2011-04-01T00:05:56 *** davidjliu has quit IRC (Quit: Page closed) 2011-04-01T00:08:16 *** phire has joined #aichallenge 2011-04-01T00:11:53 @seen Stocha 2011-04-01T00:11:53 phire: Stocha was last seen in #aichallenge 8 hours, 38 minutes, and 39 seconds ago: @later tell phire you have to hit the run button at the bottom of the window 2011-04-01T00:28:25 *** dlila has quit IRC (Quit: Leaving) 2011-04-01T00:47:14 *** Mathnerd314 has joined #aichallenge 2011-04-01T01:31:43 *** McLeopold has joined #aichallenge 2011-04-01T02:01:18 McLeopold: how are symmetric maps going? 2011-04-01T02:01:36 ok 2011-04-01T02:01:50 I've got the framework for carving symmetrically 2011-04-01T02:02:07 I just need the finish the function to create symmetric points 2011-04-01T02:02:34 are you using the same type of symmetry that antimatroid had? 2011-04-01T02:02:44 i'm pretty sure yes 2011-04-01T02:02:56 McLeopold: are you doing the check to make sure players never start in the same row or column? 2011-04-01T02:03:38 No, I'm not doing that type of check, but it might come out like that sometimes. 2011-04-01T02:03:55 you will break symmetric spawning as it currently is if you do that 2011-04-01T02:03:57 then food gen will need ot be updated 2011-04-01T02:04:07 well, don't do that :P 2011-04-01T02:04:43 if 2 starts are in the same row or column, food spawning breaks? 2011-04-01T02:05:05 possibly, because it may get a row and col translate that doesn't take n iterations to loop 2011-04-01T02:06:00 we could make row and col translate parameters at the top of the map 2011-04-01T02:06:03 then it wouldn't matter 2011-04-01T02:06:28 but it's visually less appealing if players start like that, because then the lattices are more obvious 2011-04-01T02:06:38 meh, we'll see about food gen after McLeopold finishes his maps 2011-04-01T02:06:56 yeah, just noting why it isn't really important :P 2011-04-01T02:07:04 McLeopold: can I ask you something about replays and the engine? 2011-04-01T02:07:26 yeah, also, I just replied to an email from Frontier 2011-04-01T02:07:50 That format looks really good, nice work. 2011-04-01T02:07:50 I'm thinking maybe the mapname could be a map url? so mapurl: "tcp.zeroviz.us/ants/maps/random4.map". That means the replay could be passed around and we still know where the map file resource is. Do we still need an "a" as the first element of the array if it is in an ants array? 2011-04-01T02:07:52 .. i have time for a food bonus chat if you two do? 2011-04-01T02:07:58 ^^ that was my reply 2011-04-01T02:08:06 I want the engine/playgame to be able to output replays on stdout so it can be piped into a visualiser 2011-04-01T02:08:08 I need to go, actually 2011-04-01T02:08:13 ah ok :( 2011-04-01T02:08:14 all good 2011-04-01T02:08:21 sigh: that would be the streaming format? 2011-04-01T02:08:29 either format 2011-04-01T02:08:44 main point is that current output should be to stderr 2011-04-01T02:08:52 and replay/streaming should be on stdout 2011-04-01T02:08:57 ok 2011-04-01T02:09:03 cool, will do that 2011-04-01T02:09:08 verbose puts everything to stderr then 2011-04-01T02:09:12 yup 2011-04-01T02:09:24 sigh: did you implement the food bonus yet? 2011-04-01T02:09:36 McLeopold: not yet 2011-04-01T02:09:53 could you hack that up real quick, people are starting to talk about scoring now 2011-04-01T02:10:05 ok, when I get home 2011-04-01T02:10:11 turns_left*spawn rate 2011-04-01T02:10:17 I don't really understand the justification for it that well though 2011-04-01T02:10:18 Doesn't need to be perfect 2011-04-01T02:10:27 it seems quite arbitrary 2011-04-01T02:10:29 janzert can help explain 2011-04-01T02:10:33 ok :) 2011-04-01T02:10:37 But, it is to encourage short game 2011-04-01T02:10:38 s 2011-04-01T02:10:48 yeah, I get that 2011-04-01T02:10:52 anyway, I'll let you go 2011-04-01T02:10:56 and to help make sure the last bot standing wins 2011-04-01T02:10:59 I'll harass antimatroid and janzert 2011-04-01T02:11:02 ok 2011-04-01T02:11:13 oh, and check out ants.zeroviz.us 2011-04-01T02:11:18 i saw :P 2011-04-01T02:11:26 I know django now 2011-04-01T02:11:29 :) 2011-04-01T02:11:46 McLeopold: i think the current scoring already kind of encourages a good game, but we an talk about it later 2011-04-01T02:11:55 i think we should at least simplify the food bonus 2011-04-01T02:12:00 but we can talk about that later 2011-04-01T02:12:44 sigh: it stops the farming 2011-04-01T02:12:58 if there are two players left, they can sort of continue the game and increase each others relative placing 2011-04-01T02:13:04 the food bonus makes you both fight for that? 2011-04-01T02:13:36 it might be necessary for that 2011-04-01T02:14:00 *** McLeopold has left #aichallenge 2011-04-01T02:16:42 *** Mathnerd314 has quit IRC (Read error: Connection reset by peer) 2011-04-01T02:21:04 one thing odd about the food bonus is that it depends on the length of the game 2011-04-01T02:21:54 so games that are 500 moves long are very different to games that are 2000 moves long 2011-04-01T02:21:54 *** ltriant has quit IRC (Quit: Get MacIrssi - http://www.sysctl.co.uk/projects/macirssi/) 2011-04-01T02:26:03 ^^ antimatroid 2011-04-01T02:28:06 anyway, I'm out... be back later 2011-04-01T02:28:08 *** sigh has left #aichallenge ("Leaving") 2011-04-01T02:49:55 *** Eruonen has joined #aichallenge 2011-04-01T03:01:08 *** amstan has quit IRC (Remote host closed the connection) 2011-04-01T03:24:49 *** boegel has joined #aichallenge 2011-04-01T03:25:06 *** Frontier has joined #aichallenge 2011-04-01T03:28:04 *** sigh has joined #aichallenge 2011-04-01T03:39:14 *** antimatroid has quit IRC (Ping timeout: 240 seconds) 2011-04-01T03:50:21 *** aerique has joined #aichallenge 2011-04-01T03:55:28 *** antimatroid has joined #aichallenge 2011-04-01T03:57:17 *** delt0r___ has joined #aichallenge 2011-04-01T03:58:51 *** FireFly has joined #aichallenge 2011-04-01T03:59:11 *** delt0r_ has quit IRC (Ping timeout: 276 seconds) 2011-04-01T04:06:33 I'm living in the old town center of Hannover where the houses are all connected. Since a few days there are strange noises coming from the next house. 2011-04-01T04:08:01 Today I heard that the lower-saxonian minister of econimics rented that flat, but he doesn't seem to come here often. 2011-04-01T04:08:47 It sounds like some dying machinery, rolling and scratching. 2011-04-01T04:09:04 *** Stocha has joined #aichallenge 2011-04-01T04:09:10 I'd really like to know what that is 2011-04-01T04:14:04 where is the hard evidences that counting ants at end of the game is not fair/ doesn't make for good games ? 2011-04-01T04:14:24 and that ant killed is better ? 2011-04-01T04:27:15 @seen phire 2011-04-01T04:27:15 Stocha: phire was last seen in #aichallenge 4 hours, 15 minutes, and 22 seconds ago: @seen Stocha 2011-04-01T04:27:37 lol 2011-04-01T04:27:49 @later tell McLeopold I've gotten this error a few times: http://pastebin.com/kVSQ5ch2 - suspect that changing line 38 to "if e:" will be enough but I don't know anything about sandbox.py so I'm not touching it 2011-04-01T04:27:49 sigh: Aye, aye, sir 2011-04-01T04:30:35 @seen phire 2011-04-01T04:30:35 sigh: phire was last seen in #aichallenge 2 minutes and 57 seconds ago: lol 2011-04-01T04:34:56 how you made some progres phire ? 2011-04-01T04:35:14 me, no 2011-04-01T04:35:20 I've been reading reddit all day 2011-04-01T04:35:31 stupid website 2011-04-01T04:37:42 mm, i think it is time you take your headphone and spend a few hours coding without doing anything else :p 2011-04-01T04:37:54 i don't use headphone. So that is optional. 2011-04-01T04:38:05 i guess. 2011-04-01T04:38:34 *** Naktibalda has joined #aichallenge 2011-04-01T04:38:39 sounds like a plan 2011-04-01T04:41:14 it's had enough to get anywhere when you do spend time going there :p 2011-04-01T04:41:16 hard 2011-04-01T04:49:11 *** Frontier has quit IRC (Read error: Connection reset by peer) 2011-04-01T04:50:35 *** Frontier has joined #aichallenge 2011-04-01T05:06:13 *** skypent has joined #aichallenge 2011-04-01T05:27:11 ok: headphones, check; all tabs and unreleated apps closed, check 2011-04-01T05:27:20 irc client closed 2011-04-01T05:27:24 *** phire has quit IRC (Quit: check) 2011-04-01T05:59:57 aichallenge: aerique epsilon * r9903208 / (6 files in 3 dirs): 2011-04-01T05:59:58 aichallenge: Some changes to docs, comments and layout. 2011-04-01T05:59:58 aichallenge: Remove HOST2STR. - http://bit.ly/fiKa6y 2011-04-01T06:11:40 *** Stocha has quit IRC (Quit: Page closed) 2011-04-01T06:24:19 *** Kingpin13 has quit IRC (Quit: You're a kitty!) 2011-04-01T06:25:06 *** p4p4p5 has joined #aichallenge 2011-04-01T06:26:22 @later tell McLeopold actually that's masking an even weirder problem, still having a look into it :S 2011-04-01T06:26:22 sigh: OK 2011-04-01T06:35:43 @later tell McLeopold ah got it, bad things happen if a child tries to shutdown after the main process (because .kill() is asynchronous). Have to to .wait() for children to actually die. 2011-04-01T06:35:43 sigh: As you wish. 2011-04-01T06:49:58 we should do an ai challenge for qwop :P 2011-04-01T06:50:41 aichallenge: sigh replay * r7784330 / worker/engine.py : Single point of control for logs in engine - http://bit.ly/hwcT7n 2011-04-01T06:50:43 aichallenge: sigh replay * rbd88a98 / worker/engine.py : stdout logs for stream and replay - http://bit.ly/f0kDnq 2011-04-01T06:50:44 aichallenge: sigh replay * r8e74475 / (ants/playgame.py worker/engine.py): Rename --output to --stdout to make it more obvious - http://bit.ly/gvHlGT 2011-04-01T06:53:00 aichallenge: sigh epsilon * r6f6264b / ants/ants.py : Add food bonus - http://bit.ly/fBXoE6 2011-04-01T06:53:39 sigh: :) 2011-04-01T06:56:31 *** Frontier has quit IRC (Read error: Connection reset by peer) 2011-04-01T07:21:00 *** mceier has joined #aichallenge 2011-04-01T07:44:52 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-01T07:50:02 *** p4p4p5 has joined #aichallenge 2011-04-01T08:09:10 *** Stocha has joined #aichallenge 2011-04-01T08:11:42 *** skypent has quit IRC (Quit: Fucking Bananas!) 2011-04-01T08:14:44 *** needsch has joined #aichallenge 2011-04-01T08:53:39 *** Naktibalda has quit IRC (Remote host closed the connection) 2011-04-01T08:58:45 *** sigh has quit IRC (Remote host closed the connection) 2011-04-01T09:09:22 *** Frontier has joined #aichallenge 2011-04-01T09:13:03 *** Stocha has joined #aichallenge 2011-04-01T09:38:15 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-01T09:39:58 *** p4p4p5 has joined #aichallenge 2011-04-01T09:50:57 *** Naktibalda has joined #aichallenge 2011-04-01T10:17:57 *** phire has joined #aichallenge 2011-04-01T10:18:50 Stocha, finally got some code working 2011-04-01T10:36:19 *** boegel has quit IRC (Quit: *poof!*) 2011-04-01T10:56:29 *** aerique has quit IRC (Quit: ...) 2011-04-01T10:56:45 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-01T10:58:04 *** p4p4p5 has joined #aichallenge 2011-04-01T11:09:05 phire : good :) 2011-04-01T11:09:19 *** Eruonen has quit IRC (Read error: Connection reset by peer) 2011-04-01T11:09:39 you'll explain me, or better show me in pastbin (the game data log) 2011-04-01T11:09:53 *** Stocha has quit IRC (Quit: Page closed) 2011-04-01T11:12:36 @tell Stocha, it basically expands until it doesn't see any food, then it kind of freezes, because collecting food is all it knows. 2011-04-01T11:12:36 phire: Error: 'Stocha,' is not a valid nick. 2011-04-01T11:12:50 @tell Stocha it basically expands until it doesn't see any food, then it kind of freezes, because collecting food is all it knows. 2011-04-01T11:12:50 phire: Error: I haven't seen Stocha, I'll let you do the telling. 2011-04-01T11:13:16 screw you contestbot 2011-04-01T11:22:46 @later tell Stocha it basically expands until it doesn't see any food, then all the ants just sit around and do nothing, because collecting food is all it knows. 2011-04-01T11:22:46 phire: I come to serve. 2011-04-01T11:26:45 lol@ I'll let you do the telling 2011-04-01T11:37:09 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-01T11:38:02 *** p4p4p5 has joined #aichallenge 2011-04-01T11:40:44 *** espes has quit IRC (Ping timeout: 260 seconds) 2011-04-01T11:41:37 *** espes has joined #aichallenge 2011-04-01T11:57:49 *** delt0r_ has joined #aichallenge 2011-04-01T11:59:36 *** delt0r___ has quit IRC (Ping timeout: 276 seconds) 2011-04-01T11:59:39 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-01T12:02:00 *** p4p4p5 has joined #aichallenge 2011-04-01T12:11:58 *** antimatroid has quit IRC (Ping timeout: 250 seconds) 2011-04-01T12:29:09 *** antimatroid has joined #aichallenge 2011-04-01T12:38:57 *** antimatroid has quit IRC (Ping timeout: 246 seconds) 2011-04-01T12:39:37 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-01T12:39:52 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 3.6.16/20110319135224]) 2011-04-01T12:40:43 *** espes has quit IRC (Ping timeout: 248 seconds) 2011-04-01T12:41:57 *** p4p4p5 has joined #aichallenge 2011-04-01T12:55:32 *** bhasker has joined #aichallenge 2011-04-01T13:02:38 *** antimatroid has joined #aichallenge 2011-04-01T13:03:07 *** McLeopold has joined #aichallenge 2011-04-01T13:08:05 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-01T13:10:05 *** p4p4p5 has joined #aichallenge 2011-04-01T13:21:54 *** amstan has joined #aichallenge 2011-04-01T13:21:54 *** ChanServ sets mode: +o amstan 2011-04-01T13:31:03 *** Naktibalda has joined #aichallenge 2011-04-01T13:44:08 *** Eruonen has joined #aichallenge 2011-04-01T13:51:08 *** fratern has joined #aichallenge 2011-04-01T13:52:22 *** fratern has left #aichallenge 2011-04-01T13:58:15 *** McLeopold has quit IRC (Quit: Leaving.) 2011-04-01T14:10:30 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-01T14:11:52 *** p4p4p5 has joined #aichallenge 2011-04-01T14:33:04 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-01T14:34:00 *** p4p4p5 has joined #aichallenge 2011-04-01T14:47:25 *** Cyndre has quit IRC (Read error: Connection reset by peer) 2011-04-01T14:47:39 *** Cyndre has joined #aichallenge 2011-04-01T14:59:39 *** dlila has joined #aichallenge 2011-04-01T15:03:46 how to make 0.stream ? 2011-04-01T15:10:04 contestbot: topic add The Ants Competition has started! Entries due in 2 weeks! Good luck! 2011-04-01T15:10:04 *** contestbot changes topic to "Official Google AI Challenge: http://ai-contest.com/ || Channel Logs: http://contestbot.hypertriangle.com/ || Code Repo: http://github.com/aichallenge/aichallenge || Launch Checklist: https://github.com/aichallenge/aichallenge/issues/labels/launch || Ants is on Reddit: http://goo.gl/NSpYJ || The Ants Competition has started! Entries due in 2 weeks! Good luck!" 2011-04-01T15:10:46 contestbot: topic remove 5 2011-04-01T15:10:46 *** contestbot changes topic to "Official Google AI Challenge: http://ai-contest.com/ || Channel Logs: http://contestbot.hypertriangle.com/ || Code Repo: http://github.com/aichallenge/aichallenge || Launch Checklist: https://github.com/aichallenge/aichallenge/issues/labels/launch || The Ants Competition has started! Entries due in 2 weeks! Good luck!" 2011-04-01T15:10:56 contestbot: topic remove 4 2011-04-01T15:10:56 *** contestbot changes topic to "Official Google AI Challenge: http://ai-contest.com/ || Channel Logs: http://contestbot.hypertriangle.com/ || Code Repo: http://github.com/aichallenge/aichallenge || The Ants Competition has started! Entries due in 2 weeks! Good luck!" 2011-04-01T15:16:21 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-01T15:20:57 April 1st? 2011-04-01T15:21:45 Naktibalda: what's that? 2011-04-01T15:24:03 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 3.6.16/20110323143040]) 2011-04-01T15:26:12 *** Accoun has quit IRC () 2011-04-01T15:30:36 *** p4p4p5 has joined #aichallenge 2011-04-01T15:38:10 *** Accoun has joined #aichallenge 2011-04-01T15:57:47 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-01T15:59:57 *** p4p4p5 has joined #aichallenge 2011-04-01T16:00:19 *** Accoun has quit IRC (Ping timeout: 246 seconds) 2011-04-01T16:00:41 *** Accoun has joined #aichallenge 2011-04-01T16:02:49 *** Cyndre has quit IRC (Read error: Connection reset by peer) 2011-04-01T16:03:04 *** Cyndre has joined #aichallenge 2011-04-01T16:19:59 *** Stocha has joined #aichallenge 2011-04-01T16:20:25 *** Kingpin13 has joined #aichallenge 2011-04-01T16:20:39 @later tell phire yes, i did that too. But there is a lot of way to do it :p 2011-04-01T16:20:39 Stocha: OK 2011-04-01T16:21:03 *** fgump has quit IRC (Quit: Leaving) 2011-04-01T16:22:48 what do you mean the ant competition has started ?? 2011-04-01T16:23:13 how has it started, how is the rules settled ? Where do one send entries ? 2011-04-01T16:24:09 Stocha, here is the game log if your intrested: http://pastie.org/pastes/1744958/text 2011-04-01T16:24:22 and i think its an april fools joke 2011-04-01T16:38:42 *** Stocha has quit IRC (Ping timeout: 252 seconds) 2011-04-01T16:39:52 *** Stocha has joined #aichallenge 2011-04-01T16:40:06 ah ! aprils 1 st. yeah. Depending where you live i guess :p 2011-04-01T16:40:18 what day is it over there in australia ? 2011-04-01T16:40:34 way into the 2nd 2011-04-01T16:41:05 how long did it take for this first version bot ? 2011-04-01T16:42:31 7 hours, but a lot of groundwork was layed out as well 2011-04-01T16:42:59 the map doesn't really strike as a path finding challenge :p 2011-04-01T16:43:11 7 hours; wow. 2011-04-01T16:43:31 i think i took me like 2 days, to make about the same. 2011-04-01T16:43:53 not full time 2011-04-01T16:44:14 but i also made the engine, for my simulation framework. 2011-04-01T16:44:31 yeah, I trimmed it down to actual work done 2011-04-01T16:44:37 but the map isn't very informative here. 2011-04-01T16:44:50 what does the groundwork involves ? 2011-04-01T16:45:09 *** p4p4p5 has quit IRC (Read error: Connection timed out) 2011-04-01T16:45:10 what's next also ? 2011-04-01T16:45:27 i think i went this way : Pure random bot, and some other random stuff 2011-04-01T16:45:39 then i wen : all closest ant to food, try to go to that food 2011-04-01T16:45:55 at some point i added some collision control, i don't remember when 2011-04-01T16:46:12 then i went for the either a bit of exploration, or more advance food 2011-04-01T16:46:31 yeah, collision control is next 2011-04-01T16:46:35 the more advanced food being : one food is assigned only one ant. That made one hell of a difference in food speed 2011-04-01T16:46:46 my first collision control was buggy i think 2011-04-01T16:47:26 i may have one that works now. And it is easilly plugable. Although everything is probably quite slow. I went more for expressivness rather than speed. 2011-04-01T16:47:34 i do very simples things. 2011-04-01T16:48:07 I think I might be over designing mine 2011-04-01T16:48:11 but we will see 2011-04-01T16:48:15 *** JamesMG has quit IRC (Ping timeout: 276 seconds) 2011-04-01T16:48:18 I spent some times to build my tools for evaluating the bots performances also. I always do a lot of that. 2011-04-01T16:48:36 *** JamesMG has joined #aichallenge 2011-04-01T16:48:41 you mean you have a lot of plans already ? 2011-04-01T16:48:50 My simple bots are nowhere sufficient 2011-04-01T16:48:58 i have some simple stuff to add though. 2011-04-01T16:49:18 then there is the matter of fighting 2011-04-01T16:49:31 i tryed to do some gate control. It was total fiasco 2011-04-01T16:49:32 oh yeah that 2011-04-01T16:49:39 in fact i wasn't even able to spot the gates :p 2011-04-01T16:50:17 i also tried sotftly to make the ants goes in nice formation (roman army like) That wasn't very sucessfull 2011-04-01T16:50:44 it was funny to imagine, i just didn't get it right. Maybe it is not as easy as going for food. And probably not as efficient for performances :p 2011-04-01T16:51:15 Although 3x3 ants formations moving together might do damage to simple minded bots with the Damage(x=1) rule. 2011-04-01T16:51:22 I'm thinking that I might split a small section of the board off for battle simulations 2011-04-01T16:51:42 maybe a 10x10 area 2011-04-01T16:51:47 you mean you'll build a special map ? 2011-04-01T16:51:56 with 5-10 ants on each side 2011-04-01T16:52:09 with your prefered ascii tool 2011-04-01T16:52:14 and then run detailed simulations over the next 10 turns 2011-04-01T16:52:46 what does that mean ? 2011-04-01T16:53:04 I'm not explaining this well 2011-04-01T16:53:18 i have designed a tool for build fighting maps for testing. Like yours i guess. I'm just not in the mood for implementing it. 2011-04-01T16:53:28 i'm french by the way. 2011-04-01T16:53:44 do you mean you'll move ants by hand ? 2011-04-01T16:53:48 or will they go at random ? 2011-04-01T16:53:49 my plan is, when my AI sees a group of my ants about to engage with enemy ants 2011-04-01T16:53:52 will they go straight ? 2011-04-01T16:54:15 I will create a sub simulation, with just those ants 2011-04-01T16:54:27 that would be a montecarlo solver ? 2011-04-01T16:54:28 and attempt to solve the battle like a chess game 2011-04-01T16:54:53 i have made some rough testing about a montecarlo food gatherer. Just to see. 2011-04-01T16:55:00 something that would be impossible to do over the whole map 2011-04-01T16:55:01 it was rough indeed. I wasn't confident. 2011-04-01T16:55:10 especially because the computing power is so scarce. 2011-04-01T16:55:13 *** p4p4p5 has joined #aichallenge 2011-04-01T16:55:22 will you do alpha beta ? 2011-04-01T16:55:33 or montecarlo trees ? 2011-04-01T16:55:56 no idea, haven't gotten that far yet 2011-04-01T16:55:59 alpha beta , or tree i should say. 2011-04-01T16:56:06 I will need to do some research 2011-04-01T16:56:21 you only now you want to make a solver for close fighting :p 2011-04-01T16:56:29 with limiting the number of ants considered. 2011-04-01T16:56:51 i'm wondering of the simple 3x3 formation is not enougth 2011-04-01T16:56:57 or raw double lined formation 2011-04-01T16:57:09 (well that would be simple formations) 2011-04-01T16:57:14 I don't think formations are going to really work 2011-04-01T16:57:20 why not ? 2011-04-01T16:57:51 at this point, there is great chances that formation will be my first try at combat anyway. So we might see how it does. 2011-04-01T16:57:51 the problem is that the battles are over in a single turn 2011-04-01T16:58:01 ? 2011-04-01T16:58:25 basically there is two stages in the game. 2011-04-01T16:58:53 in the begining you might have complicated combats. With arbitration between food exploration and gathering for fight. 2011-04-01T16:59:13 then in the second part of the game (depending on the map, you might only get one ) 2011-04-01T16:59:25 you send waves of ants at the ennemy 2011-04-01T16:59:32 so the combat tend to go for up until the end. 2011-04-01T16:59:51 (well you probably seen that on the ant simulator i showed) 2011-04-01T17:00:21 yeah, I think you are right 2011-04-01T17:00:34 Or do you mean the formations will break after the first turn ? 2011-04-01T17:00:58 think about the 3x3 block 2011-04-01T17:01:06 Anyway, studying simple cases on 10x10 maps, is probably the way to start 2011-04-01T17:01:20 its strong if you attack from the 4 sides 2011-04-01T17:01:23 the other way being to put up some random map, and see how it goes in there. 2011-04-01T17:01:59 but if you are clever, you can pick away the corners 2011-04-01T17:01:59 3x3 block is strong and flexible. But costly 2011-04-01T17:02:15 can you ? 2011-04-01T17:02:24 with one ant you mean ? 2011-04-01T17:02:31 or with 4 ? 2011-04-01T17:02:40 one for each corner 2011-04-01T17:02:44 ah yeah with one ant then. 2011-04-01T17:02:55 yeah the corners are weak there. 2011-04-01T17:03:15 but that would mean you are being attacked diagonally 2011-04-01T17:03:29 wich seems a weird way to get attacked 2011-04-01T17:03:29 diagonal attacks are alowed 2011-04-01T17:03:37 because diagonal movement is not efficient 2011-04-01T17:03:49 they are allowed, but not easy to set up. 2011-04-01T17:04:04 the radius of attack is pretty short also. 2011-04-01T17:04:12 I think I can manage 2011-04-01T17:04:23 it becomes all about setup 2011-04-01T17:04:39 trying to move all your ants into range at the right turn 2011-04-01T17:04:42 I think i will try very simple formations first anyway. 2011-04-01T17:05:15 but at the moment, i'm not to deeply involved in combat 2011-04-01T17:05:18 if they work, it will be a nice simple solution 2011-04-01T17:05:25 alghouth i guess the combat rule is pretty settled 2011-04-01T17:05:41 my plan is to build bot incrementally 2011-04-01T17:05:51 combat will be a lot easier when there are other bots to test it against 2011-04-01T17:06:14 so the question is : will it work agains my previous generations of bots. Those that doesn't care about formations and fighting. Just running mindlessly at the ennemy 2011-04-01T17:06:26 if it does, it is sufficient argument for me to try hard at it :p 2011-04-01T17:06:55 of course the process might be badly flawed, but it is simple to follow. And probably more fun that to try to think hard :p 2011-04-01T17:07:25 i've put a lot of thinking into food gathering. And only got out with a headache. no concrete solutions. No implementation. 2011-04-01T17:07:39 well i guess my next bot will try to improve the food gathering thing a bit. 2011-04-01T17:08:07 i'm trying to get some expressive framework to help me in that goal. It might help for other tasks too. If it work. 2011-04-01T17:09:09 Once there are other bots available for comparison, everything will be more fun. I suppose :p 2011-04-01T17:09:18 we'll need maps although. A lot of them. 2011-04-01T17:09:38 Knowing what kind of map will make it live would be good. Especially the size. 2011-04-01T17:09:43 of different types 2011-04-01T17:09:59 if map are more than 128 x128 all my bot framework might break out of acceptable speed :p 2011-04-01T17:10:22 i'll have to do it all back again from scratch. 2011-04-01T17:10:25 I had a few wraping problems 2011-04-01T17:10:40 yeah ? 2011-04-01T17:10:43 what kind ? 2011-04-01T17:11:28 I expected the mod operator work like: -1 % 100 == 99 2011-04-01T17:11:33 For combat the turn are simultaneous. Unlike chess. 2011-04-01T17:11:39 so you can't do the same sort of planning. 2011-04-01T17:11:47 mixed strategy are probable. 2011-04-01T17:12:00 it works that way. 2011-04-01T17:12:17 you do it (x+100)%100 2011-04-01T17:12:18 :p 2011-04-01T17:12:27 i think i had the same issue though 2011-04-01T17:12:31 I worked that out 2011-04-01T17:12:35 i don't remember really. 2011-04-01T17:12:49 i think i had that exact -1%100 stuff. 2011-04-01T17:13:15 I also had some brutal spelling errors 2011-04-01T17:13:21 that took me hours to debug 2011-04-01T17:13:23 but i'm not sure there is any (x+100)%100 in my code anymore. Maybe deep into the framework there is. I'm not even sure. 2011-04-01T17:13:27 hehe 2011-04-01T17:13:33 it took minutes :p 2011-04-01T17:13:38 java rocks :p 2011-04-01T17:13:53 do you have mean of producing debug output ? 2011-04-01T17:14:01 they would work 90% of the time 2011-04-01T17:14:15 also to be fair, i made my own server. So i could control everything. 2011-04-01T17:14:35 I modified the framework to trigger core dumps on a timeout 2011-04-01T17:14:44 loool 2011-04-01T17:14:48 that's nice debug output :p 2011-04-01T17:14:59 so one core dump for a, two for b ? 2011-04-01T17:15:24 just one dump at a time 2011-04-01T17:15:27 i had both err/std and some graphical output capabilities for debuging. 2011-04-01T17:15:46 yeah multiple dump or not that easy to output :p 2011-04-01T17:15:47 yeah, I think I'm going to need some graphical debugging at some point 2011-04-01T17:16:00 you might at least go for some file ouput 2011-04-01T17:16:19 i was on the point to do that, when i tryed to plug my bot into the real framework 2011-04-01T17:16:24 trying to work out what happened from just a core dump can be a little hard 2011-04-01T17:16:41 i had a silly bug : i didn't call the main function. It took me like two hours. Because i had no mean of outputing anything. 2011-04-01T17:16:58 lol 2011-04-01T17:17:25 after 2 hours of rereading the code, i finally picked up the fault. So i didn't do file output. 2011-04-01T17:17:42 but i guess for c addict, it is just easy and efficient. With a kind of ascii format 2011-04-01T17:17:56 of course i find ascii really hard to read. 2011-04-01T17:18:14 the problem I've been having is that printing debugging infomation can effect the timing 2011-04-01T17:18:55 i didn't found anyway to output anything in the real framework. i'm not very comfortable with unix or anything. 2011-04-01T17:19:11 but file output can be pretty fast. 2011-04-01T17:19:18 certainly faster than stderr 2011-04-01T17:19:19 system.err.println 2011-04-01T17:19:26