2011-05-18T00:01:33 *** bhasker has joined #aichallenge 2011-05-18T00:01:36 Frontier: you around? 2011-05-18T00:09:37 *** AlliedEnvy has quit IRC (*.net *.split) 2011-05-18T00:12:27 *** AlliedEnvy has joined #aichallenge 2011-05-18T00:12:29 *** znutar_ has quit IRC (Excess Flood) 2011-05-18T00:12:34 *** znutar has joined #aichallenge 2011-05-18T00:19:10 aichallenge: Scott Hamilton epsilon * r1edebf6 / INSTALL.md : Edited INSTALL.md via GitHub - http://bit.ly/k7lKCD 2011-05-18T00:22:05 *** smiley1983 has joined #aichallenge 2011-05-18T00:23:29 *** smiley1983 has quit IRC (Client Quit) 2011-05-18T00:32:02 @later tell janzert: unionfs-fuse causes quite a cpu hit when running bots, plus, it seems to cause my vm to be unstable when running updatedb 2011-05-18T00:32:02 McLeopold: Aye, aye, sir 2011-05-18T00:44:38 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-05-18T01:12:45 @later tell janzert: why do we download .deb packages instead of using apt-get for languages, like scala? 2011-05-18T01:12:45 McLeopold: Job's done. 2011-05-18T01:32:23 *** javagamer has quit IRC (Ping timeout: 276 seconds) 2011-05-18T01:40:17 *** needsch has joined #aichallenge 2011-05-18T01:43:56 *** needsch has quit IRC (Read error: Connection reset by peer) 2011-05-18T02:01:02 *** needsch has joined #aichallenge 2011-05-18T02:04:31 *** javagamer has joined #aichallenge 2011-05-18T02:12:20 *** antimatroid has joined #aichallenge 2011-05-18T02:42:53 *** llk1800 has joined #aichallenge 2011-05-18T02:49:24 *** amstan has quit IRC (Remote host closed the connection) 2011-05-18T02:56:49 bhasker: happy birthday! :P 2011-05-18T03:04:00 *** llk1800 has quit IRC (Quit: 离开) 2011-05-18T03:19:10 *** needsch has quit IRC (Quit: Leaving.) 2011-05-18T03:25:27 *** aerique has joined #aichallenge 2011-05-18T03:27:56 McLeopold: I'm around, yes 2011-05-18T03:39:25 Frontier: I figured out my front page issue. It was a combination of an old minified script and some other new stuff 2011-05-18T03:40:27 *** ltriant has quit IRC (Quit: Computer has gone to sleep) 2011-05-18T03:44:08 edinburgh looks pretty cool 2011-05-18T03:44:16 * antimatroid is looking at places to do a phd :P 2011-05-18T03:48:02 mit? 2011-05-18T03:52:09 Don't you need to be agressive like Gordon Freeman to get a place there? 2011-05-18T03:52:38 McLeopold: why will Makefiles not be used for official compilation? 2011-05-18T03:52:55 *** delt0r___ has joined #aichallenge 2011-05-18T03:53:02 because it's too much to implement for this upcoming contest 2011-05-18T03:53:04 McLeopold: How is the web site built right now? I have this "ant deploy" thingy in the visualizer source, but I don't know if any script is using it. 2011-05-18T03:53:24 Frontier: it's run by hand right now, I think 2011-05-18T03:53:59 Frontier: it needs root for downloading rhino, but then you have to chown stuff later. It needs work 2011-05-18T03:54:31 *** delt0r_ has quit IRC (Ping timeout: 258 seconds) 2011-05-18T03:54:43 McLeopold: I'm not overly keen on moving to the US :P 2011-05-18T03:54:59 Well I am not a root user on my machine, do you know why cvs needs root? 2011-05-18T03:55:11 i have no problem with most of the people or most of the places, but yeah, I almost want to move away from Australia based on principle, the US just isn't up there on my list :P 2011-05-18T03:55:17 otherwise mit would be my number one preference 2011-05-18T03:55:55 antimatroid: i'd chose education over politics in this case 2011-05-18T03:55:55 Frontier: yeah, MIT is one of the best uni's in the world 2011-05-18T03:56:18 aerique: there are many good universities outside the US, although funding isn't as available 2011-05-18T03:56:39 and then i'd have to go through the ordeal of gre's :P 2011-05-18T03:57:14 I'm not sure if I would move to the US, but I'm already here. 2011-05-18T03:57:25 our food sucks 2011-05-18T03:57:46 and some of the ridiculous laws one can't escape 2011-05-18T03:58:03 so yeah, i'm looking at europe 2011-05-18T03:58:05 every place has those 2011-05-18T03:58:10 McLeopold: I used the ant script on Windows and it worked there, too. I really wonder what needs root access there? Is it something with an incorrect wrong working folder? It should download all the stuff right into the source code folder. 2011-05-18T03:58:11 it would be cool to get a change 2011-05-18T03:58:12 UK has bad food too 2011-05-18T03:58:43 Frontier: I'd have to uninstall and re-run to know why. 2011-05-18T03:58:48 I didn't keep the logs 2011-05-18T03:59:30 *** smiley1983 has joined #aichallenge 2011-05-18T04:01:26 I seem to recall some talk awhile ago about the possibility of using makefiles to compile the bots. Is this option still being considered? 2011-05-18T04:02:03 smiley1983: apparently not: https://github.com/aichallenge/aichallenge/issues/112 2011-05-18T04:02:57 aerique: thank you 2011-05-18T04:03:22 McLeopold: can't we use makefiles because of worker/compiler.py? 2011-05-18T04:03:43 aerique: it doesn't support it right now 2011-05-18T04:03:58 Just use a language the can go with default compiler settings already :p 2011-05-18T04:04:13 There are some compiler flags which I would like to use for final compilation before the submission deadline, but the optimizations can be slow, and there's no reason to chew up server time mid-contest for them. Makefiles would provide a neat solution to that. 2011-05-18T04:04:14 janzert migh have got the jail working for compilation, and if so, then we may be able to do it 2011-05-18T04:04:57 yes, we all think makefiles would be good, it's a matter of time 2011-05-18T04:05:12 McLeopold: if I fix the CL starter to use MyBot.lisp, can I test the changes myself on a test server? 2011-05-18T04:05:14 Right, thankyou for clarifying 2011-05-18T04:05:24 that's a good enough reason indeed 2011-05-18T04:05:38 I did suspect it might pose some security considerations. 2011-05-18T04:05:45 aerique, yes, the compiler options are in the source... 2011-05-18T04:06:32 https://github.com/aichallenge/aichallenge/blob/epsilon/worker/compiler.py#L216 2011-05-18T04:07:19 yeah, i'm looking at that, but i meant whether i can upload a submission.zip somewhere 2011-05-18T04:07:49 we don't have one running consistently yet 2011-05-18T04:08:03 alright, i'll just make the changes then and get back to work ;-) 2011-05-18T04:10:24 aerique: although, check in with http://hypertriangle.com:13080/ every so often 2011-05-18T04:10:47 thanks 2011-05-18T04:12:20 aerique: btw, I'm new to lisp and trying to learn, but your starter bot still feels out of reach. perhaps you could make a single file starter bot, and release the full shebang somewhere else? 2011-05-18T04:13:48 *** Naktibalda has joined #aichallenge 2011-05-18T04:15:16 McLeopold: I'll see what I can do but I currently do not have a lot of time to spend on it. I also made the initial more elaborate version at the request of mega, since he wanted ASDF and proxy-bot support :-) (but that has moved elsewhere already: https://github.com/aerique/google-ai-challenge-2011-1-ants) 2011-05-18T04:16:19 but i do get the reasoning for make it easy for total beginners, so i'll try 2011-05-18T04:16:44 aerique: I still don't get proxy-bot. I need a video of watching a lisper develop in the cli. 2011-05-18T04:18:15 i'll try and find something on youtube for you (whether it's clojure, cl or scheme doesn't really matter for the general idea) 2011-05-18T04:18:54 cool 2011-05-18T04:19:41 when reading lisp books, they only ever say "lisper like to develop live", but never go into details of techniques 2011-05-18T04:20:20 it's not quite dissimilare to live coding if you've ever seen videos of that 2011-05-18T04:21:51 I've done live coding in debug mode in VS and Eclipse, but that's it. 2011-05-18T04:29:43 aichallenge: McLeopold epsilon * r3636c29 / (4 files in 2 dirs): worker updates, totally broke rankings paginations, but at least something is showing - http://bit.ly/mfyLqs 2011-05-18T04:29:43 aichallenge: McLeopold epsilon * ra52ad8b / (INSTALL.md setup/server_setup.py): Merge branch 'epsilon' of https://github.com/aichallenge/aichallenge into epsilon - http://bit.ly/lXjuMc 2011-05-18T04:29:52 *** smiley1983 has quit IRC (Quit: leaving) 2011-05-18T04:31:04 *** McLeopold has quit IRC (Quit: Leaving.) 2011-05-18T05:03:06 *** Ice_Harley has joined #aichallenge 2011-05-18T05:20:27 aichallenge: aerique epsilon * rcf2ad19 / (2 files in 2 dirs): Compile MyBot using MyBot.lisp. (issue 112) - http://bit.ly/jAQacc 2011-05-18T05:33:50 *** mcstar has joined #aichallenge 2011-05-18T06:59:12 *** antimatroid has quit IRC (Ping timeout: 248 seconds) 2011-05-18T07:12:15 McLeopold: this one's pretty old but should get the idea across: http://common-lisp.net/project/movies/movies/slime.mov (there's more here: http://www.cliki.net/Lisp%20Videos) 2011-05-18T07:12:26 oh, he isn't here 2011-05-18T07:12:52 *** antimatroid has joined #aichallenge 2011-05-18T07:26:37 the french guy's presentation? 2011-05-18T07:27:46 mcstar: are you talking to me? :) 2011-05-18T07:27:50 yep 2011-05-18T07:28:42 (it was kind of De Niro: you talking to me? huh you talking to me?) 2011-05-18T07:29:00 mcstar: no, I promised McLeopold to give him a link about a typical Lisp development session 2011-05-18T07:29:22 slime.mov is the one with the french guy on osx 2011-05-18T07:29:30 showing off marse code in slime 2011-05-18T07:29:33 morse 2011-05-18T07:30:10 *** FireFly has joined #aichallenge 2011-05-18T07:32:27 ah, like so.. then you were indeed correct :) 2011-05-18T07:35:12 do you use slime yourself? 2011-05-18T07:36:05 im using slimv, slim for vim 2011-05-18T07:37:16 yup 2011-05-18T07:37:25 how's slimv? 2011-05-18T07:38:05 im kind of a beginner 2011-05-18T07:38:09 so far so good 2011-05-18T07:38:32 better shortcuts as far as im concerned 2011-05-18T07:38:43 *** needsch has joined #aichallenge 2011-05-18T07:38:52 i'm happy there's an alternative to emacs when using slime 2011-05-18T07:39:35 if it werent i would use emacs 2011-05-18T07:39:48 it isnt that hard, but unusual 2011-05-18T07:40:27 slimv is a much younger project than slime 2011-05-18T07:40:58 is McLeopold interested in cl? 2011-05-18T07:41:23 yes, he said he was learning it 2011-05-18T07:43:05 it is kind of pointless 2011-05-18T07:43:17 but i have to share with you how much i like the loop macro :) 2011-05-18T07:43:34 it's my favourite thing as well :) 2011-05-18T07:44:28 one thing i noticed, that i almost read lisp code as i read text 2011-05-18T07:44:35 very easy on the eyes 2011-05-18T07:44:55 takes time to write it, but then it feels easy to read 2011-05-18T07:45:43 good to hear. that's quite a different viewpoint than most people have concerning lisp code :) 2011-05-18T07:45:56 i think ive read your intro to GA 2011-05-18T07:46:24 do you have some other in-depth example of it? 2011-05-18T07:47:39 you mean this one?: http://aerique.blogspot.com/2011/01/baby-steps-into-genetic-programming.html 2011-05-18T07:47:48 gp, sorry 2011-05-18T07:47:53 thats it 2011-05-18T07:48:19 what was the speed of it? 2011-05-18T07:48:43 you wrote that you run it for 300 iterations? 2011-05-18T07:48:57 i have no idea actually. it very much depends on the cl implementation 2011-05-18T07:49:08 feel free to benchmark it yourself :) 2011-05-18T07:49:41 500 total runs in that example 2011-05-18T07:50:13 i thought about doing GA for this contest 2011-05-18T07:50:26 i hoping to do GP :) 2011-05-18T07:50:39 but i decided that i shouldnt, because i cant guarantee that ill get good results 2011-05-18T07:50:53 i started to write an own visualizer and an engine 2011-05-18T07:51:01 i contributed a CL version for this GA Hello World project: https://github.com/jsvazic/GAHelloWorld 2011-05-18T07:51:02 i used lispbuilder-sdl 2011-05-18T07:51:13 cool 2011-05-18T07:51:43 ill read it 2011-05-18T07:51:56 i'm not expecting good results this time, but i'm definitely going to learn a lot 2011-05-18T07:52:11 can i ask you about that? 2011-05-18T07:52:15 have you got any screenshots of your visualizer? 2011-05-18T07:52:25 well its more like code 2011-05-18T07:52:32 its a canvas with dots 2011-05-18T07:52:40 yeah, sure. perhaps in a private channel, i don't know if the other people are interested 2011-05-18T07:52:48 ok 2011-05-18T08:19:53 aichallenge: janzert epsilon * r7f66265 / worker/compiler.py : Fix go compiler commands for 64bit - http://bit.ly/mcyESF 2011-05-18T08:56:54 aichallenge: janzert epsilon * rd0d5be6 / worker/compiler.py : Include error message when compiler exits with an error return code - http://bit.ly/lJNrvp 2011-05-18T09:35:51 *** Ice_Harley has quit IRC (Ping timeout: 253 seconds) 2011-05-18T10:03:05 *** needsch has quit IRC (Quit: Leaving.) 2011-05-18T10:03:22 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 4.0.1/20110413222027]) 2011-05-18T10:13:52 *** _0x47 <_0x47!~slash@p3EE0A10E.dip.t-dialin.net> has joined #aichallenge 2011-05-18T10:16:53 *** _0x471 <_0x471!~slash@p3EE0B919.dip.t-dialin.net> has joined #aichallenge 2011-05-18T10:17:58 *** _0x47 <_0x47!~slash@p3EE0A10E.dip.t-dialin.net> has quit IRC (Ping timeout: 246 seconds) 2011-05-18T10:30:13 *** _0x471 <_0x471!~slash@p3EE0B919.dip.t-dialin.net> has quit IRC (Ping timeout: 246 seconds) 2011-05-18T10:33:45 *** Frontier has quit IRC (Quit: Leaving.) 2011-05-18T10:36:30 *** _0x47 <_0x47!~slash@p3EE0A59A.dip.t-dialin.net> has joined #aichallenge 2011-05-18T10:37:39 *** aerique has quit IRC (Quit: ...) 2011-05-18T10:51:18 bam, one assignment done 2011-05-18T11:24:48 *** fgump has joined #aichallenge 2011-05-18T11:43:28 *** berak has joined #aichallenge 2011-05-18T11:51:43 *** orivej has joined #aichallenge 2011-05-18T11:53:05 *** delt0r___ has quit IRC (Read error: Operation timed out) 2011-05-18T11:53:23 *** delt0r___ has joined #aichallenge 2011-05-18T12:04:14 contestbot: later tell McLeopold does compiler.py compilet the java and scala starter bots for you? I get an error because the compiler exits with a return of 1 2011-05-18T12:04:14 janzert: I think that worked... 2011-05-18T12:05:28 contestbot: later tell McLeopold where we use .deb files directly is because the repositories either contain too old of a version or don't have it at all. 2011-05-18T12:05:28 janzert: Yes master! 2011-05-18T12:11:36 contestbot: later tell McLeopold Figured out the problem with java and scala compilation 2011-05-18T12:11:37 janzert: Job's done. 2011-05-18T12:13:25 aichallenge: janzert epsilon * r8dc15a2 / setup/worker_setup.py : Up the sandbox process limit so java will run - http://bit.ly/k9eu4K 2011-05-18T13:00:48 *** McLeopold has joined #aichallenge 2011-05-18T13:02:15 janzert: Scala compilation was looking for *.class files, but the starter bot included *.scala files, so I changed what extension is looked for. I'm not sure if that was correct or not. 2011-05-18T13:02:36 janzert: I'm running 32 bit ubuntu and can't get go to compile 2011-05-18T13:03:03 nor d 2011-05-18T13:04:04 *** McLeopold has left #aichallenge 2011-05-18T13:07:24 *** McLeopold has joined #aichallenge 2011-05-18T13:26:59 *** GeorgeSebastian has joined #aichallenge 2011-05-18T13:32:32 *** GeorgeSebastian has quit IRC (Ping timeout: 240 seconds) 2011-05-18T13:35:38 thanks Antimatroid! 2011-05-18T13:46:18 *** GeorgeSebastian has joined #aichallenge 2011-05-18T13:52:01 *** GeorgeSebastian has quit IRC (Ping timeout: 248 seconds) 2011-05-18T14:07:49 *** choas has joined #aichallenge 2011-05-18T14:09:10 McLeopold: my problem with java and scala turned out to be that java needs to be allowed 15 threads/processes before it will run 2011-05-18T14:09:20 and the sandbox was limiting it to 10 2011-05-18T14:10:15 go is rather a pain because it uses a different compiler name depending on if you're compiling 32bit or 64bit 2011-05-18T14:10:39 and I haven't even tried to see if D compiles yet 2011-05-18T14:26:08 i386 doesn't install 2011-05-18T14:26:18 I need gcc-multilib, but ubuntu doesn't like it 2011-05-18T14:34:34 janzert: will we be able to add a language to the workers without having to start from scratch? 2011-05-18T14:35:56 aichallenge: McLeopold epsilon * ra1c745f / (4 files): added ranking json - http://bit.ly/mGL2iE 2011-05-18T14:35:57 aichallenge: McLeopold epsilon * rdcf92f4 / (3 files in 3 dirs): Merge branch 'epsilon' of https://github.com/aichallenge/aichallenge into epsilon - http://bit.ly/khtnPa 2011-05-18T14:41:37 *** needsch has joined #aichallenge 2011-05-18T14:45:08 *** Kingpin13 has joined #aichallenge 2011-05-18T14:55:51 have anybody considered a starter pack written in bash? 2011-05-18T15:02:51 *** berak has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2011-05-18T15:13:40 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-05-18T15:24:02 *** amstan has joined #aichallenge 2011-05-18T15:24:02 *** ChanServ sets mode: +o amstan 2011-05-18T15:27:56 *** Frontier has joined #aichallenge 2011-05-18T15:39:35 *** boegel has joined #aichallenge 2011-05-18T15:56:06 *** Accoun has quit IRC () 2011-05-18T16:07:05 *** fgump has quit IRC (Quit: Leaving) 2011-05-18T16:09:41 *** boegel has quit IRC (Quit: Leaving) 2011-05-18T16:12:37 *** Accoun has joined #aichallenge 2011-05-18T16:30:05 *** Naktibalda has joined #aichallenge 2011-05-18T17:00:07 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 4.0.1/20110422205559]) 2011-05-18T17:02:09 *** _0x47 <_0x47!~slash@p3EE0A59A.dip.t-dialin.net> has quit IRC (Quit: Leaving.) 2011-05-18T17:03:18 *** chris___0076 has joined #aichallenge 2011-05-18T17:05:43 *** chris__0076 has quit IRC (Ping timeout: 246 seconds) 2011-05-18T17:26:48 McLeopold: for a new language you should be able to update compiler.py and install the new language into the base chroot and be good to go 2011-05-18T17:27:18 with the setup script, or by hand? 2011-05-18T17:27:34 at the same time, for planetwars the half life of a worker was approximately 8-10 hours so killing them all and starting fresh is what I would do on code updates 2011-05-18T17:27:37 by hand 2011-05-18T17:27:57 do you think our half-life will be higher? 2011-05-18T17:28:08 I really hope so 2011-05-18T17:28:25 should we higher Gordon? isn't he freelance? 2011-05-18T17:28:36 :) 2011-05-18T17:28:55 let me share rankings 2011-05-18T17:29:05 git push coming... 2011-05-18T17:29:30 even if the workers are stable I don't have much aversion to killing them off and starting fresh ones in order to update them 2011-05-18T17:29:45 I don't think we'll be doing that too frequently 2011-05-18T17:30:06 it takes 30 minutes to install, though 2011-05-18T17:30:15 btw, games are really slow in the jails 2011-05-18T17:30:29 games are slow out of the jail as well 2011-05-18T17:30:51 really? I can see disk access being fairly slow in the jail but a game shouldn't be doing that much at all 2011-05-18T17:31:03 our game is a lot of data to churn through 2011-05-18T17:31:28 as far as cpu or memory overhead the jail shouldn't be adding any at all 2011-05-18T17:31:36 I ran some tests at 20 turns and it was taking a lot longer in the jails 2011-05-18T17:31:51 outside of them, it was almost instant 2011-05-18T17:32:16 what could be causing it? 2011-05-18T17:32:39 *** Kingpin13 has quit IRC (Quit: this used to be the life, but I don't need another one) 2011-05-18T17:32:40 it could be my system, it keeps getting stuck and not responding 2011-05-18T17:32:51 Oo 2011-05-18T17:33:21 I haven't noticed any difference in speed here running the play_one_game.sh script when in or out of the jails 2011-05-18T17:33:42 and you both used the same setup? 2011-05-18T17:33:49 I haven't put a stopwatch to it and timed it though 2011-05-18T17:34:27 oh, good dmd has a amd64 .deb specifically for ubuntu now 2011-05-18T17:34:53 *** choas has quit IRC (Quit: leaving) 2011-05-18T17:35:27 :D 2011-05-18T17:35:27 let me try play_one_game, I was doing worker stuff 2011-05-18T17:37:49 also I'll be leaving town tomorrow for a week of work, so I'll probably not be around at all 2011-05-18T17:37:50 Would it help if I tested it, too? 2011-05-18T17:38:16 wouldn't hurt :) 2011-05-18T17:38:31 but you pretty much need a free machine or fresh vm to install on 2011-05-18T17:39:05 *** jmpespxoreax has quit IRC (Read error: Operation timed out) 2011-05-18T17:39:15 hmm, I hoped a chroot folder would be enough ^^ 2011-05-18T17:40:16 the worker is doing semi-tricky things with chroot and unionfs to get a whole bunch of separate chroots for the separate jails to run in 2011-05-18T17:40:39 *** jmpespxoreax has joined #aichallenge 2011-05-18T17:42:26 yeah I read that the real directories reside in /home/jailuserN or something 2011-05-18T17:43:55 each jail has a chroot at /srv/chroot/jailuserN/root/ but in order to keep from having 32 copies of the chroot OS on the disk they use unionfs to build an overlay from a base chroot at /srv/chroot/aic-base/ 2011-05-18T17:44:47 inside the chroot the submissions run as if they are at /home/jailuser 2011-05-18T17:47:47 *** smellyhippy has quit IRC (Ping timeout: 252 seconds) 2011-05-18T17:48:09 janzert: fuse: mountpoint is not empty 2011-05-18T17:48:42 That's a good joke when you use unionfs 2011-05-18T17:48:45 /srv/chroot/jailuser1/root [260] terminated with status 1, Filesystem could not be mounted: /srv/chroot/jailuser1/root 2011-05-18T17:49:12 there must be an option to force the mount 2011-05-18T17:49:23 I can skip or manually recover 2011-05-18T17:50:42 What about "-o nonempty"? 2011-05-18T17:50:49 I've got an empty proc and sys dir in the root folder 2011-05-18T17:51:22 If that did work, I don't think I want it 2011-05-18T17:52:02 *** ltriant has joined #aichallenge 2011-05-18T17:52:32 *** smellyhippy has joined #aichallenge 2011-05-18T17:52:46 hmm 2011-05-18T17:52:58 is that when you mount /srv/chroot/aic-base into /srv/chroot/jailuser1/root? 2011-05-18T17:53:10 my worker vm is rebuilding at the moment 2011-05-18T17:53:17 *** FireFly has quit IRC (Quit: swatted to death) 2011-05-18T17:53:29 hey ppl 2011-05-18T17:53:39 in the engine do you use space partitioning? 2011-05-18T17:53:49 quad-trees maybe? 2011-05-18T17:53:56 no 2011-05-18T17:54:18 how do you test if objects are nearby? 2011-05-18T17:54:25 did you maybe try to schroot before the unionfs-fuse filesystem was mounted? 2011-05-18T17:54:28 we don't really need to do much with nearest neighbor queries 2011-05-18T17:54:47 cause schroot would be mounting proc and sys in root/ 2011-05-18T17:54:50 McLeopold: can you explain this? 2011-05-18T17:55:01 in which case you can probably just delete them at this point 2011-05-18T17:55:04 janzert: if it happens before I log in, I have no idea :[ 2011-05-18T17:55:37 it shouldn't be schroot would be an explicit command by yourself or the sandbox 2011-05-18T17:55:39 mcstar: we do a full scan in the area of attack radius, the position offsets are cached 2011-05-18T17:55:56 *** amriedle has joined #aichallenge 2011-05-18T17:56:10 janzert: what about recovering schroots? 2011-05-18T17:56:56 excuse the stupid question, but what you do cannot be done with a bind mount? 2011-05-18T17:57:43 McLeopold: ? as in if the machine crashed while a chroot was ongoing 2011-05-18T17:58:02 yeah 2011-05-18T17:58:29 hmm, possibly 2011-05-18T17:59:10 McLeopold: so as far as i understand: you have a grid, NxM, the full dimension of the map is represented by a 2d array, and you bind objects to the squares 2011-05-18T17:59:11 Frontier: basically they need copy on write semantics and I don't think you can achieve that with a bind mount point 2011-05-18T17:59:16 I get a bunch of recovery/copyfiles messages on bootup 2011-05-18T18:00:17 mcstar: we use a grid/matrix and store ints in them, we also have dictionaries to look up objects by location 2011-05-18T18:01:32 so we can loop through a list of ants without doing a scan, or check to see if something is at a location, or use the matrix to scan for anything in the surrounding area 2011-05-18T18:01:35 so for example in [n,m] you have 12334 which is an index in another array 2011-05-18T18:01:55 ok 2011-05-18T18:02:24 i was considering space partitioning, but this sounds much simpler 2011-05-18T18:02:44 the problem with space partitioning is that the space changes each turn 2011-05-18T18:02:53 yes 2011-05-18T18:03:00 you'd have to do a lot of queries to make up for the overhead 2011-05-18T18:03:00 some fast update would be needed 2011-05-18T18:03:19 ok thx 2011-05-18T18:03:48 plus, space is wrapped, you have to rewrite standard algos with that in mind 2011-05-18T18:04:15 there is no such thing as a convex hull in ants :p 2011-05-18T18:05:01 ? 2011-05-18T18:05:01 or, at least there are many hulls, and you'd want the one with the smallest area, but that's a lot of work 2011-05-18T18:05:11 yes 2011-05-18T18:05:17 hehe, i wrote code that generates the silhouette of land in ants yesterday 2011-05-18T18:05:41 McLeopold: take a look in /var/lib/schroot/session/ if there is anything in there then that's probably schroot sessions leftover from a crash 2011-05-18T18:05:54 janzert: can I just nuke them? 2011-05-18T18:06:03 schroot -e should be used to get rid of them 2011-05-18T18:06:37 schroot -e -c session-name-which-will-probably-be-really-long 2011-05-18T18:07:02 what I'm trying to find is a command to list the session names 2011-05-18T18:07:33 schroot -l 2011-05-18T18:07:34 schroot -l might but I'm not sure 2011-05-18T18:07:40 aichallenge: McLeopold epsilon * r0ff97aa / (4 files): new ranking widget start (in progress) - http://bit.ly/itjjh5 2011-05-18T18:07:45 it doesn't give the long file names 2011-05-18T18:07:58 that'll list the available chroots but I don't know if it lists active sessions 2011-05-18T18:08:15 yeah, if there are no long names there it probably doesn't list the sessions 2011-05-18T18:08:43 oooh, it looks like I have 40 sessions in var, with the session names? 2011-05-18T18:08:46 well, schroot -e --all should end all the sessions 2011-05-18T18:09:01 so long as you don't have anything that should be running 2011-05-18T18:09:28 actually schroot -e --all-sessions would be better 2011-05-18T18:10:46 /etc/schroot/schroot.conf 2011-05-18T18:11:25 afk for a bit 2011-05-18T18:19:32 ok, back. what about the schroot.conf? 2011-05-18T18:20:43 the actual jail configs are in /etc/schroot/chroot.d/ btw 2011-05-18T18:24:29 I'm all clean now 2011-05-18T18:24:54 great 2011-05-18T18:40:11 *** antimatroid has quit IRC (Ping timeout: 240 seconds) 2011-05-18T18:43:07 *** Mathnerd314 has joined #aichallenge 2011-05-18T18:43:38 think I've got dmd install working, doing a final test before checking it in now 2011-05-18T19:06:43 *** needsch has quit IRC (Quit: Leaving.) 2011-05-18T19:10:30 contestbot: seen lavalamp 2011-05-18T19:10:30 janzert: lavalamp was last seen in #aichallenge 8 weeks, 1 day, 1 hour, 23 minutes, and 11 seconds ago: ok I'll ask him at some point 2011-05-18T19:16:30 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-05-18T19:17:30 aichallenge: janzert epsilon * r5267f3f / setup/worker_setup.py : Fix dmd installation - http://bit.ly/ilnULT 2011-05-18T19:18:16 Frontier: the worker will now install dmd 2.053, but the current starter bot has a deprecation error with it 2011-05-18T19:19:02 specifically "std.date and std.dateparse have been scheduled for deprecation. Please use std.datetime instead." 2011-05-18T19:21:22 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-05-18T19:23:00 my ants/dist/starter_bots/c is just an empty directory, which I believe is because its a git submodule, how do I get it? 2011-05-18T19:23:21 git submodule update 2011-05-18T19:23:30 right after 'git submodule init' 2011-05-18T19:24:05 otherwise all starter bots except for lolcode seem to work with compiler.py now 2011-05-18T19:24:16 :( poor kitties 2011-05-18T19:24:32 *** Mathnerd314 has quit IRC (Disconnected by services) 2011-05-18T19:24:35 I can't decide on the best lolcode implementation 2011-05-18T19:24:49 heh 2011-05-18T19:25:09 lci isn't capable of parsing strings, all the others are hacks, bfns, or special specs 2011-05-18T19:25:15 *** Mathnerd314_ has joined #aichallenge 2011-05-18T19:26:12 *** Mathnerd314_ is now known as Mathnerd314 2011-05-18T19:27:42 *** Sunhay_ has joined #aichallenge 2011-05-18T19:31:15 *** smellyhippy has quit IRC (Ping timeout: 250 seconds) 2011-05-18T19:32:40 *** Sunhay has quit IRC (*.net *.split) 2011-05-18T19:35:11 *** mcstar has left #aichallenge ("WeeChat 0.3.5") 2011-05-18T19:40:39 *** smellyhippy has joined #aichallenge 2011-05-18T19:40:39 *** smellyhippy has joined #aichallenge 2011-05-18T19:44:11 *** AlliedEnvy has quit IRC (Remote host closed the connection) 2011-05-18T19:48:04 *** jmpespxoreax has quit IRC (Changing host) 2011-05-18T19:48:04 *** jmpespxoreax has joined #aichallenge 2011-05-18T19:50:24 aichallenge: janzert epsilon * red773e1 / worker/compiler.py : Make ExternalCompiler compile subdirectories and remove JavaCompiler - http://bit.ly/mCK7Tx 2011-05-18T19:53:23 *** delt0r_ has joined #aichallenge 2011-05-18T19:54:44 *** delt0r___ has quit IRC (Ping timeout: 248 seconds) 2011-05-18T20:00:54 aichallenge: janzert epsilon * r71a2687 / ants/dist/starter_bots/c : Get latest C starter bot updates - http://bit.ly/ietCt9 2011-05-18T20:16:34 *** smellyhippy has quit IRC (Ping timeout: 264 seconds) 2011-05-18T20:24:40 aichallenge: McLeopold epsilon * rc14af1b / (5 files): update to ranking pages - http://bit.ly/kK4nDZ 2011-05-18T20:24:40 aichallenge: McLeopold epsilon * re4735cb / (3 files in 3 dirs): Merge branch 'epsilon' of https://github.com/aichallenge/aichallenge into epsilon - http://bit.ly/kQ1Sdx 2011-05-18T20:25:42 *** McLeopold has left #aichallenge 2011-05-18T20:27:49 *** smellyhippy has joined #aichallenge 2011-05-18T20:31:09 aichallenge: janzert epsilon * r0d79fe8 / (5 files): Merge branch 'epsilon' of github.com:aichallenge/aichallenge into epsilon - http://bit.ly/jyu9yb 2011-05-18T20:49:35 *** amriedle has quit IRC (Ping timeout: 276 seconds) 2011-05-18T20:58:53 *** AlliedEnvy has joined #aichallenge 2011-05-18T21:07:14 *** McLeopold has joined #aichallenge 2011-05-18T21:10:56 janzert: how is it that github says you merged the same thing as me? 2011-05-18T21:12:03 the merge always shows the previous commit 2011-05-18T21:12:35 my real commit is the parent of the merge 2011-05-18T21:13:45 ah 2011-05-18T21:22:32 hehe, the website package dependencies now end up pulling in firefox somehow 2011-05-18T21:22:39 * janzert shakes his head 2011-05-18T21:24:15 apparently through the icedtea-plugin used for the visualizer 2011-05-18T21:27:10 *** smellyhippy has quit IRC (Ping timeout: 240 seconds) 2011-05-18T21:30:20 *** smellyhippy has joined #aichallenge 2011-05-18T21:34:04 aichallenge: janzert epsilon * r400cbec / setup/server_setup.py : Have server_setup pull wiki content into website - http://bit.ly/j7g3oC 2011-05-18T21:36:55 *** smellyhippy has quit IRC (Ping timeout: 250 seconds) 2011-05-18T21:47:03 *** smellyhippy has joined #aichallenge 2011-05-18T21:54:05 *** smellyhippy has quit IRC (Ping timeout: 260 seconds) 2011-05-18T21:54:25 *** smellyhippy has joined #aichallenge 2011-05-18T21:54:30 *** smellyhippy has quit IRC (Changing host) 2011-05-18T21:54:30 *** smellyhippy has joined #aichallenge 2011-05-18T21:58:45 *** chris___0076 is now known as chris_0076 2011-05-18T22:00:26 *** ltriant has quit IRC (Read error: Connection reset by peer) 2011-05-18T22:00:27 *** Frontier has quit IRC (Quit: Leaving.) 2011-05-18T22:01:01 *** smellyhippy has quit IRC (Ping timeout: 252 seconds) 2011-05-18T22:01:31 *** ltriant has joined #aichallenge 2011-05-18T22:06:10 *** smellyhippy has joined #aichallenge 2011-05-18T22:06:11 *** smellyhippy has joined #aichallenge 2011-05-18T22:17:16 *** smellyhippy has quit IRC (Remote host closed the connection) 2011-05-18T22:17:31 *** smellyhippy has joined #aichallenge 2011-05-18T22:39:41 *** smellyhippy has quit IRC (Ping timeout: 240 seconds) 2011-05-18T22:39:42 *** smippy has joined #aichallenge 2011-05-18T22:46:15 *** smellyhippy has joined #aichallenge 2011-05-18T22:46:15 *** smellyhippy has joined #aichallenge 2011-05-18T22:46:16 *** smippy has quit IRC (Ping timeout: 246 seconds) 2011-05-18T22:47:22 *** qacek has joined #aichallenge 2011-05-18T22:50:43 *** smellyhippy has quit IRC (Ping timeout: 258 seconds) 2011-05-18T22:55:36 *** smellyhippy has joined #aichallenge 2011-05-18T22:55:36 *** smellyhippy has joined #aichallenge 2011-05-18T23:27:11 *** smellyhippy has quit IRC (Ping timeout: 240 seconds) 2011-05-18T23:45:17 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal)