2010-12-11T00:00:00 ok, they are lawyers. they cannot know open source isn't a language *rofl* 2010-12-11T00:01:00 hard to say if it is the lawyers, or just a very ignorant journalist 2010-12-11T00:01:38 *** Appleman1234 has joined #aichallenge 2010-12-11T00:04:47 hmm, seems to be the journalist. COMPUTER LANGUAGE is also a bit vague. How about programming language or computer jargon? At least we learnt that the licensing issues are murky. 2010-12-11T00:05:28 but only if the correct language was used!: Open Source. What's the lates compiler version anyway? 2010-12-11T00:05:39 GPL2? 2010-12-11T00:05:59 more importantly, how does it compare against Lisp for AI? 2010-12-11T00:07:25 *** Rubicon-|-Cross has joined #aichallenge 2010-12-11T00:07:54 Do I see that right? He had two editors reviewing the text and his name is McCool? 2010-12-11T00:09:25 that's one of my biggest problems with the legal system of today 2010-12-11T00:09:36 most of the time the lawyers and judges have nfi what they're talking about 2010-12-11T00:09:44 couldn't we have like a technical court? 2010-12-11T00:09:52 yeah, I agree 2010-12-11T00:10:47 i don't even pay attention to the law, i just try not to impose negatively on other peoples lives 2010-12-11T00:11:17 that can easily cost you a fee now and then 2010-12-11T00:11:29 more than happy to pay it :) 2010-12-11T00:12:08 i got a $120 fine for an open container of alcohol in public once, at the time speeding fines were $80 if you were just over the limit 2010-12-11T00:12:45 that was absurd! if one crime is comparably worse by any measure, the punishment should be worse, and no one can argue speeding is less dangerous to other people than an open container of alcohol in public 2010-12-11T00:12:46 Something is strange about you. 2010-12-11T00:13:05 i'm very very stubborn 2010-12-11T00:13:13 much more stubborn than your average person :) 2010-12-11T00:13:20 No, you are happy to pay fines. 2010-12-11T00:13:54 you could even say that he is fine with it 2010-12-11T00:15:04 When it is 1:58 and someone says: "I have two to two on my clock, what's on your's?" I can answer "Two to two, too." 2010-12-11T00:16:10 o_O 2010-12-11T00:16:14 i am not happy to pay them, but i value not following stupid rules more than the cost of paying them when i get hit 2010-12-11T00:16:29 I was recently supposed to pay 40€ for forgetting to pinch my train ticked. A ticket that is valid for two hours, and that I bought just before entering the train as could be checked with a look on the ticket. 2010-12-11T00:17:15 A ticket that cost 3.60. 2010-12-11T00:17:33 tell them they're dreaming 2010-12-11T00:17:55 I did. They wanted to check my id. 2010-12-11T00:18:34 is it like.. you can buy it any time.. and it's not valid until 'pinched', aka. marked as used? 2010-12-11T00:19:32 Yes burny, but it is not legit to accuse someone of a crime he could be about to commit on a later ride with the same ticket. common! 2010-12-11T00:20:00 If I had no ticket and hid on the toilet I would have payed 3.60 less. 2010-12-11T00:20:39 i was using trams in melbourne one summer without buying tickets (poor teenage skateboarder, what can I say) and the inspector got on and said right i'm checking tickets 2010-12-11T00:20:50 we got up, asked the driver to open the doors and got off, can't believe that worked aha 2010-12-11T00:21:00 I'm sure a number of tarts, just carry a ticket with them until they get cought, then make up a story, and get a new ticket 2010-12-11T00:21:26 burny: But then their ticked dates back a few days. 2010-12-11T00:21:32 thus the id.. to make sure the same person doesn't do it more than once a year or some such 2010-12-11T00:21:46 ah, ya true.. 2010-12-11T00:21:55 that is more benefit of the doubty 2010-12-11T00:22:04 The id was for that and to send me a letter with the fine. 2010-12-11T00:22:19 are you allowed to buy a ticket and use it later? 2010-12-11T00:22:28 liek a parent buying their kids tickets for school days 2010-12-11T00:23:02 Yes you are. Some tickets must be pinched other not. It has to do with the distance you travel. 2010-12-11T00:23:35 we don't have passenger trains in tasmania 2010-12-11T00:23:56 and i stopped getting busses when i got my license, although i've mostly been riding a bike the last few months 2010-12-11T00:25:27 To finish the story, I wrote the railroad company a fax telling them how unjust their regulation is and they reduced the fine to 10€. :D 2010-12-11T00:26:06 they should have just waived it, but better than original 2010-12-11T00:26:53 The 30€ I saved were only lost a few weeks later when some strange guy told me he is my neighbour and he quickly needs 30€ because the police got his car and he doesn't have the money at hand to buy it back, and he doesn't want his wife to notice... 2010-12-11T00:29:13 lol 2010-12-11T00:29:18 i've never had that happen to me 2010-12-11T00:29:20 Later I thought a lot of why humans are so helpful and what that has to do with our success in creating civilizations. I guess in all the thousands of years quite a few people lost '30€' on the way. 2010-12-11T00:29:41 someone broke into my house when i was here last year in december 2010-12-11T00:29:57 i called the cops as they were breaking in, they showed up like an hour later, but i chased them off with a (field) hockey stick in hand 2010-12-11T00:30:06 stupid bogans aren't getting my possessions, not while i'm here 2010-12-11T00:30:42 (there is very little risk of someone having a gun here in australia) 2010-12-11T00:31:19 try this: http://www.flickr.com/photos/utherpendragonsphotos/4958860354/in/pool-keep-out 2010-12-11T00:32:24 hmm, the neighbourhood i live in isn't too bad normally, so don't really worry 2010-12-11T00:32:59 *** Azrathud has quit IRC (Ping timeout: 272 seconds) 2010-12-11T00:35:23 *** bhasker has quit IRC (Quit: bhasker) 2010-12-11T00:36:21 We had someone break into our house when I was a child and we were all on holiday. They must have spied out the houses for someone who departs. The funny thing is, he broke into the toilet window, but for some strange reason this time my mother shut off the doors inside the house as well and he didn't manage to open that toilet door. 2010-12-11T00:37:09 Roumors have it there was police in the area that night for some other reason, so he thought someone saw him. 2010-12-11T00:37:43 *** delt0r_ has joined #aichallenge 2010-12-11T00:39:25 *** delt0r___ has quit IRC (Ping timeout: 276 seconds) 2010-12-11T00:39:46 omg, it seems to come from my country: http://failblog.org/ 2010-12-11T00:47:43 that's golden 2010-12-11T00:49:28 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-12-11T00:50:40 check this out: http://failblog.org/2010/12/09/epic-fail-photos-babysitter-fail-2/ - i think my girlfriend is dead, but seriously I have to get to work now, I'm late! 2010-12-11T00:54:22 *** virdo has joined #aichallenge 2010-12-11T01:05:38 *** Azrathud has joined #aichallenge 2010-12-11T01:08:08 *** Azrathud has quit IRC (Client Quit) 2010-12-11T01:25:03 *** Rubicon-|-Cross has joined #aichallenge 2010-12-11T01:37:05 *** epona has joined #aichallenge 2010-12-11T01:47:24 *** Mekanik has joined #aichallenge 2010-12-11T01:48:02 *** Azrathud has joined #aichallenge 2010-12-11T01:55:00 *** Azrathud has quit IRC (Ping timeout: 260 seconds) 2010-12-11T02:01:36 *** Azrathud has joined #aichallenge 2010-12-11T02:15:48 *** narnach has quit IRC (Quit: Leaving.) 2010-12-11T02:23:10 *** Meatkat has quit IRC (Ping timeout: 240 seconds) 2010-12-11T02:30:03 *** sigh has quit IRC (Remote host closed the connection) 2010-12-11T02:37:57 *** narnach has joined #aichallenge 2010-12-11T02:39:43 *** virdo has quit IRC (Ping timeout: 264 seconds) 2010-12-11T02:47:44 anyone about want to discuss how to handle moves with ant idea? 2010-12-11T02:48:21 delt0r: you especially 2010-12-11T02:51:51 *** Azrathud has quit IRC (Ping timeout: 250 seconds) 2010-12-11T03:03:18 *** Azrathud has joined #aichallenge 2010-12-11T03:04:53 nobody? :C 2010-12-11T03:11:51 *** jaspervdj has joined #aichallenge 2010-12-11T03:12:27 oh oh oh, we could do a number of wrappings of the board 2010-12-11T03:12:55 for a torus, mobius strip or cylindrical 2010-12-11T03:19:28 hey 2010-12-11T03:20:20 hey 2010-12-11T03:20:25 have you read the ant idea? 2010-12-11T03:20:33 i'm trying to work out how movement should work 2010-12-11T03:21:07 *** amstan has quit IRC (Ping timeout: 245 seconds) 2010-12-11T03:21:59 nope 2010-12-11T03:22:01 links? 2010-12-11T03:23:22 *** sigh has joined #aichallenge 2010-12-11T03:23:41 http://www.ai-contest.com/forum/viewtopic.php?f=21&t=1205&start=0 2010-12-11T03:23:50 sigh: down for a conversation on movement with ant game? 2010-12-11T03:24:31 what's the latest topic of discussion? 2010-12-11T03:24:41 there are 3 cases I consider problematic, i'm considering both options of friends being allowed in the same square or not 2010-12-11T03:25:27 it's simpler if they are, I think, because you don't need collision detection then 2010-12-11T03:25:30 what are the cons? 2010-12-11T03:25:34 case 1: enemy bots have an empty square between them and both move into it 2010-12-11T03:25:35 case 2: enemy bots are adjacent and try to move into each others square 2010-12-11T03:25:35 case 3: 2 friend bots and 2 enemy bots surround and empty square, all 4 bots try to move into it 2010-12-11T03:26:13 and more cases come up, but i think resolving those will cover most issues 2010-12-11T03:26:37 it's also easier to allow friend bots to be on the same square for creating new bots 2010-12-11T03:26:46 otherwise it's complicated to work out where new bots appear 2010-12-11T03:27:14 why not just isolate fights to a block? 2010-12-11T03:27:28 what do you mean? 2010-12-11T03:27:31 and what about case 2? 2010-12-11T03:27:32 *** Palmik has joined #aichallenge 2010-12-11T03:27:44 do they "fight" moving through each other, or just pass each other 2010-12-11T03:27:45 let them move 2010-12-11T03:27:52 just pass 2010-12-11T03:27:57 i think fight as they pass 2010-12-11T03:28:03 if one dies, then they're gone 2010-12-11T03:28:03 sure, that's possible too 2010-12-11T03:28:27 have you determined hwo fights are resolvged? 2010-12-11T03:28:35 *how fights are resolved? 2010-12-11T03:28:38 another issue is how to give the game state to them, if it's a grid of characters, how do you indicate a bot of player 1 with x hp 2010-12-11T03:29:07 i think points should be taken from the enemy ant with the most hp first 2010-12-11T03:29:13 then i think it works fine 2010-12-11T03:29:24 least hp first 2010-12-11T03:29:41 wouldn't the most dominant ant take on the battle? 2010-12-11T03:29:46 and protect the weak 2010-12-11T03:30:05 if you could micro then you would attack the weakest first 2010-12-11T03:30:19 but yes, if you were defending you would protect the weak 2010-12-11T03:30:26 yeah okay, i'm making a forum post, i'll leave that question open to people 2010-12-11T03:30:28 so really depends from who's side you look 2010-12-11T03:30:38 any idea on representing the board? 2010-12-11T03:31:08 map item => coord 2010-12-11T03:31:11 with field of vision, do you think you should see the radius of current ant positions, or that and any previously viewable square? with the second option the current radius can be much smaller 2010-12-11T03:32:14 I think you should be able to see map features that you have seen before, but not enemies outside fov 2010-12-11T03:32:16 *** narnach has quit IRC (Quit: Leaving.) 2010-12-11T03:32:35 so just a really big list for input and users fill up a blank board? with an example line 2010-12-11T03:32:36 A 1 4 0 0 (ant, player 1, 4 hp, x, y)? 2010-12-11T03:33:05 yeah, that's reasonable 2010-12-11T03:33:13 wait, you want to see resources you could previously, but not an enemy, but you get info from them taking resources? 2010-12-11T03:33:30 hmmm 2010-12-11T03:33:59 i quite like having full past visibility and a radius of view of like 1 2010-12-11T03:33:59 meh, I guess just fov then and force people to remember what they need 2010-12-11T03:34:05 so you have to scout out the map 2010-12-11T03:34:25 full past visibility isn't as "realistic" 2010-12-11T03:34:35 where does the visibility come from? 2010-12-11T03:34:36 tis true 2010-12-11T03:34:49 where does our telepathic communication come from? 2010-12-11T03:34:53 maybe they leave cameras :P 2010-12-11T03:35:13 sure, but then allow me to destroy enemy camera :) 2010-12-11T03:35:24 that's a valid way to go 2010-12-11T03:35:29 that's actually kind of a cool idea 2010-12-11T03:35:40 call it a marker, say 2010-12-11T03:35:48 only one person can have viewing rights of a square at once, except in current view of ants 2010-12-11T03:36:03 but possibly complicating things again too much 2010-12-11T03:36:28 i think just current field of vision 2010-12-11T03:36:39 what is your opinion on walls? 2010-12-11T03:36:47 should they be there at the start and should they be makeable? 2010-12-11T03:37:04 makeable introduces a whole new level of complexity 2010-12-11T03:37:41 I would leave makeable walls alone and only consider them if you have everything else figured out 2010-12-11T03:37:56 what about having them at the start? 2010-12-11T03:38:01 I think the combat needs improvement 2010-12-11T03:38:22 sure, you need some map features 2010-12-11T03:38:25 epona: that front page is a bad representation of the current form we're talking about 2010-12-11T03:38:34 map features already exist with resources 2010-12-11T03:38:35 I would want at least some path finding would be required 2010-12-11T03:38:48 okay, walls are in 2010-12-11T03:38:52 map features that are obstacles 2010-12-11T03:39:09 i know i probably am alone in wanting this one.. i want multiple edge wrappings 2010-12-11T03:39:24 multiple? 2010-12-11T03:39:26 we can do torus, cylindrical and mobius strip quite easily 2010-12-11T03:39:45 and klien bottle :P 2010-12-11T03:39:49 it complicates things, i just want it for awesomeness 2010-12-11T03:40:03 *klein 2010-12-11T03:40:08 i'm joking (only half) i know it's not reasonable 2010-12-11T03:40:18 whats edge wrapping? 2010-12-11T03:40:21 i think torus wrapping? 2010-12-11T03:40:35 epona: you can go out of the edges and arrive on another edge it's "wrapped" to 2010-12-11T03:40:53 if you wrap top to bottom (should be bum) and left to right, then you get a torus 2010-12-11T03:40:57 kind of 2010-12-11T03:41:07 oh ok 2010-12-11T03:41:17 personally, I don't like edge wrappings 2010-12-11T03:41:32 it makes the map smaller, and more uniform 2010-12-11T03:41:38 i think it's cool, it gives you a pesudo infinite board 2010-12-11T03:41:53 in that you can just keep walking up 2010-12-11T03:41:55 I think its a great way to make it more dynamic 2010-12-11T03:41:57 *** Mathnerd314 has quit IRC (Ping timeout: 245 seconds) 2010-12-11T03:42:02 the earth is infinite! 2010-12-11T03:42:21 you can in theory go in one direction and not ever have to stop 2010-12-11T03:42:23 that's what i mean 2010-12-11T03:42:48 assuming a flat fully spherical earth :P 2010-12-11T03:42:48 well, my opinion is that it will make the game less interesting 2010-12-11T03:42:58 but it seems everyone disagrees, so meh 2010-12-11T03:43:10 it's only a small pool of people in here 2010-12-11T03:43:16 how big do you want grids? 2010-12-11T03:43:22 i was thinking just have all 32x32? 2010-12-11T03:43:36 that seems small if you want wrapping 2010-12-11T03:43:46 64x64? 2010-12-11T03:44:07 i don't know, size will have to be tweaked once you have impl 2010-12-11T03:44:31 depends on game length too, turn duration and various other factors 2010-12-11T03:44:35 true, still something that needs working out at some point 2010-12-11T03:44:42 i think we need longer than 200 turns for this 2010-12-11T03:45:02 i would consider reducing time limits to 0.5 seconds? 2010-12-11T03:45:38 if we are getting that to ourselves and only one bot per core at a time, it's what we got now in the worst case 2010-12-11T03:46:20 possilbly 2010-12-11T03:46:29 anyway, those sort of params are easily configured later 2010-12-11T03:46:30 i don't like it, i think more interesting ideas will be used with more time 2010-12-11T03:46:33 but i'm thinking practically 2010-12-11T03:47:02 if you want more time, then engineer a game with less moves 2010-12-11T03:47:51 you can reduce the size of the board to reduce game time 2010-12-11T03:49:02 sure 2010-12-11T03:52:13 I have an idea 2010-12-11T03:52:32 what if the next challenge was a 2v2 2010-12-11T03:52:41 and teamwork was beneficial 2010-12-11T03:52:57 then its even more complicated 2010-12-11T03:53:35 ranking bots gets so much harder 2010-12-11T03:53:40 makes the game harder to balance and harder to rank... 2010-12-11T03:53:41 yeah 2010-12-11T03:53:53 it would work for this ants idea i think 2010-12-11T03:53:59 well two people would be going in it to the end 2010-12-11T03:54:02 either way, I think you need a tried and tested game to pull of 2vw 2010-12-11T03:54:05 i'd be down for it, but people wont shut up about rankings 2010-12-11T03:54:10 or mix it up 2010-12-11T03:54:17 Idt its about rankings 2010-12-11T03:54:28 this is the "ai challenge" 2010-12-11T03:54:47 i think 2 player symmetric games are fine 2010-12-11T03:55:26 bah, i've changed my mind 2010-12-11T03:55:35 why so? 2010-12-11T03:55:37 it would be a lot more interesting if bots couldn't communicate what they can see at all 2010-12-11T03:56:03 ohhhhh 2010-12-11T03:56:15 so their on their own? 2010-12-11T03:56:27 yes, but they can try to coordinate if they can see each other 2010-12-11T03:56:33 but without communication 2010-12-11T03:56:49 or actually, then add back in the changing colour idea 2010-12-11T03:56:57 okay, i'm proposing that, that's much cooler 2010-12-11T03:57:17 agreed 2010-12-11T03:57:39 now how to run lots of ants of each player? 2010-12-11T03:58:42 who wanted a better battle resolution idea? i have one 2010-12-11T03:59:44 hmm 2010-12-11T04:00:10 players can shoot bullets, only one of their bullets can exist in their field of view at a time, if it hits a wall it disappears, if it hits a player, it removes 1hp, if it goes out of the field of vision, the bullet dies or keeps going but they can fire again 2010-12-11T04:00:34 was never a fan of projectiles 2010-12-11T04:00:45 but then you can't be in the same square with them, and it makes resolution harder again 2010-12-11T04:01:15 i think fights should just be performed inside a square if you land there and maybe if you both pass through each other 2010-12-11T04:02:21 what if I keep running away like a coward? after all the map is endless :D 2010-12-11T04:02:45 surround them :P 2010-12-11T04:02:52 and walls will be randomly added to maps at the start 2010-12-11T04:03:08 we can adjust the number of walls added 2010-12-11T04:03:48 Ok, but as long as I don't run into a dead end, I will always be as fast as any other ant. except you can move diagonally 2010-12-11T04:04:42 i don't think diagonal moves are being considered 2010-12-11T04:04:53 I guess I would be running like hell until I find spice and then just sit on it and wait. 2010-12-11T04:04:54 i'm not sure about ants being separate process 2010-12-11T04:05:23 not only is it messy to set that up, the engine needs to work out information for each individual ant and send it 2010-12-11T04:05:36 it sounds like a golden idea to me 2010-12-11T04:05:49 it does doesn't it 2010-12-11T04:05:57 i don't think it should be too hard for people to write their bots either 2010-12-11T04:06:00 antimatroid: Isn't the fov just a moving bit-map? 2010-12-11T04:06:22 depends which metric you use 2010-12-11T04:06:48 *** Appleman1234 has quit IRC (Ping timeout: 276 seconds) 2010-12-11T04:06:59 are you having a square view? or squares you can access in r moves? and can you "see" through walls? 2010-12-11T04:07:27 ! implement this "cannot see through walls thing" 2010-12-11T04:07:41 I already thought out some ambush tactics for that case 2010-12-11T04:07:47 it's doable 2010-12-11T04:07:55 paint the field of view 2010-12-11T04:08:00 rather than just get it 2010-12-11T04:12:27 If ants could share squares, wouldn't it be hard to visualize and result in stacks running around to deal more damage? 2010-12-11T04:13:31 one option is to force ant fusion 2010-12-11T04:13:36 I'm just thinking if the map is symmetric, then how would I gain an advantage over the enemy in battles. 2010-12-11T04:13:39 but then we're back at the problem of where to new ants spring up 2010-12-11T04:14:20 If ants had a direction they could spawn new ants in front of them 2010-12-11T04:14:40 or just decide where they should spawn 2010-12-11T04:14:40 what if that square isn't free? 2010-12-11T04:15:02 the decision rule is messy 2010-12-11T04:15:05 then they didn't use their server information right 2010-12-11T04:15:29 what if an enemy bot moved there when you went to make the kid? 2010-12-11T04:15:55 you would have one happen after the other 2010-12-11T04:16:07 *** smellyhippy has quit IRC (Ping timeout: 245 seconds) 2010-12-11T04:16:20 i'd say child creation then child creates enemy that comes in 2010-12-11T04:17:03 for example move first, then spawn. that way, when you see an enemy two squares ahead you reconsider placing a new ant there 2010-12-11T04:17:40 and the enemy doesn't run into an ant out of nowhere 2010-12-11T04:18:40 antimatroid1: I don't understand what you said. child creates enemy?! 2010-12-11T04:18:57 child fights enemy** sorry 2010-12-11T04:19:24 you can do that too, but i think it is frustrating for the attacker 2010-12-11T04:20:27 sigh: if we do a wrapped grid, ants don't even need to be given specific coordinates to the main map, only information in their field of view 2010-12-11T04:20:35 so they don't even need to know how big the maps are 2010-12-11T04:20:38 i like that 2010-12-11T04:21:01 i guess you can do that with no wrapped, but they can feel out the map better 2010-12-11T04:21:33 that's like in RoboCom :-) 2010-12-11T04:21:56 You are getting closer to it. Now add the interpreted game language 2010-12-11T04:22:28 co-ordinates issues are not much different bw wrapped and not wrapped... wrapped almost forces you to give ant-centric info 2010-12-11T04:23:38 a friend just had a cool idea, emulate pheramone, lets ants leave a colour on a square that lasts x turns 2010-12-11T04:25:26 i think that's complicating too much again 2010-12-11T04:25:31 cool idea anyway 2010-12-11T04:25:46 that might be a bit restricting if they have to communicate that way 2010-12-11T04:26:07 the more open ended it is, the better, imo 2010-12-11T04:26:12 if you do "strings" is that viewable by both bots? what about people watching the games? 2010-12-11T04:26:36 both players* not bots 2010-12-11T04:27:05 people watching the game? 2010-12-11T04:27:53 ie. on the website, you watch your games 2010-12-11T04:28:01 do you get to see the communicating strings everyone else uses 2010-12-11T04:28:05 giving away their strategies 2010-12-11T04:28:14 it's cooler if you have to devise your own code with colours 2010-12-11T04:28:37 i'm thinking the same with bots setting their viewable message to a bot in viwing distance 2010-12-11T04:28:38 *** mega1 has joined #aichallenge 2010-12-11T04:32:50 leaving color on suqares sounds like a good option to me. you could try to pick up the enemy's color code and delude it. 2010-12-11T04:33:19 *** mega1 has quit IRC (Client Quit) 2010-12-11T04:34:15 [r] <- enemies color for "yay! spice", [W] <- wall, [A][A][A][A] <- my ant army, [ ] <- no spice at all here 2010-12-11T04:35:37 strings can be used to exchange the whole game state in a base-64 format :D 2010-12-11T04:40:42 * antimatroid1 is wondering if we can have simultaneous game battles when more than 2 bots land in the same square, run as addition stages to a turn? 2010-12-11T04:41:21 like, each player picks who to spend thir hp on, but i don't think it works 2010-12-11T04:41:30 each ant* 2010-12-11T04:42:04 you could even pass hp to friend ants in the same square that way 2010-12-11T04:42:46 and allow "run away" like in pokemon 2010-12-11T04:43:11 actually, i have a whole new idea for battling, sorry, consider this.. 2010-12-11T04:43:26 *** Azrathud has quit IRC (Ping timeout: 240 seconds) 2010-12-11T04:44:19 if you are in the same square as another ant, you can choose to throw hp at it, if the ant is still there next turn, then it loses that many hp (or gains it if it's a friend) and if hp <=0 then an ant is dead 2010-12-11T04:44:39 otherwise you just wasted hp 2010-12-11T04:44:56 nah, that sucks :) 2010-12-11T04:45:42 #ai :Cannot send to channel 2010-12-11T04:45:45 what does that mean 2010-12-11T04:48:04 *** Meatkat has joined #aichallenge 2010-12-11T04:48:10 may require you to be "identified" 2010-12-11T04:48:15 whatever that is officially on irc 2010-12-11T04:49:14 ah ok 2010-12-11T04:50:55 *** tapwater has quit IRC (Quit: tapwater) 2010-12-11T04:53:42 is winner chosen off most ants or total hp? 2010-12-11T04:54:16 well, thoughts? 2010-12-11T04:55:58 thinkin 2010-12-11T04:56:34 maybe the winning peramteters should be just as dynamic as the game 2010-12-11T04:56:59 as opposed to just a total of one thing, it could be an average 2010-12-11T04:57:21 idk if thered be a way to make a bonus for creativity 2010-12-11T04:57:44 i'm working on a large forum post atm for how i think it could be done and work 2010-12-11T04:57:49 people cna suggest variations from there 2010-12-11T04:58:19 well total hp sums up to "have i harvested more than my opponent?", right? 2010-12-11T04:59:46 but is that what is trying to be measured? 2010-12-11T05:00:07 *** sigh has quit IRC (Remote host closed the connection) 2010-12-11T05:00:39 i hope not ;) 2010-12-11T05:02:51 okay, so total number of ants? 2010-12-11T05:02:57 as that also requires harvesting 2010-12-11T05:03:03 my idea for resources is as follows: 2010-12-11T05:04:05 they get symmetrically added at the start with varying levels of hp, battle happens before harvest, harvest is automatic and goes to the bot with the least hp, if a bots hp goes above H (even), then you get two bots there with H/2 and hp-H/2 hp 2010-12-11T05:04:17 maybe, improvement of the bots throughout the game 2010-12-11T05:04:43 game(s) 2010-12-11T05:05:03 that seems to be more AI related 2010-12-11T05:05:19 actually, i kind of like delt0rs idea for harvesting 2010-12-11T05:05:38 if you don't move on a square with resources, then you harvest 1 hp after battle 2010-12-11T05:05:45 or before battle even? 2010-12-11T05:06:35 actually, here's an idea, you can see what the friends you can see can see, and so on? 2010-12-11T05:08:07 *** Azrathud has joined #aichallenge 2010-12-11T05:11:37 that idea is great, i think it's important that ants don't know "where" they are in the map, so you could have chains right next to each other, and not even know 2010-12-11T05:17:48 *** sigh has joined #aichallenge 2010-12-11T05:18:15 *** sigh is now known as Guest27623 2010-12-11T05:19:13 *** Guest27623 is now known as sigh 2010-12-11T05:20:30 my laptop power died :( 2010-12-11T05:23:03 that sucks, i'm almost ready to post an idea that i think could be implemented 2010-12-11T05:23:03 did you catch me talking about improving the grading system 2010-12-11T05:23:08 it's open to cristicism though 2010-12-11T05:24:28 cristicism? 2010-12-11T05:25:10 nah, and now I'm busy setting up my desktop which I usually only ever ssh into :P 2010-12-11T05:26:04 hmmm, here's a problem.. 2010-12-11T05:26:33 what kind of criticism do you have of it? 2010-12-11T05:26:43 if ants are separate process, when one splits into two, how do you give the new process all the information the old one was storing? do we just force teh child to start from no info? 2010-12-11T05:26:53 epona: i mean criticism for me ideas from you :) 2010-12-11T05:26:57 my* 2010-12-11T05:27:07 processes* 2010-12-11T05:28:16 i can't see any other way around that 2010-12-11T05:32:00 or... we could allow people to write a "print info for child" function, and pass that string to the bot 2010-12-11T05:33:45 no, even better, i could just pass ants each turn the map with most recent known information to it, and send that map to the new ant too 2010-12-11T05:37:14 i think were talking about different things 2010-12-11T05:43:39 we have to use wall clock if ants use separate processes too 2010-12-11T06:22:29 *** Appleman1234 has joined #aichallenge 2010-12-11T06:43:25 http://www.ai-contest.com/forum/viewtopic.php?f=21&t=1205&p=8244#p8244 2010-12-11T06:43:27 let me have it :) 2010-12-11T06:43:33 sigh cough 2010-12-11T06:43:52 *** Frontier_ has quit IRC (Remote host closed the connection) 2010-12-11T06:51:47 *** Mekanik has quit IRC (Ping timeout: 240 seconds) 2010-12-11T07:02:24 *** epona has quit IRC (Ping timeout: 255 seconds) 2010-12-11T07:04:39 *** Mekanik has joined #aichallenge 2010-12-11T07:06:26 *** Zannick has quit IRC (Ping timeout: 252 seconds) 2010-12-11T07:19:40 *** choas has joined #aichallenge 2010-12-11T07:32:24 *** Accoun has quit IRC () 2010-12-11T07:42:43 *** yasith has joined #aichallenge 2010-12-11T07:47:11 *** Utkarsh has joined #aichallenge 2010-12-11T08:06:24 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-12-11T08:11:24 *** Mekanik has quit IRC (Quit: WeeChat 0.3.3) 2010-12-11T08:30:10 *** mceier has joined #aichallenge 2010-12-11T08:30:50 *** Accoun has joined #aichallenge 2010-12-11T08:38:13 *** delt0r___ has joined #aichallenge 2010-12-11T08:39:29 *** delt0r_ has quit IRC (Ping timeout: 245 seconds) 2010-12-11T08:40:53 *** Appleman1234 has quit IRC (Ping timeout: 240 seconds) 2010-12-11T08:45:27 *** Rubicon-|-Cross has joined #aichallenge 2010-12-11T08:51:37 delt0r___: come in, come in 2010-12-11T09:14:30 YAAAS (Yet another attempt at simplification): http://www.ai-contest.com/forum/viewtopic.php?f=21&t=1205&p=8245#p8245 2010-12-11T09:14:47 I like it... but then... I would :) 2010-12-11T09:14:59 Certainly makes things more simple 2010-12-11T09:16:58 Meatkat: did you read my one above? 2010-12-11T09:17:11 it could be done without separate processes too, but i think that makes the game much more interesting 2010-12-11T09:17:57 i'm quite confident what i described could actually have a working prototype made for it 2010-12-11T09:18:36 I think you're is solid. But it's degrees of complexity above mine. 2010-12-11T09:19:20 it depends, i think we could merge them 2010-12-11T09:19:29 I would personally probably prefer yours, but the average shmo would find it intimidating 2010-12-11T09:19:32 i like your hole idea with the queen bee and taking them out by entering the hole 2010-12-11T09:19:43 but then i would want to be able to genrate walls for defense 2010-12-11T09:19:52