2011-07-10T00:05:08 *** MuTaLiSk has quit IRC (Quit: Bye) 2011-07-10T00:05:15 *** amstan has quit IRC (Ping timeout: 276 seconds) 2011-07-10T00:06:21 *** computerwiz_222 has quit IRC (Ping timeout: 276 seconds) 2011-07-10T00:13:36 *** jbrechtel has quit IRC (Ping timeout: 252 seconds) 2011-07-10T00:15:51 *** amstan has joined #aichallenge 2011-07-10T00:15:51 *** ChanServ sets mode: +o amstan 2011-07-10T00:27:25 evening 2011-07-10T00:29:04 *** computerwiz_222 has joined #aichallenge 2011-07-10T00:45:32 afternoon 2011-07-10T00:52:58 *** computerwiz_222 has quit IRC (Remote host closed the connection) 2011-07-10T00:53:53 *** jmreardon has quit IRC (Quit: jmreardon) 2011-07-10T01:02:44 *** chris__0076 has quit IRC (Read error: Connection reset by peer) 2011-07-10T01:03:54 *** Chris_0076 has joined #aichallenge 2011-07-10T01:17:36 *** kato has joined #aichallenge 2011-07-10T01:29:54 *** pcavs has quit IRC (Ping timeout: 255 seconds) 2011-07-10T01:43:04 *** bhasker has quit IRC (Quit: bhasker) 2011-07-10T02:17:14 *** bhasker has joined #aichallenge 2011-07-10T02:19:54 *** VilleH has joined #aichallenge 2011-07-10T02:32:30 *** Chris_0076 has quit IRC (Remote host closed the connection) 2011-07-10T02:34:38 *** Chris_0076 has joined #aichallenge 2011-07-10T03:23:05 *** ibdknox has joined #aichallenge 2011-07-10T03:35:35 *** Blkt has joined #aichallenge 2011-07-10T03:38:37 *** bhasker has quit IRC (Quit: bhasker) 2011-07-10T03:41:09 good day everyone 2011-07-10T03:43:57 evening 2011-07-10T03:52:12 *** sigh has quit IRC (Remote host closed the connection) 2011-07-10T03:53:23 *** sigh has joined #aichallenge 2011-07-10T03:58:35 *** AlliedEnvy has quit IRC (Remote host closed the connection) 2011-07-10T03:58:51 *** amstan has quit IRC (Ping timeout: 255 seconds) 2011-07-10T03:59:32 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-07-10T04:04:20 *** AlliedEnvy has joined #aichallenge 2011-07-10T04:35:52 good morning 2011-07-10T04:37:04 *** Blkt has quit IRC (Ping timeout: 240 seconds) 2011-07-10T04:55:45 *** sigh has quit IRC (Remote host closed the connection) 2011-07-10T05:13:38 *** Blkt has joined #aichallenge 2011-07-10T05:33:51 *** boegel has joined #aichallenge 2011-07-10T05:44:25 *** mleise has joined #aichallenge 2011-07-10T05:45:50 *** mleise has quit IRC (Client Quit) 2011-07-10T05:46:29 *** boegel has quit IRC (Quit: This computer has gone to sleep) 2011-07-10T05:47:28 *** smiley1983 has joined #aichallenge 2011-07-10T05:48:23 *** mleise has joined #aichallenge 2011-07-10T06:00:35 *** ibdknox has quit IRC (Remote host closed the connection) 2011-07-10T06:21:33 *** onensora has joined #aichallenge 2011-07-10T06:26:42 *** boegel has joined #aichallenge 2011-07-10T06:41:22 *** sigh has joined #aichallenge 2011-07-10T07:06:52 *** mceier has joined #aichallenge 2011-07-10T07:08:17 *** boegel has quit IRC (Quit: This computer has gone to sleep) 2011-07-10T07:19:15 *** jbrechtel has joined #aichallenge 2011-07-10T07:22:20 *** superflit has quit IRC (Ping timeout: 240 seconds) 2011-07-10T07:27:06 *** superflit has joined #aichallenge 2011-07-10T07:27:40 *** mcstar has joined #aichallenge 2011-07-10T07:35:10 *** Blkt has quit IRC (Remote host closed the connection) 2011-07-10T07:47:40 *** superflit has quit IRC (Read error: Operation timed out) 2011-07-10T07:49:31 *** asddas has joined #aichallenge 2011-07-10T07:55:10 *** asddas has quit IRC (Quit: Page closed) 2011-07-10T08:01:47 morning 2011-07-10T08:04:32 evening 2011-07-10T08:05:28 *** sigh has quit IRC (Remote host closed the connection) 2011-07-10T08:13:35 *** sigh has joined #aichallenge 2011-07-10T08:22:12 ineedcoffee 2011-07-10T08:30:28 i havent had mine yet 2011-07-10T08:30:49 there was a power outage in the morning, so i didnt need it 2011-07-10T08:31:34 why does the lack of power reduce your need for coffee? 2011-07-10T08:33:10 i couldnt read, or use the computer, ergo why coffee? 2011-07-10T08:33:35 *** foRei has joined #aichallenge 2011-07-10T08:36:38 mcstar: makes sense :) 2011-07-10T08:36:38 coffee is not just a reading or computing aid :P 2011-07-10T09:06:03 *** foRei has quit IRC (Read error: Connection reset by peer) 2011-07-10T09:23:39 *** bhasker has joined #aichallenge 2011-07-10T09:26:20 morning 2011-07-10T09:26:59 *** jbrechtel has quit IRC (Remote host closed the connection) 2011-07-10T09:39:43 *** boegel has joined #aichallenge 2011-07-10T09:40:09 remember: "cp x y" on system files followed by "mv y x" will cause you a lot of fun the following day if you don't have a backup 2011-07-10T09:41:03 oh that reminds me of refreshing my backup 2011-07-10T09:42:09 *** boegel has quit IRC (Client Quit) 2011-07-10T09:42:40 :) 2011-07-10T09:42:49 aren't you using an svn repository? 2011-07-10T09:42:54 or some kind of src control 2011-07-10T09:43:18 no, and especially not for my system files. that's just odd :p 2011-07-10T09:43:49 ah:p 2011-07-10T09:44:46 well windows restore:) 2011-07-10T09:44:53 enable windows system restore 2011-07-10T09:45:04 thats pretty much a src control system :) 2011-07-10T09:45:12 if you are on windows i.e 2011-07-10T09:45:32 i think even ext4 on linux has a journalling file system which you can probably use to roll back stuff 2011-07-10T09:45:39 or use ZFS on freebsd 2011-07-10T09:45:49 or ... use BTRFS on linux 2011-07-10T09:45:59 i think even that supports complete versioning of files 2011-07-10T09:46:01 bhasker: Oo not so much information 2011-07-10T09:46:18 :) 2011-07-10T09:46:26 BeTteR FileSystem? 2011-07-10T09:46:28 ZFS is an awesome filesystem 2011-07-10T09:46:42 yup 2011-07-10T09:46:43 http://en.wikipedia.org/wiki/Btrfs 2011-07-10T09:47:07 mleise: i dont get your first comment 2011-07-10T09:47:19 why is it causing problems? 2011-07-10T09:47:44 i am guessing he didn't give the complete list of statements 2011-07-10T09:47:52 he probably modifies x between cp and mv 2011-07-10T09:48:00 i am guessing 2011-07-10T09:48:11 no, that was all 2011-07-10T09:48:27 so? i dont see a problem with that 2011-07-10T09:48:42 what am i missing? 2011-07-10T09:48:44 It didn't keep symlinks and maybe some permissions. i couldn't use my desktop or mysql any longer :p 2011-07-10T09:49:30 you must have been root, so permissions would be inherited, right? 2011-07-10T09:49:38 especially gconf ran amok. it reset every setting to default that was involved at that time 2011-07-10T09:50:09 mleise: is x and y a directory? 2011-07-10T09:50:15 i was root, but trust me, that I did nothing more than copy and move back 2011-07-10T09:50:22 no x and y are files 2011-07-10T09:50:26 hm 2011-07-10T09:50:32 should have used cp -a 2011-07-10T09:50:45 can it be some missing .conf files? 2011-07-10T09:51:53 I don't know and I don't care any more. At least it gave me the opportunity to look thoroughly through my logs and I found some other minor configuration flaws, like no dri2 module loaded for my nvidia driver, or no 'keychain' application installed 2011-07-10T09:52:19 isnt dri2 loaded automatically? 2011-07-10T09:52:37 mleise: sorry for being such a smart-ass 2011-07-10T09:53:21 mcstar: It would be, but people on the net spread the word that it isn't required 2011-07-10T09:54:06 im going back to popping sour cherry and watching predator* 2011-07-10T09:54:11 mcstar: And in fact now that I gave xorg the ModulePath to the lib it loaded dri2, but nvidia unloaded and reloaded it to do some initialization stuff on it 2011-07-10T09:54:55 mcstar: have fun, I watched that movie when I was young and it was midnight, my parents were at the neighbours house 2011-07-10T09:55:03 i dont mess around with xorg since it isnt a mandatory config file 2011-07-10T09:55:09 man it was the most awesome movie at that time *g* 2011-07-10T09:55:17 it is good :) 2011-07-10T09:55:18 mleise :) i am just suggesting that having a better filesystem is helpful in such matters anyway 2011-07-10T09:55:26 but i have similar experience with aliens 2011-07-10T09:55:27 like snapshots are cheap 2011-07-10T09:55:34 btr fs, in deed 2011-07-10T09:55:37 on certain filesystems 2011-07-10T09:55:55 so modifying system files you take a snapshot/modify/snapshot restore if needed 2011-07-10T09:56:31 zfs is a pretty awesome filesystem, i wish it was the default and supported across more oses 2011-07-10T09:56:51 dynamic pooling of partitions/disks/volumes is a pretty awesome feature 2011-07-10T09:56:54 zfs seems to have a strict license 2011-07-10T09:56:57 yea 2011-07-10T09:57:06 thats why linux developed btrfs 2011-07-10T09:57:11 which is basically zfs pretty much 2011-07-10T09:57:19 zfs is supported on bsd though 2011-07-10T09:57:32 and i have used it, it takes some tuning to get it do its work but it works beautifully 2011-07-10T09:57:32 bhasker: Oracle is developing it :p 2011-07-10T09:57:35 no fscks ever 2011-07-10T09:57:37 (i believe that lvm supports snapshots) 2011-07-10T09:57:41 Oracle bought Sun 2011-07-10T09:57:46 Sun developed ZFS 2011-07-10T09:57:49 yup 2011-07-10T09:57:52 see the connection? ;) 2011-07-10T09:57:52 its part of Solaris 2011-07-10T09:58:01 its the default filesystem in solaris 2011-07-10T09:58:15 beauty of zfs is its always consistent on disk 2011-07-10T09:58:27 so a crash recovery is instant 2011-07-10T09:58:38 I wonder if this time they manage to deliver a defrag tool and transparent compression 2011-07-10T09:58:48 instead of crappy ext3/ufs where with large disks you can be sitting forever for fsck to complete 2011-07-10T09:59:12 bhasker: ok, ext4 fscheck doesn't take long to be fair 2011-07-10T09:59:15 which matters in production:) not fun when your db goes down and the box is doing fsck for 20 minutes 2011-07-10T09:59:23 well ext4 is journalling file system 2011-07-10T09:59:30 so its recovery is faster 2011-07-10T09:59:47 ext3 is journaling too, i dont get the point :p 2011-07-10T09:59:52 i haven't looked at it closely but from what i understand it also tries to maintain disk state in a consistent manner 2011-07-10T10:00:03 well ext3 state is not always consistent 2011-07-10T10:00:10 some state is in kernel buffers etc at times 2011-07-10T10:00:15 not always written to disk 2011-07-10T10:00:19 so if box crashes 2011-07-10T10:00:25 file system state maybe inconsistent 2011-07-10T10:00:37 which is why the need to do an fsck after a crash 2011-07-10T10:00:52 zfs guarantees on disk data structures to be always consistent 2011-07-10T10:01:16 and adds journalling on top 2011-07-10T10:01:25 *** Palmik has joined #aichallenge 2011-07-10T10:01:28 so in case of crashes, the box comes up , run journal entries 2011-07-10T10:01:32 and you are ready to go 2011-07-10T10:01:38 pretty much 2011-07-10T10:01:55 with ext4 you have the option to use in kernel barriers to go sure the disk writes the data in the correct order, flushing internal caches if necessary 2011-07-10T10:02:27 *** sigh has quit IRC (Ping timeout: 255 seconds) 2011-07-10T10:02:50 I really don't think there is much of a difference any more, but zfs and btrfs are surely still an improvement 2011-07-10T10:03:46 hmm, can you do online backups of these file systems with the snapshot functionality? 2011-07-10T10:18:01 *** jmreardon has joined #aichallenge 2011-07-10T10:19:32 i guess you could 2011-07-10T10:19:57 *** acieroid has quit IRC (Ping timeout: 264 seconds) 2011-07-10T10:20:21 *** Kingpin13 has joined #aichallenge 2011-07-10T10:22:09 *** acieroid has joined #aichallenge 2011-07-10T10:26:50 *** jbrechtel has joined #aichallenge 2011-07-10T10:33:43 *** pcavs has joined #aichallenge 2011-07-10T10:33:53 *** Apophis_ has joined #aichallenge 2011-07-10T10:35:58 *** mleise has quit IRC (Quit: Leaving.) 2011-07-10T10:37:06 *** Apophis has quit IRC (Ping timeout: 255 seconds) 2011-07-10T10:37:12 *** kato has quit IRC (Quit: Page closed) 2011-07-10T10:39:08 *** mleise has joined #aichallenge 2011-07-10T10:50:15 *** pcavs has left #aichallenge 2011-07-10T10:52:31 *** foRei has joined #aichallenge 2011-07-10T10:56:17 *** mleise has quit IRC (Quit: Leaving.) 2011-07-10T11:01:39 *** mleise has joined #aichallenge 2011-07-10T11:08:47 *** bhasker has quit IRC (Quit: bhasker) 2011-07-10T11:11:11 *** GreenTea has joined #aichallenge 2011-07-10T11:12:39 *** mleise has quit IRC (Quit: Leaving.) 2011-07-10T11:14:07 Hi. who knows why 5 left issues on the github are not implemented? for example Setup email sending is registering on April 23.. 2.5 monthes are passed... 2011-07-10T11:25:28 i guess just nobody has gotten around to them yet 2011-07-10T11:28:35 *** mleise has joined #aichallenge 2011-07-10T11:30:44 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-07-10T11:31:07 what exactly is the "player seed"? 2011-07-10T11:32:38 *** bhasker has joined #aichallenge 2011-07-10T11:36:51 player seed is a random seed given to the player, different from the engine seed 2011-07-10T11:37:13 previously the player was handed the random seed used by the engine for food generation :) 2011-07-10T11:37:56 haha 2011-07-10T11:38:21 i don't remember seeing the player seed in the spec 2011-07-10T11:38:25 is it new? 2011-07-10T11:38:58 " i guess just nobody has gotten around to them yet" why? this is not important issues? 2011-07-10T11:39:31 GreenTea: probably different reasons depending on who you are talking about 2011-07-10T11:39:50 some people don't have the expertise, others don't have the time, etc... 2011-07-10T11:41:02 smiley1983: yeah the seed isn't even in the game spec on the github wiki :\ 2011-07-10T11:41:58 it appears that it's already in the engine and stuff (i haven't actually checked the bot input though)... 2011-07-10T11:42:02 If so, why not to write status in the issue description: "i dont have time, i don't know how to do this.." etc? 2011-07-10T11:42:15 GreenTea: are the issues assigned to specific people? 2011-07-10T11:43:27 doesn't look like it 2011-07-10T11:43:35 that's why nobody has to make excuses here 2011-07-10T11:43:53 if you want things to speed up, maybe you could work on them 2011-07-10T11:44:10 Finding a stable trueskill implementation and making it do what is required seems to be the largest stumbling block 2011-07-10T11:44:11 2 issues are assigned. 3 - not assigned 2011-07-10T11:44:18 yeah 2011-07-10T11:45:59 "Finding a stable trueskill implementation and making it do what is required seems to be the largest stumbling block" it is so importand for launch? 2011-07-10T11:46:25 current implementation is work. we can see it on beta server 2011-07-10T11:46:49 maybe with some bugs, but anyway.. 2011-07-10T11:47:07 currently it crashes sometimes, meaning some games are not ranked 2011-07-10T11:47:27 then people complain about every individual game where they did not receive their skill points 2011-07-10T11:47:43 and what? in first month or too, it is not so important.. 2011-07-10T11:48:07 it will became really important on finals only 2011-07-10T11:48:36 you are right, they could probably proceed at this point 2011-07-10T11:48:44 for whatever reason, they've decided to wait 2011-07-10T11:48:46 it's tough to judge your bot on flaky scores, but yes, i agree, it's not the end of the world 2011-07-10T11:48:57 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Read error: Connection reset by peer) 2011-07-10T11:48:58 it does get harder to focus on a new trueskill implementation during a live contest though 2011-07-10T11:51:10 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-07-10T11:56:35 Please, some from constens developers, who knows current status, please make post on forum. with explanation of current status of development. What is ready, what is not ready, what are we waiting for. what are main stoppers before launch. I think more than 4600 people waiting this information 2011-07-10T11:56:52 Not everyone have a time to sit on this irc 2011-07-10T12:02:56 *** mleise has quit IRC (Quit: Leaving.) 2011-07-10T12:06:30 *** mleise has joined #aichallenge 2011-07-10T12:06:35 *** eashoka has joined #aichallenge 2011-07-10T12:17:00 *** mleise has quit IRC (Quit: Leaving.) 2011-07-10T12:23:55 *** mleise has joined #aichallenge 2011-07-10T12:29:27 *** mleise has quit IRC (Quit: Leaving.) 2011-07-10T12:32:19 *** mleise has joined #aichallenge 2011-07-10T12:34:20 *** smiley1983 has quit IRC (Quit: leaving) 2011-07-10T12:37:10 *** mleise has quit IRC (Client Quit) 2011-07-10T12:37:41 *** Palmik has quit IRC (Ping timeout: 240 seconds) 2011-07-10T12:41:13 *** mleise has joined #aichallenge 2011-07-10T12:44:56 *** foRei has quit IRC (Read error: Connection reset by peer) 2011-07-10T12:46:00 *** foRei has joined #aichallenge 2011-07-10T13:01:26 can someone please tell me what's wrong with the most recent ants.py in github ? I get following error after updating. (and I'm no python guy :-)) 2011-07-10T13:01:43 Traceback (most recent call last): File "../tools/playgame.py", line 14, in from ants import Ants File "../tools/ants.py", line 1117 orientations = [[(ants[0].loc, 0, {i: i for i in range(self.num_players)})]] 2011-07-10T13:02:25 SyntaxError: invalid syntax 2011-07-10T13:02:55 earlier version worked fine. I just got an update 2011-07-10T13:03:45 *** gobang has joined #aichallenge 2011-07-10T13:06:05 *** Palmik has joined #aichallenge 2011-07-10T13:22:01 pfff, should i really be the one who advises people to upgrade their python's? 2011-07-10T13:22:05 its getting boring.. 2011-07-10T13:22:16 eashoka: use at least 2.7.x 2011-07-10T13:24:55 mcstar: thanks 2011-07-10T13:26:10 spotted another problem. classes in the java standard stater package and in the tools (sample boths there) are different 2011-07-10T13:27:25 ilk class for instance has very few enum types (no UNSEEN for instance) in the starter package so the meaning of isPasable() etc are very different 2011-07-10T13:28:59 which is the correct version to use ? In the case of simpler version, how would one determine whether a given tile is a unseen one or not ? anyone who has used the java package ? 2011-07-10T13:50:37 *** ArchMonkey has joined #aichallenge 2011-07-10T13:52:42 *** Accoun has quit IRC (Read error: Connection reset by peer) 2011-07-10T13:53:15 *** Accoun has joined #aichallenge 2011-07-10T13:57:29 http://aichallengebeta.hypertriangle.com/visualizer.php?game=73016&user=17 2011-07-10T13:57:46 my bot won with 2 ants against 75 2011-07-10T13:58:38 i dont like this situation at all, if the game had run for a couple of more turns, the other ants would have eaten mine 2011-07-10T14:02:58 lol, use at least python 2.7... i have to tweak all these scripts to use python2 because arch uses python 3 as the default python command 2011-07-10T14:04:43 i symlinked python2 to /usr/bin/python 2011-07-10T14:05:07 i'd do that except that other packages are used to seeing python 3 at that path 2011-07-10T14:05:25 my only python3 aware program was ranger 2011-07-10T14:05:36 but it works fine with python2 too 2011-07-10T14:05:50 no 2011-07-10T14:05:50 uzbl and i think wicd would both break if i did that 2011-07-10T14:06:00 i changed its shebang line 2011-07-10T14:06:20 who is uzbl? 2011-07-10T14:06:43 uzbl is a web browser with a unixy philosophy 2011-07-10T14:06:48 *** sir_macelon has joined #aichallenge 2011-07-10T14:06:56 do one thing very well, use scripting and stuff for everything else 2011-07-10T14:06:58 yes, but you said "uzbl and i think" 2011-07-10T14:07:17 *uzbl and, i think, wicd 2011-07-10T14:07:25 ah, okk 2011-07-10T14:07:26 :D 2011-07-10T14:07:36 eashoka: I have created the java starter bot 2011-07-10T14:08:13 my new haskell starter is almost ready, woo! 2011-07-10T14:08:13 eashoka: sample bots just use different code 2011-07-10T14:08:24 but it can't be official because i'm using unsupported things 2011-07-10T14:08:45 such as? 2011-07-10T14:08:56 it relies on being able to exec a binary on the server 2011-07-10T14:08:58 sir_macelon : so I should use the api in starter package to upload , not the ones in sample code ? 2011-07-10T14:09:06 because it uses libraries not on the server 2011-07-10T14:09:20 and it's designed so that you can use whatever compiler version you want (i want to use ghc 7) 2011-07-10T14:09:26 jmcarthur: yeah, that probably shouldnt be in a starter bot :) 2011-07-10T14:09:45 (i mean the execing binaries part) 2011-07-10T14:10:06 eashoka: yes, the sample code is not in good quality 2011-07-10T14:10:16 it's a fine starter. just not something we would want to officially support 2011-07-10T14:10:34 sir_macelon : problem is, starter package ILK class seems not complete when compared to the sample coee ILK 2011-07-10T14:10:42 also, because of this assumption, you have to build it on a 64 bit machine 2011-07-10T14:10:54 in fact, i guess it has to be 64 bit linux, specifically 2011-07-10T14:11:10 eashoka: I removed unnecessary information, so that beginners won't get confused to much 2011-07-10T14:11:48 sir_macelon : no UNSEEN type in the enum in the starter package. I just spotted after trying to figure out why my isPassable() method doesn't work. :-) 2011-07-10T14:12:17 sir_macelon : with the current starter package api, how do u determine a tile is a unseen one or not ? 2011-07-10T14:13:02 *** amstan has joined #aichallenge 2011-07-10T14:13:02 *** ChanServ sets mode: +o amstan 2011-07-10T14:13:10 eashoka: anyway starter package didn't offer a functionality to detect unseen tiles 2011-07-10T14:13:33 eashoka: you have to take care by your own, it's a taks for a programmer 2011-07-10T14:14:12 eashoka: but basically you just check if the tile is in view radius of any of your ants 2011-07-10T14:15:25 eashoka: so this UNSEEN was just there without any purpose or background implementation, you can check it in the sample bots, they don't use it at all 2011-07-10T14:16:20 eashoka: yes, it is compared in the isPassable() method, but it is not set anywhere in the code 2011-07-10T14:16:24 sir_macelon : ah ok. thanks. so when the server gives you an ants object, it's 2d map for tiles can have unseen tiles with different status (and a wrong one) marked on ? 2011-07-10T14:16:46 that reminds me... this view radius check... is a square visible if *any* part of it is in range, if the *center* is in range, or the *entire* square is in range? 2011-07-10T14:16:56 same for food and attack ranges 2011-07-10T14:17:48 eashoka: no, server does not send you any objects, just strings, you create the objects in your code 2011-07-10T14:17:56 actually maybe that's in the spec and i just missed it 2011-07-10T14:18:14 but if not and somebody clarifies for me then i'll go ahead and add it 2011-07-10T14:18:51 nevermind 2011-07-10T14:18:53 the spec is quite clear 2011-07-10T14:19:03 sorry for the noise 2011-07-10T14:20:51 sir_macelon : I meant this : "public void do_turn(Ants ants) {" ok, in the string the server sends, how would it mark the unseen tiles ? I mean the server is supposed to give you info only for the tiles you can see right ? so what would be the default enum for a unseen tile ? LAND ? 2011-07-10T14:21:15 jmcarthur: if you visualize it, its best to put a squares center into the origo 2011-07-10T14:21:27 eashoka: the spec for starters is that unseen squares are assumed to be land until they have been seen 2011-07-10T14:21:31 so the integer coordinates are centers as well 2011-07-10T14:21:40 eashoka: i disagree with the wisdom of that choice, but there it is 2011-07-10T14:21:44 but of course it doesnt matter 2011-07-10T14:22:20 *** Accoun has quit IRC (Read error: Connection reset by peer) 2011-07-10T14:22:28 eashoka: https://github.com/aichallenge/aichallenge/wiki/Ants-Starter-Pack-Guide 2011-07-10T14:22:54 *** Accoun has joined #aichallenge 2011-07-10T14:23:22 eashoka: yes, the default tile is LAND 2011-07-10T14:23:45 jmcarthur and sir_macelon : thanks for the info both of you! I was really trying to find the 'defect' in my code why the hell it's not working as expected. (isPassable()) now I know why my bot is doing so poorly :-) 2011-07-10T14:24:07 mcstar: it's just that i once wrote a rogue-like rendered, so i'm painfully aware of a million ways to compute vision 2011-07-10T14:24:11 *renderer 2011-07-10T14:25:16 i suppose this is easier since we only care about range and not about walls blocking your vision 2011-07-10T14:25:22 *** Kingpin13 has quit IRC (Ping timeout: 252 seconds) 2011-07-10T14:25:41 yes, and since its finally cooling down i can think 2011-07-10T14:25:45 *** amriedle has joined #aichallenge 2011-07-10T14:25:47 sir_macelon : is there any plan to implement playGame.py, game.py, engine.py etc in java as was in PW ? 2011-07-10T14:26:09 eashoka: i think noway is the answer 2011-07-10T14:26:29 eashoka: i think we are trying to move away from a java infrastructure 2011-07-10T14:29:00 mcstar, jmcarthur : ok 2011-07-10T14:29:45 eashoka: you would have to do it by your own and share with us :) 2011-07-10T14:30:13 sir_macelon : will try :-) 2011-07-10T14:30:14 currently pythoners are favoured only 2011-07-10T14:30:44 jython anyone? 2011-07-10T14:30:50 then you can link the code and stuff 2011-07-10T14:31:06 if you wanna make a java gui for example 2011-07-10T14:31:15 eashoka: your bot is not that bad at all, made it into top 20 2011-07-10T14:31:18 sir_macelon : I almost like python, except for the white space thingy..... 2011-07-10T14:32:21 amstan: but it's rather a matter of changing the code, so that it won't be killing the bots after timeout while debugging, than just running it in java 2011-07-10T14:32:28 sir_macelon : it once came to 10th as well, then dropped sudednly after i added that cool feature. Tx anyway, but it's no where near ur one. I'm full of owe on how slyly you hunt ither ants relentlessly 2011-07-10T14:33:21 eashoka: I haven't tried python, I prefer to stick to my main language 2011-07-10T14:33:48 *** gobang has quit IRC (Ping timeout: 255 seconds) 2011-07-10T14:34:03 eashoka: thanks, but I still see room for improvement :) 2011-07-10T14:34:25 sir_macelon : same here. as for the bots, I put some extra code specifially avoid getting eaten by ur hunter ants, but for no awail. manged to run away for a while though but they eventually getting me. very cool bot indeed! 2011-07-10T14:35:19 better would be to go on the offensive 2011-07-10T14:35:30 so long as you can figure out a nice battle strategy 2011-07-10T14:35:45 i suspect that a little game tree search would do nicely for that 2011-07-10T14:36:03 sir_macelon : ur bot has made the contest really interesting and very challenging, taking it to a different levle where others are forced to react. 2011-07-10T14:36:17 eashoka: thanks, the offensive module is most advanced in my bot, still I struggle with exploration and food collection subtelities 2011-07-10T14:36:28 jmcarthur : yeah I was thinking that too. no use running for sir_macelon all the time :-) 2011-07-10T14:36:40 being offensive gives you map control :) 2011-07-10T14:38:23 jmcarthur : but in a game of say 10 players, be offensive and getting killed while others are peacefully collecting food is bit risky, this is the part i'm trying to balance out :-) 2011-07-10T14:39:04 may be for a initial little period, one need to gather and then go hunting. 2011-07-10T14:39:35 i haven't explored the game deeply enough yet, but i suspect you might be wrong 2011-07-10T14:39:49 of course that doesn't apply to sir_macelon. his ants seem more safe when they are attacking! they hardly get killed 2011-07-10T14:39:59 if you can keep enemies away from a patch of land then you can monopolize the food in that area 2011-07-10T14:40:10 exactly 2011-07-10T14:40:21 you have to actually be able to do you offense well, not just throw ants at the problem 2011-07-10T14:40:30 the mythical ant month? :P 2011-07-10T14:40:45 *your offense 2011-07-10T14:41:34 eashoka: if you attack 2v1 you will be safe :) 2011-07-10T14:45:22 sir_macelone : yeah.. I'm getting a little offensive module going, my one was designed more like a gatherer from the beginning, so bit hard with that design to get individual ants to grouyp together etc. that's the part u r doing beautifully. it's a treat to watch how ur 2-3 pairs going after prey :-) 2011-07-10T14:46:32 is anybody trying to guess things they can't see, yet? 2011-07-10T14:47:19 i guess it's to early for that as map symmetries are not settled yet 2011-07-10T14:48:58 i also mean things like enemy ants and such 2011-07-10T14:49:40 like if you already have seen the area, know where water is, and know where some enemy ants were previously, where might they be now 2011-07-10T14:50:05 i'm tracking food out of sight, i'm planning to do enemy ant info but haven't done that yet 2011-07-10T14:50:24 i'm undecided on whether i want to go all out with markov chains or just expand out where an enemy ant could possibly be 2011-07-10T14:50:32 tracking as in guessing or tracking as in "i saw it here before, maybe it will still be here when i get back"? 2011-07-10T14:51:34 say a location was viewable last turn as an enemy ant and i can't match that up to a visible enemy now, i'd mark that and adjacent non-visible squares as possibly containing an ant, then each turn i'd push that information outwards in the non-visible parts of my map 2011-07-10T14:51:49 you could use actual probabilities with markov chains if you thought you could make use of the information 2011-07-10T14:52:41 you could alternatively track how long ago the information of an ant that could reach the square was seen 2011-07-10T14:53:57 yeah i've been thinking about this stuff too 2011-07-10T14:54:07 probabilities are tougher than they sound 2011-07-10T14:54:20 nah, markov chains :) 2011-07-10T14:54:21 because you also have to consider this isn't just a random ant. it might be smart 2011-07-10T14:54:29 oh yeah, i'd assume just randomness 2011-07-10T14:54:42 then i'd say your probabilities are misleading at best 2011-07-10T14:54:48 yeah true 2011-07-10T14:55:20 i will probably just go with something like tracking both possible ant information and the most recently seen ant that could have made it there 2011-07-10T14:55:23 the propogation method is the easiest one 2011-07-10T14:55:33 i'm also keen to play around with "avoiding enemies" while path finding 2011-07-10T14:55:35 well, i think, at least 2011-07-10T14:55:46 i'm more interested in seeking them out ;) 2011-07-10T14:56:09 i want to do both, navigate around them in some instances, hunt them down in others :) 2011-07-10T14:57:36 avoiding enemies is clearly disadvantageous when you are gathering food. you can die together with a competing enemy for the resource so that ur second ant may get to the food. wheras if u try to avoid the enemy, he will just walk away with the food. 2011-07-10T14:58:50 eashoke: you've just revealed a bot feature publicly :) 2011-07-10T14:59:07 eashoka: you've just revealed a bot feature publicly :) 2011-07-10T14:59:38 *** choas has joined #aichallenge 2011-07-10T14:59:50 my bot seems to be happier when it encounters bots who try to avoid. (except for sir_macelon, he seems to avoid 1-1 clashes at all time only because he want to come at you with two!) 2011-07-10T15:00:03 eashoka: i think there can be instances where the path you choose will determine whether you can collect a piece of food or whether an enemy could potentially engage you in a battle before reaching the food 2011-07-10T15:00:14 and there's a few other things i'd play around with 2011-07-10T15:00:35 sir_macelon : just voiced an observation when seeing games. :-) 2011-07-10T15:01:40 eashoka: some people would spent hours before discovering that, you've just helped them freely :) 2011-07-10T15:02:02 sir_macelon: what's wrong with sharing strategy ideas? :P 2011-07-10T15:05:13 antimatroid: you're spoiling the possibility to discover them for other contestants :P 2011-07-10T15:06:56 sir_macelone : actually there is a good point there. you get so much satisfaction when you figure (ir thought u figured) something out by urself. taking that fun away from ppl would be very cruel to them :-) 2011-07-10T15:11:24 antimatroid : IMHO, I think markov chains might be an overkill. I would rather spend time on building formations trying to survive bit longer from sir_macelon when I meet him in next version :-) 2011-07-10T15:12:00 honestly, i hope the games will get more intricate and interesting than theyre now 2011-07-10T15:14:39 mcstar: they should be, I am still awaiting competition :) 2011-07-10T15:14:49 BTW, the server is very slow these days, it take ages for you to come back up in the ranking for a new version. to make it worse, there are lot of good bots at skill level 70s. not easy getting through. there was a time I could get into first 20 in about 10 games, now it takes about 30. think that means competition is getting heated up. 2011-07-10T15:15:03 sir_macelon: after my snake is ready, you'll get my full attention :) 2011-07-10T15:15:53