T15:40:22 that's fine 2011-04-25T15:41:06 It is because of Java and the required fonts and other libs. 2011-04-25T15:41:58 hmm I wonder if 'openjdk-6-jre-headless' is the right JRE for the job 2011-04-25T15:47:21 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T15:48:47 *** berak has joined #aichallenge 2011-04-25T15:55:23 *** boegel has quit IRC (Quit: Leaving) 2011-04-25T15:56:05 *** Accoun has quit IRC () 2011-04-25T15:57:21 *** jaycobcoleman has joined #aichallenge 2011-04-25T16:00:57 *** mcstar has joined #aichallenge 2011-04-25T16:01:03 *** sids_aquarius has quit IRC (Ping timeout: 248 seconds) 2011-04-25T16:01:51 *** jaycobcoleman has quit IRC (Ping timeout: 252 seconds) 2011-04-25T16:06:31 amstan: I cannot find the Sun JDK in the packages list. 2011-04-25T16:06:50 oh man. don't tell me i have to get the other repos now 2011-04-25T16:07:01 I don't know 2011-04-25T16:07:33 *** Accoun has joined #aichallenge 2011-04-25T16:07:57 The default-jdk on the server doesn't come with a plugin.jar, unfortunatly. 2011-04-25T16:08:38 I could try icedtea, but you are using the Sun JDK, correct? 2011-04-25T16:09:11 ah you use openjdk 2011-04-25T16:09:16 apparently i used openjdk 2011-04-25T16:09:25 and there's no sun java in the repo it seems 2011-04-25T16:10:08 so why do YOU have a plugin.jar and the server doesn't? Is it because the server is configured as 'headless'? 2011-04-25T16:10:27 icedtea6-plugin seems to be the package providing plugin.jar 2011-04-25T16:10:37 wait.. no 2011-04-25T16:10:39 nvm 2011-04-25T16:11:46 openjdk-6-jre-lib 2011-04-25T16:12:14 *** salvatore_ has joined #aichallenge 2011-04-25T16:12:51 is installed 2011-04-25T16:13:21 oh man... now what, lol 2011-04-25T16:13:24 yet "find / -name plugin.jar" yields no results 2011-04-25T16:13:38 that's what brings it in 10.10 2011-04-25T16:13:44 it seems like 11.04 is different 2011-04-25T16:13:52 *** sids_aquarius has joined #aichallenge 2011-04-25T16:13:58 the server runs 11.04? 2011-04-25T16:14:04 yes 2011-04-25T16:14:36 I bet it has to do with the headless config, but what do I know about Ubuntu ^^ 2011-04-25T16:15:24 it doesn't, the repository is exactly the same with the desktop 2011-04-25T16:15:36 but it seems they took out the plugin.jar from the new version 2011-04-25T16:15:39 lol man 2011-04-25T16:15:41 what is this plugin.jar again? 2011-04-25T16:15:43 switch to arch 2011-04-25T16:16:07 mcstar: then we'll talk another 3 months to launch the contest 2011-04-25T16:16:10 take* 2011-04-25T16:16:15 :) 2011-04-25T16:16:17 *** salvatore_ has quit IRC (Client Quit) 2011-04-25T16:16:21 plugin.jar contains the Java classes required to run applets (or to compile some) 2011-04-25T16:16:37 Frontier1: ok, what do i search for on google? 2011-04-25T16:16:56 *** berak has quit IRC (Read error: Connection timed out) 2011-04-25T16:17:23 lol, no idea. we could try to fix the issue with openjdk, try icedtea instead or grap some repo for sun jdk :p 2011-04-25T16:17:34 *grab 2011-04-25T16:18:52 *** berak has joined #aichallenge 2011-04-25T16:18:54 isn't it oracle now also? 2011-04-25T16:19:11 I'll try icedtea now. Yes, it is Oracle now, you are right 2011-04-25T16:19:15 Frontier1: you installing stuff? dpkg is locked 2011-04-25T16:19:26 i want to install apt-file, search for the jar that way 2011-04-25T16:19:53 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 3.6.16/20110323143040]) 2011-04-25T16:19:54 ok i'm uninstalling openjdk 2011-04-25T16:20:18 *** mcstar has quit IRC (Quit: WeeChat 0.3.4) 2011-04-25T16:20:30 *** mcstar has joined #aichallenge 2011-04-25T16:20:59 I installed apt-file 2011-04-25T16:21:55 *** berak has quit IRC (Client Quit) 2011-04-25T16:22:58 apt-file search plugin.jar 2011-04-25T16:23:00 pick one 2011-04-25T16:23:17 icedtea-plugin seems interesting 2011-04-25T16:23:21 now I install icedtea-plugin. 2011-04-25T16:24:21 it may behave differently from openjdk and the oracle jdk. 2011-04-25T16:24:40 it just installed firefox, lol 2011-04-25T16:28:00 dude.. 2011-04-25T16:30:30 *sigh* it provides a plugin.jar, but not in a lib directory. I just assumed an Oracle JDK to be installed. Looks like that was a bad idea. 2011-04-25T16:33:29 amstan: There is no way I could possibly find plugin.jar with the plentora of operating systems and Java runtimes. I should make the full path to plugin.jar a property for the build process defaulting to /lib/plugin.jar. 2011-04-25T16:40:18 amstan: wtf? icedtea doesn't use sun.javascript, but sun.applet instead. 2011-04-25T16:40:49 oh man.. java is so much fun, lol 2011-04-25T16:41:52 ok, I'm supposed to use "netscape.javascript.JSObject", but I remember that I wasn't able to use Rhino and this together. 2011-04-25T16:46:55 *** Mathnerd314 has joined #aichallenge 2011-04-25T16:48:04 I'll have to reboot to Linux to investigate that further. I changed a lot since then. Maybe it works now. 2011-04-25T16:50:24 *** Frontier1 has quit IRC (Read error: Connection reset by peer) 2011-04-25T16:52:24 *** Frontier has joined #aichallenge 2011-04-25T17:07:19 aichallenge: Marco Leise epsilon * ra118cf0 / (3 files in 2 dirs): visualizer: made path to plugin.jar configurable - http://bit.ly/genOrv 2011-04-25T17:11:52 amstan: I just ran "ant -Djava.plugin=/usr/share/icedtea-web/plugin.jar jar" and it compiled sucessfully. Maybe you can run ant deploy now. JAVA_HOME is no longer needed. 2011-04-25T17:13:04 If you don't want to type it again and again, you can create a local.properties file with "java.plugin: " as its content. 2011-04-25T17:15:13 amstan: what is a widget? 2011-04-25T17:15:31 Frontier: so we can embed it in other pages, like the main page 2011-04-25T17:16:50 ah ok. I should look over it then. I don't think there is a way to load a game without the game_id on the address line right now 2011-04-25T17:20:47 hmm. looks fine. it should be usable on the front page as long as it is used as in visualizer.php 2011-04-25T17:21:16 seems to work with ant jar -Djava.plugin=/usr/lib/jvm/java-6-openjdk/jre/lib/plugin.jar too 2011-04-25T17:21:44 maybe we can just replace it with `locate plugin.jar` or something similar 2011-04-25T17:22:06 maybe not.. hmm 2011-04-25T17:22:07 that seems too fuzzy to me. 2011-04-25T17:22:12 yeah 2011-04-25T17:22:38 there may be different plugin.jars and on a slow disk it could take several minutes 2011-04-25T17:22:50 it seems to be pretty different in 10.10 2011-04-25T17:22:58 locate uses a database, doesn't actually do find|grep 2011-04-25T17:23:14 hmm ok then :) 2011-04-25T17:26:02 *** Eruonen has quit IRC (Ping timeout: 258 seconds) 2011-04-25T17:28:25 amstan: I had to install some packages: ant, cvs, default-jdk and icedtea-plugin. Since you already had a plugin.jar the package list may vary depending on the Ubuntu version used :) 2011-04-25T17:28:51 *** Eruonen has joined #aichallenge 2011-04-25T17:29:33 contestbot: learn websitedeps as ant cvs default-jdk icedtea-plugin markdown 2011-04-25T17:29:33 amstan: Job's done. 2011-04-25T17:32:58 amstan: Do you checkout and compile the web site on the target machine? 2011-04-25T17:33:11 amstan: just add the packages needed to the appropriate list in server_setup.py 2011-04-25T17:33:20 janzert: yes, that's why i'm saving it 2011-04-25T17:33:28 i g2g in a minute to eat 2011-04-25T17:33:31 ok 2011-04-25T17:36:02 *** jaycobcoleman has joined #aichallenge 2011-04-25T17:46:09 *** jaycobcoleman has quit IRC (Quit: Page closed) 2011-04-25T17:46:30 *** mceier has quit IRC (Quit: leaving) 2011-04-25T17:46:32 oh i see on the github wiki that no common lisp starter package is provided 2011-04-25T17:46:38 is this so? 2011-04-25T17:51:40 mcstar: look at this list https://github.com/aichallenge/aichallenge/tree/epsilon/ants/dist/starter_bots 2011-04-25T17:52:28 ty 2011-04-25T17:56:17 *** sids_aquarius has quit IRC (Ping timeout: 246 seconds) 2011-04-25T18:13:03 let me ask a stupid question 2011-04-25T18:13:16 since u have the source code of the bots 2011-04-25T18:13:33 isnt it easier to check for some prob. hazardous function calls 2011-04-25T18:13:38 at compile time 2011-04-25T18:13:44 than to use sandboxing? 2011-04-25T18:13:52 or before compile time sry 2011-04-25T18:22:08 *** Blkt has quit IRC (Read error: Operation timed out) 2011-04-25T18:23:08 *** Blkt has joined #aichallenge 2011-04-25T18:25:55 *** amstan_ has joined #aichallenge 2011-04-25T18:25:55 *** ChanServ sets mode: +o amstan_ 2011-04-25T18:31:09 mcstar: you would have to do it for separately for every language and for many (most? all?) languages I'm pretty sure it is equivalent to the halting problem 2011-04-25T18:36:42 im not an expert programmer but im pretty sure that if you exclude certain calls, for example open(file|socket) u can greatly cut down harmful possibilities 2011-04-25T18:36:54 mcstar: yes, but there's 20 languages 2011-04-25T18:37:21 easier to just make a sandbox that does exactly this, but for all of them 2011-04-25T18:37:29 mcstar: try googling for a discussion on making it safe to run untrusted code in python for a sample of how complex it can be 2011-04-25T18:37:53 ok im not trying to be a wiseguy here 2011-04-25T18:37:56 just asking :) 2011-04-25T18:37:58 and why the restricted exectution module was removed from the standard library 2011-04-25T18:38:21 i just read on the wiki that sandboxing can be slow 2011-04-25T18:38:24 np, it really is quite difficult though 2011-04-25T18:39:00 the current method doesn't incur any overhead generally 2011-04-25T18:39:24 i.e. the sandboxing info on the wiki is quite old 2011-04-25T18:39:36 ok 2011-04-25T18:40:13 are u participating or just making the challenge possible? 2011-04-25T18:40:34 just helping set it up 2011-04-25T18:40:54 not participating for sure, I'm disqualified by virtue of having access to the servers 2011-04-25T18:41:02 ok right 2011-04-25T18:41:07 havent thought of that :) 2011-04-25T18:46:11 btw ur halting problem only applies to asm w/o virtual memory manager i think 2011-04-25T18:46:53 if the language somehow allows self modification of code I think you end up at the halting problem 2011-04-25T18:47:17 well i think the OS is already is a sandbox 2011-04-25T18:47:26 so any language that includes an exec 2011-04-25T18:47:30 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-04-25T18:47:59 u cant access the parts of memory responsible for IO 2011-04-25T18:48:05 only if the OS allows you 2011-04-25T18:48:23 it has nothing to do with turing completeness i think 2011-04-25T18:48:42 I thought we were talking about scanning source code for malicious actions? 2011-04-25T18:48:48 yes 2011-04-25T18:49:18 *** Blkt has quit IRC (Ping timeout: 240 seconds) 2011-04-25T18:49:23 but a process is managed by the operating system as i understand it 2011-04-25T18:49:45 I'm saying if the language allows self modified code to run then you can't tell if the program will perform a malicious action without solving the halting problem 2011-04-25T18:50:05 im saying that you can :) 2011-04-25T18:50:28 u cant make heads or tails of what its trying to do 2011-04-25T18:50:40 but to make harm it must ask the OS for certain IO 2011-04-25T18:50:42 you can go outside of the program and prevent it (that is what a sandbox does) but you can't just scan it before hand and say yes it is safe 2011-04-25T18:51:09 but you can't know what the program will do be simply scanning it 2011-04-25T18:51:27 that is the gist of the halting problem 2011-04-25T18:52:16 but i know if it accesses the outside world, and i can inspect that system call even in the source code 2011-04-25T18:52:40 it must use glibc for example 2011-04-25T18:52:44 the system call may not be apparent in the original source code 2011-04-25T18:53:01 it can build it by using the self modifying code 2011-04-25T18:53:16 anyway I've gotta run 2011-04-25T18:53:20 ok 2011-04-25T18:53:26 see you around 2011-04-25T18:53:33 i think i see what u mean 2011-04-25T18:53:34 bb 2011-04-25T18:55:07 you might be interested in looking up a competition that ran (I believe) last year where the goal was to make a program that would black out portions of an image but still surreptitiously leave the original information embedded in the image 2011-04-25T18:55:20 and make it so the source of the program looked completely innocent 2011-04-25T18:55:44 whats it called? 2011-04-25T18:55:49 unfortunately I don't remember the name or much else about the competition 2011-04-25T18:56:03 once i saw a C source 2011-04-25T18:56:17 which printed a thanksgiving song 2011-04-25T18:56:18 I mostly remember a write up on the winners 2011-04-25T18:56:28 but u couldnt find it in the source 2011-04-25T18:56:45 ill try to find it ty 2011-04-25T18:57:32 yeah, there are lots of obfuscated source competitions the really interesting part of this one was that source code was meant to look completely innocent and correct 2011-04-25T18:57:44 * janzert really goes away 2011-04-25T19:09:56 i've seen an ioccc entry that used lots of #defines before what appeared to be a java hello world program 2011-04-25T19:10:07 but it actually drew a smiley face on the screen 2011-04-25T19:10:47 :) 2011-04-25T19:10:50 like this? 2011-04-25T19:10:53 :) 2011-04-25T19:11:03 no, a large ascii art thing 2011-04-25T19:11:08 *** amstan_ has quit IRC (Ping timeout: 276 seconds) 2011-04-25T19:11:13 joking 2011-04-25T19:11:35 i saw a line of C 2011-04-25T19:11:39 which printed itslef 2011-04-25T19:12:20 these things are good for they point out how little i know of the language 2011-04-25T19:12:25 *** FireFly has quit IRC (Quit: swatted to death) 2011-04-25T19:12:51 im not a programmer 2011-04-25T19:13:21 havent really decided to participate in the contest 2011-04-25T19:14:35 i intended originally to help run the contest 2011-04-25T19:14:45 then i got behind on just about everything ever 2011-04-25T19:15:07 so i am in this channel fairly infrequently 2011-04-25T19:16:11 interesting, this applies to me as well 2011-04-25T19:16:42 the behind and the infrequently part 2011-04-25T19:18:26 its my lucky day 2011-04-25T19:18:41 i found the contest janzert talked about 2011-04-25T19:18:58 god bless google 2011-04-25T19:25:17 *** needsch has quit IRC (Quit: Leaving.) 2011-04-25T19:25:46 *** _0x47 <_0x47!~slash@91.37.227.86> has quit IRC (Quit: Leaving.) 2011-04-25T19:27:55 *** sigh has joined #aichallenge 2011-04-25T19:35:18 *** nann has joined #aichallenge 2011-04-25T19:39:44 *** Apophis_ has quit IRC (Ping timeout: 276 seconds) 2011-04-25T19:48:05 *** Kingpin13 has quit IRC (Quit: money can't buy you love, but it can if you have enough) 2011-04-25T19:48:27 *** Kingpin13 has joined #aichallenge 2011-04-25T19:56:06 ai-contest.com says that account creation is closed 2011-04-25T19:56:26 yes, since the contest is currently closed 2011-04-25T19:56:52 the forum accounts are separate from the contest accounts btw 2011-04-25T19:57:16 *** amriedle has joined #aichallenge 2011-04-25T19:57:16 i was reading Ants Five Minute Quickstart Guide 2011-04-25T19:57:30 and the first thing is to register an account 2011-04-25T19:57:43 yes, but the contest isn't open yet 2011-04-25T19:57:59 actually the current site is still the planetwars site 2011-04-25T19:58:35 ok, so everything i need is on github 2011-04-25T19:58:53 everything that exists is on github ;) 2011-04-25T19:59:44 that may or may not be everything you need 2011-04-25T20:00:00 when does the contest take place? 2011-04-25T20:00:27 after we finish writing everything needed to run it... 2011-04-25T20:00:38 hopefully that occurs within the next few weeks 2011-04-25T20:00:57 *** Redgis has quit IRC (Quit: ... mains libres) 2011-04-25T20:01:24 then starts the 2 months period of continuous contests 2011-04-25T20:01:30 right? 2011-04-25T20:01:47 yes, although 2 month length could be subject to change 2011-04-25T20:02:01 although I don't think it will at this point 2011-04-25T20:02:26 before u finish up everything the bots cant compete online 2011-04-25T20:02:52 i guess without registration or something i cant try my code 2011-04-25T20:03:11 you can run games locally 2011-04-25T20:03:42 this was that i didnt get 2011-04-25T20:03:51 i thought i can already upload the code or something 2011-04-25T20:04:00 and try if it works 2011-04-25T20:04:07 not until the contest is live 2011-04-25T20:04:10 ok 2011-04-25T20:04:48 is there a mailing list where u send out notifications? 2011-04-25T20:05:14 no mailing list, just the forums 2011-04-25T20:05:46 I believe there is a way to get a newsfeed (rss or atom) from them though 2011-04-25T20:06:07 ill check 2011-04-25T20:06:36 amstan would be able to tell you how 2011-04-25T20:06:55 then amstan pls tell 2011-04-25T20:07:18 he's not actually here at the moment :) 2011-04-25T20:07:29 i see that now 2011-04-25T20:07:46 not a bad idea thou 2011-04-25T20:07:57 after 2 oclock 2011-04-25T20:14:24 well 2011-04-25T20:14:26 good nite 2011-04-25T20:14:29 *** mcstar has quit IRC (Quit: WeeChat 0.3.4) 2011-04-25T20:29:31 *** bhasker has quit IRC (Quit: bhasker) 2011-04-25T20:44:02 *** amstan_ has joined #aichallenge 2011-04-25T20:44:02 *** ChanServ sets mode: +o amstan_ 2011-04-25T20:49:19 contestbot: seen j3camero 2011-04-25T20:49:19 amstan_: j3camero was last seen in #aichallenge 1 day, 20 hours, 11 minutes, and 42 seconds ago: computerwiz_222 haha yes. I remember fiddling with those damn lines before the last launch. Grr. 2011-04-25T20:49:25 *** amstan_ is now known as amstan 2011-04-25T20:58:04 *** Chris_0076 has joined #aichallenge 2011-04-25T20:59:47 Yo yo 2011-04-25T21:00:09 *** amstan_ has joined #aichallenge 2011-04-25T21:00:10 *** ChanServ sets mode: +o amstan_ 2011-04-25T21:00:48 *** amstan_ has quit IRC (Remote host closed the connection) 2011-04-25T21:01:32 *** chris__0076 has joined #aichallenge 2011-04-25T21:01:52 *** Chris_0076 has quit IRC (Read error: Operation timed out) 2011-04-25T21:02:07 *** McLeopold has joined #aichallenge 2011-04-25T21:03:08 *** amstan_ has joined #aichallenge 2011-04-25T21:03:08 *** ChanServ sets mode: +o amstan_ 2011-04-25T21:03:18 Sup amstan 2011-04-25T21:03:23 yo 2011-04-25T21:03:39 I am contemplating trying Minecraft. Good idea? Bad idea? 2011-04-25T21:03:51 j3camero: just wait until after we release the contest 2011-04-25T21:03:53 *** amstan has quit IRC (Ping timeout: 250 seconds) 2011-04-25T21:03:57 *** amstan_ is now known as amstan 2011-04-25T21:04:01 then it's a good idea 2011-04-25T21:04:57 *** chris___0076 has joined #aichallenge 2011-04-25T21:05:10 *** chris__0076 has quit IRC (Read error: Connection reset by peer) 2011-04-25T21:05:38 amstan: I hope to look at conflicts and leaderboards later tonight 2011-04-25T21:05:51 McLeopold: leaderboards, decent 2011-04-25T21:05:55 McLeopold: but what are conflicts? 2011-04-25T21:06:39 *** computerwiz_222 has joined #aichallenge 2011-04-25T21:07:13 I thought I saw an email about merge conflicts 2011-04-25T21:07:28 *** Eruonen has quit IRC () 2011-04-25T21:07:43 McLeopold: email wasn't, it was just a conflict from a comitter, he merged and didn't delete the conflict lines from the message 2011-04-25T21:07:48 there's no conflicts 2011-04-25T21:08:13 it was about this: https://github.com/aichallenge/aichallenge/pull/67 2011-04-25T21:08:23 ah 2011-04-25T21:09:14 *** Kingpin13 is now known as KP13 2011-04-25T21:09:15 Oh man. I am psyched for leaderboards. 2011-04-25T21:09:26 *** KP13 is now known as Kingpin13 2011-04-25T21:10:37 i think i'm going to work on visualizer integration 2011-04-25T21:11:50 j3camero: 2 things from you, do we have money? and where are we hosting? 2011-04-25T21:12:10 j3camero: we should get 11.04 on our servers btw 2011-04-25T21:12:30 amstan: does that auto-config script work on anything else than 11.04? 2011-04-25T21:12:39 it was made on 10.10 2011-04-25T21:12:44 decent 2011-04-25T21:12:48 it only works on 10.10 so far as I've tested :P 2011-04-25T21:13:21 server_setup.py works nicely on 11.04 2011-04-25T21:13:40 didn't try the worker yet, i'll need to know more details about how it works 2011-04-25T21:13:46 yeah, I would hope that there wouldn't be much change needed for 11.04 2011-04-25T21:14:29 the worker setup is in flux right now as I get chroots integrated 2011-04-25T21:14:49 janzert: tell me when you want me to test it, i'll be ready 2011-04-25T21:15:14 amstan: it would be nice if the wiki could render ant map text in a custom section to a png or insert a canvas. Then we could make the pages easy to explain the game. 2011-04-25T21:15:28 amstan is going to test the hall of aichallenge on a sempron :D 2011-04-25T21:15:42 well, I think I'm going to be able to finish up the chroot setup tonight, then I have to rewrite sandbox.py to actually do sandboxing again 2011-04-25T21:15:42 computerwiz_222: it's not like our pw server was any better 2011-04-25T21:18:17 also after doing a bit of browsing on current server pricing I'm a little less leery about using Ec2 for all our server (including the main server) 2011-04-25T21:19:09 the main thing that still concerns me is instance stability on Ec2 for the main server, especially if we're going to jump right on 11.04 2011-04-25T21:20:20 The main web server does need to be pretty solid. I have no problem having it not on EC2. 2011-04-25T21:20:30 Especially given the Ec2 fail as of late. 2011-04-25T21:20:30 i'm ok with that, i was all for using dedicated servers before 2011-04-25T21:20:55 Hell, we could even host it with the CSC. Not sure that's actually more stable though... 2011-04-25T21:21:28 j3camero: Well, the EC2 failure was EBS. 2011-04-25T21:21:41 Local storage was minimally affected. 2011-04-25T21:21:51 CSC has more frequent outages, really. 2011-04-25T21:22:00 (I say that as the previous sysadmin.) 2011-04-25T21:22:14 jbroman: yeah, the recent outage shouldn't have been much of a problem for us 2011-04-25T21:22:17 I was really impressed by the two days of "scheduled power outage" in the MC building a few months back. 2011-04-25T21:22:20 *** jaycobcoleman has joined #aichallenge 2011-04-25T21:23:16 the version of ubuntu we used in the last contest would randomly hard lock up about every ~6 hours, that's the part that worries me 2011-04-25T21:23:34 especially if we're jumping right into an unproven version 2011-04-25T21:24:10 I assume that we will be able to switch OS versions pretty easily. Is it just a matter of updating all the package names and so on? 2011-04-25T21:24:26 But that was just workers? 2011-04-25T21:24:27 depends on what has changed in the OS :/ 2011-04-25T21:24:52 McLeopold: for planetwars yes and isn't really a problem for workers to do so 2011-04-25T21:25:05 it would be a problem if the main server was though 2011-04-25T21:26:23 as an example of changes between OS versions that would effect us, the location of chroot config files seems to have changed with every version of ubuntu for the three previous versions leading up to and including 10.10 2011-04-25T21:26:39 so far me and Frontier had a huge problem with just the difference between 10.10 and 11.04 2011-04-25T21:26:43 I haven't seen any reference that it changes in 11.04 though 2011-04-25T21:26:48 it's a totally different package that we need and stuff like that 2011-04-25T21:27:30 i just think we should do stuff on 11.04 2011-04-25T21:27:54 the beta was out for a while, and we'll regret if we don't have the latest packages for something because we're not having the most recent ubuntu 2011-04-25T21:28:13 I just like the extra 6 months of stabilizing :} 2011-04-25T21:28:56 I've had no problems at all with Meerkat maverick :P If that's at all comforting 2011-04-25T21:29:23 if 11.04 proves itself stable as released though I don't really have a problem with jumping to it 2011-04-25T21:29:24 or those two words switched 2011-04-25T21:29:46 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-04-25T21:30:23 What reason is there to go to 11.04? 2011-04-25T21:30:37 the benefit is newer language versions 2011-04-25T21:30:47 at least I believe that is the main benefit 2011-04-25T21:31:09 and of course that doesn't matter so much for older stable languages 2011-04-25T21:31:12 Why not pick an older LTS version that has had some time to work out all the kinks? 2011-04-25T21:31:39 that caused more problems because of old language versions in the last contest 2011-04-25T21:32:07 LTS is overrated if you ask me 2011-04-25T21:32:13 everyone wanted their new language version because the one included was "ancient" 2011-04-25T21:32:22 lol 2011-04-25T21:32:54 *** smellyhippy has quit IRC (Ping timeout: 258 seconds) 2011-04-25T21:33:19 *** McLeopold has quit IRC (Ping timeout: 252 seconds) 2011-04-25T21:33:24 j3camero: What are you doing here? Go buy minecraft 2011-04-25T21:33:25 even the haskellers which from the outside I would have thought was a fairly stable language at this point 2011-04-25T21:33:30 The people who were unable to express their algorithm in Python 2.6? :-) 2011-04-25T21:33:41 jaycobcoleman: don't tempt him 2011-04-25T21:33:46 we didn't have 2.6 :P 2011-04-25T21:33:57 amstan: no no I ust bought the game. But I have self control don't worry. 2011-04-25T21:33:59 :) 2011-04-25T21:34:03 11.04 comes with 2.7, everyone will be overjoyed 2011-04-25T21:34:12 :) 2011-04-25T21:34:13 j3camero: you can come to my server later 2011-04-25T21:34:27 j3camero: you think you have self control... but then you play minecraft 2011-04-25T21:34:37 this ^ 2011-04-25T21:35:05 for some reason minecraft always seems more interesting during exams 2011-04-25T21:35:19 or really anytime you should be doing something 2011-04-25T21:35:22 like sleeping 2011-04-25T21:36:19 so.. i'm sitting idle hanging out amstan's place, are there any housekeeping jobs that anyone needs done? 2011-04-25T21:37:00 So has anyone played through portal 2? 2011-04-25T21:37:06 jaycobcoleman: I have. 2011-04-25T21:37:16 Have you played co-op? 2011-04-25T21:37:49 Yes. 2011-04-25T21:38:26 Blast. I want to play it, but not with a completely random person. Not that I know you 2011-04-25T21:39:06 Ah, well. It's good; you'll enjoy it. 2011-04-25T21:39:15 MOOON!!! 2011-04-25T21:39:54 *** smellyhippy has joined #aichallenge 2011-04-25T21:40:26 yeah, just wish the co-op play had been a little longer 2011-04-25T21:41:11 On the other hand there was less of a story to co-op and trying to find five hours (how long it took me) to play with a friend is tricky enough. 2011-04-25T21:41:12 played it with my brother and there were times where I've never laughed that hard in a game before :) 2011-04-25T21:41:23 aichallenge: Alexandru Stan epsilon * r4c87c77 / (.gitignore INSTALL.md): added more info to the INSTALL instructions, added stuff to gitignore - http://bit.ly/i5LVWr 2011-04-25T21:41:26 The single-player was hilarious. 2011-04-25T21:42:19 I was actually discussing it today, but I don't think there can be as strong of a story line in a co-op mode as you can get in single player 2011-04-25T21:42:28 contestbot: forget newdesign 2011-04-25T21:42:28 amstan: Job's done. 2011-04-25T21:42:28 So what does everyone think of the new design? 2011-04-25T21:42:29 http://hypertriangle.com:13080/ 2011-04-25T21:42:36 I really liked the opening sequence in the single player 2011-04-25T21:42:47 janzert: I had the same discussion with a person that said "not adding co-op is because they hate people with friends" (or similar) 2011-04-25T21:43:23 I would really like ideas from people. I'm planning on writing up some strategy guides, but would like some input as to what theories should be covered. I plan on splitting them into large groups like pathfinding, decision making, etc. and then sub-groups of difficulty; beginner, intermediate, advanced. 2011-04-25T21:43:47 the thing I realized playing the co-op was that I just couldn't concentrate on the story as much because I was involved with the person I was playing with 2011-04-25T21:43:55 and the site is looking beautiful 2011-04-25T21:43:56 Frontier: you still here? i might need some guidance 2011-04-25T21:45:11 the new design is really coming together nice 2011-04-25T21:45:29 good to hear 2011-04-25T21:46:00 I can't wait to see how the visualizer embedded on the front page looks 2011-04-25T21:48:08 janzert: working on it with amstan 2011-04-25T21:48:42 * janzert nods 2011-04-25T21:49:40 problem is that frontier is not here 2011-04-25T21:50:03 *** McLeopold has joined #aichallenge 2011-04-25T21:51:19 nawww 2011-04-25T21:51:21 denied for gsoc 2011-04-25T21:51:42 antimatroid: yeah, same 2011-04-25T21:51:44 i thought i might have had a decent shot with that 2011-04-25T21:51:51 ah well 2011-04-25T21:51:58 to both 2011-04-25T21:51:59 sorry :P 2011-04-25T21:52:09 :) s'all good 2011-04-25T21:52:16 there is always next year.. i was helping amstan compile his entry - too bad he didn't get in 2011-04-25T21:52:27 it means i get to do something more interesting to me for my honours thesis and with something i can bare programming for 2011-04-25T21:52:29 but no $5k 2011-04-25T21:52:39 so i got this so far.. http://home.hypertriangle.com/~alex/aichallenge/website/visualizer.php?game_id=0 2011-04-25T21:52:39 sigh: was the new vision code put into starter bots, sample bots, or both? 2011-04-25T21:52:54 http://home.hypertriangle.com/~alex/aichallenge/website/visualizer.php?game_id=1 2011-04-25T21:52:56 there we go 2011-04-25T21:52:59 0 didn't like me 2011-04-25T21:53:05 jaycobcoleman: it was just put into the python and c++ starter bots i think 2011-04-25T21:53:14 but all old bots should work fine the way they did 2011-04-25T21:53:15 amstan: that is decent :D 2011-04-25T21:54:03 amstan: that looks good, although when i full screen it the right menus are slightly hanging off the screen 2011-04-25T21:54:24 jaycobcoleman: I know that mcleopold put it into the starter packs... I don't think the sample bots use vision 2011-04-25T21:54:27 antimatroid: yes, good point 2011-04-25T21:54:39 I would have to check though, but I'm sure you can do that :) 2011-04-25T21:54:43 antimatroid: yeah, i guess, that's because i have some css to always display the scrollbar 2011-04-25T21:54:47 i should fix this with frontier 2011-04-25T21:55:01 sigh: You're definitely right 2011-04-25T21:55:03 yep, just noting it 2011-04-25T21:55:07 I forgot about that 2011-04-25T21:55:08 i much prefer the full screen to the other 2011-04-25T21:55:09 j3camero: http://home.hypertriangle.com/~alex/aichallenge/website/visualizer.php?game_id=1 2011-04-25T21:56:00 sigh: can the engine take a parameter from the map file about which type of symmetry it's using? 2011-04-25T21:56:08 i'll update the map generator and maps for that at some point 2011-04-25T21:56:12 the coding behind it is flexible too, the games are stored inside games 2011-04-25T21:56:22 antimatroid: not at the moment, no 2011-04-25T21:56:29 nothing is ready for it anyway :) 2011-04-25T21:56:37 http://home.hypertriangle.com/~alex/aichallenge/website/games/0/1.replay 2011-04-25T21:59:27 *** McLeopold1 has joined #aichallenge 2011-04-25T21:59:50 *** McLeopold has quit IRC (Quit: Page closed) 2011-04-25T22:00:06 *** McLeopold1 is now known as McLeopold 2011-04-25T22:01:32 McLeopold: http://home.hypertriangle.com/~alex/aichallenge/website/visualizer.php?game_id=1 2011-04-25T22:01:34 McLeopold1: are you okay with us giving people first seen non-water (land) square information too? 2011-04-25T22:01:38 to keep stocha happy :P 2011-04-25T22:01:51 -1 2011-04-25T22:01:53 antimatroid: you guys should be looking to finalize the game mechanics 2011-04-25T22:02:03 the game mechanics are finalised 2011-04-25T22:02:06 i'm not happy with you changing stuff like protocol so late 2011-04-25T22:02:17 we're still in devlopment stage yes? 2011-04-25T22:02:20 *** kaemo has quit IRC (Ping timeout: 246 seconds) 2011-04-25T22:02:28 we're not even up to beta testing 2011-04-25T22:02:35 still... 2011-04-25T22:02:39 forgive my git incompetence, but how to I overwrite any files I've altered for the updated ones? 2011-04-25T22:02:40 *** kaemo0 has quit IRC (Ping timeout: 248 seconds) 2011-04-25T22:02:55 amstan: we don't plan to do anything to cause starter bot re-writes 2011-04-25T22:03:11 amstan: I think the actual game mechanics are as finalized as they can be until a beta test (/me not counting protocol as game mechanics) 2011-04-25T22:03:27 it kinda keeps us unfocused 2011-04-25T22:03:44 jaycobcoleman: first, you must understand that reverting updates will affect the entire repo, not just specific directories 2011-04-25T22:04:00 that's fine, I've been keeping my things elsewhere 2011-04-25T22:04:19 jaycobcoleman: if you haven't changed much, use "git checkout " 2011-04-25T22:04:26 you can do files 1 at a time 2011-04-25T22:04:28 i think stocha has a fair argument, he didn't care for removing the implicit information that we did for food, but he's asking for that information, i think we should just provide it 2011-04-25T22:04:37 starter bots could easily mark seen/unseen information then 2011-04-25T22:04:43 can't do something like checkout *? 2011-04-25T22:04:52 it wont break any pre-existing bots 2011-04-25T22:05:13 jayacobcoleman: probably you can 2011-04-25T22:05:27 amstan: i too would also like to get these changes done sooner rather than later though :P 2011-04-25T22:06:09 aichallenge: Alexandru Stan epsilon * r209b6e6 / (.gitignore website/visualizer.php): resized visualizer to fit the new design - http://bit.ly/go0PyF 2011-04-25T22:06:16 antimatroid: we've looked at that before and it was 10x more text to parse 2011-04-25T22:06:32 wasn't that for displaying currently visible land? 2011-04-25T22:06:45 maybe it was... 2011-04-25T22:06:51 otherwise i can just tell stocha no dice 2011-04-25T22:07:32 i could easily give the information he's asking for in my bot 2011-04-25T22:07:38 just mark a square as seen in ones vision code 2011-04-25T22:07:48 we could make vision code a requirement of starter bots and not send foor nor land 2011-04-25T22:07:58 i think that might be easiest 2011-04-25T22:08:42 when I was benchmarking my java bot, string parsing was taking most of the processing time 2011-04-25T22:09:11 contestbot: later tell Stocha no dice on first seen land information (we're concerned about the size of input to bots, particularly when being sent for tcp games), but we are willing to make it a requirement in the starter bots that the information is calculated without any competitor programming necessary 2011-04-25T22:09:11 antimatroid: As you wish. 2011-04-25T22:09:41 I forgot about tcp, so yeah, let's not do it 2011-04-25T22:10:03 So...would you be okay with removing 'r' lines as well? Can't we calc that? 2011-04-25T22:10:04 i do think we should kindly ask people to update the starter bots with the information 2011-04-25T22:10:17 if some of the less used starter packages don't get updated, i don't really care 2011-04-25T22:10:30 you can calculate that but it's significantly more work 2011-04-25T22:10:33 McLeopold. Says I'm ahead 1 commit, then listed a few files. I did git checkout * and when I tried again it still said ahead 1 commit but didn't list any files. The vision code seems to be the same as before though and I know that was updated 2011-04-25T22:10:34 i really really want to keep those 2011-04-25T22:11:08 working out the r lines yourself consists of tracking what is new and old food information after passing state information and deleting any visible old food information 2011-04-25T22:11:13 I like the 'r' lines as weel 2011-04-25T22:11:13 *well 2011-04-25T22:11:14 jaycobcoleman: you can go "git reset" I think. amstan may be able to give a better command 2011-04-25T22:11:17 working out seen squares is a hell of a lot easier 2011-04-25T22:12:12 jaycobcoleman: if you ahead by 1 commit then you have a local commit 2011-04-25T22:12:32 antimatroid: loop through previous food locations and then check visibility, if not food and visible, then remove 2011-04-25T22:12:49 jaycobcoleman: the others will laugh at me for this, but sometimes if i really mess things up locally, i just grab my changes, delete everything, grab it again and start over :P 2011-04-25T22:13:11 antimatroid: ha ha (voice of nelson) 2011-04-25T22:13:17 haha I think I'll take your method animatroid 2011-04-25T22:13:18 :) 2011-04-25T22:13:22 McLeopold: I think you mean if food and visible 2011-04-25T22:13:30 but what if it's a food location now? 2011-04-25T22:13:34 I've done that a few times as I've been getting to know git :) 2011-04-25T22:13:37 you need to track the new and old food information 2011-04-25T22:13:42 antimatroid: if not reported as food this turn and visible 2011-04-25T22:13:49 that's quite a bit of effort as opposed to just marking a square as not food if you receive an 'r' line 2011-04-25T22:14:53 but would be consistent with not sending 'l' lines 2011-04-25T22:15:30 *** kaemo0 has joined #aichallenge 2011-04-25T22:15:33 *** kaemo has joined #aichallenge 2011-04-25T22:16:08 meh, I'm done with this argument. It honestly isn't that difficult either way. Just make a decision soon :P 2011-04-25T22:16:10 i don't see how that's consistent? 2011-04-25T22:16:17 antimatroid: are you distinguishing between food that is visible and not visible? 2011-04-25T22:16:32 i'm giving information about visible squares having food or not 2011-04-25T22:16:55 and non-visible squares that don't have food? 2011-04-25T22:17:04 i mean *do* have food 2011-04-25T22:17:11 lol yeah i'm kind of glad i'm not bothering with this starter anymore! 2011-04-25T22:17:13 no r lines or f lines are given for a square that isn't visible 2011-04-25T22:17:29 jmcarthur: the change is really trivial 2011-04-25T22:17:37 like not clearing food information and handling an r line 2011-04-25T22:17:46 right, but you're given an 'f' line, then that space goes out of vision, do you flag it differently? 2011-04-25T22:18:02 i stop assuming it's no longer there 2011-04-25T22:18:14 janzert: i was just sifting through some old forum posts and I saw that you had keepalives enabled on the server. I'm a bit of a server noob.. why would you need keepalives? I am curious as to their application. 2011-04-25T22:18:16 if food is there when it goes out of vision, you now have the option of assuming whether it stays or goes 2011-04-25T22:18:44 why not just store the last time food was seen there? 2011-04-25T22:18:47 and you don't have to implicitly work out the state of a visible square once it comes back into range 2011-04-25T22:18:59 *** bhasker has joined #aichallenge 2011-04-25T22:19:15 yeah, but what do you do in the data structure when it goes out of vision? 2011-04-25T22:19:17 if you always report the last time food was seen in a particular place you don't even have to distinguish between visible and invisible 2011-04-25T22:19:28 jmcarthur: you can, but the different is processing an r line vs distinguishing between new and old food information after processing input and calculating visible squares so that you can delete visible squares with just old food information 2011-04-25T22:19:52 depending on what assumption you want to make about food out of range 2011-04-25T22:19:58 huh? 2011-04-25T22:20:08 if you want to assume food that goes out of range is still food 2011-04-25T22:20:19 what i just proposed means you don't have to assume anything 2011-04-25T22:20:29 in order to determine whether the location doesn't have food when you get back in view of the square, you need to do the above 2011-04-25T22:20:47 oh you mean because of the crazy input format? 2011-04-25T22:20:50 you still need to know when a space in general was last seen so you can compare to the food list last seen 2011-04-25T22:20:51 yes 2011-04-25T22:20:54 oh 2011-04-25T22:21:03 well you know i won't defend that :P 2011-04-25T22:21:11 my problem is with how hard it is to calculate that implicit information which should be trivial 2011-04-25T22:21:21 the 'r' line fixed it perfectly 2011-04-25T22:22:00 you get an 'r' line if a visible square without food contained food the last time you saw it 2011-04-25T22:22:04 including if that was last turn 2011-04-25T22:22:19 I think animatroid's solution would take out a little bit of unnecessary computing time on something trivial, apart from that I don't care anymore 2011-04-25T22:22:22 so people don't need to wipe food information if they want to track non-visible food, and can just process those 'r' lines 2011-04-25T22:22:29 I don't really see a reason not to implement it 2011-04-25T22:22:46 i'm more worried about having messy bots with code prone to bugs 2011-04-25T22:22:58 i'm more interested in people competing on strategies, not how well they can handle information 2011-04-25T22:23:24 but you're okay with making them calc unseen land? 2011-04-25T22:23:45 ,... aren't we suggesting people do it for people in the starter bots? 2011-04-25T22:23:47 i actually don't see how an 'r' line is all that helpful. it only makes sense if you're making assumptions in the first place, doesn't it? 2011-04-25T22:23:48 as it's so trivial 2011-04-25T22:24:01 jmcarthur: the other option is also an assumption 2011-04-25T22:24:18 this way people can easily make whichever assumption they like 2011-04-25T22:24:21 it also doesn't help you with squares you can't see anyway 2011-04-25T22:24:29 I think to be consistant, we either include both 'r' and 'l' lines, or neither. 2011-04-25T22:24:35 so you will keep making the same assumption about them 2011-04-25T22:24:36 no, but it gives you explicit information about visible squares 2011-04-25T22:24:54 what do 'i' lines indicate? 2011-04-25T22:24:57 If we include neither, the spec is simpler 2011-04-25T22:25:01 'L' lines 2011-04-25T22:25:10 l lines indicate a non-water square the first time it's seen 2011-04-25T22:25:20 and my vote is for including both if it's both or none 2011-04-25T22:25:21 jaycobcoleman: you need a better font for your IRC :P 2011-04-25T22:25:23 'l' sorry 2011-04-25T22:25:38 by the way lets not use L 2011-04-25T22:25:41 for that very reason 2011-04-25T22:25:48 but it's land 2011-04-25T22:25:55 g for ground? 2011-04-25T22:26:11 that's ridiculous, it's not OCR'd or anthing 2011-04-25T22:26:48 McLeopold: if all starter bots have to do vision, they can pseudo explicitly provide the l line information for people 2011-04-25T22:26:56 it's a matter of marking a square as seen if it's visible 2011-04-25T22:27:01 antimatroid: ant the 'r' info as well 2011-04-25T22:27:07 not so simple 2011-04-25T22:27:10 as i've explained 2011-04-25T22:27:16 It seems just as simple or hard in my mind 2011-04-25T22:27:32 what's the difference in calculating the 2? 2011-04-25T22:27:33 it's much messier imo and requires quite a bit of changing stuff 2011-04-25T22:27:39 evening folks 2011-04-25T22:27:45 bhasker: evening? 2011-04-25T22:27:48 calculating the 2? 2011-04-25T22:27:52 bhasker: boy or girl? 2011-04-25T22:27:54 erm, ignore that ? 2011-04-25T22:27:55 yup evening :) 2011-04-25T22:27:56 girl 2011-04-25T22:28:02 sleeping next to me like a little angel 2011-04-25T22:28:03 congrats 2011-04-25T22:28:04 akshara? 2011-04-25T22:28:08 yup :) 2011-04-25T22:28:10 akshara! 2011-04-25T22:28:10 i'm determined to remember a name for once 2011-04-25T22:28:21 They sure look creepy when the first come out, don't they? 2011-04-25T22:28:25 I believe it was akshara you said? :) 2011-04-25T22:28:26 she is a handful and has a personality 2011-04-25T22:28:28 i hear cone headish? 2011-04-25T22:28:43 not really ours was a health 3.4 kgs :) 2011-04-25T22:28:44 more like alien head 2011-04-25T22:28:50 so she was actually quite plump 2011-04-25T22:28:52 but yea the head 2011-04-25T22:28:54 yeah, gotta get that thing through :P 2011-04-25T22:28:55 is more conical 2011-04-25T22:29:00 and they are grey and purple until the start breathing 2011-04-25T22:29:03 it rounds up 2011-04-25T22:29:07 yup and a little blue 2011-04-25T22:29:20 its an awesome feeling to hold her though 2011-04-25T22:29:34 i also hear some babies need to have a new a hole ripped for them, but maybe i shouldn't go there 2011-04-25T22:29:36 It only gets better 2011-04-25T22:29:48 :) 2011-04-25T22:29:58 yes, i'be been given him hell about the lack of sleep that's coming, but all in all it'd be awesome having a kid 2011-04-25T22:30:17 last 4 nights have been quite sleepless:) 2011-04-25T22:30:22 we got lucky and he slept 6-8 hours at night right away 2011-04-25T22:30:31 ours seems to be a day sleeper 2011-04-25T22:30:44 she's been sleeping like an angel for last 3-4 hrs 2011-04-25T22:31:02 yesterday night it was crazy she cried like nonstop 1-2 hrs 2011-04-25T22:31:09 and it was almost impossible to console her 2011-04-25T22:31:29 took like 4 people trying in rounds, and finally she settled down a bit with my father in law 2011-04-25T22:31:35 cryers gonna cry 2011-04-25T22:31:49 * jmcarthur facepalms 2011-04-25T22:31:49 mcleopold how old is your son 2011-04-25T22:31:51 animatroid: speaking of strategy (it was a while ago). Now that you're talking, what theories do you want included in strategy guides? 2011-04-25T22:31:58 1 year, 1 month 2011-04-25T22:32:05 ah nice must have started walking 2011-04-25T22:32:10 fun times 2011-04-25T22:32:11 yep 2011-04-25T22:32:12 jaycbocoleman: are you planning to add to them, feel free to add anything 2011-04-25T22:32:18 things move higher in the house now 2011-04-25T22:32:31 yeah, you adjust slowly 2011-04-25T22:32:33 jmcarthur: was that directed at me? 2011-04-25T22:32:40 <-- has lotsa RTS experience 2011-04-25T22:32:46 yes 2011-04-25T22:32:50 :p 2011-04-25T22:32:58 animatroid: What's in there so far? 2011-04-25T22:33:16 erm.. good question, hang on 2011-04-25T22:34:05 jaycobcoleman: 1) how to not collide your own ants 2) how to find a path in a maze 2011-04-25T22:34:14 warcraft 1/2/3, starcraft 1/2, TA, TACC, Supcom 1/2, AOE 1/2/3, C&C 1/2, C&C Red Alert1/2, Kohan, Total War, Ground Control,UAW 2011-04-25T22:34:29 alright, anything else? 2011-04-25T22:34:31 mcleopold i have a nice c++ function to extract a path 2011-04-25T22:34:39 https://github.com/aichallenge/aichallenge/wiki/Ants-Strategy-Guide 2011-04-25T22:34:41 bhasker: no CC Generals? 2011-04-25T22:34:48 no never liked it 2011-04-25T22:35:05 i kind of got bored of C&C franchise 2011-04-25T22:35:07 too static 2011-04-25T22:35:11 and not much change in game play 2011-04-25T22:35:12 that is a more general strategy guide 2011-04-25T22:35:18 i think Supcom is awesome 2011-04-25T22:35:20 Supcom2 is okay 2011-04-25T22:35:22 i can't find one if i've done it yet for a few simple improvements 2011-04-25T22:35:32 Supcom Forged Alliance is still one of the best RTS 2011-04-25T22:35:49 I always liked Warcraft 2's split between land and sea. I can't think of another game that has that 2011-04-25T22:35:57 supcom has land/sea/air 2011-04-25T22:36:02 one of the few that has a proper air 2011-04-25T22:36:07 not like the hovering types 2011-04-25T22:36:12 but real planes/bombers etc 2011-04-25T22:36:34 and full fleet of destroyers/subs/antisubs/frigates/air craft carriers/battleships 2011-04-25T22:36:37 McLeopold: what do you think about starting people out with like 3 seconds per turn? 2011-04-25T22:36:41 I used to create maps where you had to build ships to cross a river to get more resources, then had to do it again to get to the enemy 2011-04-25T22:36:42 if you haven't played Supcom you should 2011-04-25T22:36:48 animatroid: I'll try to add some more body to the guides 2011-04-25T22:36:58 *** computerwiz_222 has quit IRC (Remote host closed the connection) 2011-04-25T22:37:02 jaycobcoleman: i call dibs on writing the "battling enemies" part 2011-04-25T22:37:02 mcleopold have you played Total Annihilation 2011-04-25T22:37:07 yes 2011-04-25T22:37:14 what about Supcom 2011-04-25T22:37:19 I really liked the resources as flow rates idea 2011-04-25T22:37:21 if you liked TA 2011-04-25T22:37:24 you would love supcom 2011-04-25T22:37:30 supcom2 is more like starcraft 2011-04-25T22:37:31 Grrrr. You can have it 2011-04-25T22:37:32 haven't played supcom 2011-04-25T22:37:38 antimatroid: no on 3 seconds 2011-04-25T22:37:39 Try it 2011-04-25T22:37:42 even though its 4yrs old now 2011-04-25T22:37:45 bhasker: did you ever play future cop lapd on sony playstation? 2011-04-25T22:37:49 it still kicks starcraft2s ass 2011-04-25T22:37:55 i hate playing games on the ps3 2011-04-25T22:37:56 McLeopold: people are going to be really limited on what they can do 2011-04-25T22:38:05 hate having to use the gamepad 2011-04-25T22:38:12 keyboard + mouse> gamepad 2011-04-25T22:38:13 anyday 2011-04-25T22:38:27 my ps3 nowdays is more of a netflix player/blueray player 2011-04-25T22:38:29 sony playstation 1 2011-04-25T22:38:34 it is an old awesome game 2011-04-25T22:38:38 doh never had one, i was never much of a console gamer 2011-04-25T22:38:45 always been more of a pc gamer 2011-04-25T22:38:46 netiher 2011-04-25T22:38:54 i don't really play video games now at all 2011-04-25T22:39:05 animatroid: I'll probably just try writing a guide as a whole and then give you a link to pastebin and you can add your thoughts/edit 2011-04-25T22:39:31 antimatroid: I think any naive algo for this game will quickly surpase time and memory limits 2011-04-25T22:39:34 i have quite a few ideas for ants , wonder how many i will be able to implement 2011-04-25T22:39:34 pcs are way to hard to keep at good gaming standards 2011-04-25T22:39:51 *too 2011-04-25T22:39:55 pcs games are way more complex and nicer 2011-04-25T22:39:55 McLeopold: sure, but i'm talking about top bots being able to do anything cool 2011-04-25T22:40:05 it's just going to be who can do the most in the shortest amount of time 2011-04-25T22:40:14 not who can do the coolest stuff 2011-04-25T22:40:17 i hate playing multiple player games on ps3 because team coordination is a mess 2011-04-25T22:40:18 bhasker: this was true, not so sure anymore 2011-04-25T22:40:24 without voice 2011-04-25T22:40:29 ps3 games still suck compared to pc 2011-04-25T22:40:37 i wouldn't touch a single mmo designed for ps3 2011-04-25T22:40:45 or a first person shooter 2011-04-25T22:40:50 i'm also a bit concerned about my possible strategy idea of taking up memory at the start of the game and not releasing it 2011-04-25T22:40:52 antimatroid: convince janzert, then I'll listen on the 3 seconds 2011-04-25T22:40:53 sony has a crappy community 2011-04-25T22:41:00 an FPS on ps3 is like shooting oneself 2011-04-25T22:41:06 or any console for that matter 2011-04-25T22:41:15 keyboard + mouse > controller 2011-04-25T22:41:18 for an FPS 2011-04-25T22:41:19 or an RTS 2011-04-25T22:41:23 or an MMMO 2011-04-25T22:41:38 is there a way to limit how much memory a bot can use? 2011-04-25T22:41:45 I would echo your sentiments 2011-04-25T22:41:45 but still more money :P 2011-04-25T22:42:02 animatroid: Do you want to? 2011-04-25T22:45:20 *** amriedle has quit IRC (Quit: Lost terminal) 2011-04-25T22:45:53 want to what? 2011-04-25T22:46:29 I limit memory usage 2011-04-25T22:46:36 no I 2011-04-25T22:47:52 i do 2011-04-25T22:48:06 i would like to give each bot a fair share of the servers memory resources 2011-04-25T22:48:25 not make it a memory wars contest in turn 0 2011-04-25T22:48:33 where people fight to take up the most memory and don't release any of it 2011-04-25T22:49:15 I should have done that in planet wars :( 2011-04-25T22:49:15 I don't think the allotted time allows us to take up more memory than there is to go around 2011-04-25T22:49:28 or does it? 2011-04-25T22:49:30 aichallenge: Alexandru Stan epsilon * r9aa8ad4 / (5 files in 2 dirs): made a visualizer widget that can be included from any page, added one to the index - http://bit.ly/hgKLDa 2011-04-25T22:50:04 McLeopold: me too, but seriously, it think that might be an issue if we have lots of bots on the one machine 2011-04-25T22:50:17 and it's not really a memory wars, as one bot gets to go first with turn 0 i believe? 2011-04-25T22:51:10 I think we can limit memory, I remember reading some python jail code that did that, but I'm not sure how robust it was 2011-04-25T22:52:46 It's certainly possible 2011-04-25T22:54:14 we'd need to pass bots their memory limit as a parameter 2011-04-25T22:54:39 and tracking how much memory you have used probably isn't the easiest thing in the world? 2011-04-25T22:54:49 how many bots are we planning to run on a server at a time? 2011-04-25T22:55:13 *** delt0r_ has joined #aichallenge 2011-04-25T22:55:55 animtroid: You're just trying to come up with reasons for me to use c++ instead of python, aren't you? :P 2011-04-25T22:56:09 *animatroid 2011-04-25T22:56:19 erm, i don't think i need to give anyone more reasons than there are to make such a choice for this game :P 2011-04-25T22:56:33 *** delt0r___ has quit IRC (Ping timeout: 250 seconds) 2011-04-25T22:57:58 by all, I'm off to play portal 2 with dev comentary 2011-04-25T22:58:30 bye 2011-04-25T22:58:50 *** jaycobcoleman has quit IRC (Quit: Page closed) 2011-04-25T22:59:04 McLeopold: how's it going? 2011-04-25T22:59:29 just getting started for the night 2011-04-25T23:00:08 I'm wondering if we can use ulimit to limit a bot's memory usage 2011-04-25T23:01:29 amstan: i noticed the specs page looks a lot nicer now :) 2011-04-25T23:01:44 the LaTeX stuff is still wrong, but we can work out what to do with that later 2011-04-25T23:01:46 antimatroid: yes, bullets and code stuff 2011-04-25T23:02:03 antimatroid: i got a pointer from sigh about it, i think i'll use pandoc instead of simple markdown 2011-04-25T23:02:25 you also managed to avoid to dreaded space on the first line of "code" stuff 2011-04-25T23:02:26 *** computerwiz_222 has joined #aichallenge 2011-04-25T23:02:28 aichallenge: Alexandru Stan epsilon * rdad1159 / website/css/layout.css : added some margins under the visualizer - http://bit.ly/hCJTiH 2011-04-25T23:02:36 on github wikis it forces a space there 2011-04-25T23:02:51 it's terribly annoying :P 2011-04-25T23:03:00 are we using pandoc in the github wiki? is it a markup language? 2011-04-25T23:03:13 McLeopold: it's a thing that reads markdown files 2011-04-25T23:03:21 oh 2011-04-25T23:03:28 antimatroid: i cannot change th rendering engine from the wiki, it's github's stuff 2011-04-25T23:03:33 amstan: did you see my note on including ant maps in wiki text? 2011-04-25T23:03:41 McLeopold: nope, where? 2011-04-25T23:03:50 on irc earlier 2011-04-25T23:04:00 amstan: i know, i'm just pointing out how much nicer it is on the site without the space 2011-04-25T23:04:17 antimatroid: oh, cool 2011-04-25T23:04:18 lol 2011-04-25T23:04:22 the question is, can pan doc use plugins, and can markdown allow us to markup pluging stuff? 2011-04-25T23:04:39 McLeopold: yes pandoc is written in haskell, can do latex and stuff 2011-04-25T23:05:01 is there a way to carry across old user accounts for the new contest? 2011-04-25T23:05:15 so people have rights to their account? 2011-04-25T23:05:45 least some troll decide to go through and just register all the "top" usernames 2011-04-25T23:06:12 antimatroid: we changed the password hash, so it would take some work 2011-04-25T23:06:19 antimatroid: that's no, because we store the passwords in a diff format now 2011-04-25T23:06:27 okay 2011-04-25T23:06:30 unless we code something to migrate 2011-04-25T23:06:36 i'll add that to my list of things i want for zeta 2011-04-25T23:06:49 it wouldn't be to hard, try hash 1, then try hash 2 and force a password reset 2011-04-25T23:07:06 antimatroid: can we see that list? 2011-04-25T23:07:17 i have lots of hypothetical lists 2011-04-25T23:07:19 McLeopold: yeah 2011-04-25T23:07:22 i just refer to them as though they actually exist 2011-04-25T23:07:31 i have a very long to learn list 2011-04-25T23:07:37 we could switch to openid now and not worry 2011-04-25T23:07:43 antimatroid: stuff like these should be issues 2011-04-25T23:07:52 one issue is people not wanting their old username and someone else wanting that 2011-04-25T23:08:19 i would expect such users to come over and ask us to do stuff manually for them 2011-04-25T23:08:25 including cases of stolen usernames 2011-04-25T23:08:28 or people not remembering their old password and can't correctly verify it's them 2011-04-25T23:08:33 i don't think we have to worry too much about that 2011-04-25T23:08:41 btw, bots are limited to just over 1GB of memory 2011-04-25T23:08:56 janzert: but how much memory do we have if we run 10+ player games? 2011-04-25T23:09:10 depends on the servers we use :P 2011-04-25T23:09:33 I'm looking at I believe it was 17.5GB instances 2011-04-25T23:09:58 17.1 2011-04-25T23:09:59 so memory has really never been a problem? 2011-04-25T23:10:03 that's cool 2011-04-25T23:10:26 if we can use serialized running and that instance type it should work out fine 2011-04-25T23:10:40 let's cap players at 16 and promise 1gb? 2011-04-25T23:10:45 there are only two cores on that instance type so it won't work without serialized running 2011-04-25T23:10:55 antimatroid: why is memory such an issue? 2011-04-25T23:11:03 amstan: i just want to make sure it isn't 2011-04-25T23:11:16 and i expect game tree's to be popular again in this, so memory will be used 2011-04-25T23:11:29 although i decided not to use games trees last night 2011-04-25T23:11:35 i'm going to iteratively make normal form games 2011-04-25T23:11:39 if we can't give them 1gb, then it'll start swapping 2011-04-25T23:11:59 and if we count cpu time, it shouldn't disadvantage the player because of that 2011-04-25T23:12:18 if we have 17.5gb and can cap them at 1gb then we can guarantee them that? 2011-04-25T23:12:29 nobody has come up with a cpu timer yet 2011-04-25T23:12:38 most will probably not use 1 full gb 2011-04-25T23:12:38 or we can guarantee 500mb and keep the 26 player limit 2011-04-25T23:12:44 we have SIGSTOP now yeah? 2011-04-25T23:12:55 sigh has it in a branch 2011-04-25T23:13:02 I don't know how much it was tested 2011-04-25T23:13:15 good thing for the beta period imo 2011-04-25T23:13:21 i would really like to have that used 2011-04-25T23:13:43 i.e. the first question I have is whether a sub-process is stopped when the parent is? 2011-04-25T23:13:48 i'm mostly waiting until someone tells me the code is alright, because I haven't done this stuff before 2011-04-25T23:14:56 sigh: also I was wondering how hard does it get for you two run some number of players in parallel, say 2 at a time? 2011-04-25T23:15:06 that would be really awesome if possible 2011-04-25T23:15:20 we could get really good resource usage then 2011-04-25T23:15:35 what do you mean by hard? 2011-04-25T23:15:37 s/two/to 2011-04-25T23:15:55 it becomes harder to make it "fair" 2011-04-25T23:16:13 e.g. say there are 12 players in the game, instead of fully serialized I want to run 2 at a time 2011-04-25T23:16:18 that's easy 2011-04-25T23:16:31 janzert: what are your thoughts about a >1 second time limit per bot? 2011-04-25T23:16:33 for each turn the two that run simultaneously should be shuffled 2011-04-25T23:16:37 there's a function which takes a list of bots and runs them in parallel 2011-04-25T23:16:43 great 2011-04-25T23:16:46 would mean the games take quite a lot longer, but would also mean people can actually do some cool stuff 2011-04-25T23:16:56 so for parallel mode it does all of them, for serial it is given one at a time 2011-04-25T23:17:00 another alternative might be less ants / smaller maps 2011-04-25T23:17:04 antimatroid: we're already much more resource intensive per game I'd rather not add to it 2011-04-25T23:17:36 we expect the games to run longer and be able to play less games at a time 2011-04-25T23:18:13 really we've so far doubled the amount of time it takes for 2 bots to take a turn yeah? 2011-04-25T23:18:25 before it was 2 bots every 1 second, now it's 1 bot every 1 second 2011-04-25T23:18:42 the instance type I mentioned with the 17GB of ram is 4+ times as expensive as the last contest workers 2011-04-25T23:19:16 we were running 4 bots in parallel per worker, now it is 1 or 2 2011-04-25T23:19:44 but 2 bots had a processor to themselves yeah? 2011-04-25T23:19:47 the max game length was 200 now it is 1000 2011-04-25T23:19:58 yeah and there's that 2011-04-25T23:19:59 :P 2011-04-25T23:20:31 good thing is it was 2 bots per game now it is 4-16 or something 2011-04-25T23:20:58 i think really 3-6 players will be plenty to get multiplayer 2011-04-25T23:21:10 but I believe still on balance this is going to be less bots playing per unit of time per worker 2011-04-25T23:21:54 well, per dollar spent on a worker 2011-04-25T23:23:08 we probably don't want to run the contest for any longer than we did too 2011-04-25T23:24:06 anyway back to the engine, I'd like to see the engine have a mode where it will run n players in parallel with basically shuffling the player list every move so the players that are paired together get randomized and n set to the number of cores in the worker 2011-04-25T23:24:23 I think 2 months is probably a good length 2011-04-25T23:26:02 i meant the final contest sorry :P 2011-04-25T23:26:12 ahh 2011-04-25T23:26:44 so each player gets a core each? 2011-04-25T23:26:49 yes 2011-04-25T23:27:15 which is more than they got before ;) 2011-04-25T23:27:21 that would be ideal 2011-04-25T23:27:27 i know :P 2011-04-25T23:28:47 if someone is looking for something to do for the contest, another thing that would be nice is to see if there is a way to make java run with less threads 2011-04-25T23:29:00 loljava :P 2011-04-25T23:29:11 seems that it won't even start up with a limit of 10 threads :( 2011-04-25T23:29:19 which is rather ridiculous 2011-04-25T23:30:46 what isn't very trivial is finding sets of ants that might affect each other in battle 2011-04-25T23:31:01 water information can make it complicated if you want to get the correct partitions 2011-04-25T23:44:52 McLeopold: tell me if the css gets in the way, i haven't been able to test the table css because i didn't have any rankings anywhere 2011-04-25T23:45:17 you can use my sql script to generate fake info 2011-04-25T23:45:42 McLeopold: yay, where? 2011-04-25T23:46:00 amstan: sql/create_test_bot.py actually calls the create test data function 2011-04-25T23:46:17 I have another script to randomize mu and sigma data... 2011-04-25T23:46:52 McLeopold: what db info does it use? 2011-04-25T23:47:37 it pulls from the server_info.py, so just copy it to the directory 2011-04-25T23:47:58 *** sigh has quit IRC (Ping timeout: 241 seconds) 2011-04-25T23:48:17 it creates 10,000 users and 30,000 games by default, so you may want to turn that down a little 2011-04-25T23:49:17 hehe 2011-04-25T23:49:18 5k users and 1mil games would be a good realistic ballpark 2011-04-25T23:49:42 1mil? really? 2011-04-25T23:49:47 i'm considering trying to email daniel 2011-04-25T23:49:50 *** computerwiz_222 has quit IRC (Remote host closed the connection) 2011-04-25T23:49:53 McLeopold: shouldn't this be in integration testing? 2011-04-25T23:49:53 it was 700k for the finals in planetwars 2011-04-25T23:50:03 just the 3 days? 2011-04-25T23:50:05 see if he'd be interested in writing a tcp server to run on the websites server and actually just be part of hte website 2011-04-25T23:50:10 yep 2011-04-25T23:50:33 McLeopold: matchup_count? 2011-04-25T23:50:34 were any of you the ones to email him to get more maps added to his tcp? 2011-04-25T23:50:35 what's that? 2011-04-25T23:50:41 i wouldn't mind an email address... :P 2011-04-25T23:50:51 matchups are different from games, there shouldn't be too many in the queue 2011-04-25T23:50:57 antimatroid: get me his forum username 2011-04-25T23:51:06 McLeopold: so 10 is good? 2011-04-25T23:51:17 0 is good to, for your leader board tests 2011-04-25T23:51:35 amstan: dhartmei 2011-04-25T23:54:07 McLeopold: 702112 games in the finals of planetwars 2011-04-25T23:54:27 i would expect less games with this? 2011-04-25T23:54:31 then crap, I may have to review the matchmaking sql 2011-04-25T23:54:41 antimatroid: link would be nice 2011-04-25T23:54:57 antimatroid: yeah, most likely in the finals 2011-04-25T23:55:03 http://ai-contest.com/forum/memberlist.php?mode=viewprofile&u=181&sid=9278a392153143fa2f714503e24cef30 2011-04-25T23:55:04 McLeopold: can i move this in integration testing? 2011-04-25T23:55:09 antimatroid: the overall contest number will be much higher of course 2011-04-25T23:55:10 I'm putting my tcpserver on github right now... 2011-04-25T23:55:21 oh you have one? 2011-04-25T23:55:25 *** sigh has joined #aichallenge 2011-04-25T23:55:39 or the old one? 2011-04-25T23:55:39 amstan: sure, a copy of the top function is in manager right now, they should be merged and cleaned up 2011-04-25T23:55:46 yes, a new python one 2011-04-25T23:55:53 oh joy :) 2011-04-25T23:56:19 <_flag> Does the "beta launch" objective on the issues page on github mean the contest will be officially announced on the site? 2011-04-25T23:56:37 _flag: no 2011-04-25T23:56:54 _flag: ideally we should have stuff done on our test server, a fully working competition there 2011-04-25T23:57:36 McLeopold: "db_password": "alert_decay", 2011-04-25T23:57:38 McLeopold: people would flip their poop if a tcp server was up :P 2011-04-25T23:57:50 McLeopold: you shouldn't include stuff like that in git 2011-04-25T23:57:58 <_flag> amstan: So would the actual launch soon follow? Do we have an expected start date for the contest? 2011-04-25T23:58:13 _flag: no, we'll decide after we see the state of beta 2011-04-25T23:58:29 _flag: but yes.. it should follow soon 2011-04-25T23:58:51 _flag: unfortunately it'll be ready once everything is done, it's hard to say how long 2011-04-25T23:59:24 for reference beta for planetwars was roughly 2 weeks I believe 2011-04-25T23:59:37 <_flag> amstan: okay, thanks -- I also noticed there was an issue about my example bot in my starter pack, that was the original starter bot but I wasn't aware the starter bot had to follow a spec so you can just delete it. 2011-04-25T23:59:49 planetwars just beta'd in competition if i recall? :P