2011-11-17T00:03:25 *** bmh has joined #aichallenge 2011-11-17T00:03:40 mleise: I think there's an inductive proof that the subdivisions are all convex. 2011-11-17T00:09:01 *** Antimony has quit IRC (Ping timeout: 240 seconds) 2011-11-17T00:12:29 *** ChrisH_ has quit IRC (Quit: Page closed) 2011-11-17T00:13:58 http://paste.aichallenge.org/BGFUQ/ 2011-11-17T00:14:29 now that would be an obnoxious free-for-all 2011-11-17T00:14:57 it's weird how with the symmetry, the map looks bigger than it is 2011-11-17T00:15:20 how was that generated? 2011-11-17T00:15:54 magic 2011-11-17T00:16:00 and it's not really.. symmetrical by the ants point of view 2011-11-17T00:16:01 and some python code 2011-11-17T00:16:07 McLeopold: That's not symmetric though, is it? 2011-11-17T00:16:19 ignore the hills I presume 2011-11-17T00:16:23 yeah 2011-11-17T00:16:24 they are just there for markers 2011-11-17T00:16:28 ah 2011-11-17T00:16:41 oh, they're like voronoi points 2011-11-17T00:16:45 yep 2011-11-17T00:16:53 voronoi maze carving 2011-11-17T00:17:05 clever 2011-11-17T00:17:15 that's the offset grid you were talking about 2011-11-17T00:17:51 good ol' Worley noise 2011-11-17T00:17:54 I don't remember talking about that... 2011-11-17T00:18:14 hmm. i guess i'm hallucinating 2011-11-17T00:21:12 oh right, I can paste stuff... 2011-11-17T00:21:19 http://paste.aichallenge.org/xPJ22/ -- I think I forage alright 2011-11-17T00:24:16 *** cyphase has quit IRC (Ping timeout: 240 seconds) 2011-11-17T00:27:19 *** ALplus has joined #aichallenge 2011-11-17T00:27:51 *** Areks has joined #aichallenge 2011-11-17T00:35:24 *** JorgeB has joined #aichallenge 2011-11-17T00:36:13 bmh, maybe if you do it voronoi diagram style. (select 100 points on the map and bfs outward from those simultaneously) 2011-11-17T00:36:25 *** yoden has quit IRC (Quit: Leaving.) 2011-11-17T00:36:30 mleise: That's the idea. 2011-11-17T00:36:58 oh, i didn't get that at first. now i can claim that I discovered it on my own ^^ 2011-11-17T00:37:15 *** Antimony has joined #aichallenge 2011-11-17T00:37:40 *** cyphase has joined #aichallenge 2011-11-17T00:41:17 *** mleise has quit IRC (Ping timeout: 245 seconds) 2011-11-17T00:41:21 *** bmh has quit IRC (Quit: Page closed) 2011-11-17T00:45:18 *** ALplus has quit IRC (Ping timeout: 252 seconds) 2011-11-17T00:57:58 *** gcflymoto has joined #aichallenge 2011-11-17T01:04:05 :) 2011-11-17T01:17:52 *** Antimony has quit IRC (Ping timeout: 240 seconds) 2011-11-17T01:23:31 *** ltriant has quit IRC (Quit: Computer has gone to sleep) 2011-11-17T01:24:38 *** epicmonkey has joined #aichallenge 2011-11-17T01:30:26 *** avdg1 has joined #aichallenge 2011-11-17T01:33:56 *** avdg has quit IRC (Ping timeout: 244 seconds) 2011-11-17T01:36:12 *** raemde has joined #aichallenge 2011-11-17T01:41:49 *** analyst74 has quit IRC (Ping timeout: 240 seconds) 2011-11-17T01:42:40 *** darius_ has joined #aichallenge 2011-11-17T01:43:17 Hi, all! 2011-11-17T01:47:11 *** epicmonkey has quit IRC (Ping timeout: 252 seconds) 2011-11-17T02:12:37 *** b0rder has quit IRC (Quit: 离开) 2011-11-17T02:15:28 *** Mooloo has quit IRC (Ping timeout: 240 seconds) 2011-11-17T02:21:09 *** lollacaust has joined #aichallenge 2011-11-17T02:23:25 *** djr_ has joined #aichallenge 2011-11-17T02:28:19 *** Mooloo has joined #aichallenge 2011-11-17T02:28:48 ? 2011-11-17T02:32:39 *** djr__ has joined #aichallenge 2011-11-17T02:33:04 *** ikaros has joined #aichallenge 2011-11-17T02:33:39 *** datachomper has quit IRC (Quit: Leaving.) 2011-11-17T02:34:18 *** bhasker has joined #aichallenge 2011-11-17T02:34:37 hi 2011-11-17T02:35:52 *** djr_ has quit IRC (Ping timeout: 240 seconds) 2011-11-17T02:38:12 *** Harpyon has joined #aichallenge 2011-11-17T02:41:56 *** lampahalogenowa has joined #aichallenge 2011-11-17T02:43:47 bhasker! 2011-11-17T02:43:53 :) 2011-11-17T02:43:55 how goes it 2011-11-17T02:44:15 good. just had baby #2, off work for a while 2011-11-17T02:44:23 back from india? 2011-11-17T02:44:45 oh Congrats!!, yup back from India , been a month and joined Shazam 2011-11-17T02:45:00 so been busy was wondering if i should update my bot to play with the new rules 2011-11-17T02:45:10 only change i think will be to add support to defend/attack hills 2011-11-17T02:45:29 how is the kid doing? when was he/she born? 2011-11-17T02:45:30 Hello, I've got problem with my bot.. He should play game yesterday but he still has "Next Game: 0 players are ahead" in my profile. Can someone help me? Where should I report this ? 2011-11-17T02:45:30 yeah 2011-11-17T02:45:34 which is kind of a big change 2011-11-17T02:46:06 he's doing well. sleeping now, getting very chubby. born 10/15 2011-11-17T02:46:13 nice:) how many pounds 2011-11-17T02:46:28 7 lbs 8 oz or so. he's probably over 10 now. 2011-11-17T02:47:13 wow nice 2011-11-17T02:47:30 looks like your bot is doing quite well 2011-11-17T02:47:39 except you seem to have uploaded a new version 2011-11-17T02:47:43 yeah, it'll be a week or so before it shows up on the official leaderboards 2011-11-17T02:48:04 ants.fluxid.pl is the tcp server most people are using these days (there's also tcpants.com) 2011-11-17T02:48:46 *** HaraKiri has joined #aichallenge 2011-11-17T02:48:48 now that there are hills there are interesting tradeoffs with defense/offense 2011-11-17T02:49:20 lampahalogenowa, check that your account is not deactivated 2011-11-17T02:49:28 you should find it in your profile 2011-11-17T02:49:43 I've checked this. It is still active 2011-11-17T02:49:59 O! Status has changed 2011-11-17T02:50:10 =) 2011-11-17T02:50:10 bot is playing now, so nevermind 2011-11-17T02:50:12 :) 2011-11-17T02:50:13 there are six workers up now but games are going pretty slow 2011-11-17T02:50:37 yea sometimes it very slow 2011-11-17T02:50:41 it's.. 2011-11-17T02:52:55 *** lampahalogenowa has quit IRC (Quit: Page closed) 2011-11-17T02:53:53 too slow 2011-11-17T02:54:02 hmm k will see when i can upload my new bot 2011-11-17T02:54:09 maybe over the weekend i might find sometime 2011-11-17T02:54:26 *** Harpyon has quit IRC (Ping timeout: 276 seconds) 2011-11-17T02:55:54 *** Palmik has joined #aichallenge 2011-11-17T03:03:34 *** djr__ has quit IRC (Quit: Leaving) 2011-11-17T03:11:44 *** aerique has joined #aichallenge 2011-11-17T03:12:26 *** jordy has joined #aichallenge 2011-11-17T03:12:58 *** jordy has left #aichallenge 2011-11-17T03:15:10 *** sigh has joined #aichallenge 2011-11-17T03:17:08 *** mj41 has joined #aichallenge 2011-11-17T03:18:35 *** g0llum has joined #aichallenge 2011-11-17T03:22:48 *** gcflymoto has quit IRC (Ping timeout: 265 seconds) 2011-11-17T03:25:19 *** amstan has quit IRC (Ping timeout: 248 seconds) 2011-11-17T03:26:50 *** Fandekasp has joined #aichallenge 2011-11-17T03:48:59 *** pairofdice has joined #aichallenge 2011-11-17T03:49:03 *** mviel_ has quit IRC (Read error: Connection reset by peer) 2011-11-17T03:50:38 *** mviel has joined #aichallenge 2011-11-17T04:08:19 *** savaron has joined #aichallenge 2011-11-17T04:14:00 *** praveeen has joined #aichallenge 2011-11-17T04:15:25 *** imp5imp5 has joined #aichallenge 2011-11-17T04:16:49 *** kaemo has joined #aichallenge 2011-11-17T04:16:58 *** liberforce has joined #aichallenge 2011-11-17T04:24:38 *** ikaros_ has joined #aichallenge 2011-11-17T04:27:18 *** mviel_ has joined #aichallenge 2011-11-17T04:30:50 *** mviel has quit IRC (Ping timeout: 252 seconds) 2011-11-17T04:39:17 *** praveen_ has joined #aichallenge 2011-11-17T04:45:59 *** replore__ has quit IRC (Remote host closed the connection) 2011-11-17T04:47:37 *** replore_ has joined #aichallenge 2011-11-17T04:51:31 *** savaron has quit IRC (Quit: Page closed) 2011-11-17T04:52:10 *** replore_ has quit IRC (Remote host closed the connection) 2011-11-17T04:56:00 *** praveen_ has quit IRC (Quit: Page closed) 2011-11-17T05:10:02 *** Fandekasp has quit IRC (Ping timeout: 245 seconds) 2011-11-17T05:22:09 *** Fandekasp has joined #aichallenge 2011-11-17T05:22:52 *** kaemo has quit IRC (Ping timeout: 276 seconds) 2011-11-17T05:25:24 *** Extrarius has quit IRC (Ping timeout: 244 seconds) 2011-11-17T05:26:14 *** Extrarius has joined #aichallenge 2011-11-17T05:26:54 *** NightExcessive has joined #aichallenge 2011-11-17T05:27:19 *** grwip has joined #aichallenge 2011-11-17T05:28:48 Fluxid: ratings on your server seem to be working now 2011-11-17T05:29:58 *** ikaros_ has quit IRC (Quit: Ex-Chat) 2011-11-17T05:30:53 *** maysam81 has quit IRC (Ping timeout: 265 seconds) 2011-11-17T05:30:56 *** ikaros_ has joined #aichallenge 2011-11-17T05:31:25 *** Areks has quit IRC (Ping timeout: 240 seconds) 2011-11-17T05:31:37 Garf: good. seems phpskills works the best 2011-11-17T05:37:05 *** kaemo has joined #aichallenge 2011-11-17T05:38:25 *** ikaros_ has quit IRC (Quit: Ex-Chat) 2011-11-17T05:44:56 *** Areks has joined #aichallenge 2011-11-17T05:46:01 *** mcstar has joined #aichallenge 2011-11-17T05:56:15 *** sir_macelon has joined #aichallenge 2011-11-17T05:56:47 hi 2011-11-17T05:56:56 hiiii 2011-11-17T05:58:23 *** ikaros has quit IRC (Remote host closed the connection) 2011-11-17T06:08:15 how come all my 4 opponents timed out in one game? http://aichallenge.org/visualizer.php?game=99585 2011-11-17T06:08:26 *** avdg has joined #aichallenge 2011-11-17T06:09:18 Looks like they're the same bot 2011-11-17T06:09:27 one of the starters, probably GreedyBot 2011-11-17T06:10:35 they make the same moves 2011-11-17T06:10:40 Yeah 2011-11-17T06:11:06 but it should not have happened 2011-11-17T06:11:13 those bots are too simple to time out 2011-11-17T06:11:22 Why not? GreedyBot times out all the time 2011-11-17T06:11:27 why? 2011-11-17T06:11:33 thats stupid 2011-11-17T06:11:36 It's a bad python bot 2011-11-17T06:11:40 :D 2011-11-17T06:12:02 *** avdg1 has quit IRC (Ping timeout: 276 seconds) 2011-11-17T06:13:04 sir_macelon: do you have a tool with which you can quickly decide what will be the outcome of a battle situation? 2011-11-17T06:13:29 decide->discover 2011-11-17T06:13:33 *** hrls has joined #aichallenge 2011-11-17T06:13:35 *** Anilm3 has joined #aichallenge 2011-11-17T06:13:41 Hi to everyone 2011-11-17T06:13:55 Hi 2011-11-17T06:14:13 Any lisp coders here? 2011-11-17T06:14:26 mcstar: yes, paper and pen :) 2011-11-17T06:14:35 sir_macelon: do you want something better? 2011-11-17T06:15:30 mcstar: not necessarilly I haven't been working on my battle code since months 2011-11-17T06:15:37 :( 2011-11-17T06:15:38 *** hrls has quit IRC (Client Quit) 2011-11-17T06:15:42 mcstar: do you have something better? 2011-11-17T06:15:46 yes! 2011-11-17T06:16:00 what;s that? 2011-11-17T06:16:06 are you on linux? 2011-11-17T06:16:13 *** kaemo has quit IRC (Ping timeout: 240 seconds) 2011-11-17T06:16:18 win 2011-11-17T06:16:28 well, its a c++ + qt app 2011-11-17T06:16:42 its easier to compile it on linux 2011-11-17T06:16:45 i guess 2011-11-17T06:16:55 what does it do? 2011-11-17T06:17:11 you can draw ants on its board quickly, and see what will happen 2011-11-17T06:17:29 sounds neat 2011-11-17T06:18:01 yeah, but so far only amstan tried it, and id welcome some input 2011-11-17T06:18:08 I have ubuntu also so I could try it out the other day 2011-11-17T06:18:23 https://github.com/liquid-phynix/ants-fight.git 2011-11-17T06:18:27 it on github 2011-11-17T06:18:28 s 2011-11-17T06:19:26 sir_macelon: do you use unity? 2011-11-17T06:20:06 mcstar: ok, thanks, i'll look into it, but I don't have much time recently 2011-11-17T06:20:21 Anilm3: what do you mean with unity? 2011-11-17T06:20:42 The new UI ubuntu uses 2011-11-17T06:20:43 *** kaemo has joined #aichallenge 2011-11-17T06:21:21 with a sidebar and other stuff 2011-11-17T06:21:37 Anilm3: I have some old install, so I suppose not, been working on the win part mostly now 2011-11-17T06:21:46 mcstar someone should compile linux and windows binaries for that. at least i don't have qt installed :x 2011-11-17T06:21:58 sir_macelon: ok, thanks anyway 2011-11-17T06:22:05 HaraKiri: well, would you trust a binary? 2011-11-17T06:22:13 i can compile a linux one for you 2011-11-17T06:22:17 and upload it 2011-11-17T06:22:29 mcstar: 2011-11-17T06:22:34 Anilm3: 2011-11-17T06:22:38 he needs qt installed 2011-11-17T06:22:45 even with a binary 2011-11-17T06:22:55 unless you compile with static libraries 2011-11-17T06:22:59 yes 2011-11-17T06:23:01 i'm pretty naive, yes :D. but im on windows right now anyways. also, i'm sure many ppl on the tools forum would appreciate that tool 2011-11-17T06:23:14 oh ok 2011-11-17T06:23:17 i dont have windows right now 2011-11-17T06:23:51 but its possible to cross compile i think 2011-11-17T06:24:47 HaraKiri: do you want to be my test subject? 2011-11-17T06:25:35 Anilm3: btw, qt is a 25MBs download for me 2011-11-17T06:25:49 Download Size : 24814.97 K 2011-11-17T06:25:56 its not that big 2011-11-17T06:26:33 for a windows binary? 2011-11-17T06:26:50 oh god this unity thing stucks 2011-11-17T06:26:53 sucks* 2011-11-17T06:27:04 Anilm3: Indeed. I like GNOME 3 better. 2011-11-17T06:27:07 join the club 2011-11-17T06:27:14 lol 2011-11-17T06:27:18 mcstar: But static compilations often comes with headaches... 2011-11-17T06:27:21 ditched kde/gnome for awesome 2011-11-17T06:27:27 time to use a tiling wm guys 2011-11-17T06:27:28 NightExcessive: how do I get rid of this stuff 2011-11-17T06:27:38 people should just use fluxbox 2011-11-17T06:27:44 Minthos: really? 2011-11-17T06:27:48 and not worry about the new fancy wms 2011-11-17T06:27:50 mcstar: I tried it for a while, but is crazy 2011-11-17T06:27:50 yes really 2011-11-17T06:27:52 Anilm3: Don't use Ubuntu... or figure out how to use a different desktop environment with Ubuntu. I personally use openSUSE. 2011-11-17T06:27:54 have you tried a good tiling wm? 2011-11-17T06:27:59 Minthos: I agree 2011-11-17T06:28:05 no, but I'm very happy with fluxbox 2011-11-17T06:29:24 *** ikaros has joined #aichallenge 2011-11-17T06:29:47 mcstar: test subject for windows binary? 2011-11-17T06:29:52 yes 2011-11-17T06:30:04 sure! should i install qt? or without? 2011-11-17T06:30:14 well, id like without 2011-11-17T06:30:17 NightExcessive: I'm going to give it a try for a few days and if my brain explodes I'll get rid of ubuntu.. 2011-11-17T06:30:24 to make it self-contained 2011-11-17T06:30:30 but i never did cross compilation before 2011-11-17T06:30:32 would be awesome 2011-11-17T06:30:35 so i have to read up 2011-11-17T06:30:39 Anilm3: My whole body exploded last time I tried... not just my brain. :) 2011-11-17T06:30:54 sure, just hit me if you need me 2011-11-17T06:31:02 k thx 2011-11-17T06:31:20 kubuntu was my second longest-used distro 2011-11-17T06:31:33 i could use it for a half year 2011-11-17T06:33:51 I was using ubuntu 10.04 but it decided to die 2011-11-17T06:33:59 and I couldn't revive it :( 2011-11-17T06:34:45 for me ubuntu and kubuntu became unusable imho. xubuntu works ok 2011-11-17T06:35:01 I managed to murder openSUSE a few times by being impatient and killing the package manager in the middle of an update. 2011-11-17T06:35:19 fluxbox looks somewhere between fancy and insane 2011-11-17T06:35:46 I used fluxbox for a while and loved it 2011-11-17T06:36:03 why only for a while then? ;p 2011-11-17T06:36:20 *** sir_macelon has quit IRC (Quit: Page closed) 2011-11-17T06:36:49 *** ikaros has quit IRC (Ping timeout: 244 seconds) 2011-11-17T06:37:02 anyway, im using arch for over 4 years i think 2011-11-17T06:37:23 I was using arch with fluxbox and I had an incompatibility with my laptop and just installed ubuntu 2011-11-17T06:37:34 sad 2011-11-17T06:37:43 I didn't have time to experiment :( 2011-11-17T06:37:58 arch is a great distro btw 2011-11-17T06:38:23 yeah, but not as ffriendly as ubuntu, but thats not a problem for me 2011-11-17T06:38:25 *** ikaros has joined #aichallenge 2011-11-17T06:38:52 the only problem I see with arch is the package system 2011-11-17T06:39:06 but if a needed to do something more systemic in ubuntu i had a harder time then i have on arch 2011-11-17T06:39:24 Anilm3: thats one of its advantages 2011-11-17T06:39:31 mcstar: works well 2011-11-17T06:39:37 but packages are unsigned 2011-11-17T06:39:52 *** liberforce has quit IRC (Quit: Leaving.) 2011-11-17T06:39:54 I have trust issues 2011-11-17T06:40:13 k 2011-11-17T06:42:00 it isn't too hard to get vanilla ubuntu to run fluxbox 2011-11-17T06:42:34 *** olexs has joined #aichallenge 2011-11-17T06:42:34 then you get a great package system, a friendly and very googlable os, and an awesome wm 2011-11-17T06:42:56 ubuntu forums are closed 2011-11-17T06:43:01 you need to register to see a thread 2011-11-17T06:43:10 insane 2011-11-17T06:43:19 Minthos: why not just go to debian? 2011-11-17T06:43:19 really? 2011-11-17T06:43:40 Anilm3: why would I? I'm happy with ubuntu 2011-11-17T06:44:16 Minthos: allright then 2011-11-17T06:44:38 mcstar: I don't know which forums you're referring to, ubuntuforums.org seems to work fine without registration 2011-11-17T06:45:34 frequently when i was googling a problem llinux related i ended up on an ubuntu forum but i couldnt see its content 2011-11-17T06:46:40 *** goffrie_ has joined #aichallenge 2011-11-17T06:46:40 *** goffrie has quit IRC (Disconnected by services) 2011-11-17T06:46:41 *** goffrie_ is now known as goffrie 2011-11-17T06:46:51 well, i did a random search, and i can see the forum's content now 2011-11-17T06:47:00 but i swear i run into this several times 2011-11-17T06:47:09 I hate unity I'm going to try xubuntu 2011-11-17T06:47:23 NightExcessive: didn't took long... 2011-11-17T06:47:42 Anilm3: Did you explode yet? 2011-11-17T06:47:47 I did 2011-11-17T06:47:58 :) 2011-11-17T06:48:10 Pff, I'm sure Unity is awesome 2011-11-17T06:48:28 weird, before i switched to tiling wm, i experimented with old gnome + dockbarx 2011-11-17T06:48:46 its was graphically pleasing and i could do my job relatively fast 2011-11-17T06:49:02 this was almost a year ago 2011-11-17T06:49:23 but id like to stress the usefulness of a tiling wm, it really frees your mind 2011-11-17T06:49:36 from all those clicks and graphics 2011-11-17T06:49:44 and its FAST 2011-11-17T06:50:39 GNOME 3 + X11 is extremely fast on my hardware... and has never gobbled my mind up. :P 2011-11-17T06:50:48 mcstar: I tried awesome and xmonad and didn't really got that excited... 2011-11-17T06:50:53 Graphics are GOOD! 2011-11-17T06:51:09 im in the minority 2011-11-17T06:51:12 its ok 2011-11-17T06:51:18 We sound like a bunch of internet hippies :) 2011-11-17T06:51:31 no, i used osx, and was all for looks at one time 2011-11-17T06:51:33 i should try a tiling wm. the idea sounds logic 2011-11-17T06:51:39 but awesome just changed me 2011-11-17T06:51:50 Well osx and Apple in general is too much about looks 2011-11-17T06:52:06 i mean i tried a new style every months or so 2011-11-17T06:52:07 games don't work very good with the unity stuff 2011-11-17T06:52:22 at one point i gave up and went minimalistic 2011-11-17T06:52:25 never came back 2011-11-17T06:52:28 osx has a pretty worthless workspace system 2011-11-17T06:52:29 story ends 2011-11-17T06:52:59 mcstar: one day I will go back to arch+fluxbox and never come back, but I don't find the time :( 2011-11-17T06:53:32 I don't feel like experimenting with OS's much when I only have one computer 2011-11-17T06:54:00 I used gentoo + fluxbox earlier, time constraints made me switch to ubuntu + fluxbox, it will take a lot to get me to switch to anything else now :) 2011-11-17T06:54:36 once youre comfortable you hardly change 2011-11-17T06:54:48 *** kaemo has quit IRC (Quit: Lost terminal) 2011-11-17T06:54:49 i went nuts yesterday when i had to use webchat 2011-11-17T06:55:15 :) 2011-11-17T07:00:26 mcstar: why? 2011-11-17T07:01:41 Because a server went down and he didn't pick another ;) 2011-11-17T07:02:11 *** Conorach has joined #aichallenge 2011-11-17T07:03:53 pairofdice: i did 2011-11-17T07:04:01 not right away 2011-11-17T07:04:22 but even freenode.net was down, and i had to google for another server 2011-11-17T07:05:01 and there was that ghosting isssue and i wasnt sure if the two things were unrelated 2011-11-17T07:05:35 weechat says "connection refused" but it should say "no response" or simialr 2011-11-17T07:06:07 I use the webchat and I didn't have any issue yesterday 2011-11-17T07:06:30 Maybe a local dns issue, who knows 2011-11-17T07:06:47 *** ace1010 has joined #aichallenge 2011-11-17T07:07:18 Anilm3: of yourse you didnt 2011-11-17T07:07:27 webchat was up 2011-11-17T07:07:37 the site was down as well as my favourite node 2011-11-17T07:07:49 y->c 2011-11-17T07:07:54 you have a favourite node? 2011-11-17T07:07:58 yes 2011-11-17T07:08:15 there is this novel the neuromancer that i want to read 2011-11-17T07:08:18 oh that was the whole drama about 2011-11-17T07:08:20 it is from gibson 2011-11-17T07:08:31 oh, I see 2011-11-17T07:08:37 ive seen the movie, but id like to read it too 2011-11-17T07:08:52 so i wont forget it if i connect to gibson.freenode.net 2011-11-17T07:09:06 :D 2011-11-17T07:09:13 why don't you use a post-it note on the side of your screen? 2011-11-17T07:09:17 That's a bit convoluted :) 2011-11-17T07:09:39 my screen is full of post-it's since the start of ants :x 2011-11-17T07:09:49 hahaha 2011-11-17T07:09:52 mine too 2011-11-17T07:10:17 i had a lot of "notes" in basket 2011-11-17T07:10:30 last time i switched hdd's i forgot to back it up 2011-11-17T07:10:46 basket is a qt note teaker 2011-11-17T07:10:48 -e 2011-11-17T07:11:00 horrible feeling, right? for the next few months, you wonder what important things were in there, but you cant remember 2011-11-17T07:11:43 those were ideas 2011-11-17T07:11:52 when i think up something i wrote it down 2011-11-17T07:11:55 why are the servers so slow? less than 1 game per minute each 2011-11-17T07:11:56 I'm too cool for real post-it notes. I use virtual post-it notes. :) 2011-11-17T07:12:11 Anilm3: Because bots are taking longer 2011-11-17T07:12:16 I prefer real ones 2011-11-17T07:12:29 mcstar, shame your nice qt tool did not compile on win ( stone-age mingw c++) qt 4.5.3 2011-11-17T07:13:10 g0llum: it needs c++0x 2011-11-17T07:13:18 maybe thats the problem 2011-11-17T07:13:19 fancy 2011-11-17T07:13:20 How often do you people upload a bot to the site? 2011-11-17T07:13:23 i know.. 2011-11-17T07:13:27 ah k 2011-11-17T07:13:34 g0llum: what is the error? 2011-11-17T07:13:38 id like to know 2011-11-17T07:13:41 mingw does not have it 2011-11-17T07:13:45 oh 2011-11-17T07:13:59 but im sure the latest one does doesnt it? 2011-11-17T07:14:04 unordered_maps.. 2011-11-17T07:14:11 i dont want you to try it, but thats the problem right? 2011-11-17T07:14:31 mcstar, again ? 2011-11-17T07:14:32 yeah, i pulled that out of my bot's code 2011-11-17T07:14:47 g0llum: im guessing with the latest mingw id be fine 2011-11-17T07:14:50 wont it? 2011-11-17T07:14:56 it'd 2011-11-17T07:15:15 *** liberforce has joined #aichallenge 2011-11-17T07:15:28 let me take a look 2011-11-17T07:16:40 i think i will make the active color button title boldfaced 2011-11-17T07:18:20 can't you just replace it with a normal map? 2011-11-17T07:18:30 it'll use more memory and be slower, but who cares? 2011-11-17T07:20:17 im using a pos struct as key, its basically a pair 2011-11-17T07:20:46 id need to implemen the "less" operation for them to use them in std::map 2011-11-17T07:21:08 you can do it pal! 2011-11-17T07:21:24 but you can use the new standard 2011-11-17T07:21:32 so I wont have to change anything :D 2011-11-17T07:21:52 Usually it's a bad practice assuming the user is who has to take care of the portability 2011-11-17T07:22:48 i think it is portable 2011-11-17T07:22:58 doesnt use any os specific thing 2011-11-17T07:23:17 how can you consider portable an application that uses an unfinished standard and an unfinished library? 2011-11-17T07:23:18 and in the linux world you distribute sources 2011-11-17T07:23:25 ? 2011-11-17T07:23:34 since when is it unfinished? 2011-11-17T07:23:43 since when it is finished? 2011-11-17T07:23:51 WHAT is unfinished? 2011-11-17T07:23:59 c++11 2011-11-17T07:24:01 neither c++11 or qt 2011-11-17T07:24:13 wait, ++11 is done, right!? 2011-11-17T07:24:17 of course 2011-11-17T07:24:26 oh all right, september 2011-11-17T07:24:42 so you can assume it has been already widely implemented 2011-11-17T07:24:50 no 2011-11-17T07:24:58 but it is widely implemented 2011-11-17T07:25:20 the features i use are available in msvs and gcc and intel too 2011-11-17T07:25:20 suppose I use ubuntu 8.10 on an old computer 2011-11-17T07:25:31 Anilm3: what is your point? 2011-11-17T07:25:43 you can write whatever software and it will break somewhere 2011-11-17T07:25:48 if it is not imperative to use new features don't use them 2011-11-17T07:25:59 it is imperative 2011-11-17T07:26:02 stick to the widely implemented standard 2011-11-17T07:26:05 Anilm3: suppose I use A/UX on my old apple IIsi 2011-11-17T07:26:06 why? because 2011-11-17T07:26:28 you can also use boost 2011-11-17T07:26:37 Anilm3: i dont WANT you to use it, but you can, im pleased if you want and can 2011-11-17T07:26:50 but above that i dont care 2011-11-17T07:27:08 ok then people g0llum won't be able to use your app 2011-11-17T07:27:28 i can live with that 2011-11-17T07:27:38 but i think he needs a newer mingw 2011-11-17T07:27:39 because it's what you want and not what the user needs 2011-11-17T07:27:53 i can stii fire up a vm and od from ubuntu 2011-11-17T07:28:10 Anilm3: its like asking the user to install python 2011-11-17T07:28:26 why on earth would he do that? just to participate in a contest? 2011-11-17T07:28:40 actually yes, just for that 2011-11-17T07:29:05 do you think the contest's engine should be in native i386 pre .net binary? 2011-11-17T07:29:22 I'm just arguing about good software engineering practices 2011-11-17T07:29:47 you can 2011-11-17T07:29:59 but i will stop now im not a software engineer 2011-11-17T07:30:43 mcstar, tried with mingw4.5.2 - no auto ;( 2011-11-17T07:31:11 hm 2011-11-17T07:31:15 really? 2011-11-17T07:31:28 the server has gcc which version? 2011-11-17T07:31:32 4.5.X 2011-11-17T07:31:49 it must have been added in the 4.5 branch 2011-11-17T07:32:06 because my code compiles on the server 2011-11-17T07:32:09 mingw versions != usual gcc version 2011-11-17T07:32:16 oh 2011-11-17T07:32:30 and they're always some behind 2011-11-17T07:34:01 g0llum: but is that the latest version? 4.5.2? 2011-11-17T07:34:02 those qt ppl invented their own rtt and containers, so they don't feel the nned to adopt toc++10 so much 2011-11-17T07:34:11 yup 2011-11-17T07:34:13 *** Jak_o_Shadows has quit IRC (Remote host closed the connection) 2011-11-17T07:34:19 i can get rid of auto 2011-11-17T07:35:58 probably just needs a typedef for that iterator 2011-11-17T07:36:11 g0llum: can you try it again? 2011-11-17T07:37:31 just trying that.. 2011-11-17T07:41:35 apparently it freezed you machine 2011-11-17T07:41:37 r 2011-11-17T07:41:57 no, puzzled by errors. 2011-11-17T07:42:31 :D 2011-11-17T07:42:41 can you pastebin some? 2011-11-17T07:42:44 well in short, no such luck. still no unordered_map, the hash-specialization wa 2011-11-17T07:42:53 aham 2011-11-17T07:43:19 *** savaron has joined #aichallenge 2011-11-17T07:44:32 since im in a good mood ill try to get rid of unordered_map 2011-11-17T07:44:54 http://pastebin.com/vVXwTxim // mcstar 2011-11-17T07:45:43 did you try the suggestion? 2011-11-17T07:46:08 standard, C++0x. This support is currently experimental, and must be enabled with the -std=c++0x 2011-11-17T07:46:25 ++ -c -O2 -frtti -fexceptions -mthreads -Wall -DUNICODE -DQT_LARGEFILE_SUPPORT -DQT_DLL -DQT_NO_DEBUG -DQT_GUI_LIB -DQT_CORE_LIB -DQT_THREAD_SUPPORT -DQT_NEEDS_QMAIN 2011-11-17T07:46:34 you didnt enable c++0x support at all 2011-11-17T07:46:53 me ? no, forgot that ;) 2011-11-17T07:47:06 delete the makefiel 2011-11-17T07:47:11 just to be sure and run qmake 2011-11-17T07:47:27 after that edit the Makefile's CXXFLAGS var 2011-11-17T07:47:29 k? 2011-11-17T07:47:49 VICTORY! 2011-11-17T07:47:58 :O 2011-11-17T07:48:43 pairofdice: on dec the 18th 2011-11-17T07:48:57 My ants are using the "act like a crazy person" defence 2011-11-17T07:50:00 *** darius_ has quit IRC (Quit: Page closed) 2011-11-17T07:50:17 pairofdice: how about a replay? 2011-11-17T07:50:30 http://aichallenge.org/visualizer.php?game=100719 2011-11-17T07:50:47 pairofdice: lol, your ants are naked 2011-11-17T07:51:48 What 2011-11-17T07:52:11 nothing 2011-11-17T07:53:14 *** u_ has joined #aichallenge 2011-11-17T07:53:50 *** u__ has joined #aichallenge 2011-11-17T07:53:54 does anyone know good rule of tump what strength difference a mu difference compares to? 2011-11-17T07:54:02 i.e. 75% winrate -> 2.0 mu difference 2011-11-17T07:54:04 *** g0llum has quit IRC (Read error: Connection reset by peer) 2011-11-17T07:54:09 rule of thumb, rather 2011-11-17T07:54:23 *** g0llum has joined #aichallenge 2011-11-17T07:54:46 mcstar, getting closer, original code compiled http://pastebin.com/RwWmuYUK 2011-11-17T07:56:15 Anyone know what this whole activate/deactivate button your profile is all about? Do I have to activate my bots every 3 days or get pulled form the contest? 2011-11-17T07:56:28 yes 2011-11-17T07:56:30 collect2: ld returned 1 exit status 2011-11-17T07:56:43 g0llum: seems it needs an -l switch 2011-11-17T07:56:47 or something, weird 2011-11-17T07:56:48 Anilm3: that yes for me? 2011-11-17T07:56:59 yes to your second question 2011-11-17T07:57:04 Anilm3: thanks 2011-11-17T07:57:09 elderwolf: not active -> your bot will only play games very occasionally 2011-11-17T07:57:20 elderwolf: not the same as "pulled from the contest" 2011-11-17T07:57:25 gotcha 2011-11-17T07:57:46 http://aichallenge.org/visualizer.php?game=70433 winner for the most living ants contest. 2011-11-17T07:57:53 *** u_ has quit IRC (Ping timeout: 252 seconds) 2011-11-17T07:57:53 *** u__ is now known as u_ 2011-11-17T07:58:30 Also are any of you using the Ant engine academically? I'm comparing different multi-agent algorithms using the ant engine. Curious if anyone out there has some academic bots out there. 2011-11-17T07:59:08 I used to use Robocup Rescue, but that damn thing is about as reliable as.. well something not very reliable. 2011-11-17T07:59:15 winning seems to be academic for my bot :) 2011-11-17T07:59:17 bugnuts2: oh my god, it's full of stars 2011-11-17T08:00:19 Heh. Winning the contest doesn't seem to be an excercise in AI, but an excercise in strategy to me. 2011-11-17T08:01:33 In my academic bots I don't allow such things as shared memory between ants unless there is a communications channel. All knowing does not benefit research. 2011-11-17T08:02:03 g0llum: ill try something 2011-11-17T08:02:09 what is the win32 define? 2011-11-17T08:02:15 __WIN32__ ? 2011-11-17T08:02:24 mcstar: predef.sf.net 2011-11-17T08:02:45 http://sourceforge.net/apps/mediawiki/predef/index.php?title=Operating_Systems#Windows 2011-11-17T08:02:47 Zao: thx 2011-11-17T08:03:05 yi think it does not like the precompiled libqtmain.a 2011-11-17T08:03:19 elderwolf: applying agent technology is severe overkill, or are there interesting `epistemic' elements to this environment? 2011-11-17T08:03:38 If you're using Qt, they have some Q_OS_ASDF defines too. 2011-11-17T08:04:00 Zao, yep, qt problem 2011-11-17T08:05:29 digger3: I'm a Ph.D. student doing his dissertation. As far as the contest, yes it's overkill. As far as having a nice simulation with randomness and seeds built in with a community behind it, it suits my needs just fine. I work with task allocation in AI, so it's a nice environment for me to do some comparisons. 2011-11-17T08:06:35 g0llum: can you try it again? 2011-11-17T08:06:54 wait.. 2011-11-17T08:06:58 elderwolf: ah ok, I'm also a Ph.D. student with too many GOOFAI-trauma's to truly care about agent technology :D 2011-11-17T08:07:00 Also, do the games always rotate so slowly? 2011-11-17T08:07:10 *** ace1010 has quit IRC (Quit: Page closed) 2011-11-17T08:07:18 elderwolf: they do :( 2011-11-17T08:07:22 elderwolf: you mean 600 mins between games? yes, the official servers are way underpowered 2011-11-17T08:07:33 elderwolf: everybody is testing on tcp servers instead 2011-11-17T08:07:37 elderwolf: depends on the number of allocated workers: http://aichallenge.org/server_stats.php 2011-11-17T08:08:17 tcpants 2011-11-17T08:08:24 digger3: i'm not familiar with the term GOOFAI. 2011-11-17T08:08:26 err, http://tcpants.com/howto 2011-11-17T08:08:39 You can get games at a much fsater rate there 2011-11-17T08:08:57 pairofdice: nice, thanks. 2011-11-17T08:09:17 elderwolf: whoops, I meant GOFAI http://en.wikipedia.org/wiki/GOFAI 2011-11-17T08:09:23 pairofdice: i'm tired of pitting my bot against greedy, lefty, etc... and against itself. 2011-11-17T08:09:40 elderwolf: prepare to be severly slaughtered though :) 2011-11-17T08:10:20 elderwolf: it will be nice to have a bot with an original methodology on there 2011-11-17T08:10:29 I'm looking forward to slaughtering you :) 2011-11-17T08:10:29 mcstar, compiles, but http://pastebin.com/jJvzVS4m 2011-11-17T08:10:53 digger3: ahh, cool. I'm interested in GOFAI in my own projects and research, however my university is heavy into multi-agents so it's easier to work in that environment. 2011-11-17T08:11:24 digger3: Also apparently my first dissertation ideas have been done to death recently. 2011-11-17T08:11:32 g0llum: weird .c.o' 2011-11-17T08:11:38 shouldnt be c 2011-11-17T08:11:55 after a bit of googling it might be a c vs. c++ issue 2011-11-17T08:12:01 i dont know 2011-11-17T08:12:04 sad 2011-11-17T08:12:30 digger3: My passion is designing AI systems that do not act fully logical, but mimic the more irrational human mind from time to time. Sometimes you just need to rush the enemey with your rag tag team cause they killed your favorite dog. 2011-11-17T08:12:46 *** antimatroid has joined #aichallenge 2011-11-17T08:12:54 elderwolf: GOFAI is mostly a dead methodology though, at least that's the general opinion :) 2011-11-17T08:13:13 elderwolf: naturally, otherwise it gets into boring alpha-beta/minimax stalemates 2011-11-17T08:13:19 *** twymer has joined #aichallenge 2011-11-17T08:13:42 digger3: if i only had time during undergrad to double major in statistics. mmmmmm 2011-11-17T08:14:38 elderwolf: the scary part is, my major was AI and I never had any statistics. The dominant environment were sucky logic-based systems. 'We don't need no stinking statistics here' was the general opinion at the time I guess 2011-11-17T08:15:23 "stinking statistics" 2011-11-17T08:15:28 Since we're on the subject, what people use for things such as "decide whether to attack this anthill or not"? 2011-11-17T08:15:34 could be the name of a progressive rock band 2011-11-17T08:15:42 mcstar: lol, it probably is :) 2011-11-17T08:16:36 retybok: most bots that I've seen perform some for of Kami-kaze when finding an enemy ant hill 2011-11-17T08:16:48 retybok: my decision making is pretty much along the lines of: attack it 2011-11-17T08:17:14 of course the individual ants try not to suicide too blatantly when they get close 2011-11-17T08:17:15 mcstar, there's a rule for c.o, and one for C.o, and win is not case sensitive, i just removed it 2011-11-17T08:17:16 digger3: Bah. I wanted to create a stastical human-like chat bot that would maintain a statistical emotion matrix. Sometimes input would evoke illogical responses. 2011-11-17T08:17:28 elderwolf: good luck with that :) 2011-11-17T08:17:40 digger3: Minthos that's what I do as well, but it's not always the best decision :) 2011-11-17T08:17:50 g0llum: i see, does it work now 2011-11-17T08:18:15 I'm sure there are some good algorithms for that problem, but I don't know any :-/ 2011-11-17T08:18:26 retybok: agreed, you and the poor lad who you're attacking usually end up > 3rd in larger matches, it's basically suicide while getting points for razing 1 hill 2011-11-17T08:18:34 digger3: yah, that was canned due to too many people doing AI with emotions and the need for a huge group to determine what was true emotional responses and opinion and such. 2011-11-17T08:18:43 digger3: exactly 2011-11-17T08:18:49 Damn philosophy. 2011-11-17T08:18:55 plus it's not very nice for the poor lad in question :) 2011-11-17T08:18:56 nope, it boils down to undefined reference to `__gxx_personality_sj0' when linging libqtmain.a 2011-11-17T08:19:19 retybok: agreed, I should rephrase that.... s/lad/ENEMY TERRORST!!!!!/ 2011-11-17T08:19:56 i bet it works, if irecompie qt, but i won't ;) 2011-11-17T08:20:05 ok 2011-11-17T08:20:15 thx 2011-11-17T08:20:16 Enemy CombatAnt 2011-11-17T08:20:34 ill ask if somebody else has qt + gcc on windows 2011-11-17T08:20:51 retybok: triggering a suicide run when you've got some percentage of map controll makes sense. Given the relation between map control, food gathering ability and number of ants 2011-11-17T08:20:56 guys does anybody else have qt + mingw or something on windows? 2011-11-17T08:21:08 pairofdice: nice1 :) 2011-11-17T08:21:10 nope :p 2011-11-17T08:21:33 mcstar: i might start writing a bot tonight 2011-11-17T08:21:43 digger3: exactly, but your limited knowledge of ennemy players makes it a lot harder 2011-11-17T08:21:45 antimatroid: hey 2011-11-17T08:21:48 you back? 2011-11-17T08:21:50 my ride got cut short, rack broke :( 2011-11-17T08:21:53 how was the trip? 2011-11-17T08:21:59 oh 2011-11-17T08:21:59 *** savaron has quit IRC (Ping timeout: 265 seconds) 2011-11-17T08:22:01 good, we managed over 200km in the end 2011-11-17T08:22:14 thats 1 day right? 2011-11-17T08:22:17 4 2011-11-17T08:22:25 we had lots of weight on the back of the bikes and LOTS of up hill 2011-11-17T08:22:25 hm 2011-11-17T08:22:30 ah ok 2011-11-17T08:22:49 retybok: calculating the % of map control is relatively easy though 2011-11-17T08:23:02 antimatroid: do you have qt? 2011-11-17T08:23:03 retybok: assuming decent exploration capabilities 2011-11-17T08:23:07 mcstar: nope 2011-11-17T08:23:18 sure? are you all gtk? 2011-11-17T08:23:26 our tron state evaluation function depended on map control 2011-11-17T08:23:30 it's very easy to calculate 2011-11-17T08:23:45 qt is what people use to make guis from what i know? 2011-11-17T08:23:54 i don't think i've ever used it other than quickly installing it 2011-11-17T08:23:55 yes, but you probably have it 2011-11-17T08:24:08 digger3: yeah, that's pretty easy, but it doesn't protect you against one guy piling up a huge army in an unexplored corner of the map 2011-11-17T08:24:12 i also use ubuntu these days? i have no idea about these things :P 2011-11-17T08:24:15 mcstar: I've built Qt with gcc on Windows. It's great fun. 2011-11-17T08:24:26 Whatever you do, make sure you're using configure.exe and not ./configure. 2011-11-17T08:24:41 The latter is only for Real Operating Systems. 2011-11-17T08:24:46 Zao: do you have that environment still? 2011-11-17T08:24:51 antimatroid: qt is used to make Guis, but also includes other functionalities 2011-11-17T08:25:01 retybok: right, so any unexplored part is full with enemy ants of opponent X with some probability P 2011-11-17T08:25:02 mcstar: It's not really usable, it's 64-bit Qt 4.8.0 2011-11-17T08:25:17 mcstar: A garden variety install of TDM-gcc ought to do the trick. 2011-11-17T08:25:29 Zao: could you try to compile my source? 2011-11-17T08:25:39 mcstar: I'm at work, sadly. 2011-11-17T08:25:44 ah ok 2011-11-17T08:25:44 Didn't leave my home box on. 2011-11-17T08:25:48 * antimatroid is going for a shower, it's been like 5 days :\ 2011-11-17T08:25:49 What source? 2011-11-17T08:25:57 antimatroid: GLHF! 2011-11-17T08:26:14 Zao: https://github.com/liquid-phynix/ants-fight.git 2011-11-17T08:26:22 Zao: ? i've been camping/riding :P 2011-11-17T08:26:29 mcstar: -.git 2011-11-17T08:26:32 i'm not that scummy 2011-11-17T08:26:40 *** g0llum has quit IRC (Ping timeout: 240 seconds) 2011-11-17T08:26:48 ? 2011-11-17T08:27:00 (in the link, github URLs do not contain .git as suffix) 2011-11-17T08:27:02 its the link for cloning, of 2011-11-17T08:27:08 Ah. 2011-11-17T08:27:13 course if you want to browse delete back :D 2011-11-17T08:28:28 digger3: but even then, the problem is what to do with all this info 2011-11-17T08:28:45 suppose there are five players on the map, with roughly equal forces. Do you attack your neighbor? 2011-11-17T08:30:19 ideally you would attack whomever is getting attacked by someone else 2011-11-17T08:30:25 good luck figuring that out 2011-11-17T08:30:55 *** twymer has quit IRC (Ping timeout: 258 seconds) 2011-11-17T08:31:42 I suppose game theory has some interesting answers, but I unfortunately don't know much about game theory 2011-11-17T08:32:23 or actually the best target is the player attacking someone else, not the player getting attacked 2011-11-17T08:32:32 but again, good luck figuring it out :) 2011-11-17T08:32:54 the best target is the target with the lowest degrees of freedom 2011-11-17T08:42:04 *** yoden has joined #aichallenge 2011-11-17T08:46:49 *** twymer has joined #aichallenge 2011-11-17T08:47:04 *** Antimony has joined #aichallenge 2011-11-17T08:57:45 *** imp5imp5 has quit IRC (Ping timeout: 265 seconds) 2011-11-17T08:59:49 *** Areks has quit IRC (Ping timeout: 240 seconds) 2011-11-17T09:02:35 *** Anilm3 has quit IRC (Ping timeout: 265 seconds) 2011-11-17T09:03:04 *** twymer has quit IRC (Ping timeout: 240 seconds) 2011-11-17T09:04:45 *** ltriant has joined #aichallenge 2011-11-17T09:15:59 *** Antimony has quit IRC (Ping timeout: 252 seconds) 2011-11-17T09:20:10 *** TheLinker has joined #aichallenge 2011-11-17T09:23:56 *** Antimony has joined #aichallenge 2011-11-17T09:26:07 *** twymer has joined #aichallenge 2011-11-17T09:27:51 *** UncleVasya has joined #aichallenge 2011-11-17T09:30:08 *** amstan has joined #aichallenge 2011-11-17T09:30:08 *** ChanServ sets mode: +o amstan 2011-11-17T09:31:25 *** Antimony_ has joined #aichallenge 2011-11-17T09:34:17 *** praveen_ has joined #aichallenge 2011-11-17T09:34:55 *** Antimony has quit IRC (Ping timeout: 248 seconds) 2011-11-17T09:35:05 *** Antimony_ is now known as Antimony 2011-11-17T09:39:12 *** Antimony has quit IRC (Ping timeout: 245 seconds) 2011-11-17T09:52:53 thankyou 24 hours mcdonalds 2011-11-17T09:54:28 *** Anilm3 has joined #aichallenge 2011-11-17T09:55:26 +1 2011-11-17T09:55:39 * digger3 is too scared to even consider mcstar 2011-11-17T09:55:40 Hi 2011-11-17T09:55:46 *mcdonalds 2011-11-17T09:55:55 haha 2011-11-17T09:56:02 am i not sweet and edible? 2011-11-17T09:56:07 damn you [TAB] 2011-11-17T09:56:12 McLeopold: is sweeter... 2011-11-17T09:56:42 mcstar: maybe... do you have high blood sugar? ;p 2011-11-17T09:56:51 digger3: why on earth would you try to tab-complete macdonalds: 2011-11-17T09:56:54 ? 2011-11-17T09:57:08 probably i do, i drink too much tea with too much sugar 2011-11-17T09:57:50 mcstar: excellent question, I guess the only sensible explanation would be that I frequently visit channels with mcdonalds as a nick. Either that or I did something I cannot explain myself. 2011-11-17T09:58:53 :) 2011-11-17T09:59:06 Let's face it. You are l a lazy persone and hope that Universe will finish your job. Even if there is no reasons to hope for that :) 2011-11-17T09:59:48 lazy programmers, are good programmers 2011-11-17T10:00:42 Yep, russians say: 'laziness is an engine of the progress'. 2011-11-17T10:01:00 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-11-17T10:02:41 It means that in case man must but doesn't want to do something he invents a device for this job. 2011-11-17T10:03:29 I need to invent a device that earns me money so I can program for fun instead of for money 2011-11-17T10:06:37 You and everybody else :) 2011-11-17T10:06:58 Just replace program 2011-11-17T10:07:43 *** NightExcessive has quit IRC (Ping timeout: 255 seconds) 2011-11-17T10:11:24 I need to invent a device that earns me money so I can conjugate_imperative($PREFERRED_ACTIVITY) for fun instead of for money 2011-11-17T10:14:19 how sad 2011-11-17T10:14:19 Rank: 6462 Skill: -0.00 2011-11-17T10:14:24 -0 2011-11-17T10:14:28 that's got to be bad 2011-11-17T10:14:59 means no games played 2011-11-17T10:15:23 I was just joking Mr. Obvious 2011-11-17T10:15:43 Doctor Obvious to you 2011-11-17T10:16:19 actually just Mr. 2011-11-17T10:16:26 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-11-17T10:16:39 even though I'm arrogant enough for at least one phd 2011-11-17T10:16:55 *** Rinum has joined #aichallenge 2011-11-17T10:17:21 *** moconne1 has joined #aichallenge 2011-11-17T10:17:31 I'm trying to add ants to a map, how do I do that? 2011-11-17T10:18:03 as in manually place ants 2011-11-17T10:20:09 letters 2011-11-17T10:20:19 or numbers? 2011-11-17T10:20:26 yep numbers 2011-11-17T10:20:31 or letters? 2011-11-17T10:20:49 ive writen at least 2 map loader functions but i ignore anything that isnt % 2011-11-17T10:21:31 *** sigh has quit IRC (Remote host closed the connection) 2011-11-17T10:23:13 g0lllum 2011-11-17T10:23:16 you left 2011-11-17T10:23:27 i think i have a fix 2011-11-17T10:24:03 @later tell g0llum ping me when youre back 2011-11-17T10:24:04 mcstar: Job's done. 2011-11-17T10:24:52 *** bishoco has joined #aichallenge 2011-11-17T10:34:11 *** bishoco has quit IRC (Quit: Page closed) 2011-11-17T10:35:03 if I put in #s or capital letters, I get hills with ants? 2011-11-17T10:35:15 is it possible to just get ants 2011-11-17T10:35:51 oh 2011-11-17T10:35:55 righty 2011-11-17T10:36:02 i forgot we have hills now 2011-11-17T10:36:15 im not sure 2011-11-17T10:36:35 have a look at the test map 2011-11-17T10:36:42 I think it has some 2011-11-17T10:39:05 I think A, B, ... are hills and 1, 2, ... are ants 2011-11-17T10:39:28 judging from test.map and tutorial.map 2011-11-17T10:39:34 http://www.youtube.com/watch?v=tCRPUv8V22o&feature=player_embedded 2011-11-17T10:39:49 theres a great number after 5 minutes 2011-11-17T10:40:00 I mean 0, 1, ... of course, not 1, 2, ... 2011-11-17T10:40:13 *** Nbooo has joined #aichallenge 2011-11-17T10:41:00 Anilm3: We're tied for last place!!! WOOOO 2011-11-17T10:41:27 *** okan_ has joined #aichallenge 2011-11-17T10:44:05 *** praveen_ has quit IRC (Ping timeout: 265 seconds) 2011-11-17T10:50:28 *** kapoc has joined #aichallenge 2011-11-17T10:55:00 *** ikaros has joined #aichallenge 2011-11-17T11:01:44 *** roflmao has joined #aichallenge 2011-11-17T11:02:02 *** xathis has joined #aichallenge 2011-11-17T11:02:51 *** moconne1 has quit IRC (Quit: Page closed) 2011-11-17T11:03:44 *** Harpyon has joined #aichallenge 2011-11-17T11:04:59 *** kapoc has left #aichallenge 2011-11-17T11:05:21 *** olexs has quit IRC (Quit: Leaving.) 2011-11-17T11:05:29 *** UncleVasya has quit IRC (Ping timeout: 244 seconds) 2011-11-17T11:08:28 *** smiley1983 has joined #aichallenge 2011-11-17T11:12:27 *** mleise has joined #aichallenge 2011-11-17T11:15:53 *** aerique has quit IRC (Quit: ...) 2011-11-17T11:18:02 *** olexs has joined #aichallenge 2011-11-17T11:19:57 *** olexs has quit IRC (Client Quit) 2011-11-17T11:21:20 *** roflmao has quit IRC (Remote host closed the connection) 2011-11-17T11:25:23 anyone knows if theres a way to download maps from fluxid? the links are only html files 2011-11-17T11:25:57 I made a perl script to import them, can send it to you if you want, pretty straight forward 2011-11-17T11:26:20 sure that would be nice pastebin or email? 2011-11-17T11:26:44 never used pastebin, sounds like a nice occasion to start 2011-11-17T11:27:03 just paste it @ pastebin.com :) 2011-11-17T11:27:42 ikaros: those maps are in aichallenge repo 2011-11-17T11:27:50 and to download on aichallenge.org 2011-11-17T11:28:01 so no need for the script then... 2011-11-17T11:28:05 hm 2011-11-17T11:28:14 on aichallenge.org i didnt find them 2011-11-17T11:28:16 I should have asked here before writing it 2011-11-17T11:28:27 http://aichallenge.org/maps.php 2011-11-17T11:28:34 http://aichallenge.org/maps.php 2011-11-17T11:28:36 yup 2011-11-17T11:28:39 well 2011-11-17T11:28:43 anyway, it had been a long time since I touched perl so it was good practice ;) 2011-11-17T11:28:58 but, if you have aichallenge repo cloned locally 2011-11-17T11:28:58 theres no mmaze_02p_01.map e.g. 2011-11-17T11:29:02 i have 2011-11-17T11:29:05 you have those maps in maps directory ;P 2011-11-17T11:29:42 no :/ 2011-11-17T11:29:54 tools/maps right? 2011-11-17T11:30:00 mmaze are from multimaze directory 2011-11-17T11:30:07 ok so just names are different? 2011-11-17T11:30:23 yeah, because there are to directories with same-name map files 2011-11-17T11:30:42 hm ill check out which map this should be.. 2011-11-17T11:30:43 but on tcp those are put in one directory, so from multi have just added one letter 2011-11-17T11:30:54 strange that its multi hill because it isnt 2011-11-17T11:30:58 very irritating 2011-11-17T11:31:16 i mean, mmaze are from multi_hill_maze directory 2011-11-17T11:31:41 yea i understand 2011-11-17T11:31:49 but http://ants.fluxid.pl/map/mmaze_02p_01.map 2011-11-17T11:31:50 ikaros: then maybe it't other way arouns ;) 2011-11-17T11:31:54 where is that multi hill? 2011-11-17T11:31:55 lol 2011-11-17T11:32:03 heh 2011-11-17T11:32:16 *** NightExcessive has joined #aichallenge 2011-11-17T11:32:41 ikaros: yeah, maze* are multihill, and mmaze are single-hill, sorry 2011-11-17T11:32:47 *** roflmao has joined #aichallenge 2011-11-17T11:33:32 well then i dont have that map 2011-11-17T11:33:54 guess ill have to try a git pull on the aichallenge thingy 2011-11-17T11:34:00 ok maybe this will be faster after all 2011-11-17T11:34:01 http://pastebin.com/yqUxA0ui 2011-11-17T11:34:21 I hope there isn't any perl purist around 2011-11-17T11:34:27 I guess it's poorly written 2011-11-17T11:34:48 yeah, many times just writing a script is faster than looking for real solution, i agree ;) 2011-11-17T11:35:13 and it help to keep practicing ;) 2011-11-17T11:35:15 https://github.com/aichallenge/aichallenge/tree/epsilon/ants/maps/multi_hill_maze 2011-11-17T11:35:20 you ugly scripters 2011-11-17T11:36:04 https://github.com/aichallenge/aichallenge/blob/epsilon/ants/maps/multi_hill_maze/maze_02p_01.map http://ants.fluxid.pl/map/maze_02p_01.map look the same 2011-11-17T11:36:13 wow fluxid thx i have a whole different directory structur 2011-11-17T11:36:19 guess ill just clone this again 2011-11-17T11:36:23 ikaros: no, just pull 2011-11-17T11:36:32 theres something wrong :) 2011-11-17T11:36:33 there were many changes in repo sooooo 2011-11-17T11:36:36 with my git repo 2011-11-17T11:36:38 you should pull frequently 2011-11-17T11:36:43 yea 2011-11-17T11:36:43 ah 2011-11-17T11:37:03 i need to fork it anyways for submitting a pull request soon 2011-11-17T11:37:13 so theres finally a reason to do it :) 2011-11-17T11:39:59 *** roflmao has quit IRC (Read error: Connection reset by peer) 2011-11-17T11:41:27 *** roflmao has joined #aichallenge 2011-11-17T11:41:55 *** mj41 has quit IRC (Ping timeout: 260 seconds) 2011-11-17T11:42:19 *** dauryg has joined #aichallenge 2011-11-17T11:42:39 *** Antimony has joined #aichallenge 2011-11-17T11:42:54 *** tobym has quit IRC (Remote host closed the connection) 2011-11-17T11:45:09 *** JorgeB has quit IRC (Ping timeout: 244 seconds) 2011-11-17T11:45:38 my ant was juuuust about to raze A's hill http://ants.fluxid.pl/replay.3003 2011-11-17T11:46:50 *** dauryg has quit IRC (Client Quit) 2011-11-17T11:47:46 Rinum++ :) 2011-11-17T11:53:35 *** Antimony has quit IRC (Ping timeout: 248 seconds) 2011-11-17T11:54:40 *** smiley1983 has quit IRC (Ping timeout: 260 seconds) 2011-11-17T11:55:04 *** smiley1983 has joined #aichallenge 2011-11-17T11:55:52 *** dvladim has joined #aichallenge 2011-11-17T11:55:53 *** roflmao has quit IRC (Ping timeout: 276 seconds) 2011-11-17T11:57:20 *** JorgeB has joined #aichallenge 2011-11-17T11:58:28 *** JorgeB has quit IRC (Client Quit) 2011-11-17T11:58:45 *** JorgeB has joined #aichallenge 2011-11-17T11:58:49 *** agentcobra has joined #aichallenge 2011-11-17T11:58:54 *** mleyen has joined #aichallenge 2011-11-17T11:59:10 hi 2011-11-17T11:59:14 *** Antimony has joined #aichallenge 2011-11-17T11:59:47 *** smiley1983 has quit IRC (Ping timeout: 276 seconds) 2011-11-17T12:01:51 *** dvladim has quit IRC (Quit: Konversation terminated!) 2011-11-17T12:05:23 *** smiley1983 has joined #aichallenge 2011-11-17T12:05:29 *** roflmao has joined #aichallenge 2011-11-17T12:11:17 *** agentcobra has quit IRC (Ping timeout: 252 seconds) 2011-11-17T12:13:10 hm with the newest version playgame.py fails to run my old bot.. 2011-11-17T12:13:16 never change a running system :) 2011-11-17T12:19:26 http://wurstcaptures.untergrund.net/music/ 2011-11-17T12:19:46 *** NotABug has joined #aichallenge 2011-11-17T12:20:56 mcstar: yesterday I was looking for that to show it to a friend, what a coincidence, thanks 2011-11-17T12:21:05 :D 2011-11-17T12:21:17 (t>>8)*(-t>>2)&(0.02*t) 2011-11-17T12:21:19 try this 2011-11-17T12:21:26 *** roflmao has quit IRC (Ping timeout: 244 seconds) 2011-11-17T12:21:49 nice 2011-11-17T12:23:23 *** mj41 has joined #aichallenge 2011-11-17T12:23:55 mcstar: police ((t>>8)+(t>>12))*(-t>>2) 2011-11-17T12:24:11 wait, ive got to save these 2011-11-17T12:26:21 *** g0llum has joined #aichallenge 2011-11-17T12:28:13 mcstar,Anilm3: what are those? 2011-11-17T12:28:35 grwip: fun 2011-11-17T12:28:35 http://wurstcaptures.untergrund.net/music/ this? 2011-11-17T12:28:42 it's music!! 2011-11-17T12:28:42 :D 2011-11-17T12:28:56 try out equations and generate sound 2011-11-17T12:28:56 sorry got it, didn't think the link was related... 2011-11-17T12:28:59 nice 2011-11-17T12:29:35 so cool!!! I used to make experimental music with my laptop, this brings a lot of nice memories 2011-11-17T12:29:59 blee bloo blee 2011-11-17T12:30:04 left: (t>>9)&(0.02*t+(t-1000000)%23) 2011-11-17T12:30:10 right: (t>>9)&(0.02*t+t%22) 2011-11-17T12:30:46 *** amstan has joined #aichallenge 2011-11-17T12:30:46 *** jasox has joined #aichallenge 2011-11-17T12:30:46 *** ChanServ sets mode: +o amstan 2011-11-17T12:31:08 *** lorill has joined #aichallenge 2011-11-17T12:31:14 *** TheLinker has quit IRC (Ping timeout: 252 seconds) 2011-11-17T12:31:16 I have a dumb question 2011-11-17T12:31:38 is there any guarantee to have one ant per hill on the first turn ? 2011-11-17T12:31:45 yes 2011-11-17T12:31:51 only one, and always on the hill ? 2011-11-17T12:32:08 yes 2011-11-17T12:32:13 *** __SPIRIT__ <__SPIRIT__!515f1b02@gateway/web/freenode/ip.81.95.27.2> has joined #aichallenge 2011-11-17T12:32:35 I didn't see this in the game spec 2011-11-17T12:33:22 Three shall be the number thou shalt count, and the number of the counting shall be three. Four shalt thou not count, neither count thou two, excepting that thou then proceed to three. Five is right out. Once the number three, being the third number, be reached, then lobbest thou thy Holy Hand Grenade of Antioch towards thy foe, who being naughty in My sight, shall snuff it 2011-11-17T12:33:32 *** JorgeB_ has joined #aichallenge 2011-11-17T12:33:34 *** JorgeB has quit IRC (Read error: Operation timed out) 2011-11-17T12:33:34 *** JorgeB_ is now known as JorgeB 2011-11-17T12:34:26 mcstar: that's a guarantee? 2011-11-17T12:34:32 no 2011-11-17T12:34:56 the only logical guarantee that you should have is that you have at least 1 ant and at least 1 hill 2011-11-17T12:35:15 if it's on the hill or not i don't think it matters 2011-11-17T12:35:21 the always on the hill is probably very true 2011-11-17T12:35:31 not with a scenario 2011-11-17T12:35:37 whats that? 2011-11-17T12:35:39 but if none of your ants see your hill you don't get the location 2011-11-17T12:35:52 *** Antimony has quit IRC (Ping timeout: 245 seconds) 2011-11-17T12:35:57 amstan: something changed again? 2011-11-17T12:36:01 so it'd be annoying, sinc eyou can't be certain about the amount of hills you have until a new ant spawns twice on your initial hill 2011-11-17T12:36:05 Yexo: that's right 2011-11-17T12:36:20 argh. So I can't assume that my ant locations at first turn is the same as my hill locations ? 2011-11-17T12:36:39 *** Antimony has joined #aichallenge 2011-11-17T12:36:53 lorill: why are you doing that assumption? the engine tells you where your hills are 2011-11-17T12:36:53 lorill, i assume that 2011-11-17T12:37:13 only if they are in sight 2011-11-17T12:37:41 btw amstan i did a pull request for that lualib 2011-11-17T12:37:42 but each hill has an ant on it on the first turn 2011-11-17T12:37:49 ikaros: link? 2011-11-17T12:37:52 i hope so.. else im in trouble 2011-11-17T12:37:58 ehm mom :) 2011-11-17T12:38:03 moment 2011-11-17T12:38:08 <@amstan> not with a scenario 2011-11-17T12:38:20 oh true 2011-11-17T12:38:30 that would be cool actually 2011-11-17T12:38:32 there arent scenarios played in the contest arent they 2011-11-17T12:38:37 have a game where you have to find your hill first 2011-11-17T12:38:54 ikaros: nothing in the rules or the specs says that that cannot change though ;) 2011-11-17T12:38:58 amstan, https://github.com/aichallenge/aichallenge/pull/373 2011-11-17T12:39:08 @amstan: the starter kit only tells me there's an ant there 2011-11-17T12:39:09 lorill: I'm sorry Dave, err lorill; I cannot 'amstan:'. 2011-11-17T12:39:17 hehe bqf true 2011-11-17T12:39:27 lorill: what starter kit? 2011-11-17T12:39:28 Arggg. I'm winning my matches, but I'm timing out despite my failsafes. 2011-11-17T12:39:35 the alternative java one I guess 2011-11-17T12:39:43 but I can change it, so it doesn't matter 2011-11-17T12:39:54 if you have hill support you should already know where your hill is, if you see it 2011-11-17T12:39:57 so on the first turn, I get ants + hills, even if ants are on my hills ? 2011-11-17T12:40:16 what if you start with 3 ants? do you assume you have 3 hills? 2011-11-17T12:40:17 the point being, if the ant is on it, I don't see it (at least that was my understanding) 2011-11-17T12:40:20 *** grwip has quit IRC (Quit: Leaving) 2011-11-17T12:40:26 for the moment, yes 2011-11-17T12:40:30 if the ant is on it, it's a capital letter 2011-11-17T12:40:31 that's why I wanted to be sure 2011-11-17T12:40:34 read the map specs 2011-11-17T12:40:52 if you're player 1 2011-11-17T12:40:57 0 is your hill without an ant on it 2011-11-17T12:41:00 a is a normal ant 2011-11-17T12:41:10 A is an ant on top of a hill you have 2011-11-17T12:42:07 contestbot: later tell janzert please merge this: https://github.com/aichallenge/aichallenge/pull/373 2011-11-17T12:42:07 amstan: OK 2011-11-17T12:42:09 aha 2011-11-17T12:42:21 this shuld be in the "Turn Information:" section I guess, but ok 2011-11-17T12:42:31 turn information is another matter 2011-11-17T12:42:41 I'm lost then 2011-11-17T12:43:07 If I don't get the info at setup neither at turn info, when do I get it ? 2011-11-17T12:43:34 oh, i see how this is done 2011-11-17T12:43:37 you mean different things by turn info 2011-11-17T12:43:39 you don't get the map at all 2011-11-17T12:43:43 I know 2011-11-17T12:43:58 lorill: if you have an ant and a hill on the same spot, and assuming you don't see any water or food in first turn 2011-11-17T12:44:02 what you'll get is 2011-11-17T12:44:03 *** TheLinker has joined #aichallenge 2011-11-17T12:44:12 a 0 0 0 2011-11-17T12:44:13 h 0 0 0 2011-11-17T12:44:35 ok, so both are sent then 2011-11-17T12:44:45 only on the first turn, or always ? 2011-11-17T12:45:04 you always get sent what you see 2011-11-17T12:45:35 except water 2011-11-17T12:45:36 and do I always see my hill if one of my ant is on it ? 2011-11-17T12:45:52 lorill: yes, what gave you the impression you can't see it? 2011-11-17T12:47:04 g0llum: you there? 2011-11-17T12:48:07 yea back, connection problems 2011-11-17T12:48:22 can you check something for me? 2011-11-17T12:48:24 the fact that I can't see enemy hills if ants are on the same tile 2011-11-17T12:48:28 i'll try 2011-11-17T12:48:30 --enable-sjlj-exceptions 2011-11-17T12:48:36 try adding this to the flags 2011-11-17T12:48:39 so I guessed the same would be true for mine 2011-11-17T12:48:45 linker ? 2011-11-17T12:48:53 no compile 2011-11-17T12:48:59 no, compile 2011-11-17T12:49:12 lorill: if your ants are on their hills, you def can't see them, becasue the hills won't exist anymore 2011-11-17T12:49:34 man, so many misunderstandings 2011-11-17T12:49:35 mingw and your qt is compiled differently 2011-11-17T12:49:40 that was not at all what I meant 2011-11-17T12:50:25 if an enemy ant is on his own hill, and this tile is in my view, do I see ant + hill or only ant ? 2011-11-17T12:50:53 both 2011-11-17T12:51:03 (sorry to sound so dumb, english is not my mother tongue) 2011-11-17T12:51:17 *** liberforce has left #aichallenge 2011-11-17T12:51:36 dont worry, neither is for a lot of us 2011-11-17T12:51:56 then I don't understand this forum post 2011-11-17T12:51:57 http://forums.aichallenge.org/viewtopic.php?p=10073#p10073 2011-11-17T12:52:22 mcstar, made no diff, but my qt version is from 2009 2011-11-17T12:52:30 lorill: there's nothing that i said in there about hills 2011-11-17T12:52:33 erm, no, this one 2011-11-17T12:52:34 http://forums.aichallenge.org/viewtopic.php?p=10175#p10175 2011-11-17T12:52:35 amstan: nice idead 2011-11-17T12:52:38 -s 2011-11-17T12:52:40 -d 2011-11-17T12:52:42 2 posts later 2011-11-17T12:53:08 lorill: i'm not sure how he wants to do that 2011-11-17T12:53:26 yeah sure, if we can always see hills even if an ant is on it :) 2011-11-17T12:53:33 that changes a lot of my plans 2011-11-17T12:53:48 I'll check if my bot can actually use this info 2011-11-17T12:53:49 g0llum: im not sure whether it is enabled i your qt or in the mingw compiled binary 2011-11-17T12:53:55 well, don't make your plans based on something a guy misunderstood about the contest 2011-11-17T12:53:59 and i'll remove my tentative to hide my hills 2011-11-17T12:54:02 so i guess maybe you need the opposite? 2011-11-17T12:54:06 try disabling it? 2011-11-17T12:54:13 heh, that's why I checked it here first :) 2011-11-17T12:55:14 thanks for the info ! 2011-11-17T12:55:54 *** praveen_ has joined #aichallenge 2011-11-17T12:56:25 Guess I made a record high of 750 ants in the hive. anybody got more than this? 2011-11-17T12:57:01 imagine the gravitational pull of those ants 2011-11-17T12:57:12 they move the moon out of its orbit 2011-11-17T12:59:01 aichallenge: Alexandru Stan epsilon * rbdc09ca / website/starter_packages.php : c starter package does have hill support - http://git.io/QVeX1A 2011-11-17T12:59:15 okan_: fixed 2011-11-17T12:59:26 okan_: next time try asking in here, other people can help too 2011-11-17T13:00:19 ok i will ask in here next time , thanks again 2011-11-17T13:01:34 mcstar, made an experiment and tried to compile one of the samples, got lots of errors related to SjLj, so i guess my qt version does work with that version of mingw 2011-11-17T13:01:46 *** Antimony_ has joined #aichallenge 2011-11-17T13:02:16 yep could be 2011-11-17T13:02:42 *** smiley1983 has quit IRC (Ping timeout: 258 seconds) 2011-11-17T13:02:57 *** smiley1983 has joined #aichallenge 2011-11-17T13:03:32 *** Antimony has quit IRC (Ping timeout: 252 seconds) 2011-11-17T13:03:41 guess you better ignore my errors, and wait until someone with a more recent version sents issues to github 2011-11-17T13:03:42 *** Antimony_ is now known as Antimony 2011-11-17T13:05:29 *** NoxiaZ^ has joined #aichallenge 2011-11-17T13:06:08 *** lorill has quit IRC (Quit: Page closed) 2011-11-17T13:06:13 g0llum: thx 2011-11-17T13:06:46 nvm, had my fun with it, and it works from ubuntu! 2011-11-17T13:07:02 any idea how this can happen: http://pastebin.com/S6CWgTeh if the path to the bot is correct? 2011-11-17T13:07:04 *** JorgeB has quit IRC (Quit: Textual IRC Client: http://www.textualapp.com/) 2011-11-17T13:07:14 ah, ok, so you could try it 2011-11-17T13:07:19 i have this issue with playgame.py since pulling a new version 2011-11-17T13:08:33 *** praveen_ has quit IRC (Quit: Page closed) 2011-11-17T13:08:54 Anilm3: btw, Arch packages are not unsigned anymore 2011-11-17T13:09:06 pacman v4 has package signing and over 60% of packages in the repos are already signed 2011-11-17T13:10:04 ['ikaros_bot_v8/Release/ikaros_bot'] 2011-11-17T13:10:13 ikaros: why is that in [] brackets? 2011-11-17T13:10:16 thestinger: thanks great news!! 2011-11-17T13:10:34 i have no idea 2011-11-17T13:10:45 may not be the problem, but i remember an error like that and i think it shuoldnt be in brackets 2011-11-17T13:10:48 is that python list stuff? 2011-11-17T13:10:59 but it shouldnt be a list 2011-11-17T13:11:10 'Failed to start {0}'.format(shell_command) 2011-11-17T13:11:16 shell command hsould be a string 2011-11-17T13:11:19 i just pass it as "ikaros_bot_v8/Release/ikaros_bot" 2011-11-17T13:11:26 tthat is inlined directly to {0} 2011-11-17T13:11:28 to playgame.py such as the other bot .. which works 2011-11-17T13:11:42 Anilm3: the new pacman is still in the [testing] repository, but mostly only because they haven't made a keyring package yet (pacman retrieves the keys from a keyserver when you install, and asks you to trust them - but you can manually download them using the signatures on the site instead) 2011-11-17T13:11:46 ok, im just guessing here 2011-11-17T13:11:56 very strange issue 2011-11-17T13:12:05 thestinger: but you dont use it right? 2011-11-17T13:12:29 I use Arch and I use [testing] :P 2011-11-17T13:12:35 i know 2011-11-17T13:12:40 but you dont use pacman v4 2011-11-17T13:12:42 I'll have to take a look 2011-11-17T13:12:52 or do you? 2011-11-17T13:12:54 mcstar: I do 2011-11-17T13:13:04 so what is th ebenefit? 2011-11-17T13:13:19 was there any incident that somebody made a malicious package? 2011-11-17T13:13:25 well, there are many improvements on the version in [extra], package signing is just one 2011-11-17T13:13:51 mcstar: kernel.org was hacked, and that's one of the Arch mirrors - someone could have tampered with packages 2011-11-17T13:13:58 there was no way to know if someone was doing stuff before 2011-11-17T13:14:05 when? 2011-11-17T13:14:25 http://www.h-online.com/open/features/Kernel-Log-more-details-on-the-kernel-org-hack-1371641.html 2011-11-17T13:14:47 well, that's a bad article (was just the first one on google news) 2011-11-17T13:15:17 theres a link in it 2011-11-17T13:15:21 so in august? 2011-11-17T13:15:26 bad news 2011-11-17T13:16:20 the git repositories are safe, b/c they are basically signed even if you aren't using gpg 2011-11-17T13:16:29 each commit has a hash based on every previous commit 2011-11-17T13:16:50 and there are lots of mirrors (every dev has at least one, yay DVCS) 2011-11-17T13:16:51 arch's weakness is aur 2011-11-17T13:17:00 you can easily make a malicious pkgbuild 2011-11-17T13:17:18 it would be pretty obvious i think 2011-11-17T13:17:32 i remember i read something in the forums that it happened once 2011-11-17T13:18:06 https://projects.archlinux.org/pacman.git/tree/NEWS there's the changelog, lots of new stuff in v4 2011-11-17T13:21:42 *** nle85 has joined #aichallenge 2011-11-17T13:24:05 *** epicmonkey has joined #aichallenge 2011-11-17T13:24:43 *** Antimony has quit IRC (Ping timeout: 260 seconds) 2011-11-17T13:25:34 *** __SPIRIT__ <__SPIRIT__!515f1b02@gateway/web/freenode/ip.81.95.27.2> has quit IRC (Quit: Page closed) 2011-11-17T13:26:29 god.. can't get this to work .. 2011-11-17T13:26:52 ikaros: why dont you clone again? 2011-11-17T13:27:00 hm? 2011-11-17T13:27:09 dont try to fix something if its broken 2011-11-17T13:27:38 ikaros: did you try a fresh clone of the rep? 2011-11-17T13:29:17 i forked it one hour ago.. 2011-11-17T13:32:36 working for me 2011-11-17T13:33:33 its just one bot which i cant start.. my last version 2011-11-17T13:35:59 oh 2011-11-17T13:36:05 is it executable? 2011-11-17T13:36:17 Failed to start 2011-11-17T13:36:21 *** ALplus has joined #aichallenge 2011-11-17T13:36:21 so the file is there 2011-11-17T13:36:48 *** goldcaddy77 has joined #aichallenge 2011-11-17T13:37:10 yea its compiled c++ binary 2011-11-17T13:37:12 Anybody know if there is any way to donate so that we can get quicker games? 2011-11-17T13:37:19 i even compiled again and started it manually to be sure 2011-11-17T13:38:12 i also printed the parameter from the python script and it looks good.. no list .. correct path 2011-11-17T13:39:33 goldcaddy77: talk to amstan 2011-11-17T13:40:56 weird... my new version always takes 4ms to 10ms for pre-combat stuff, then ~30ms for combat (but it will bail out if turn time is running out) - but it now times out on tcp :( 2011-11-17T13:40:59 ikaros: try full path 2011-11-17T13:41:05 or use ./ at least 2011-11-17T13:41:06 aichallenge: janzert epsilon * r4557e6d / worker/engine.py : Remove unused uninitialized variable - http://git.io/SjrG-Q 2011-11-17T13:41:32 thestinger: i had the same issue 2011-11-17T13:41:47 my bot runs very fast locally, and just hangs over tcp 2011-11-17T13:41:52 but not always 2011-11-17T13:42:37 http://ants.fluxid.pl/replay.3111 2011-11-17T13:42:43 it can definitely deal with lots of ants/enemies 2011-11-17T13:42:43 *** Antimony has joined #aichallenge 2011-11-17T13:42:54 *** patrisk has joined #aichallenge 2011-11-17T13:42:54 not sure why it times out now 2011-11-17T13:42:55 *** Fandekasp has quit IRC (Ping timeout: 248 seconds) 2011-11-17T13:43:17 aichallenge: Janzert epsilon * r456540e / setup/worker_setup.py : 2011-11-17T13:43:17 aichallenge: Merge pull request #373 from dbriemann/epsilon 2011-11-17T13:43:17 aichallenge: added liblua5.1-socket-dev for proper lua timer support - http://git.io/rGKqwA 2011-11-17T13:43:22 thestinger: are you using all the available time? Perhaps it's just a bit of network lag that causes your timeouts? 2011-11-17T13:43:44 *** Fandekasp has joined #aichallenge 2011-11-17T13:43:57 *** patrisk has quit IRC (Client Quit) 2011-11-17T13:44:03 but the turntime is set to 5 secs on tcp 2011-11-17T13:44:05 I think the timing out has more to do with the tcp client/server communication than with your bot. 2011-11-17T13:44:06 *** ihtisham has joined #aichallenge 2011-11-17T13:44:11 yes 2011-11-17T13:44:26 someone should give tcpclient.py a good hard look 2011-11-17T13:44:40 mcstar: turn time is 5 seconds, but your bot also gets 5000 ms as maximum turn time 2011-11-17T13:44:52 Yexo: but our bots finish very soon 2011-11-17T13:44:54 so if it uses say 4990 ms there is only 10ms available for network overhead 2011-11-17T13:45:01 mcstar: it definitely leaks threads sometimes 2011-11-17T13:45:10 well, maybe just sockets 2011-11-17T13:45:23 I've been checking with 'lsof -i' 2011-11-17T13:45:49 it sometimes leaks sockets, and for each one I think it also leaves a thread running and an instance of the bot 2011-11-17T13:45:52 >.< 2011-11-17T13:46:41 oh? bad script, go to your room 2011-11-17T13:46:57 mcstar, ./ didnt work but full path does it 2011-11-17T13:47:03 no idea why but thx 2011-11-17T13:47:04 =) 2011-11-17T13:47:06 ikaros: weird 2011-11-17T13:47:09 yea 2011-11-17T13:47:18 If anyone has suggestiosn for http://forums.aichallenge.org/viewtopic.php?f=25&t=1925 (Sometimes orders are ignored) I'd be very grateful 2011-11-17T13:47:32 there was a similar error couple of days ago, didnt work with relative paths 2011-11-17T13:47:44 the engine is still broken in this regard 2011-11-17T13:47:48 Yexo: food can block your moves 2011-11-17T13:47:55 thestinger: it's not that 2011-11-17T13:48:08 I've checked the logs. there is no food spawn, the tile is free 2011-11-17T13:48:52 well now that it works.. time to go productive again.. 2011-11-17T13:49:12 *** NoxiaZ^ has quit IRC (Ping timeout: 245 seconds) 2011-11-17T13:50:18 amstan, thx for merging .. is the lib already available on the server? 2011-11-17T13:51:26 ikaros: i'm not sure, janzert knows 2011-11-17T13:51:31 are people actually storing the state of their ants and comparing their state with what they get as input? 2011-11-17T13:51:41 they might not be unless the workers restart 2011-11-17T13:51:49 Yexo: nope 2011-11-17T13:52:23 Yexo: yep 2011-11-17T13:52:54 ikaros: there seems to be a bunch of workers booting right now 2011-11-17T13:52:56 of course i will not have assert's in my code 2011-11-17T13:53:07 lol, imagine you leave it there at the end of the contest 2011-11-17T13:53:13 I can easily disable that assert (my bot doesn't depend on it) 2011-11-17T13:53:17 and you dont win because your assert fails 2011-11-17T13:53:30 but currently it indicates either a bug in my bot or in the engine 2011-11-17T13:53:30 of cours,e but people make mistakes 2011-11-17T13:53:42 Yexo: yup 2011-11-17T13:53:54 and I've checked the input/ouput, which leads me (however unlikely) to believe it's a bug in the engine 2011-11-17T13:54:08 but then I'd expect that other people would've hit it already 2011-11-17T13:54:14 its not impossible just unllikely 2011-11-17T13:54:18 im checking it btw 2011-11-17T13:54:25 amstan, ok so when i upload my next version it will probably work :) thx 2011-11-17T13:54:28 ok, thanks :) 2011-11-17T13:55:05 yeah, the workers are updating now 2011-11-17T13:55:24 Yexo: if you compare the dead ant feedback you get as input with code that tracks all ants and removes - is it consistent? 2011-11-17T13:55:44 amstan: the worker list from worker_ssh includes a lot of machines that aren't actually workers 2011-11-17T13:55:50 janzert: i was just about to say that 2011-11-17T13:56:03 goldcaddy77: I don't track the daed ant input. for every ant in my input I check that it's either on a tile an ant moved to last turn or on a hill 2011-11-17T13:56:06 janzert: but there's no way of knowing, i would do the same thing manually 2011-11-17T13:56:07 I can't think of an easy way to fix it though 2011-11-17T13:56:16 janzert: unless we get a last_seen time in the db 2011-11-17T13:56:19 than I throw away all ants I didn't get in the input 2011-11-17T13:56:50 well in some days they all should have the lib installed right? that would be enough ;) 2011-11-17T13:56:50 really the only way to tell for sure would be to try logging in 2011-11-17T13:56:59 janzert: want me to try that? 2011-11-17T13:57:57 that's pretty heavy handed 2011-11-17T13:58:02 ah, I found a weird bug that causes an infinite loop :) 2011-11-17T13:58:12 was it an ant? 2011-11-17T13:58:23 xD 2011-11-17T13:58:30 yes actually 2011-11-17T13:59:01 Yexo: gotcha... i mean that's fine, but sometimes you want to do something when an ant dies, too 2011-11-17T13:59:07 that's why you'd want to keep track 2011-11-17T13:59:16 if an enemy ant spawns in a hill and I have only 1 ant next to that hill and there is no way for that ant to survive, a loop never ends :P 2011-11-17T13:59:39 goldcaddy77: I might do that when I need it. For now I don't need the state from the previous turn at all, but I still want to know why it isn't consistent 2011-11-17T13:59:50 janzert: hmm? 2011-11-17T14:00:31 Yexo: your log is weird, turn 152 info in intertwined with turn 153 info 2011-11-17T14:00:34 *** ihtisham has quit IRC (Quit: Page closed) 2011-11-17T14:00:40 but otherwise i see your problem 2011-11-17T14:00:48 *** NotABug has quit IRC (Ping timeout: 265 seconds) 2011-11-17T14:00:55 are you sure, the engine actually received that command? 2011-11-17T14:01:05 janzert: what was heavy handed? 2011-11-17T14:01:19 attempting an ssh login just to create a list of active servers 2011-11-17T14:01:27 mcstar: the command "o 65 77 N" is between two other commands that were executed 2011-11-17T14:01:39 i'm already doing almost that, with a tcp connection on port 22 2011-11-17T14:01:40 Yexo: how about a stream output? i think that catches everything the engine recieves 2011-11-17T14:01:44 the game was played on a tcp server, so I cannot be entirely sure what happened 2011-11-17T14:02:04 Yexo: but you log that, not the engine 2011-11-17T14:02:16 that log is from your bot right? 2011-11-17T14:02:17 http://paste.aichallenge.org/CoKNs/raw/ This was logged by me 2011-11-17T14:02:18 yes 2011-11-17T14:02:25 that isnt necessarily what the engine seen of that 2011-11-17T14:02:29 I write every line I read/write to input.txt / output.txt 2011-11-17T14:02:29 janzert: how about we apply that only to the active ssh list, as another filter? 2011-11-17T14:02:42 Yexo: so the problem i think is that that line got lost 2011-11-17T14:02:43 well, nvm, that would be the same thing 2011-11-17T14:02:48 and the engine is in fact correct 2011-11-17T14:03:03 the tcp connection is quite a small part of the full ssh setup 2011-11-17T14:03:07 mcstar: that's possible. Is there someway to replay a game given the html replay? 2011-11-17T14:03:14 but I also don't have a better solution 2011-11-17T14:03:53 janzert: do we care too much? I could pass ssh some flags to just stop sshing if there's a password 2011-11-17T14:04:00 janzert: when you try to login 2011-11-17T14:04:08 that way it's easier for -a 2011-11-17T14:04:22 btw, I didn't look too closely at the code but I think it is checking the 20 highest worker ids? 2011-11-17T14:04:25 just check the logs for workers phoning home 2011-11-17T14:04:26 Yexo: you can get your input from the visualizer 2011-11-17T14:04:29 janzert: it is 2011-11-17T14:04:31 theres a button called bot-in 2011-11-17T14:04:39 mcstar: yes, I've checked that input 2011-11-17T14:04:45 amstan: it should check the workers with the most recent activity 2011-11-17T14:04:45 that input is consistent with my own logged input 2011-11-17T14:04:46 but thats not enough i think 2011-11-17T14:04:58 for instance we've had worker 13 come back several times 2011-11-17T14:05:01 janzert: but unless i check the logs, i can't know that 2011-11-17T14:05:04 what I would actually need is my own output (I have my own log of that, but not what the server received) 2011-11-17T14:05:13 query by recent games 2011-11-17T14:05:16 Yexo: yeah i know, but i think that is not possible 2011-11-17T14:05:28 hmm, that could work 2011-11-17T14:05:33 ok 2011-11-17T14:05:36 i'll work on that today 2011-11-17T14:05:38 or rather sort your query by max game id 2011-11-17T14:05:50 just use the server_stats query with a different time frame if needed 2011-11-17T14:06:08 mcstar: I could in theory recreate it from the html, but that would mean I get the version with that move not executed,and I still wouldn't know if the engine received it 2011-11-17T14:06:24 McLeopold: ideally we should get workers that haven't played games yet in there too 2011-11-17T14:06:28 hmm, except it should include workers that don't have a game yet 2011-11-17T14:06:32 Yexo: but you could try looking at the json file, that is in the html too, but it would be the same as watching the replay 2011-11-17T14:06:35 right :) 2011-11-17T14:06:37 McLeopold: like workers that just phoned in for the install 2011-11-17T14:06:45 yes eaxctl 2011-11-17T14:06:45 i think it would be best to try the ssh stuff 2011-11-17T14:06:55 i don't think it'll be that much of a problem 2011-11-17T14:07:29 we're already tcping to port 22, ssh will act the same way when it fails, except it'll be better when it makes a connection 2011-11-17T14:07:31 one thing with the ssh stuff if you don't have it do a full login it doesn't really tell you if it's a worker does it? 2011-11-17T14:07:51 amstan: you could setup a ping for each server that calls the setup 2011-11-17T14:08:04 McLeopold: i already have that, problem is we're getting false positives 2011-11-17T14:08:08 oh 2011-11-17T14:08:28 so, we would need our own program to respond to pings 2011-11-17T14:08:28 *** roflmao has joined #aichallenge 2011-11-17T14:08:33 janzert: idk, i'm just thinking fail if public key was not accepted, or if password is required 2011-11-17T14:08:39 nice, my crappy pre-combat version on the site made it to rank 18 2011-11-17T14:08:42 ahh, yeah 2011-11-17T14:08:43 did fluxid figure out that sigma issue? 2011-11-17T14:08:48 janzert: that should only work for workers 2011-11-17T14:09:00 ok, doing that 2011-11-17T14:09:25 Fluxid: do you by chance keep more logs than the html replays from your games? 2011-11-17T14:09:35 *the games on your server 2011-11-17T14:11:02 amstan: btw, mysql says the planetwars playback table is corrupt 2011-11-17T14:11:04 this new visualizer is great 2011-11-17T14:11:26 *** Surya has joined #aichallenge 2011-11-17T14:11:27 janzert: ok then, i guess we'll have to do it again 2011-11-17T14:11:33 Hey everyone 2011-11-17T14:11:37 Hi 2011-11-17T14:11:39 janzert: i was hoping to copy it without taking jeff's site offline for 2 hours 2011-11-17T14:12:02 i guess i could stop mysql, move the db, then start it 2011-11-17T14:12:04 then download it 2011-11-17T14:12:05 http://pastebin.com/9AKJcRrc 2011-11-17T14:12:19 mcstar: I still have my console open frmo that game, so I am sure that tcpclient.py receive that line and send it to the server 2011-11-17T14:12:19 yeah, that might be best 2011-11-17T14:12:34 gotta run cya later 2011-11-17T14:12:35 Yexo: ? 2011-11-17T14:12:40 janzert: maybe use rsync while transferring, idk if ssh does checksums 2011-11-17T14:12:47 it only echos your bot's output 2011-11-17T14:12:53 you cant be sure that liine got to the tcp server 2011-11-17T14:12:55 I'm trying to set up a tcpServer and everything's going well (except that i had to uninstall python 3.2 to go back to 2.7) BUT I have an issue, I don't know where to find the website generated nor the games'data. If someone could help me with that. Thx 2011-11-17T14:13:04 or whatever happened to it after it got there 2011-11-17T14:13:17 mcstar: "print( client_mess ); sock.sendall( client_mess )" <- I'm pretty sure it got send to the server 2011-11-17T14:13:20 *** Lithosphere has quit IRC (*.net *.split) 2011-11-17T14:13:20 *** MikeyG has quit IRC (*.net *.split) 2011-11-17T14:13:20 *** jstemmer has quit IRC (*.net *.split) 2011-11-17T14:13:21 *** Minthos has quit IRC (*.net *.split) 2011-11-17T14:13:21 *** isbric has quit IRC (*.net *.split) 2011-11-17T14:13:21 *** aliceinwire has quit IRC (*.net *.split) 2011-11-17T14:13:25 *** MikeyG has joined #aichallenge 2011-11-17T14:13:27 *** isbric has joined #aichallenge 2011-11-17T14:13:29 *** Minthos has joined #aichallenge 2011-11-17T14:13:40 Yexo: i dont see it 2011-11-17T14:13:59 Surya, start webserver.py, and look on port 2080 2011-11-17T14:13:59 on what do you base your statement? 2011-11-17T14:14:04 *** durarara has joined #aichallenge 2011-11-17T14:14:40 if it's a tcp connection (as the name of the script implies), than either all that content got received or it it didn't 2011-11-17T14:14:45 there can't be random lines missing 2011-11-17T14:14:54 whathever the server did with it I can't be sure of indeed 2011-11-17T14:15:30 *** roflmao has quit IRC (Ping timeout: 244 seconds) 2011-11-17T14:16:53 Yexo: what is the content of client_mess? 2011-11-17T14:17:01 is it line based? or buffered? 2011-11-17T14:17:16 it's all lines including the final "go" line 2011-11-17T14:17:29 so it contains on whole turn 2011-11-17T14:17:33 yes 2011-11-17T14:17:57 *** Hexren has joined #aichallenge 2011-11-17T14:18:00 sorry, should've said that before. But that's basically why I was certain the server received the line 2011-11-17T14:18:00 *** Hexren has left #aichallenge 2011-11-17T14:18:04 *** analyst74 has joined #aichallenge 2011-11-17T14:18:06 you have to take a look at the other end of the itnerface 2011-11-17T14:18:07 it executed liens before and after 2011-11-17T14:18:43 where would I be able to find that? 2011-11-17T14:18:45 i guess there is another script that pipes from the socket to the engine? 2011-11-17T14:18:54 its in berak's tcp erver repo 2011-11-17T14:19:06 i think, dont take my word for it 2011-11-17T14:19:13 I'll try and find that, thanks 2011-11-17T14:20:08 *** Surya has quit IRC (Ping timeout: 265 seconds) 2011-11-17T14:24:48 hmm, while the normal repository usea a thread-safe Queue object for reading from bots, the tcp server uses a standard python list 2011-11-17T14:25:12 I'm not sure if that's actually thread-safe 2011-11-17T14:26:40 I don't think you can actually have a non-thread safe object in CPython due to the GIL 2011-11-17T14:26:56 global * lock? 2011-11-17T14:27:01 what is *? 2011-11-17T14:27:04 global interpreter lock 2011-11-17T14:27:09 thx 2011-11-17T14:27:23 magiik: berak has added phpskills, and i use it now instead of python version i used before (java onecrashed often), people said t works well now 2011-11-17T14:28:12 aichallenge: McLeopold epsilon * r7c50c5e / (4 files in 4 dirs): Fix HILL_ANT string - http://git.io/HinaBw 2011-11-17T14:28:18 Yexo: why do you mean? 2011-11-17T14:28:36 I'm trying to find a bug (explained here: http://forums.aichallenge.org/viewtopic.php?f=25&t=1925 ) 2011-11-17T14:28:51 basically an order of my bot was not executed by the engine 2011-11-17T14:28:59 without a valid reason not to execute it 2011-11-17T14:29:07 *** Surya has joined #aichallenge 2011-11-17T14:29:18 Fluxid: yeah seems like the sigma is getting better, mine dropped from the 3+ it was hanging around to like 1.2 now 2011-11-17T14:29:26 so far I've determined that my bot did output the mvoe and that tcpclient.py send it to the server 2011-11-17T14:29:31 *** GeorgeSebastian has joined #aichallenge 2011-11-17T14:29:40 G0llum, Thank you for your attention. I tried to specify the port (I launched the tcpServer on port 2081) so I tried 2081 (and then 2080 as you said). My game is played well, but then I have the "ERROR - php - writing ranks : [Error 2] The system cannot find the file specified (which is ants_skills.php and located at the correct spot) 2011-11-17T14:29:42 same here. my bots now have stable ratings 2011-11-17T14:29:51 Yexo: can you give me an example? 2011-11-17T14:29:55 in replay? 2011-11-17T14:29:58 http://ants.fluxid.pl/replay.1627 2011-11-17T14:30:03 turn 152 / 153 2011-11-17T14:30:21 both on 65 77 was ordered N in turn 152 but didn't move 2011-11-17T14:30:39 both? 2011-11-17T14:31:44 Yexo: no idea, all you can get is json replay and bot input from "BOT IN" button 2011-11-17T14:32:03 yes, I got that already. Thanks though 2011-11-17T14:33:15 Yexo: I assume you logged the next input showing the ant didn't move from 65 77? 2011-11-17T14:33:29 McLeopold: yes 2011-11-17T14:33:38 oh, I see it now 2011-11-17T14:33:40 Yexo: prolly better off trying to replicate it with playgame.py and seeds 2011-11-17T14:33:55 be wont be able to locally 2011-11-17T14:33:57 magiik: that's hard when you don't have the other bots that played 2011-11-17T14:34:09 i meant try to reproduce it w/o those bots 2011-11-17T14:34:14 Yexo: you have the other bots' input 2011-11-17T14:34:16 well - you could play the replay 2011-11-17T14:34:19 back to your bot 2011-11-17T14:34:24 Anyone who could help with the TcpServer/Webserver.py? I got my games done but I got an error on "ERROR - php - writing ranks : [Error 2] The system cannot find the file specified (which is ants_skills.php and located at the correct spot). Isn't it supposed to generate a cute website and everything? I can't find it 2011-11-17T14:34:28 and use the same seed as the game 2011-11-17T14:34:29 McLeopold: yes, I can get that from the replay 2011-11-17T14:34:58 if there are no more ideas I could extract all the moves done by the other bots from the replay and write a "bot" that executes those moves 2011-11-17T14:35:11 Surya, it can't find php 2011-11-17T14:35:17 or just write a psuedo playgame 2011-11-17T14:35:23 that plays your ant's replay back to it 2011-11-17T14:35:29 Yexo: if you don't see this issue on official, and you don't see it often on tcp, I would ignore it and work on more productive things 2011-11-17T14:35:33 all you need is for your bot to make the moves - all the other ant positions that your bot will see 2011-11-17T14:35:36 is contained in your bots replay 2011-11-17T14:35:45 magiik: I know, that's trivial 2011-11-17T14:35:57 I can reproduce the problem, but the ant won't magically move 2011-11-17T14:36:09 then you would have to debug the engine 2011-11-17T14:36:15 if you can reproduce it in playgame.py 2011-11-17T14:36:26 McLeopold: I see it about 1 in 10 games on tcp 2011-11-17T14:36:44 bot is not playing on official yet because we're with a group and haven't yet created an account 2011-11-17T14:36:51 as long as the engine isn't making the mistake... 2011-11-17T14:37:00 my bot certainly isn't 2011-11-17T14:37:05 to it's either in the tcp server or in the engine 2011-11-17T14:37:35 which tcp server? 2011-11-17T14:37:39 the tcp server is quite simple, so I'm more and more strongly suspecting the engine 2011-11-17T14:37:42 https://github.com/berak/ants-tcp/blob/master/tcpserver.py 2011-11-17T14:37:44 and thats why i say replicate it in playgame.py 2011-11-17T14:38:06 not the code, the server :) 2011-11-17T14:38:08 magiik: and how would I go about doing that? playgame.py needs more than my own bot 2011-11-17T14:38:14 running http://ants.fluxid.pl/replay.1627 2011-11-17T14:38:16 any idea how long into your turn it happened? 2011-11-17T14:38:20 well, run with test bots 2011-11-17T14:38:26 could it be that issue with the engine reading input in 50ms blocks or something 2011-11-17T14:38:42 Fluxid: can you log bot inputs and outputs until Yexo gets another assert? 2011-11-17T14:39:06 Yexo: i do have the code for one of those bots m3 heh 2011-11-17T14:39:12 but that doesn't help with replicating the other bots 2011-11-17T14:40:01 my diffusion map bot's emergent behavior seems to be sneak attack the hive 2011-11-17T14:40:04 g0llum, I added the php.exe folder in the PATH and I still got the error. Are you sure it's because of the php.exe and not the file itself? 2011-11-17T14:40:51 because i havn't implemented proper combat - just a sort of kludged together diffusion map based thing with enemy ants vs my ants weights 2011-11-17T14:41:06 McLeopold: i can try.. 2011-11-17T14:41:40 *** bishoco has joined #aichallenge 2011-11-17T14:42:10 Fluxid: https://github.com/aichallenge/aichallenge/blob/epsilon/ants/playgame.py#L379 2011-11-17T14:42:22 here is the options you need to pass into the run_game() call 2011-11-17T14:42:24 Yexo: you could also stress-test the engine with random ant configurations 2011-11-17T14:42:45 I'll try that now too 2011-11-17T14:42:46 i mean with massive load 2011-11-17T14:42:51 Surya, well that was my problem there, no idea now 2011-11-17T14:43:07 creating massive load might be a problem 2011-11-17T14:43:08 just open a file for each bot and pass that in the 'input_logs' option 2011-11-17T14:43:11 unless I run the engine twice or so 2011-11-17T14:43:24 Yexo: i meant with many ants 2011-11-17T14:43:31 g0llum, ok I'm gonna keep looking, thank you very much 2011-11-17T14:43:38 https://secure.wikimedia.org/wikipedia/en/wiki/Argument-dependent_name_lookup *sigh* 2011-11-17T14:43:40 can you generate maps with ants already on it right? 2011-11-17T14:43:59 like the test map right? 2011-11-17T14:44:05 magiik: hmm, that's an idea indeed 2011-11-17T14:44:12 Fluxid: so this line, in the last parameter, just needs a list of file descriptors https://github.com/berak/ants-tcp/blob/master/tcpserver.py#L201 2011-11-17T14:44:13 *** Jumbala has joined #aichallenge 2011-11-17T14:44:14 could probably kludge something together to generate a quick test map each turn 2011-11-17T14:44:20 on each *round 2011-11-17T14:44:27 i heard there are scenarios, i dont know what they are but they sound connected to this 2011-11-17T14:44:41 a scenario is basically a map file with ants included 2011-11-17T14:44:43 g0llum, last question: isn't it supposed to generate a website? I mean, it updates the sqlLite db but no website 2011-11-17T14:44:45 what techniques are people using to walk through the map? 2011-11-17T14:44:58 Anilm3: probably any ones u can think of? 2011-11-17T14:45:14 Yexo: perfect, no have a bot that just randomly goes somewhere, but check if the move was possible and the engine atually moved the ant 2011-11-17T14:45:20 now* 2011-11-17T14:45:57 i bet a thousand (imaginary) dollars you will not tbe able to replicate the bug locally 2011-11-17T14:46:03 If anyone finds that this is a bug in the engine, put in a github issue 2011-11-17T14:46:36 Surya, i think, webserver.py is generating webpages frm that sqlite db 2011-11-17T14:46:46 magiik: I was just asking for proper AI techniques, not your secrets, don't worry 2011-11-17T14:47:05 Anilm3: what do you mean by walk? pathfinding? 2011-11-17T14:47:08 Anilm3: yeah, i'm just saying, pretty much anything you can think of will/already is implemented 2011-11-17T14:47:29 Anilm3: following walls, using pathfinding ( a lot of a* and BFS around here ) 2011-11-17T14:47:46 I just have a quick question... I'm starting to implement some things and I was wondering if it was possible for an ant to attack another ant on the other side of the map: example: I'm on the lowest tile and the enemy ant is on the highest tile, can I attack it since it is technically 1 tile away? 2011-11-17T14:47:52 diffusion maps had a bit of popularity as well ( but i think it's just a tweaked subset of BFS ) 2011-11-17T14:47:57 g0llum, You installed one right? When I launch webserver.py it creates ressource files but no actual website. Did you make it until the end? 2011-11-17T14:48:01 Jumbala: yes 2011-11-17T14:48:09 Thank you! 2011-11-17T14:48:12 magiik: thanks I'll take a look! 2011-11-17T14:48:15 the ants cross the edges of the map 2011-11-17T14:48:45 magiik: I have pathfinding (A*) just for known targets but not for exploring the map 2011-11-17T14:49:11 *** Jumbala has quit IRC (Quit: Page closed) 2011-11-17T14:49:20 Anilm3: for exploring the map - i use a diffusion map populated with explore goals - explore goals are generated based on the last seen values of each tile i record 2011-11-17T14:49:32 Surya, i sever seen anay resource files. just start it , and point your browser to localhost:2080 2011-11-17T14:49:36 Anilm3: but you could easily do the same thing replacing dissuion maps with a* 2011-11-17T14:49:44 maybe assign ants to explore goals, etc... 2011-11-17T14:49:53 *diffusion 2011-11-17T14:50:14 g0llum, ok and the tcpServer port should be the same or different? 2011-11-17T14:50:41 hrm... mebe i should run a local tcp server and play my bots against each other 2011-11-17T14:50:43 and do your ants form groups as they walk? 2011-11-17T14:50:52 different, since you can't have two servers on the same machine on the same port 2011-11-17T14:51:08 Yexo: that might be another idea to explore - set up a local tcpserver and try to replicate it that way? 2011-11-17T14:51:47 Anilm3: my ants actually spread as they walk - they only group when enemies are present - and that's sort of emergent behavior - they want to walk past the enemny ants - but the enemy ants prevent them from going, so they bunch up near the enemies 2011-11-17T14:51:57 then once there are enough of my ants, they ignore the enemies and charge in 2011-11-17T14:53:17 I'm watching your replays, quite nice 2011-11-17T14:53:59 m2 or m3 are mine right now - m2 seems to perform better, early version tho 2011-11-17T14:54:05 BRB AFK 2011-11-17T14:54:09 g0llum, ok the problem was that I launched wamp in the same time i guess, i now have a webpage with the following error : you need to send INFO: you need to send: [USER name password]. end go (thank you again for your help, I feel kind of stupid to ask you so many questions!) 2011-11-17T14:54:15 *** mj41 has quit IRC (Ping timeout: 244 seconds) 2011-11-17T14:54:18 Surya, change the 'trueskills' option in tcpserver.py to 'py' for starters, try to get games running, late get your php up and running, and use that 2011-11-17T14:55:17 Anilm3: I've had a lot of success with BFS pathfinding. I used this as a guide to implement in JAVA: http://www.peachpit.com/articles/article.aspx?p=101142 2011-11-17T14:57:16 bishoco: thanks 2011-11-17T14:57:41 okay 2011-11-17T14:58:29 Yexo: connect, yo 2011-11-17T14:58:41 You're GotAnts, right? 2011-11-17T14:58:44 yeah 2011-11-17T14:58:45 yes 2011-11-17T14:58:49 it's playing 2011-11-17T14:59:04 I might actually have reproduced it locally in the meantime, still checking that 2011-11-17T14:59:40 g0llum, ok I don't have the php error now, thanks! I still can't access to the webpage (localhost:2081) because of the "INFO: you need to send: [USER name password]. end go" that I got on it. Do you have any idea? 2011-11-17T14:59:41 hum, it doesn't look like it logs stuff to file 2011-11-17T15:00:52 damn, file is empty 2011-11-17T15:02:31 Surya: i dont believe u use the tcp playing port as the webserver 2011-11-17T15:02:31 Yexo, can it be that simple, that your game lastetd longer than 4 minutes ? i think the timeout in tcpclient.py is ridicilous low (line 49) 2011-11-17T15:02:37 2081 sounds like the tcp port 2011-11-17T15:02:53 but i really dont know how the tcp server works... 2011-11-17T15:02:58 McLeopold: anything else except setting input_logs and output_logs? 2011-11-17T15:03:17 g0llum: isn't that timeout reset every turn? 2011-11-17T15:03:28 *** Antimony has quit IRC (Ping timeout: 240 seconds) 2011-11-17T15:03:53 McLeopold: it doesn't write to files at all 2011-11-17T15:03:55 Yexo, i think, it's for the whole session 2011-11-17T15:04:08 that wouldn't lead to this problem though 2011-11-17T15:04:43 that's the socket timeout 2011-11-17T15:04:52 i.e. if the other end doesn't respond in X, then terminate connection 2011-11-17T15:04:54 Fluxid: well, that's how the normal local game logs work, so it should work for you 2011-11-17T15:05:25 Fluxid: the option should be set to a list of open files, one for each bot 2011-11-17T15:05:31 http://docs.python.org/library/socket.html#socket.socket.settimeout 2011-11-17T15:05:43 McLeopold: ok, my fault, it works :) 2011-11-17T15:05:56 you could try setting it to 0 if you think that's the problem - but if the server stops responding, you wont ever tiem out the connection 2011-11-17T15:06:17 changed options but gave to run_game old ones 2011-11-17T15:06:19 no, that timeout is fine 2011-11-17T15:06:30 it's for blocking operations, so it's basically per turn 2011-11-17T15:06:31 k. 2011-11-17T15:07:13 and anyways - you can see the server got the move, cuz it's in the replay 2011-11-17T15:08:25 Yexo: it logs your stuff now, tell me when you need input or output of game 2011-11-17T15:08:33 ok, thanks 2011-11-17T15:09:58 *** GeorgeSebastian has quit IRC (Quit: popped out) 2011-11-17T15:10:17 Fluxid: you might also want to log 'error_logs' as well, to check for invalid or duplicate orders 2011-11-17T15:10:35 baaaah 2011-11-17T15:10:38 :) 2011-11-17T15:10:40 that's of minor concern 2011-11-17T15:10:46 I might not be 2011-11-17T15:10:53 I can do that manually if an error is triggered 2011-11-17T15:10:55 if the engine fails to parse your order, we need to know 2011-11-17T15:10:57 McLeopold: i've interrupted many games already, later 2011-11-17T15:10:58 ;) 2011-11-17T15:11:06 *** daboomguy has joined #aichallenge 2011-11-17T15:11:35 *** daboomguy has left #aichallenge 2011-11-17T15:11:49 Fluxid: it's the right thing to do, interrupt the games now, don't wait until we get inconclusive results and need to change it again 2011-11-17T15:12:15 Fluxid: so, get input, output and error logs 2011-11-17T15:12:45 Fluxid, we'll all come back .. 2011-11-17T15:13:43 *** Transformer has joined #aichallenge 2011-11-17T15:13:47 'kay 2011-11-17T15:15:24 <_flag> Has the tcp server been shut off? 2011-11-17T15:15:43 *** Rinum has quit IRC (Ping timeout: 265 seconds) 2011-11-17T15:16:35 _flag: restarted. sorry, we're debugging stuff for Yexo ;) 2011-11-17T15:16:40 it works now again 2011-11-17T15:16:42 _flag, maintenance, second 2011-11-17T15:16:56 Fluxid: my last game 2011-11-17T15:16:59 <_flag> Okay, no problem :) 2011-11-17T15:17:00 perhaps unfinished 2011-11-17T15:17:13 Yexo: i killed the server, bot did you gett assetion error? 2011-11-17T15:17:17 but* 2011-11-17T15:17:19 *** Transformer has left #aichallenge 2011-11-17T15:17:21 yes, just before you restarted 2011-11-17T15:17:35 ok 2011-11-17T15:18:00 not game 3242 though 2011-11-17T15:18:05 so probably an unfinished game 2011-11-17T15:18:23 Yexo: http://fluxid.pl/game_3249.input http://fluxid.pl/game_3249.output 2011-11-17T15:19:01 *** NoxiaZ^ has joined #aichallenge 2011-11-17T15:22:25 Fluxid: that input/output is correct. If I rerun my bot locally with that input I get different output 2011-11-17T15:23:52 *** epicmonkey has quit IRC (Ping timeout: 240 seconds) 2011-11-17T15:25:03 *** bishoco has quit IRC (Quit: Page closed) 2011-11-17T15:25:23 Yexo: using the same player seed? 2011-11-17T15:25:50 not relevant, I don't do anything random. I basiclaly got the same output but with one extra line 2011-11-17T15:25:58 Fluxid: just happened again 2011-11-17T15:28:30 Yexo: what id of game 2011-11-17T15:28:38 3265 2011-11-17T15:29:44 http://fluxid.pl/game_3265.input http://fluxid.pl/game_3265.output http://fluxid.pl/game_3265.errors 2011-11-17T15:29:44 sorry, that's not right 2011-11-17T15:30:01 the game hasn't finished yet, so I don't actually know the id 2011-11-17T15:30:07 ............ 2011-11-17T15:30:22 ;) 2011-11-17T15:30:39 really sorry, looking at too many things at once 2011-11-17T15:32:54 i shouldnt have bet that much on this 2011-11-17T15:34:47 Yexo, 3266, probably 2011-11-17T15:34:58 yes 2011-11-17T15:35:17 although I've been careless and overwritten my own logs of that game 2011-11-17T15:37:24 *** jstemmer has joined #aichallenge 2011-11-17T15:38:31 *** Lithosphere has joined #aichallenge 2011-11-17T15:38:31 *** Lithosphere has joined #aichallenge 2011-11-17T15:39:31 *** xathis has quit IRC () 2011-11-17T15:44:07 *** Antimony has joined #aichallenge 2011-11-17T15:44:11 so? 2011-11-17T15:44:30 to bug or not to bug? 2011-11-17T15:44:43 Ants 2011-11-17T15:45:01 wave your antennae 2011-11-17T15:45:30 still not sure. had some problems with reproducing (and made some errors in that) 2011-11-17T15:46:36 *** Chris_0076 has quit IRC (Remote host closed the connection) 2011-11-17T15:53:15 Fluxid: I hate to be picky, but shouldn't you have put a bot identifier in the filename? 2011-11-17T15:53:39 McLeopold: i log only Yexo 2011-11-17T16:02:07 magiik, thank you, it was the problem (the port wasn't the TCP playing port) Thank you very much (and g0llum too!) Have a great evening 2011-11-17T16:03:28 ;) 2011-11-17T16:03:32 *** choas has joined #aichallenge 2011-11-17T16:05:01 ug, my new bot is taking over a second per move 2011-11-17T16:05:09 *** pezzotto has joined #aichallenge 2011-11-17T16:05:28 hi guys 2011-11-17T16:05:43 just a question of which I didn't find a good answer on the forums 2011-11-17T16:05:52 how is time calculated? 2011-11-17T16:06:12 is it the wall clock or the CPU time? 2011-11-17T16:06:21 walltime 2011-11-17T16:06:38 ok 2011-11-17T16:06:51 so which exactly is the C (or python) function to get the time? 2011-11-17T16:07:26 I have a C++ clock you can use if you want 2011-11-17T16:07:55 I am using python, but I will gladly see the c++ code to get the right function 2011-11-17T16:08:26 in python, just time(), in c it's more complicated, since your os is mixing in 2011-11-17T16:08:56 *** Accoun has quit IRC () 2011-11-17T16:09:53 so is it not the time() function in c? (the one provided by time.h) 2011-11-17T16:10:08 pezzotto: gettimeofday 2011-11-17T16:10:15 that should do it 2011-11-17T16:11:34 ok... so it's gettimeofday in C and time.time() in python, right? 2011-11-17T16:11:40 pezzotto: http://docs.python.org/py3k/library/time.html 2011-11-17T16:11:43 *** Jak_o_Shadows has joined #aichallenge 2011-11-17T16:11:51 great, thanks! 2011-11-17T16:12:16 however think about this choice of using python 2011-11-17T16:12:30 ? 2011-11-17T16:12:37 Indeed 2011-11-17T16:12:53 gettimeofday() is the old POSIX way to do it, the new way is clock_gettime(), but I think you need -lrt passed to gcc for that 2011-11-17T16:13:15 anyway, either will get you all the precision you need 2011-11-17T16:13:27 mcstar what does it mean "think about this choice of using python" ? 2011-11-17T16:13:35 pezzotto: only that 2011-11-17T16:13:40 i dont want to step on toes here 2011-11-17T16:13:47 so i remain unbiased 2011-11-17T16:13:48 it's ok :-) 2011-11-17T16:13:58 I know python has problems with speed 2011-11-17T16:14:34 I have actually "outsourced" a few important bits to C and wrapped in python to gain an edge in speed 2011-11-17T16:14:46 sometimes it's painful, but worthwhile 2011-11-17T16:14:59 only you can be the judge of that 2011-11-17T16:15:36 I doubt many things I am doing with python could be done as easily with C++ 2011-11-17T16:15:50 like what? 2011-11-17T16:16:11 there is no c++ interpreter, thats for sure 2011-11-17T16:16:19 there is 2011-11-17T16:16:23 (now dont enybody tell me otherwise) 2011-11-17T16:16:26 install cint from the aur :P 2011-11-17T16:16:39 I am eavily relying on the introspection of python to change the behaviour of the ants in real time 2011-11-17T16:16:41 thestinger: there isnt im sure, you are kidding me 2011-11-17T16:16:59 doing this in C++ will require a lot of if-then-else 2011-11-17T16:17:18 mcstar: http://paste.pocoo.org/show/509413/ 2011-11-17T16:17:19 introspection might cost you more than an if/else 2011-11-17T16:17:27 the results are quite good so far... I'll see how the competition goes 2011-11-17T16:17:45 It costs me on time, but again I can outsource heavy bits in C++ 2011-11-17T16:18:02 thestinger: i was kidding, but do you think its a viable alternative to gcc? 2011-11-17T16:18:10 no xD 2011-11-17T16:18:14 k 2011-11-17T16:18:16 you'll be surprised by seeing that 90% of the computation is spent in 2-3% of your code 2011-11-17T16:18:17 i rest my case 2011-11-17T16:18:18 I think it's intended for development, it's slower than CPython 2011-11-17T16:18:44 pezzotto: why would i be surprised? 2011-11-17T16:19:06 it's a generic "you".. people would be surprised :-) 2011-11-17T16:19:14 pezzotto: yeah, but that 2% is 60x slower when you're using python 2011-11-17T16:19:16 im not people 2011-11-17T16:19:22 :D 2011-11-17T16:19:27 instead of my combat code taking 10ms, it would take 600ms 2011-11-17T16:19:40 yes, so that 2% is outsourced in C++ and wrapped around 2011-11-17T16:19:48 in pypy that would be a lot better, but it would still be 5x-10x slower 2011-11-17T16:19:53 well, how are you outsourcing it? 2011-11-17T16:19:56 ctypes is slow 2011-11-17T16:20:07 pezzotto: if you find it satisfying to write c when you started out with python, thats ok 2011-11-17T16:20:11 i wouldnt really do that 2011-11-17T16:20:11 I've had some problems as the servers don't support boost.. so I had to hand code it 2011-11-17T16:20:15 no for this kind of project 2011-11-17T16:20:36 I love C++... but not as much as I love python 2011-11-17T16:20:49 it's simply elegant 2011-11-17T16:21:01 yes, but python is ugly 2011-11-17T16:21:10 c'mon! 2011-11-17T16:21:10 *** Chris_0076 has joined #aichallenge 2011-11-17T16:21:11 oh, you meant the other way around 2011-11-17T16:21:13 not at all 2011-11-17T16:21:13 Python isn not ugly 2011-11-17T16:21:24 python is beautiful 2011-11-17T16:21:39 pezzotto: i was trolling, but my real opinion is that its not pretty 2011-11-17T16:21:42 lisp is pretty 2011-11-17T16:21:51 yeah 2011-11-17T16:22:01 I might learn lisp one day 2011-11-17T16:22:10 or maybe haskell... I don't really fancy all those brackets 2011-11-17T16:22:19 what brackets? 2011-11-17T16:22:28 parentheses 2011-11-17T16:22:28 lisp brackets 2011-11-17T16:22:33 oh 2011-11-17T16:22:36 yes 2011-11-17T16:22:43 I thought you meant in haskell, I read that wrong xD 2011-11-17T16:23:02 good old time of the IRC chat.. haven't used for ages 2011-11-17T16:23:16 anyway... thanks for your help guys, see you in the arena! 2011-11-17T16:23:18 pezzotto: newer IM's dont have this problem? 2011-11-17T16:23:25 thats why i go with lua.. very short development time and kinda fast with luajit.. and well i just like it 2011-11-17T16:23:40 IM? 2011-11-17T16:23:42 *** mikewintermute has joined #aichallenge 2011-11-17T16:23:46 instant messenger 2011-11-17T16:23:47 *** GeorgeSebastian has joined #aichallenge 2011-11-17T16:23:47 *** GeorgeSebastian has joined #aichallenge 2011-11-17T16:23:48 i dunno what IM has to do with irc 2011-11-17T16:23:55 ah 2011-11-17T16:24:05 i guess it's sort of instant messaging 2011-11-17T16:24:11 but who uses an IM client for irc? 2011-11-17T16:24:12 Pretty much 2011-11-17T16:24:27 magiik: wtf are you talking about? 2011-11-17T16:24:27 *** pezzotto has quit IRC (Quit: Page closed) 2011-11-17T16:24:40 btw many IM's support irc 2011-11-17T16:24:48 *** oggu has joined #aichallenge 2011-11-17T16:24:54 and you can do IM from irc :P 2011-11-17T16:24:56 i dont think i could ever get away from sreen + irssi 2011-11-17T16:25:01 *screen 2011-11-17T16:25:02 https://secure.wikimedia.org/wikipedia/en/wiki/BitlBee 2011-11-17T16:25:05 ^ so awesome 2011-11-17T16:25:28 *** amstan has quit IRC (Ping timeout: 240 seconds) 2011-11-17T16:25:31 interesting..... 2011-11-17T16:25:35 *** Accoun has joined #aichallenge 2011-11-17T16:25:40 it's really great if you like weechat/irssi 2011-11-17T16:26:14 wonder if there's anyway to get multiple logins on AIM though 2011-11-17T16:26:18 *** okan_ has quit IRC (Quit: Page closed) 2011-11-17T16:26:23 if i like weechat why would i want to use beetlbee 2011-11-17T16:26:26 or some sort of plugin that i could connet to my server running bitlbee to connect my other IM clients 2011-11-17T16:26:31 *** mikewintermute has quit IRC (Client Quit) 2011-11-17T16:26:42 oh wait aim supports multiple logins... what dosen't though? yahoo or msn? 2011-11-17T16:27:01 *** cyphase has quit IRC (Ping timeout: 258 seconds) 2011-11-17T16:29:05 mcstar: it lets you use weechat for regular IMing 2011-11-17T16:29:19 google/facebook chat, etc. 2011-11-17T16:29:28 so its a proxy thing? 2011-11-17T16:29:59 yeah, it's like an irc->IM protocol server 2011-11-17T16:34:36 *** g0llum has quit IRC (Read error: Connection reset by peer) 2011-11-17T16:35:12 im stuck with battle 2011-11-17T16:35:30 *** ltriant has quit IRC (Quit: Get MacIrssi - http://www.sysctl.co.uk/projects/macirssi/) 2011-11-17T16:36:13 the only possible way i see is too messy and ugly 2011-11-17T16:36:28 but the proper way i stoo damn expensive 2011-11-17T16:36:30 *** bhasker has quit IRC (Quit: bhasker) 2011-11-17T16:36:39 mcstar: yeah, I haven't figured it out yet 2011-11-17T16:36:56 I do have combat now, but it assumes enemies don't move 2011-11-17T16:37:04 hm 2011-11-17T16:37:15 i have never thought of that 2011-11-17T16:37:19 would have* 2011-11-17T16:37:30 quite a simplification 2011-11-17T16:37:36 Mine assume they move 2011-11-17T16:37:46 mine too 2011-11-17T16:37:49 Gets them dancing sometimes 2011-11-17T16:38:02 but im havin trouble recognizing sudden death situations with that 2011-11-17T16:38:28 and i havent found any enlightening ideas the hole day 2011-11-17T16:39:30 *** cyphase has joined #aichallenge 2011-11-17T16:40:09 i even have trouble with the definition of "success" 2011-11-17T16:40:21 yeah, I'm having trouble deciding that atm 2011-11-17T16:40:23 I've found many, not any that would help my ants though 2011-11-17T16:41:08 I simulate what would happen with ants in a certain location (after moving), and I know which allies/enemies will die 2011-11-17T16:41:24 i have move generation too 2011-11-17T16:41:37 i resolve every such battle situation 2011-11-17T16:41:42 but i have trouble with the scoring 2011-11-17T16:41:52 and it doesnt really work for ~10 ants 2011-11-17T16:41:55 letting my ants trade 2 enemy deaths for 1 death doesn't work since I assume the enemies stay still - it makes them too aggressive 2011-11-17T16:41:58 *** ltriant has joined #aichallenge 2011-11-17T16:42:14 so I've just left them really cautious for now 2011-11-17T16:42:28 I think I'll adjust the losses they are willing to take based on % of the map I control or something 2011-11-17T16:42:36 if i make them cautious, they wont attack cooperatively even if they could win 2011-11-17T16:42:53 mine do since I'm pretending the enemy is stationary 2011-11-17T16:43:05 but it makes them do stupid stuff a lot 2011-11-17T16:43:16 like moving 1 ant towards where 2 enemy ants are about to move 2011-11-17T16:43:50 maybe i should just use a neural net library and train it with random input 2011-11-17T16:43:54 I think for the big battles this way is going to work, since the lines of combat are fairly static 2011-11-17T16:43:57 Lots to tune with my ants. But my next bot will be c++ 2011-11-17T16:44:29 It's funny though, I looked at your #18 bot and no combat at all :) 2011-11-17T16:44:29 for the small battles, you really need to see what would happen if the enemies move 2011-11-17T16:44:43 yeah, that's an old version 2011-11-17T16:44:54 it also has really crappy foraging and other problems 2011-11-17T16:45:02 And still #18 2011-11-17T16:45:21 http://ants.fluxid.pl/player/strcat_rewrite this is what I'm working on atm 2011-11-17T16:46:05 lots of new problems to deal with - I need to adjust aggressiveness (they are too cautious now) and need to reteach my ants to actively strafe/encircle enemies 2011-11-17T16:46:57 my bot sucks because it loses its hill very early 2011-11-17T16:47:10 *** ChrisH_ has joined #aichallenge 2011-11-17T16:47:18 i didnt want to think about defending my hill until ive done battle 2011-11-17T16:47:24 maybe i should have done that first 2011-11-17T16:47:46 I have various defense methods implemented in branches, but they all make my bot worse 2011-11-17T16:47:48 :( 2011-11-17T16:48:03 but still you dont lose your hill 2011-11-17T16:48:10 I often do 2011-11-17T16:48:11 you would not have gotten to #18 2011-11-17T16:48:26 that version only got that high b/c it rushes at enemy hills and gets early points 2011-11-17T16:48:38 it can't survive long against the top 25 bots 2011-11-17T16:49:33 I rotate direction preferences, so there are usually ants leaving my hills from a few directions which sort of defends against scouts 2011-11-17T16:49:40 my rewrite doesn't actually do that atm 2011-11-17T16:49:49 and it has new movement issues now >.< 2011-11-17T16:52:16 are those wiskers? 2011-11-17T16:52:17 *** meatroaf_ has joined #aichallenge 2011-11-17T16:52:20 are you a cat? 2011-11-17T16:52:22 oh i see 2011-11-17T16:52:25 strcat 2011-11-17T16:52:32 street cat 2011-11-17T16:52:32 >_< 2011-11-17T16:52:33 *** NoxiaZ^ has quit IRC (Read error: Connection reset by peer) 2011-11-17T16:53:29 *** McLeopold has left #aichallenge 2011-11-17T16:55:29 *** dr- has quit IRC (Read error: Connection reset by peer) 2011-11-17T16:56:50 Whew, got a new version posted just before my next game. Hopefully I've finally licked timeouts. 2011-11-17T16:59:01 *** NightExcessive has quit IRC (Ping timeout: 255 seconds) 2011-11-17T17:00:34 *** ChrisH_ has quit IRC (Quit: Page closed) 2011-11-17T17:03:33 *** Palmik has quit IRC (Remote host closed the connection) 2011-11-17T17:03:50 ChrisH_ did you see http://forums.aichallenge.org/viewtopic.php?f=24&t=1900 2011-11-17T17:04:23 pguillory mentioning inconsistent performance for Go 2011-11-17T17:05:02 not something I am seeing (except maybe with a lot of logging...) 2011-11-17T17:05:54 bugnuts2: the regular Go compilers don't really optimize beyond turning your code into efficient assembly 2011-11-17T17:06:10 they don't inline, do loop unrolling, or any of the fancy optimizations that gcc and llvm can do 2011-11-17T17:06:17 i dont think running gc manually guarantees that it doesnt run on its own?! 2011-11-17T17:06:19 they're meant to compile stuff stuff really fast 2011-11-17T17:06:47 not concerned about slow really more unpredictable performance 2011-11-17T17:07:18 well, the garbage collector is also really immature - there's a complete rewrite of it in the development branch and weekly release now I think 2011-11-17T17:07:24 I track the per turn GC pauses and they are very short for my bot 2011-11-17T17:08:06 the immaturity of the gc is more in the context of mulithreaded apps. 2011-11-17T17:08:26 that makes sense 2011-11-17T17:09:18 and in any case I see 2-3 GC invocations per turn with an aggregate pause of typically 1500us 2011-11-17T17:10:51 I guess the goroutine scheduler could be doing something weird 2011-11-17T17:12:12 hard to imagine it causing a swing that big. 2011-11-17T17:12:39 a context switch could totally screw up caching I guess? 2011-11-17T17:13:51 I think there's a compiler flag to disable bounds checking which should be a speed gain without breaking any bots 2011-11-17T17:13:58 context switches in go are basically free for goroutines in a single threaded environment I think. 2011-11-17T17:14:11 yeah, b/c there's no locking 2011-11-17T17:14:27 but you're still switching to different code, so the CPU cache is going to suffer 2011-11-17T17:15:32 yeah but cache misses are orders of magnitude cheaper than the 50ms swings in timing he is mentioning. 2011-11-17T17:16:37 Hey guys, I'm trying to update my ExploreMap function so my ants spread out the maximum. So far it's a disaster, any advice? 2011-11-17T17:18:13 Surya: make it better 2011-11-17T17:18:24 Ok! :) I'll try§ 2011-11-17T17:18:25 ! 2011-11-17T17:18:56 try googling for fast terrain exploration 2011-11-17T17:19:20 Oh, thank you for the key words 2011-11-17T17:19:34 *** mleyen has quit IRC (Quit: ~ Trillian Astra - www.trillian.im ~) 2011-11-17T17:19:34 *** jasox has quit IRC (Quit: Leaving) 2011-11-17T17:20:13 *** choas has quit IRC (Ping timeout: 240 seconds) 2011-11-17T17:20:24 *** NightExcessive has joined #aichallenge 2011-11-17T17:21:37 ... O_o 2011-11-17T17:22:08 surya is there a replay somewhere that shows your bot? 2011-11-17T17:23:35 bugnuts2: No, I'm only playing local, the official server is way too slow 2011-11-17T17:23:47 and not on the tcp servers? 2011-11-17T17:23:48 *** Jolan-S-J-Linux- has joined #aichallenge 2011-11-17T17:23:55 Adam smith? 2011-11-17T17:23:58 I haven't tried yet 2011-11-17T17:24:09 Are you there adam smith? 2011-11-17T17:24:16 I'm making mine, so me and my friends can compete directly at first 2011-11-17T17:24:28 if($adamSmith){echo "allooooo";} 2011-11-17T17:24:51 I've done everything except open the ports so my friends can participate (I have a hard time with this actually) 2011-11-17T17:25:02 *** Antimony has quit IRC (Ping timeout: 245 seconds) 2011-11-17T17:25:38 *** treeform has joined #aichallenge 2011-11-17T17:25:54 who is adam smith? 2011-11-17T17:26:00 economist? 2011-11-17T17:26:02 hey guys, I just realized that my local players in tools is not showing any colours in the buttons ont he side :/ 2011-11-17T17:26:03 *** Jolan-S-J-Linux- has quit IRC (Client Quit) 2011-11-17T17:26:25 my local visualizer I mean 2011-11-17T17:26:32 *** Antimony has joined #aichallenge 2011-11-17T17:26:45 yeah, with chromium 2011-11-17T17:27:14 aaaaah. 2011-11-17T17:27:17 thanks mcstar 2011-11-17T17:27:32 for what? np 2011-11-17T17:28:30 *** Harpyon_ has joined #aichallenge 2011-11-17T17:28:30 *** Jak_o_Shadows has quit IRC (Remote host closed the connection) 2011-11-17T17:28:43 *** Harpyon has quit IRC (Quit: Computer has gone to sleep.) 2011-11-17T17:28:43 *** Harpyon_ is now known as Harpyon 2011-11-17T17:29:18 *** Garf has quit IRC (Read error: Connection reset by peer) 2011-11-17T17:31:26 *** twymer has quit IRC (Ping timeout: 244 seconds) 2011-11-17T17:34:27 *** userjjb has joined #aichallenge 2011-11-17T17:36:44 *** Jak_o_Shadows has joined #aichallenge 2011-11-17T17:39:59 yeah, my tcp timeouts are just network issues or something 2011-11-17T17:40:18 I started printing the turn time to stderr at the end of each turn and it was 6ms with my last timeout 2011-11-17T17:41:13 lol 2011-11-17T17:41:49 how many ants? 2011-11-17T17:41:59 not many http://ants.fluxid.pl/replay.3432 2011-11-17T17:43:16 maybe ill add a timer too 2011-11-17T17:43:22 just for fun 2011-11-17T17:43:37 yeah, I've just been optimizing for the hell of it when I'm out of ideas to implement 2011-11-17T17:43:55 :D 2011-11-17T17:44:00 next up is testing unordered_set v. set v. abusing vectors/list 2011-11-17T17:44:14 vs? 2011-11-17T17:44:19 yeah 2011-11-17T17:44:20 they server different purposes 2011-11-17T17:44:40 id never search in a list for example 2011-11-17T17:44:50 I know, but for some of the small sets I have, the hash-based unordered_set is probably overkill 2011-11-17T17:44:50 i just dont like double loops 2011-11-17T17:44:53 you might want to check that the tcp client shuts off nagle congestion control when creating the connection to the server 2011-11-17T17:45:00 even if they are implicit 2011-11-17T17:45:01 janzert: it doesn't 2011-11-17T17:45:07 ouch 2011-11-17T17:45:09 I tried setting NODELAY, but it seemed to make it worse 2011-11-17T17:45:18 hmm 2011-11-17T17:45:22 sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM) 2011-11-17T17:45:25 #sock.setsockopt(socket.IPPROTO_TCP, socket.TCP_NODELAY, 1) 2011-11-17T17:45:30 that's what I tried (the second line is what I added) 2011-11-17T17:45:47 weird 2011-11-17T17:45:55 it just seemed to cause non-stop timeouts so I commented it out for now 2011-11-17T17:46:07 thestinger: are you building up the output, or do you flush it at once? 2011-11-17T17:46:20 oh, that's a good point 2011-11-17T17:46:21 *** bishoco has joined #aichallenge 2011-11-17T17:46:31 actually its a bad quesiton 2011-11-17T17:46:33 the c++ starter bot flushes the input over and over, and I'm still using that code 2011-11-17T17:46:53 so my better qestion: 2011-11-17T17:47:00 maybe it was ending up in a bunch of tcp packets when I turned on NODELAY 2011-11-17T17:47:14 are you collecting the orders than flash or you flush after every computed order 2011-11-17T17:47:24 flush* 2011-11-17T17:47:31 *** pairofdice has quit IRC (Quit: Just keep breathing) 2011-11-17T17:47:33 atm, I flush after every order but it all happens at the end of my turn 2011-11-17T17:47:40 I'll try changing that to one flush at the end 2011-11-17T17:47:46 i do it by outputting them at the very end 2011-11-17T17:48:10 actually im not sure now 2011-11-17T17:48:21 btu i think i flush after every line 2011-11-17T17:48:27 std::endl; flushes the stream 2011-11-17T17:48:40 ah, no \n 2011-11-17T17:49:03 but my go() flushes 2011-11-17T17:49:07 ah 2011-11-17T17:49:20 \n will still flush line-buffered streams, I'm not sure what the tcpclient does 2011-11-17T17:49:38 thestinger: tcpclient.py collects all your output until you print "go" and only than sends everything to the server 2011-11-17T17:49:51 oh, so that's not the problem >.< 2011-11-17T17:49:56 I'll turn NODELAY back on then 2011-11-17T17:50:32 My bot takes 11 seconds to move 13 ants. This isn't looking good 2011-11-17T17:56:12 Antimony: tree-generation? 2011-11-17T17:57:19 *** Antimony has quit IRC (Ping timeout: 248 seconds) 2011-11-17T18:00:37 I think his bot melted his computer :P 2011-11-17T18:04:25 *** amstan has joined #aichallenge 2011-11-17T18:04:25 *** ChanServ sets mode: +o amstan 2011-11-17T18:07:22 *** Antimony has joined #aichallenge 2011-11-17T18:09:27 he has some network problems 2011-11-17T18:09:59 im not familiar with the country code .edu 2011-11-17T18:10:25 edutopia 2011-11-17T18:10:31 educuador 2011-11-17T18:10:46 *** oggu has quit IRC (Quit: oggu) 2011-11-17T18:18:00 *** HaraKiri has quit IRC () 2011-11-17T18:20:33 mcstar: it belongs to united states, for education 2011-11-17T18:20:46 most universities have a .edu tld 2011-11-17T18:21:17 amstan: you are helpful but you should not have answered that question 2011-11-17T18:21:32 ? 2011-11-17T18:21:36 why not? 2011-11-17T18:22:00 obviosly i was mocking his contants disconnections 2011-11-17T18:22:03 -s 2011-11-17T18:22:18 he's not that bad 2011-11-17T18:23:37 13 ants with one tree? well of course 2011-11-17T18:23:42 amstan: lol that was clearly a joke 2011-11-17T18:23:53 *** yoden has quit IRC (Read error: Connection reset by peer) 2011-11-17T18:23:57 edussia 2011-11-17T18:24:03 :) 2011-11-17T18:24:28 *** yoden has joined #aichallenge 2011-11-17T18:32:23 *** Fandekasp has quit IRC (Ping timeout: 276 seconds) 2011-11-17T18:34:10 i think im gonna install win7 into a vm 2011-11-17T18:34:52 now what is the best choice for image format? 2011-11-17T18:35:05 lets say i later want to write it out to a real partition 2011-11-17T18:42:21 *** bishoco has quit IRC (Quit: Page closed) 2011-11-17T18:42:29 no ideas? 2011-11-17T18:42:49 i think i had success with a qemu image once 2011-11-17T18:43:25 i think those can be directly written to a partition, if they arent compressed and fully allocated ones 2011-11-17T18:43:37 and are* 2011-11-17T18:51:01 *** fpcfan has quit IRC (Ping timeout: 240 seconds) 2011-11-17T18:51:11 *** Antimony has quit IRC (Ping timeout: 248 seconds) 2011-11-17T18:52:04 *** fpcfan has joined #aichallenge 2011-11-17T18:58:46 mcstar: why are you and I both making windows 7 vm's? 2011-11-17T18:59:01 *** Surya has quit IRC (Ping timeout: 265 seconds) 2011-11-17T18:59:04 * Zao eyes mcstar 2011-11-17T18:59:09 Forgot all about you, heh. 2011-11-17T18:59:21 *** McLeopold has joined #aichallenge 2011-11-17T18:59:22 aarossig: are we? 2011-11-17T18:59:23 mcstar: I need the vm to run AVR Studio, sadly there are no simulators that are as nice as the one included with AVR Studio 2011-11-17T18:59:30 *** nle85 has quit IRC (Ping timeout: 265 seconds) 2011-11-17T18:59:41 mcstar: you mentioned it at 6:34 :P 2011-11-17T18:59:45 i just thought id compile my app on win 2011-11-17T19:00:05 aarossig: There's nothing nice with anything AVR Studio. 2011-11-17T19:00:18 aarossig: Time zones are nice. You should try them some time. 2011-11-17T19:00:22 Even if it's just Zulu. 2011-11-17T19:00:24 yeah 2011-11-17T19:00:41 its 2.00 on my clock 2011-11-17T19:00:46 which isnt dst corrected 2011-11-17T19:00:48 Zao: I myself really enjoy using vi, avr-gcc and Makefiles but my class is a little annoying and they want screenshots 2011-11-17T19:01:07 what is this avr studio? 2011-11-17T19:01:13 aarossig: that was one assignment 2011-11-17T19:01:14 Zao: also, the Port "simulators" are pretty sweet, but that's about all I need from there 2011-11-17T19:01:38 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-11-17T19:01:43 mcstar: AVR Studio is for developing on the AVR microcontroller platform 2011-11-17T19:01:53 mcstar: Horribad IDE for the AVR series of Atmel microcontrollers. 2011-11-17T19:01:55 ah 2011-11-17T19:02:04 Comes with some silly gcc version and a horrible UX :) 2011-11-17T19:02:12 Scroll wheels? Never heard of 'em. 2011-11-17T19:02:16 Zao: have you used the most recent version? 2011-11-17T19:02:32 so aarossig which image format do you prefer? 2011-11-17T19:02:45 i might want to later write it to a real partition 2011-11-17T19:02:53 mcstar: I'll just be using VirtualBox and their vdi's 2011-11-17T19:03:16 mcstar: I don't want to waste a lot of time on this -- default settings all the way 2011-11-17T19:03:16 ive actually done something similar but i dont remember it anymore 2011-11-17T19:03:52 aarossig: ive found open-vm-tools 2011-11-17T19:03:58 *** fpcfan has quit IRC (Ping timeout: 260 seconds) 2011-11-17T19:04:02 have this replaced vmware? 2011-11-17T19:04:27 mcstar: I haven't used vmware in ages... VirtualBox is in the repo for Ubuntu so I just installed it 2011-11-17T19:04:39 *** traMKer has joined #aichallenge 2011-11-17T19:04:51 ah sorry 2011-11-17T19:04:57 so many vm's i get confused 2011-11-17T19:05:04 mcstar: yeah.. lol 2011-11-17T19:05:15 *** ikaros has joined #aichallenge 2011-11-17T19:05:32 mcstar: i hate doing this sort of thing, it's painful 2011-11-17T19:05:36 *** fpcfan has joined #aichallenge 2011-11-17T19:05:38 aarossig: like that one time you had screen in screen? 2011-11-17T19:05:43 amstan: exactly. 2011-11-17T19:05:53 mcstar: You're using the 'and' spelling of && without including 2011-11-17T19:06:15 ? 2011-11-17T19:06:16 amstan: which one does the command go to?! 2011-11-17T19:06:18 and? 2011-11-17T19:06:22 mcstar: Tried building your saue. 2011-11-17T19:06:23 *sauce 2011-11-17T19:06:26 ah 2011-11-17T19:06:32 i dont need that include 2011-11-17T19:06:40 mcstar: If you're standard-compliant, you do. 2011-11-17T19:06:51 really? 2011-11-17T19:07:03 it isnt in the "language"? 2011-11-17T19:07:17 i read those are all defined and i dont need anything else 2011-11-17T19:07:29 and, not, xor, and_eq, or_eq 2011-11-17T19:07:30 and such 2011-11-17T19:07:37 mcstar: They're keywords, but they're not required to take effect unless you include the header. 2011-11-17T19:08:30 Can't find the particular chapter and verse right now. 2011-11-17T19:08:45 QtWin's gcc 4.4 needs it anyway. 2011-11-17T19:08:48 http://en.cppreference.com/w/cpp/keywords 2011-11-17T19:09:06 mcstar: I know that they're keywords. They're treated specially. 2011-11-17T19:09:10 Trust me, I'm not making shit up. 2011-11-17T19:09:21 im just saying it isnt mentioned 2011-11-17T19:09:25 Also, you're using std::hash. Is this supposed to be 0x? 2011-11-17T19:09:35 yes of course 2011-11-17T19:09:41 it in the readme 2011-11-17T19:09:49 is* 2011-11-17T19:09:49 People don't read readmes :D 2011-11-17T19:09:54 Zao: bad zao 2011-11-17T19:10:08 It's you that asked to build it, not I. 2011-11-17T19:10:13 well, i dont read standard or comform to them so ... 2011-11-17T19:10:24 hey ok 2011-11-17T19:10:27 not complaining 2011-11-17T19:10:39 but my source of c++/c++11 features doesnt mention this 2011-11-17T19:10:46 mcstar: http://en.cppreference.com/w/cpp/language/operator_alternative they have a code example there wth no include 2011-11-17T19:11:10 yeah, i know it is working without it 2011-11-17T19:11:16 but zao might still be right 2011-11-17T19:11:36 *** rajanaresh has joined #aichallenge 2011-11-17T19:11:47 whoa. they added *more* digraphs? wtf 2011-11-17T19:11:49 In C++11 the headers have no effect. 2011-11-17T19:12:14 Zao: if you succeed in building it i wont install win7 2011-11-17T19:12:14 In C++03 the behaviour is essentially that of C89's definition of iso646.h 2011-11-17T19:12:45 mcstar: I can't build it, I don't have the 4.8.0 Qt build lying around anymore. 2011-11-17T19:12:51 Disk space isn't free :) 2011-11-17T19:13:02 And more importantly, it's time to sleep. 2011-11-17T19:13:09 Work is on the horizon. 2011-11-17T19:13:21 Zao: and depending on your timezone, the sun might be too :P 2011-11-17T19:14:11 Zao: oh k 2011-11-17T19:14:15 thx for trying 2011-11-17T19:14:44 mcstar: I'll try building if you like... though this will be a bare win7 install and I have absolutely 0 experience with compiling in a windows environment 2011-11-17T19:15:01 mcstar: tell me what to do and I'll provide lots of feedback :) 2011-11-17T19:15:30 aarossig: i guess you need mingw and the windows qt installer from qt.nokia.something 2011-11-17T19:15:52 mcstar: alright, well that's a start -- I'll let you know when my vm is ready 2011-11-17T19:16:05 aarossig: okey dokey 2011-11-17T19:16:14 its 1:15 here though 2011-11-17T19:16:22 does it mean i dont sleep tonight? 2011-11-17T19:16:33 mcstar: heh yeah.. it will definitely be tomorrow 2011-11-17T19:16:39 mcstar: I need about 3 more hours 2011-11-17T19:17:03 tomorrwo then 2011-11-17T19:18:16 *** grom358 has quit IRC (Ping timeout: 240 seconds) 2011-11-17T19:21:01 *** heinrich5991 has quit IRC (Ping timeout: 240 seconds) 2011-11-17T19:26:00 *** twymer has joined #aichallenge 2011-11-17T19:27:09 oops, kernel upgrade 2011-11-17T19:27:12 *** mcstar has quit IRC (Quit: WeeChat 0.3.6) 2011-11-17T19:29:04 *** heinrich5991 has joined #aichallenge 2011-11-17T19:32:03 *** grom358 has joined #aichallenge 2011-11-17T19:33:36 wow: http://forums.aichallenge.org/viewtopic.php?f=21&t=1648&start=30#p11701 2011-11-17T19:34:26 *** Antimony has joined #aichallenge 2011-11-17T19:34:52 *** bhasker has joined #aichallenge 2011-11-17T19:37:08 "Here's the EC2 instance we are running. Right now it is $0.168 per hour, $4.032 per day or $120.96 per month." 2011-11-17T19:37:35 what? are our costs really $121/month? that's absurdly low for what i expected. 2011-11-17T19:38:30 Zannick: where's that from? 2011-11-17T19:38:45 McLeopold said that on the first page of that thread you just linked 2011-11-17T19:43:29 *** meatroaf_ has quit IRC (Ping timeout: 265 seconds) 2011-11-17T19:44:44 *** gnutella has joined #aichallenge 2011-11-17T19:48:14 *** rajanaresh has quit IRC (Ping timeout: 244 seconds) 2011-11-17T19:48:24 *** gnutella has quit IRC (Client Quit) 2011-11-17T19:49:46 *** Anilm3 has quit IRC (Ping timeout: 265 seconds) 2011-11-17T19:51:27 *** Weird has joined #aichallenge 2011-11-17T19:52:09 lo people... say I have a strange situation hehe... 2011-11-17T19:52:25 ok 2011-11-17T19:52:26 I log all actions etc I do (made a java bot :p) 2011-11-17T19:53:02 but in a 'game vs bots' of say a few hundred moves.. it just stops somewhere with an errorcode of - 2011-11-17T19:53:21 number? 2011-11-17T19:53:23 anyone have a clue what that might be about? (as I don't ever use error codes to throw stuff) 2011-11-17T19:53:28 ye like -41 2011-11-17T19:53:31 or -34 2011-11-17T19:53:40 it seems kind of random to me :s 2011-11-17T19:53:44 what error code? 2011-11-17T19:53:48 where do you see that? 2011-11-17T19:53:59 in the logfile that I myself keep ;) 2011-11-17T19:54:08 and because the game 'randomly crashed' 2011-11-17T19:54:14 I put a try / catch around the whole thing 2011-11-17T19:54:31 how about letting the errors go? 2011-11-17T19:54:32 and save the exception that gets thrown in my log file a swell 2011-11-17T19:54:41 ye but I don't throw it Oo 2011-11-17T19:54:42 and see what the actual error is? 2011-11-17T19:54:54 via stderr 2011-11-17T19:55:11 *** Chris_0076 has quit IRC (Quit: Leaving) 2011-11-17T19:55:25 stderr? 2011-11-17T19:55:47 Weird: http://en.wikipedia.org/wiki/Standard_streams 2011-11-17T19:57:04 ah ye well what I do is... at the top of the 'doTurn' function I open my try... and at the end I close it... and I save the thrown exception to a file (in the catch) 2011-11-17T19:57:21 so that 'should' have the standard error in there innit Oo 2011-11-17T19:57:25 yes, comment out your code for the saving the exception for now 2011-11-17T19:57:28 and let it go to stderr 2011-11-17T20:00:13 *** analyst74 has quit IRC (Ping timeout: 240 seconds) 2011-11-17T20:01:49 ye so according to wikipedia I'd see it in the cmd promt when I just run the program innit... nothing strange there except that it kills my bot for no real reason :p 2011-11-17T20:03:11 the tutorial thingy works without any issue though... 2011-11-17T20:03:12 hmmm 2011-11-17T20:03:32 Weird, so you dont see your output in the console? 2011-11-17T20:03:45 I see lots of output which is normal 2011-11-17T20:03:50 aka the updated game states etc 2011-11-17T20:03:53 i mean the error :) 2011-11-17T20:04:20 Weird: did you enable -e? 2011-11-17T20:04:26 nope... it just sais that it 'eliminates me' (while there are no enemies anywhere it's like between move 1 and 10) 2011-11-17T20:05:17 ahh that might make a difference ;) 2011-11-17T20:06:17 hmmm it blocks some move, but that shouldn't really make it stop doing stuff right? 2011-11-17T20:06:27 dont think so 2011-11-17T20:06:52 *** NightExcessive_ has joined #aichallenge 2011-11-17T20:07:02 but no... I don't see any error there 2011-11-17T20:07:04 *** NightExcessive_ has quit IRC (Client Quit) 2011-11-17T20:07:14 except again that it's strange that it just eliminates me :s 2011-11-17T20:07:25 i wanted to point you at the -e flag like amstan 2011-11-17T20:07:30 I'm using that 'play_one_game.cmd' thingy 2011-11-17T20:07:37 you said you wrote an error to file 2011-11-17T20:07:38 ye that gives some more output... but no error :s 2011-11-17T20:07:44 yer 2011-11-17T20:07:46 and -E? 2011-11-17T20:07:49 with try / catch in java 2011-11-17T20:08:01 well what error is it? copy to pastebin 2011-11-17T20:08:26 Weird: remove the try catch 2011-11-17T20:08:50 ye I removed the try / catch 2011-11-17T20:09:00 it's just : Error: -41 2011-11-17T20:09:04 or -34 2011-11-17T20:09:11 and you have -E and it crashes without anything while saying that on stderr? 2011-11-17T20:09:15 which seem like random numbers to me :s 2011-11-17T20:09:28 well it doesn't actually crash... it just kills the bot :s 2011-11-17T20:09:34 *** datachomper has joined #aichallenge 2011-11-17T20:09:35 but no theres nothing on there 2011-11-17T20:09:48 not sure then 2011-11-17T20:09:50 like the game just plays and somewhere my ants stop moving ;) 2011-11-17T20:09:57 while the others still continue 2011-11-17T20:10:16 *** NightExcessive has quit IRC (Ping timeout: 255 seconds) 2011-11-17T20:10:30 *** datachomper has quit IRC (Client Quit) 2011-11-17T20:11:42 didnt hear that one before :O 2011-11-17T20:13:50 so tired.. gnight 2011-11-17T20:13:53 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-11-17T20:14:05 :< 2011-11-17T20:24:55 *** antimatroid has quit IRC (Quit: Leaving.) 2011-11-17T20:25:07 *** antimatroid has joined #aichallenge 2011-11-17T20:35:25 *** Antimony has quit IRC (Ping timeout: 260 seconds) 2011-11-17T20:38:05 *** roflmao has joined #aichallenge 2011-11-17T20:38:53 thestinger 2011-11-17T20:40:43 hi 2011-11-17T20:40:57 machine has been not working ALL DAY 2011-11-17T20:41:03 but I saw that replay -- awesome! 2011-11-17T20:41:12 :) 2011-11-17T20:41:50 and that was for non-moving targets? 2011-11-17T20:42:00 so you just rolled with that and it was still relatively betteR? 2011-11-17T20:42:09 yeah 2011-11-17T20:42:35 nice! 2011-11-17T20:42:39 now get it working for moving targets 2011-11-17T20:42:43 and it'll be a beast 2011-11-17T20:42:51 my bot now has some flaws that don't exist in my old version so I'm going to fix those first 2011-11-17T20:42:54 mostly movement stuff 2011-11-17T20:43:16 wow skill of 44 2011-11-17T20:43:17 pro 2011-11-17T20:43:32 http://aichallenge.org/profile.php?user=2255 look at my old version o_o 2011-11-17T20:43:37 no combat and crappy foraging 2011-11-17T20:44:13 I think it should be around rank 25-30, but it got lucky in a few games 2011-11-17T20:45:01 can't wait to try my code -- once i get this stupid machine working lol 2011-11-17T20:46:38 atm, I'm switching around container types and refactoring 2011-11-17T20:46:57 I ended up abusing some vectors/lists with my new code (searching through linked lists...) 2011-11-17T20:48:35 are you going to try to get combat working for moving targets 2011-11-17T20:48:41 or are you just going to keep your current new combat code 2011-11-17T20:49:25 I think I'll end up using this code for _all_ my ants, because it's fast 2011-11-17T20:49:31 ah 2011-11-17T20:49:31 *** smiley1983 has quit IRC (Read error: Operation timed out) 2011-11-17T20:49:41 then I'll use remaining time to do better combat for a few ants 2011-11-17T20:49:42 *** smiley1983 has joined #aichallenge 2011-11-17T20:49:57 btw if you ever want a semi-decent bot to compete against locally 2011-11-17T20:50:01 i'd be happy to send you my jar 2011-11-17T20:50:16 big battles seem to mostly be big static battle lines, so I should be fine with those (and trying to use minimax or something there would be way too slow) 2011-11-17T20:52:20 true 2011-11-17T20:52:37 what i want to implement though is zapping for the ants that are NOT part of the battle line 2011-11-17T20:52:40 to hurt the enemies food production 2011-11-17T20:52:43 and win the war of attrition 2011-11-17T20:52:52 hence the need to be able to manage moving targets 2011-11-17T20:53:22 yeah, my combat code doesn't protect my foragers because they don't realize they're stepping into a place 2 enemies are about to be able to attack 2011-11-17T20:54:29 yeah it's tricky 2011-11-17T21:01:37 *** GeorgeSebastian has quit IRC (Read error: Connection reset by peer) 2011-11-17T21:01:42 *** traMKer has quit IRC (Read error: Connection reset by peer) 2011-11-17T21:04:17 *** mviel__ has joined #aichallenge 2011-11-17T21:05:03 *** smiley1983 has quit IRC (Ping timeout: 248 seconds) 2011-11-17T21:06:36 *** Antimony has joined #aichallenge 2011-11-17T21:07:33 *** smiley1983 has joined #aichallenge 2011-11-17T21:08:14 *** mviel_ has quit IRC (Ping timeout: 252 seconds) 2011-11-17T21:11:47 I need to pick a good ratio of kills:losses to accept 2011-11-17T21:12:18 *** smiley1983 has quit IRC (Ping timeout: 276 seconds) 2011-11-17T21:12:34 *** smiley1983 has joined #aichallenge 2011-11-17T21:12:47 *** Chris_0076 has joined #aichallenge 2011-11-17T21:17:18 *** GeorgeSebastian has joined #aichallenge 2011-11-17T21:21:26 *** smiley1983 has quit IRC (Ping timeout: 252 seconds) 2011-11-17T21:23:03 Fluxid: reproduced the problem in game 3636. My own logs confirm it again, could you please post the input/output/error logs for that game? 2011-11-17T21:23:53 thestinger: kills > losses works for me 2011-11-17T21:24:55 *** smiley1983 has joined #aichallenge 2011-11-17T21:25:18 *** treeform has quit IRC (Read error: Operation timed out) 2011-11-17T21:28:56 *** Chris_0076 has quit IRC (Quit: Leaving) 2011-11-17T21:29:31 *** number_prophet is now known as modafinil 2011-11-17T21:32:59 *** smiley1983 has quit IRC (Ping timeout: 252 seconds) 2011-11-17T21:33:19 *** smiley1983 has joined #aichallenge 2011-11-17T21:36:41 *** roflmao1 has joined #aichallenge 2011-11-17T21:36:55 * avdg wonders when he finally learns c++ :/ 2011-11-17T21:40:16 The good news: I reduced my bot's move time from 11 seconds to 4 2011-11-17T21:40:23 The bad news: It's still takes 4 seconds to move! 2011-11-17T21:40:37 wow, is it really that long? :p 2011-11-17T21:40:45 does that include game trees or something? 2011-11-17T21:41:01 That's the itme for my new constraint solver 2011-11-17T21:41:02 well, I quited my js bot when it needed 200ms to move 50 bots :/ 2011-11-17T21:41:08 My old one could solve it in like 50ms 2011-11-17T21:41:20 I'm not sure why my new version is slower 2011-11-17T21:41:38 did you add features? 2011-11-17T21:41:46 no 2011-11-17T21:41:56 I think the reason is that I changed from DFS to UCS 2011-11-17T21:42:19 But UCS should be faster 2011-11-17T21:42:47 well, I hope you can easily revert your choices, so you can easily compare both 2011-11-17T21:42:55 there are probably some constrains 2011-11-17T21:43:30 My old system was a mess 2011-11-17T21:43:47 *** fpcfan has quit IRC (Ping timeout: 245 seconds) 2011-11-17T21:44:12 my new system makes it possible to add features. For example, I can now prevent ants from switching places 2011-11-17T21:44:46 and it was supposed to be a lot faster too 2011-11-17T21:44:47 heh, I wished I could use my js bot without much delay 2011-11-17T21:45:10 but I failed with it after I added coverage trackers 2011-11-17T21:45:28 which wasn't that much code, but still making the bot 10 times slower 2011-11-17T21:45:37 *** emiel_ has joined #aichallenge 2011-11-17T21:45:52 probably because of the size of the data, making it slower to be processed 2011-11-17T21:46:04 hey there, it seems the coffeescript starter package is empty 2011-11-17T21:46:10 well I'm doing C++, so there's no reason for it to ever be that slow 2011-11-17T21:46:24 and especially not sloewr than the older unoptimized version 2011-11-17T21:46:28 I'm sucking with having c++ arrays I want :p 2011-11-17T21:46:36 ? 2011-11-17T21:46:46 2-bits 2d arrays 2011-11-17T21:46:57 *** roflmao1 has quit IRC (Read error: Connection reset by peer) 2011-11-17T21:47:05 Just make your own class 2011-11-17T21:47:25 yeah, I was thinking about it 2011-11-17T21:47:33 but I was worried about making it inefficient 2011-11-17T21:48:04 but anyway 2011-11-17T21:48:06 just code it first and worry about efficiency later 2011-11-17T21:48:09 that's what I did 2011-11-17T21:48:25 unless you have measurable delay, there's no reason to spend effort optimizing it 2011-11-17T21:48:30 yeah, but I don't know c++ that well, thats why I was worried 2011-11-17T21:48:41 especially since you have no way of knowing whether the change acutally worked 2011-11-17T21:49:30 yeah, let me dig into the do-it-yourself direction (despite I hoped c++ had a little bit more to offer) 2011-11-17T21:49:53 why not just use the existing array implementations? 2011-11-17T21:49:59 *** smiley1983 has quit IRC (Ping timeout: 252 seconds) 2011-11-17T21:50:09 std::vector is fine for a 2d array (or std::array if you really care about stack vs. heap) 2011-11-17T21:50:13 like what? boost? 2011-11-17T21:50:15 well if you can make something more speed efficient by design it's usualy a good idea to start with that ;p 2011-11-17T21:50:24 I want it fixed size once set up 2011-11-17T21:50:29 so, std::array 2011-11-17T21:50:32 dunno if a vector fits in that 2011-11-17T21:50:38 Weird: careful with premature optimization 2011-11-17T21:51:03 *** fpcfan has joined #aichallenge 2011-11-17T21:51:04 ah amstan I'm still just working on making this thing run :s 2011-11-17T21:51:05 My problem right now is postmature optimization 2011-11-17T21:51:17 amstan: premature abstraction is worse :) 2011-11-17T21:51:19 (dunno if I'm doing premature optimization, I just know what I want) 2011-11-17T21:51:24 I've sped it up by a factor of 4 already, but it's still 100 times slower than it should be 2011-11-17T21:51:24 Weird: so.. why did you come up with your own logging system? 2011-11-17T21:51:31 but tbh I have a pretty nice (very fast) solution because I designed it to be really fast ;) 2011-11-17T21:51:58 amstan ehhh I just always log stuff by just making some 'key functions' return a string which I can write to disk ;) 2011-11-17T21:52:29 *** userjjb has quit IRC (Quit: Page closed) 2011-11-17T21:53:44 it's not something I spent any time on making hehe ;p 2011-11-17T21:53:57 designed to be fast? you have to write code.. 2011-11-17T21:54:07 why didn't you just use the logging functions from the engine? 2011-11-17T21:54:20 all you have to do is system.err.println 2011-11-17T21:54:25 or w/e 2011-11-17T21:54:26 *** emiel_ has quit IRC (Quit: emiel_) 2011-11-17T21:54:46 ye dunno... I once started it like this :P 2011-11-17T21:54:49 kind of a habbit 2011-11-17T21:55:23 mainly cuz I dislike stuff other than notepad haha :P 2011-11-17T21:56:03 *** smiley1983 has joined #aichallenge 2011-11-17T21:56:09 avdg: std::vector is just an array (single block of memory) allocated on the heap, std::array is on the stack like a C array 2011-11-17T21:56:55 std::vector> is a 2D-array of ints, etc. - you don't need to reinvent it :P 2011-11-17T21:57:34 meh, I still don't know c++ enough 2011-11-17T21:57:49 so my ideas and c++ aren't quite matching that well 2011-11-17T21:58:03 I learned C++ for this too :) 2011-11-17T21:58:13 http://en.cppreference.com/w/cpp <- that's a great reference for the standard lib 2011-11-17T21:58:29 yeah, I already discovered that site 2011-11-17T21:59:07 *** emiel_ has joined #aichallenge 2011-11-17T21:59:09 its great, though unfinished and not for education (ohwell, I don't want to be hard) 2011-11-17T21:59:39 its a great reference though 2011-11-17T22:00:15 *** datachomper has joined #aichallenge 2011-11-17T22:00:45 (now squeezing the int in 4 parts and I'm happy :p) 2011-11-17T22:00:49 *** callahan has quit IRC (Ping timeout: 260 seconds) 2011-11-17T22:01:08 hm, what do you mean? 2011-11-17T22:01:24 well, I don't want to use 8 bits when I only use 2 2011-11-17T22:01:36 especially not when I'm scaling it 2011-11-17T22:01:47 bool is 1-byte and not 1 bit for a reason :) 2011-11-17T22:01:55 using individual bits will be slower, but saves memory 2011-11-17T22:02:31 oh well, then I probably want to take that cost as well 2011-11-17T22:03:13 the C way would be to make a struct to hold them (you can do bit fields in a struct, but just using 2 bools is better) 2011-11-17T22:03:29 C++ has std::pair and std::tuple which are basically just anonymous structs 2011-11-17T22:03:39 std::pair looks like what you want 2011-11-17T22:03:54 *** Scryer has joined #aichallenge 2011-11-17T22:04:16 yeah, it gets close 2011-11-17T22:04:36 but my brains don't believe it yet :p not that I don't trust you ;-) 2011-11-17T22:05:46 std::vector uses only 1 bit per entry 2011-11-17T22:06:09 otherwise it's almost certainly faster to use at least one byte (and possible a double-word) per tile 2011-11-17T22:06:17 Yexo: I think they fixed that in C++11 2011-11-17T22:06:58 the std::vector thing? 2011-11-17T22:07:02 *** roflmao has quit IRC (Ping timeout: 265 seconds) 2011-11-17T22:07:51 maybe not 2011-11-17T22:08:49 yeah, I guess they didn't fix it 2011-11-17T22:09:25 you can just use uint8_t from cstdint though 2011-11-17T22:09:31 http://www.cplusplus.com/reference/stl/bitset/ 2011-11-17T22:10:18 *** bhasker has quit IRC (Quit: bhasker) 2011-11-17T22:12:30 *** GeorgeSebastian has quit IRC (Ping timeout: 244 seconds) 2011-11-17T22:12:40 yeah, they want to fix std::vector _after_ the now finished C++11 standard >.< 2011-11-17T22:13:59 *** yoden has quit IRC (Quit: Leaving.) 2011-11-17T22:14:54 *** callahan has joined #aichallenge 2011-11-17T22:15:41 http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2007/n2160.html 2011-11-17T22:19:28 *** mleise has quit IRC (Ping timeout: 240 seconds) 2011-11-17T22:21:19 *** GeorgeSebastian has joined #aichallenge 2011-11-17T22:23:23 *** mviel_ has joined #aichallenge 2011-11-17T22:24:41 *** smiley1983 has quit IRC (Ping timeout: 252 seconds) 2011-11-17T22:26:40 *** mviel__ has quit IRC (Ping timeout: 240 seconds) 2011-11-17T22:31:27 *** Antimony has quit IRC (Ping timeout: 248 seconds) 2011-11-17T22:36:23 *** smiley1983 has joined #aichallenge 2011-11-17T22:45:51 *** smiley1983 has quit IRC (Ping timeout: 248 seconds) 2011-11-17T22:47:52 *** smiley1983 has joined #aichallenge 2011-11-17T22:52:16 *** smiley1983 has quit IRC (Ping timeout: 258 seconds) 2011-11-17T23:00:55 *** treeform has joined #aichallenge 2011-11-17T23:02:14 *** Antimony has joined #aichallenge 2011-11-17T23:04:04 *** Scryer has quit IRC (Ping timeout: 265 seconds) 2011-11-17T23:04:23 *** BenJackson has joined #aichallenge 2011-11-17T23:05:02 *** goldcaddy77 has quit IRC (Ping timeout: 265 seconds) 2011-11-17T23:08:35 *** avdg has quit IRC (Quit: Leaving.) 2011-11-17T23:10:58 *** avdg has joined #aichallenge 2011-11-17T23:15:59 *** rajanaresh has joined #aichallenge 2011-11-17T23:32:04 *** bhasker has joined #aichallenge 2011-11-17T23:34:09 *** smiley1983 has joined #aichallenge 2011-11-17T23:38:16 *** smiley1983 has quit IRC (Ping timeout: 244 seconds) 2011-11-17T23:40:20 *** Conorach has quit IRC (Ping timeout: 244 seconds) 2011-11-17T23:46:58 *** analyst74 has joined #aichallenge 2011-11-17T23:52:29 *** djr_ has joined #aichallenge 2011-11-17T23:58:25 *** treeform has quit IRC (Remote host closed the connection) 2011-11-17T23:59:02 *** Weird has quit IRC (Quit: Page closed) 2011-11-17T23:59:09 *** JorgeB has joined #aichallenge