2011-01-16T01:25:30 deque's are my new vectors! 2011-01-16T01:40:48 hashtables are my new deques! 2011-01-16T01:40:52 am I doing this right? 2011-01-16T01:58:28 *** Accoun has joined #aichallenge 2011-01-16T02:07:56 *** amstan_ has quit IRC (Ping timeout: 240 seconds) 2011-01-16T02:35:34 *** chris___0076 is now known as chris_0076 2011-01-16T02:38:17 *** sigh has quit IRC (Remote host closed the connection) 2011-01-16T02:38:42 *** sigh has joined #aichallenge 2011-01-16T02:44:54 *** contestbot has joined #aichallenge 2011-01-16T02:44:54 *** ChanServ sets mode: +o contestbot 2011-01-16T02:45:13 *** amstan_irssi has joined #aichallenge 2011-01-16T02:48:01 *** Meatkat_ has quit IRC (Ping timeout: 260 seconds) 2011-01-16T03:24:23 sigh: :P to you too 2011-01-16T03:26:27 here's a link to that map generator for snowmen if anyone wants it 2011-01-16T03:26:33 http://codepad.org/2QCb1qMp 2011-01-16T03:26:48 you'll need to uncomment some stuff in main when running it locally to actually write the maps to file 2011-01-16T04:16:03 *** Accoun has quit IRC () 2011-01-16T05:01:39 *** Frontier has joined #aichallenge 2011-01-16T05:07:11 *** mega1 has joined #aichallenge 2011-01-16T05:15:45 I just cracked some software and wherever I would talk about it I would be banned I guess ^^. 2011-01-16T05:17:30 * Frontier supresses the urge to tell someone about it. 2011-01-16T05:32:51 *** dr- has joined #aichallenge 2011-01-16T05:50:55 Frontier: i don't think people really care that much :P 2011-01-16T05:51:16 well freenode have rules about it 2011-01-16T05:51:45 *** Accoun has joined #aichallenge 2011-01-16T05:51:54 because they are policing every channel and all :P 2011-01-16T05:52:09 we don't want to be the first... 2011-01-16T05:52:18 so the channels enforce these rules 2011-01-16T05:53:12 with the amount of actual illegal activity going on on irc, i doubt they'd make an example out of one person mentioning a program they cracked, but sure :) 2011-01-16T05:53:35 we like our free channels 2011-01-16T05:53:51 Most channels get out the banhammer 2011-01-16T05:54:11 warz is on other irc networks 2011-01-16T05:54:26 warez 2011-01-16T05:54:33 not really bad, but i've known of a lot of private bt sites that would use freenode 2011-01-16T05:54:49 only warez i ever got from irc was skatevideos back in the day 2011-01-16T05:55:22 we also like them warez free... its really just another form of spam 2011-01-16T05:55:43 best kind of spam? :P 2011-01-16T05:56:08 well for the person receiving it, not for those pushing it through 2011-01-16T05:56:27 *** Accoun has quit IRC (Client Quit) 2011-01-16T05:56:29 even now.. we are not discussing the next AI thing 2011-01-16T05:56:39 hence spam :P 2011-01-16T05:57:29 meh, this channel has always had more than strictly ai conversations 2011-01-16T05:57:46 it isn't impeding people from talking about the main topic, so I don't personally care 2011-01-16T05:58:21 delt0r___: have you looked at the snowballs game at all? 2011-01-16T05:58:28 no 2011-01-16T05:58:38 I have been busy with work 2011-01-16T05:58:52 Just trying to get my game out now 2011-01-16T05:59:54 the snowballs game is funny, I think I'll do it 2011-01-16T06:00:24 i'm frustrated with how messy incrementing states is, but meh 2011-01-16T06:00:49 delt0r___: it's an actual acm contest if you haven 2011-01-16T06:00:50 t seen 2011-01-16T06:00:55 http://queue.acm.org/icpc/ 2011-01-16T06:01:16 yep that's puzzling at the beginning 2011-01-16T06:01:43 *** mceier has joined #aichallenge 2011-01-16T06:02:24 I rewrote the python player, just to be clearer ... 2011-01-16T06:03:36 i wrote my c++ bot from scratch 2011-01-16T06:03:42 it seems strange... 2011-01-16T06:03:43 did you get my map generator link before? 2011-01-16T06:04:02 I typically don't make a snowman in the middle of a snowball fight 2011-01-16T06:04:13 i haven't got that far yet 2011-01-16T06:04:17 i am still setting everything up 2011-01-16T06:04:19 http://i.imgur.com/Z8UPO.png 2011-01-16T06:04:33 the move algorithm works, but it doesn't make any intuitive sense 2011-01-16T06:05:55 do you understand which linear path a snowball gets when you throw it if not horyzontal or vertical ? 2011-01-16T06:06:10 dr-: http://codepad.org/2QCb1qMp that's my map gen, you will find 5 maps at the bottom and you need to uncomment lines 204-206 to have them save to files in the directory you run it from 2011-01-16T06:06:24 okay, thanks antimatroid1 :) 2011-01-16T06:06:44 read the linear interpolation part, it gives you functions for working out what square an object will be in at each step between moves 2011-01-16T06:07:05 as well as the height of a snowball as it moves 2011-01-16T06:07:44 i really hate the move cancellation, you can swap squares, but not through an empty middle one, and all sorts of other weird things that are just unintuitive 2011-01-16T06:07:54 okay, i read it but found it confusing 2011-01-16T06:08:08 it makes more sense once you have done some programming 2011-01-16T06:08:34 i've got 11 files already, and my bot is just walking around 2011-01-16T06:08:38 did you consider the 10% limit for trees in your generator ? 2011-01-16T06:08:54 yes, 0.1*31*31 = 96.something 2011-01-16T06:09:05 yep, i just see it, sorry 2011-01-16T06:09:11 all good :) 2011-01-16T06:11:57 i've sort of worked out how to do move selection and a decision/game tree depending on whether i can see any of their kids 2011-01-16T06:12:02 i just need to get all the messy stuff programmed 2011-01-16T06:12:09 then just play around with evaluation 2011-01-16T06:13:11 same for me, i think the strategy is clear in my head 2011-01-16T06:13:42 will you use the catch action ? 2011-01-16T06:14:04 doubtful 2011-01-16T06:14:13 actually, yes probably 2011-01-16T06:14:22 that's true, it's useful to share small snowball between our children 2011-01-16T06:14:30 just to spare time 2011-01-16T06:14:38 yeah, or if you think they're going to hit you, might be worth trying to catch 2011-01-16T06:15:09 i'm going to do a pretty wide tree when children can see each other 2011-01-16T06:15:32 then a much deeper kind of decision tree at other points 2011-01-16T06:16:58 at the moment, i improve my donothing strategy :D 2011-01-16T06:17:47 i pretty much need to rewrite their engine to get my tree going though 2011-01-16T06:17:50 so that will take a while 2011-01-16T06:18:01 i may get impatient and write a quick heuristicy bot at some point 2011-01-16T06:19:40 otherwise, what's up with your ant game antimatroid1 ? 2011-01-16T06:20:01 doesn't look like my variation will be used, so it's mostly dead 2011-01-16T06:20:11 i'm fairly uninterested in the ant game they've got going atm 2011-01-16T06:20:39 i still think they should use a variation of frontiers visualiser though, that was awesome 2011-01-16T06:20:54 yours is good, that's true 2011-01-16T06:21:34 the one they show isn't a ant game but a zombie game i think 2011-01-16T06:21:40 the current one is sort of get grains and multiply, i don't think there's even much strategy wise to the fighting 2011-01-16T06:21:53 yep i saw that 2011-01-16T06:22:08 i made mine with strategy in mind, and really liked the home base invasion/defense part, which was my main objective 2011-01-16T06:22:16 hence the two games are fairly different 2011-01-16T06:22:40 zombie ants want to eat your GRAAAAINS 2011-01-16T06:22:59 one of the better random calls :P 2011-01-16T06:26:58 have to go, see ya ;) 2011-01-16T06:27:03 *** dr- is now known as dr`away 2011-01-16T06:38:42 *** dr`away has quit IRC (Ping timeout: 240 seconds) 2011-01-16T06:49:52 *** choas has joined #aichallenge 2011-01-16T06:54:07 what? my viszualizer? well... that was created with a game of 32x32 in mind, not 128x128 :D 2011-01-16T06:56:21 yes, i really don't like that ants game anyway 2011-01-16T06:56:39 i'd rather 3v3 tron 2011-01-16T07:03:05 to be honest I liked it more then walls were blocking sight and we were contemplating one process per ant with actual communication between te ants through color signals, scents or something like that. 2011-01-16T07:04:13 We miss a real killer game idea for next term. Something addictive like Minecraft. 2011-01-16T07:06:17 It has few rules, but creates a lot of possibilities. 2011-01-16T07:11:04 We always think of games with grids or production rates. Maybe we should have looked at HaxBall http://www.youtube.com/watch?v=yGsDEizkXJU a little closer. 2011-01-16T07:18:04 *** delt0r_ has joined #aichallenge 2011-01-16T07:18:28 haxball is just a cut down version of the server soccer league 2011-01-16T07:20:07 *** delt0r___ has quit IRC (Ping timeout: 276 seconds) 2011-01-16T07:34:18 nothing has really been decided anyway has it? 2011-01-16T07:34:28 does anyone actually like the current ants game they have done up? 2011-01-16T07:34:29 *** Frontier has quit IRC (Ping timeout: 260 seconds) 2011-01-16T07:34:48 Frontier: i agree about the original idea being good, but individual processes really isn't feasible 2011-01-16T07:46:57 *** dr1 has joined #aichallenge 2011-01-16T07:47:14 *** dr1 is now known as dr- 2011-01-16T07:47:15 *** Obeleh has joined #aichallenge 2011-01-16T07:47:23 I just don't really like ants... 2011-01-16T07:48:22 it doesn't feel ai enough and feels more like ecology... and stuff... My colleagues in Munich even have a ant simulator for studying ants. 2011-01-16T07:48:44 i breed ants, and it's very interesting ;) 2011-01-16T07:48:52 but not ai 2011-01-16T07:49:09 ants are completely cool... real ones that is 2011-01-16T07:49:23 s/ants/anything else/g 2011-01-16T07:52:47 is this simulator available, delt0r_ ? 2011-01-16T07:53:08 delt0r_: I'm still hanging out for your game :P 2011-01-16T07:53:22 i quite like snowballs game 2011-01-16T07:53:35 the strategy space is huge at least 2011-01-16T07:53:35 dr-: I don't know 2011-01-16T07:54:14 and the game mechanics aren't entirely trivial, particularly with throwing snow 2011-01-16T07:55:14 dr-: its this group http://evol.bio.lmu.de/index.html 2011-01-16T07:55:26 can't find a more specfic link right now 2011-01-16T07:55:55 they are modeling slave maker ants. They live in acorns in the US 2011-01-16T07:59:48 antimatroid1: I am tring to finish it... but i keep getting real work to do 2011-01-16T08:00:10 delt0r_: thanks, i'll check this out 2011-01-16T08:01:01 delt0r_: atm it sounds better to me than any of the other alternatives 2011-01-16T08:01:10 but i doubt it'll get in for the next contest 2011-01-16T08:01:56 I think the stratego is the best idea so far 2011-01-16T08:01:59 nice 2011-01-16T08:02:00 simple 2011-01-16T08:02:05 complicated 2011-01-16T08:02:08 fun 2011-01-16T08:02:14 and easy to code for a engine 2011-01-16T08:02:31 Mine is pretty difficult for the engine 2011-01-16T08:02:45 unless you know your complex math properly 2011-01-16T08:03:27 In fact even if my idea was finished... i would still vote for that 2011-01-16T08:04:36 stratego is flags yeah? 2011-01-16T08:04:56 my ants invasion could potentially have been seen like that, but you just had to "claim their flag" to win 2011-01-16T08:05:54 antimatroid1: yea... the American name for the game i think. It was originally dutch as i understand 2011-01-16T08:06:28 ant have complicated moves. stratego does not.. 2011-01-16T08:06:38 ie everything but how to win is simple 2011-01-16T08:07:03 ants was pretty basic at every step 2011-01-16T08:07:07 my ants* 2011-01-16T08:07:31 compared to this snowballs game, writing a bot would have been a walk in the park 2011-01-16T08:09:18 i agree, snowballs game : 10 differents moves whereas ants games still have less moves 2011-01-16T08:10:12 10 different moves? where did you get that? 2011-01-16T08:10:15 mmm 2011-01-16T08:10:19 *** mega1 has quit IRC (Remote host closed the connection) 2011-01-16T08:10:25 oh, 10 different kinds? 2011-01-16T08:10:29 yep 2011-01-16T08:10:32 the more i think about it... the more i want to play stratego 2011-01-16T08:10:42 what would the deal with copyright be? 2011-01-16T08:10:50 just call it something else 2011-01-16T08:10:53 just call it bombs and flags 2011-01-16T08:10:55 flags :P 2011-01-16T08:11:12 probably need to not use "bombs" and "flags" however 2011-01-16T08:11:23 perhaps zombies and nijias 2011-01-16T08:11:35 can't go wrong with a name like that 2011-01-16T08:12:05 instead of a spy --its a ninja 2011-01-16T08:12:42 instead of a bomb... a zombie... and we all know that zombies can't really move, but still beat a ninja 2011-01-16T08:12:52 *** sigh has quit IRC (Remote host closed the connection) 2011-01-16T08:13:00 and you have to protect your.... 2011-01-16T08:13:01 cow 2011-01-16T08:13:11 oh, did you ever play that primary school game where you each got a playing card, and if you met with an enemy, you were captured if you had a lower card and could be escorted back to a jail or whatever? 2011-01-16T08:13:18 *** mega1 has joined #aichallenge 2011-01-16T08:13:19 or a pink pony 2011-01-16T08:13:38 omg --ponies... 2011-01-16T08:13:48 unicorns > ponies 2011-01-16T08:14:06 fluffy unicorns... 2011-01-16T08:14:20 fluffy pink zombie unicorns 2011-01-16T08:14:39 actually i like this idea 2011-01-16T08:15:44 *** mega1 has quit IRC (Remote host closed the connection) 2011-01-16T08:16:40 take a discrete grid, fog of war etc., you each start with x kids/ants/unicorns/whatever, assigned numbers 1, ..., x, which is which is your own private information, if you land on a square adjacent to one of their bots, the bots with the lowest rankings get removed first until no more bots need to be removed 2011-01-16T08:18:53 *** mega1 has joined #aichallenge 2011-01-16T08:19:01 you can also say that your 1 can kill their x winning you the game 2011-01-16T08:21:16 *** Frontier has joined #aichallenge 2011-01-16T08:25:09 *** boand_ has joined #aichallenge 2011-01-16T08:36:47 *** Migi32 has joined #aichallenge 2011-01-16T08:39:05 the complexity is that each team sets up its peices anyway they want 2011-01-16T08:47:49 So my "official" vote is for simultaneous stratego. 2011-01-16T08:48:21 my rts one will be better for next round anyway... get some play time with balancing everything 2011-01-16T08:53:08 *** Sylph has joined #aichallenge 2011-01-16T08:54:06 Also want to finish a andriod version of the game 2011-01-16T08:56:30 bah, it's too hot 2011-01-16T08:56:45 yeah, -13 2011-01-16T08:56:53 well it 10C here... so the snow is melting.... not what i want 2011-01-16T08:56:55 it's summer here 2011-01-16T08:57:05 it's mid-winter here 2011-01-16T08:57:06 :) 2011-01-16T08:57:16 yesterday was something .. -25 2011-01-16T08:57:20 yet australia has hardly any bushfires and plenty of floods 2011-01-16T08:57:24 not the typical summer 2011-01-16T08:57:30 yes it is 2011-01-16T08:57:41 you guys have floods every second year 2011-01-16T08:57:53 for at least as long as i can remeber 2011-01-16T08:57:59 we usually get some pretty bad bushfires during the summer 2011-01-16T08:58:28 i live nowhere near the bush though, so i'm all good 2011-01-16T09:07:54 *** dr- has quit IRC (Quit: WeeChat 0.3.0) 2011-01-16T11:22:29 i'm also beginning to lean away from the tron game in favor of stratego. i'm finding that the line of sight stuff is quite involved. the efficient ways to do it are too complicated for a specification. without the line of sight stuff i think the game loses some appeal. simple range-based fog of war doesn't really cut it 2011-01-16T11:22:59 I just really like stratego 2011-01-16T11:23:17 and given time frames... if a feb/march date is the target 2011-01-16T11:23:38 I think its more realistic that some games that really need a bit more trashing out 2011-01-16T11:24:00 in fact I was looking for the board game here... 2011-01-16T11:24:20 are you thinking simultaneous move-all-units or simultaneous move-one-unit? 2011-01-16T11:24:38 heh yeah, i've played a few games with my wife since i thought of using a variation of it for the contest 2011-01-16T11:24:57 *move-one-unit-each 2011-01-16T11:25:02 i think i still like the former 2011-01-16T11:25:08 i'm less certain about fog of war 2011-01-16T11:25:47 I would have the same as the board game regarding FOG 2011-01-16T11:25:55 the strategy space is already huge, and there's already hidden information 2011-01-16T11:26:00 so... none? 2011-01-16T11:26:00 you only know what something is when you attack... 2011-01-16T11:26:09 its up to the bot to keep track 2011-01-16T11:26:11 oh you are calling that fog of war 2011-01-16T11:26:14 *** Blkt has joined #aichallenge 2011-01-16T11:26:34 well you don't get to see what things are right.. till attacked? 2011-01-16T11:26:38 i was talking about knowing *where* units are only when in sight range, as in the traditional RTS sense 2011-01-16T11:26:49 Oh hasbro has a different version out? with a dragon 2011-01-16T11:26:51 and then only knowing *what* they are when fighting 2011-01-16T11:26:59 yeah they do. i haven't seen it anywhere 2011-01-16T11:27:08 oh ic 2011-01-16T11:27:30 that way if you lose sight of a unit after fighting it, you might not be able to associate its unit type with it again later 2011-01-16T11:27:33 might make it too hard 2011-01-16T11:27:36 Don't know about that... would tend to stick closely to the board game... as it has pretty good balance 2011-01-16T11:27:41 yeah 2011-01-16T11:27:48 another thing though 2011-01-16T11:28:18 if you have full vision, it is very easy to see what's not moving. the bots might have to avoid moving a *lot* of their units just to not give away the bombs and flag 2011-01-16T11:28:29 since this is move-everything-at-once 2011-01-16T11:28:30 yea 2011-01-16T11:28:39 thats the point of stratego 2011-01-16T11:28:47 yeah, but normally you only move one thing at a time 2011-01-16T11:28:54 so it's harder to see what's not moving 2011-01-16T11:29:06 well I guess one unit per turn is ok too 2011-01-16T11:29:29 I would do either all units can only move on square... and more than one unit can move in a turn 2011-01-16T11:29:50 or let units have different movement as per stratego 2011-01-16T11:29:56 and only one per turn 2011-01-16T11:30:04 i think i prefer the former still 2011-01-16T11:30:05 *** mceier has quit IRC (Quit: leaving) 2011-01-16T11:30:38 "attacks" would need to be resolved with simultaneous movement.. not sure how that works with things like scouts 2011-01-16T11:31:02 the only problem is that some units are usefull because of their movement... ie scouts 2011-01-16T11:31:29 in my code, i just have it to where moving onto the same tile is an attack, and also if two units are moving into each other it's an attack where the winner moves onto the loser's old position 2011-01-16T11:31:38 scouts are useful for other reasons too 2011-01-16T11:31:46 they are basically fodder for feeling things out 2011-01-16T11:31:54 their movement is just bonus 2011-01-16T11:32:37 i've also been assuming that it's simply illegal for a bot to send two of its own units to the same position 2011-01-16T11:32:38 their movement is/was very important 2011-01-16T11:32:50 scouts that is 2011-01-16T11:32:59 sure. i'm just saying i don't think it's the end of the world if they are weakened a bit 2011-01-16T11:33:04 keep one till end game was a big deal IIRC 2011-01-16T11:33:16 it would really simplify the game if they moved the same as everything else 2011-01-16T11:33:23 this is where i think its easy to break the ballance... 2011-01-16T11:33:25 heck, the new stratego has tons of new rules :\ 2011-01-16T11:33:26 ballance is hard 2011-01-16T11:33:49 i've played games of stratego where neither player moved their scouts more then on square 2011-01-16T11:34:05 officially you can't attack and move at the same time anyway. only in variations can you do that 2011-01-16T11:34:23 with scouts' special movement i mean 2011-01-16T11:35:24 also officially when a unit attacks abother, both players get to learn something, but there are also interesting variations where the one being attacked only has to reveal the unit if the unit is dying anyway, but the attacker still has to reveal 2011-01-16T11:35:28 *another 2011-01-16T11:35:54 the latter makes scouts far less useful 2011-01-16T11:36:18 except as fodder... and if they move more than one square there's no incentive to attack them at all 2011-01-16T11:36:40 anyway, i'm just saying there are lots of challenging variations and i don't think breaking scouts breaks the game 2011-01-16T11:36:44 the rule set i played 2011-01-16T11:37:02 was that both attacked and attacker had to be "revealed" 2011-01-16T11:37:12 yes, that's the "official" rules 2011-01-16T11:37:16 scouts couldn't attack unless adjcent 2011-01-16T11:37:32 but i had a set that also suggested the variant i meantioned, too 2011-01-16T11:37:37 *mentioned 2011-01-16T11:38:14 with simultaneous movement, the first think i would do if i saw a unit do a scout-movement is move away from the scout so it can't reveal anything 2011-01-16T11:38:17 *thing 2011-01-16T11:38:19 well I think that keeping as close, while make turns simultaneous.. would be the best idea 2011-01-16T11:38:27 okay, i'll start checking my typing before i hit enter now :) 2011-01-16T11:38:42 talking about next AI challenge? 2011-01-16T11:38:51 and i would favour the one unit per turn 2011-01-16T11:39:03 ah, then that's different 2011-01-16T11:39:24 i just don't want to be *too* much like the board game 2011-01-16T11:39:47 well simulationus turns its a pretty big difference 2011-01-16T11:39:57 but yea 2011-01-16T11:40:04 there is the copyright thing 2011-01-16T11:40:20 well, it's hard to claim copyright on gameplay 2011-01-16T11:40:27 good evening people 2011-01-16T11:40:35 lawyers get paid by the hour 2011-01-16T11:40:41 not for good advice 2011-01-16T11:41:14 so move-all-units then? ;) 2011-01-16T11:41:31 well its your idea... i am happy enough either way 2011-01-16T11:42:01 if its all units move... then i would perhaps even change the units... perhaps drop scouts... 2011-01-16T11:42:04 or something 2011-01-16T11:42:34 keep a miner, spy, marshell and flag 2011-01-16T11:42:34 without scouts the lowest ranking unit is miners, which kind of sucks because you need to hold on to them 2011-01-16T11:42:47 yea i guess 2011-01-16T11:42:47 i assume you also mean keep bombs 2011-01-16T11:42:52 yes 2011-01-16T11:43:17 bombs are awesome when a Marshal finds one 2011-01-16T11:43:44 and a scout that never moves just became a taboo unit 2011-01-16T11:43:55 i'm afraid of having too few unit types because then there's really just not much point in having micro and shuffling things around so that the fights are optimal, and the guessing game is a lot easier 2011-01-16T11:44:22 this is why i would favour one unit per turn... 2011-01-16T11:44:27 keep the turns short 2011-01-16T11:44:32 ie .2 sec 2011-01-16T11:44:35 or something 2011-01-16T11:44:36 miner/spy/marshall would be rock/paper/scissors, except for the bomb thing 2011-01-16T11:44:58 Na.. you are right... just keep the units as is 2011-01-16T11:45:51 we could theme it after ants 2011-01-16T11:45:55 or something 2011-01-16T11:46:17 Zombies 2011-01-16T11:46:20 anything but what stratego does 2011-01-16T11:46:31 ninja zombies 2011-01-16T11:46:34 and unicorns 2011-01-16T11:46:40 blind people 2011-01-16T11:46:44 fluffy unicorns 2011-01-16T11:47:06 so a zombie is a bomb... 2011-01-16T11:47:10 ie does not move. 2011-01-16T11:47:24 but kill pretty much anything 2011-01-16T11:47:44 except for the redneck 2011-01-16T11:47:47 i am only half joking 2011-01-16T11:47:59 redneck == zombie 2011-01-16T11:48:08 rednecks are the miners 2011-01-16T11:48:13 lol 2011-01-16T11:48:19 they carry around shotguns and kill zombies 2011-01-16T11:48:24 I was going to go for buffys 2011-01-16T11:48:37 i'm running with the zombie scenario 2011-01-16T11:48:43 the marshal is the town mayor 2011-01-16T11:48:56 this makes no sense 2011-01-16T11:48:57 I think the marshal should be Alice... 2011-01-16T11:49:04 why are two towns fighting during a zombie apocalypse? 2011-01-16T11:49:14 why not 2011-01-16T11:49:26 zombies don't need a reason 2011-01-16T11:49:30 the virus that made the zombies drove them crazy 2011-01-16T11:49:32 that's it 2011-01-16T11:49:55 Alice from resident evil is a Marshal 2011-01-16T11:50:09 the flag? mm. perhaps that can be the cure 2011-01-16T11:50:43 scouts are normal citizens... 2011-01-16T11:51:08 well anyway... on a more serious note 2011-01-16T11:51:27 I would think given time frames etc.. this is perhaps the best idea. 2011-01-16T11:51:34 the only thing left is maps 2011-01-16T11:51:50 do we have a fixed map... or not... or even really big maps 2011-01-16T11:51:51 *** amstan_ has joined #aichallenge 2011-01-16T11:51:51 *** ChanServ sets mode: +o amstan_ 2011-01-16T11:52:05 i want really big maps 2011-01-16T11:52:10 *** dr- has joined #aichallenge 2011-01-16T11:52:11 simultaneous movement begs for that 2011-01-16T11:52:23 the hard part here is unit placement 2011-01-16T11:52:28 yea 2011-01-16T11:52:42 so give 10 sec or something for unit placement... 2011-01-16T11:52:59 fixed "starting" grid 2011-01-16T11:53:32 rather than a over sized starting zone that you don't even have to fully occupy? 2011-01-16T11:53:42 i would say that the maps should all be top vs. bottom, and the bot always sees himself as playing the bottom, and the bot knows that he should always just place his units on the bottom half or so of the map 2011-01-16T11:53:56 sounds good 2011-01-16T11:54:09 I would not have any fog of war 2011-01-16T11:54:16 yeah i think it's unnecessary 2011-01-16T11:54:17 so you see units. 2011-01-16T11:54:23 just not what they are 2011-01-16T11:54:25 bots will just have to move fewer units than possible 2011-01-16T11:54:30 to hide things 2011-01-16T11:54:35 indeed 2011-01-16T11:54:51 lakes or obstructions in the map? 2011-01-16T11:54:55 sure 2011-01-16T11:55:02 fixed/random? 2011-01-16T11:55:05 i'd just say the same sort of thing as the walls in the last tron game 2011-01-16T11:55:17 so random 2011-01-16T11:55:22 fixed/pregenerated is probably best 2011-01-16T11:55:26 not new for each game 2011-01-16T11:55:30 oh 2011-01-16T11:55:36 disagree? 2011-01-16T11:55:36 k 2011-01-16T11:55:41 i like random 2011-01-16T11:55:49 i'm rather indifferent 2011-01-16T11:55:51 did quantum physics 2011-01-16T11:55:54 heh 2011-01-16T11:56:04 random and prime numbers are cool 2011-01-16T11:56:12 random is fine. but i want to at least be able to feed fixed maps to the engine 2011-01-16T11:56:20 yes 2011-01-16T11:56:24 of course 2011-01-16T11:56:43 and so we might be best off with our normal architecture of engine + map generator 2011-01-16T11:56:52 and we could just tie them together for tournaments 2011-01-16T11:56:58 I would have the obstructions only distributed in the center of the map 2011-01-16T11:57:02 if we want randomness 2011-01-16T11:57:22 yes... completely agree 2011-01-16T11:57:30 i wouldn't. i think obstructions that hinder movement might lead to some interesting placement strategies 2011-01-16T11:57:42 yea... 2011-01-16T11:57:43 as long as there are no closed areas 2011-01-16T11:57:55 that are not filled in with walls, at least 2011-01-16T11:58:01 *** amstan_ has quit IRC (Ping timeout: 240 seconds) 2011-01-16T11:58:06 easy enough to check at least.. 2011-01-16T11:58:38 and they would need to be symmetric 2011-01-16T11:58:49 again--easy to do 2011-01-16T11:59:13 but if 2 units are next to each other... both move "past" --what happens? 2011-01-16T11:59:18 ie 2011-01-16T11:59:39 in my code they fight and the winner moves to the loser's old position 2011-01-16T11:59:53 units in that scenario fight before all other fights 2011-01-16T12:00:02 brb 2011-01-16T12:00:06 right...and since there are no diag moves... thats the only case to consider 2011-01-16T12:02:11 *** Naktibalda has joined #aichallenge 2011-01-16T12:06:04 well, sometime in the next few days i'll dust off the stratego code and get it working 2011-01-16T12:06:19 i had already decided i wasn't liking the direction of this tron game due to the complexity 2011-01-16T12:06:36 and what do you think of ants? (I am not a fan) 2011-01-16T12:06:47 it would suck to have to describe a symmetric field of view algorithm in the specification 2011-01-16T12:06:54 lol 2011-01-16T12:06:59 i mostly dislike the ants game 2011-01-16T12:07:13 there are ideas for it that i like, but i'm not sure if they would add enough 2011-01-16T12:07:23 it lacks depth 2011-01-16T12:07:27 they feel to "a-life" to me... and not enough ai 2011-01-16T12:07:50 scavenging tends to feel that way 2011-01-16T12:08:42 I could write a strageto game engine/starter pack/etc combo next weekend.. 2011-01-16T12:08:58 what kind of coms format did you use? 2011-01-16T12:09:09 that's the main thing that i haven't done 2011-01-16T12:09:29 that and testing of any existing code... 2011-01-16T12:09:42 which means there's a ton of stuff i haven't done :P 2011-01-16T12:10:42 not test driven development ;) 2011-01-16T12:11:04 not even very good type-driven development 2011-01-16T12:11:10 lol 2011-01-16T12:11:15 what lang... 2011-01-16T12:11:17 haskell 2011-01-16T12:11:22 of course 2011-01-16T12:11:25 :D 2011-01-16T12:11:27 Well i would use java 2011-01-16T12:11:45 since lisp is still not viable for me as a day job lang 2011-01-16T12:12:08 you're treating this like a day job? :P 2011-01-16T12:12:13 no 2011-01-16T12:12:22 but since i am all set up to use java 2011-01-16T12:12:26 do a lot of java 2011-01-16T12:12:33 and know java pretty well 2011-01-16T12:12:53 it is easyer to just use java for something so small 2011-01-16T12:13:11 I really like CLOS etc... but i don't really know them... and haven't used it much 2011-01-16T12:13:41 I still bork lisp macros from time to time... and macros of macros don't work the way i generally expect 2011-01-16T12:14:40 no need to make excuses for using a language you are reliable with instead of one that's more fun 2011-01-16T12:15:02 that about sums it up 2011-01-16T12:15:30 either way... I will be happy to provide a stater pack and a visualizer etc... 2011-01-16T12:16:42 i'm still going to keep coding my engine. i have a vision that i want to see 2011-01-16T12:17:11 I meant i would conform to your engine... spec 2011-01-16T12:17:28 ha 2011-01-16T12:18:01 nothing in stone. if you come up with a comm format before me i could go with that unless there are things i just *really* don't like about it 2011-01-16T12:19:24 i probably need to start over anyway... with a more principled, test-driven approach 2011-01-16T12:19:40 i was just bashing it out, previously... for prototyping, basically 2011-01-16T12:19:54 turns out it's pretty simple though :) 2011-01-16T12:23:37 for the comm format i'm thinking we should stick with the same overall format as what i was going for with the tron game 2011-01-16T12:23:43 extensible, simple 2011-01-16T12:24:57 yea 2011-01-16T12:25:11 well i won't get to to at least next weekend 2011-01-16T12:25:38 to->it 2011-01-16T12:27:40 i *shouldn't* get to it until... 13 days from now 2011-01-16T12:27:44 that doesn't mean i won't 2011-01-16T12:31:41 *** boegel has joined #aichallenge 2011-01-16T12:43:45 how does the marshal/spy thing work if the both attack at the same time? 2011-01-16T12:49:28 i'm just saying the spy always beat the marshal 2011-01-16T12:49:42 which... kind of stinks i guess 2011-01-16T13:10:57 *** Naktibalda has quit IRC (Remote host closed the connection) 2011-01-16T13:11:34 *** Naktibalda has joined #aichallenge 2011-01-16T13:19:04 how about if the marshal attacks it wins 2011-01-16T13:19:14 but if both attack... the spy wins 2011-01-16T13:19:53 and finally what about "chained" movement 2011-01-16T13:19:56 ie 2011-01-16T13:20:01 A,b 2011-01-16T13:20:09 sorry 2011-01-16T13:20:17 A->B->C-> 2011-01-16T13:20:34 A can move because B moves etc 2011-01-16T13:20:48 in the same turn.. of course... if one is in a fight..... 2011-01-16T13:24:31 this works out fine as long as two units moving directly at other fight before units that will simply share a square after moving 2011-01-16T13:24:46 ?? 2011-01-16T13:24:56 so if we have positions X, Y, and Z where A1 and A2 are on X and Y respectively and B is on Z and A1 is moving to Z, A2 is moving to X, and B is moving to X... A1 and B will fight first. if B wins, B moves to X and then B and A2 fight on X 2011-01-16T13:25:16 i think i got that right. lemme read over it 2011-01-16T13:25:21 ... might require making a diagram 2011-01-16T13:25:45 why do A1 and B fight first... won't the order make a difference 2011-01-16T13:25:46 oh wron 2011-01-16T13:25:47 g 2011-01-16T13:26:09 i set up the scenario wrong and but then explained what would happen for a different scenario 2011-01-16T13:26:14 let me just write up a fresh one 2011-01-16T13:26:23 i'll do diagrams and make a gist 2011-01-16T13:28:27 this means every unit that can move... could move in a turn 2011-01-16T13:28:36 right 2011-01-16T13:29:08 but fights would break this... and order does matter (im pretty certain ) 2011-01-16T13:29:29 fights won't break it if the order is well-defined, and i have a well-defined order 2011-01-16T13:29:34 i'm trying to write up an explanation 2011-01-16T13:29:44 This can be resovled in a order independnat manner as i did in my ants engine 2011-01-16T13:30:35 I don't think their should be a "specified" order... I think at that stage a single unit per turn works better again 2011-01-16T13:33:32 this makes sense though. if two units are moving toward each other it makes sense that they would fight each other "as they move", before the fights that happen when units finally arrive at their destinations 2011-01-16T13:34:07 delt0r_: https://gist.github.com/782013 2011-01-16T13:34:33 yes... but what can and can't move depends on the fight results... this can even effect *what* fights if you think about units move "sideways" into a wall of moving units 2011-01-16T13:34:34 *** boegel has quit IRC (Quit: *poof!*) 2011-01-16T13:34:49 that's not true 2011-01-16T13:34:57 the movement rules are consistent regardless of the fight results 2011-01-16T13:35:13 How so 2011-01-16T13:35:21 what is the scenario you have in mind? 2011-01-16T13:35:53 if two unit "move past each other" then how the fight turns out, changes which square is occupied after the fight 2011-01-16T13:35:58 the rules here are just that mutual attacks are carried out before movements. the winner always gets to move though, and no other units' movements are affected by it 2011-01-16T13:36:06 yes, but not how movement works 2011-01-16T13:36:16 it just changes the outcomes of other fights 2011-01-16T13:36:59 which is now order dependent... 2011-01-16T13:37:04 no it's not 2011-01-16T13:37:16 concrete scenario? 2011-01-16T13:37:20 In your example 2011-01-16T13:37:31 the movements are the same either way 2011-01-16T13:37:51 it makes sense that the fights could have different outcomes depending on the units. that's... just how stratego works 2011-01-16T13:37:52 if A1 loses agaist B 2011-01-16T13:37:55 then A2 fights 2011-01-16T13:38:00 as well 2011-01-16T13:38:02 yes 2011-01-16T13:38:11 i don't see that as a negative 2011-01-16T13:38:28 so now with more than 3 you can construct a loop where the result depends on where you start 2011-01-16T13:38:36 it's just that units that fight while moving get to fight before units that arrive at their destinations. it makes physical and practical sense to me 2011-01-16T13:38:39 I am pretty sure 2011-01-16T13:38:39 no 2011-01-16T13:38:40 you can 2011-01-16T13:38:47 i'll demonstrate 2011-01-16T13:38:55 with intermediate steps shown 2011-01-16T13:39:13 if you can come up with the scenario you think would not work 2011-01-16T13:40:02 doing so now 2011-01-16T13:40:11 this hinges on not allowing a bot to send two of its own units to the same square on the same turn 2011-01-16T13:40:18 * delt0r_ gets white board markers 2011-01-16T13:42:47 *** Accoun has joined #aichallenge 2011-01-16T13:45:21 delt0r_: for the record, i know that i'm write because the code i wrote is comonadic, which just means that if squares also contain bot orders, the future state of each square can be uniquely determined in isolation of the future states of the other squares 2011-01-16T13:45:39 delt0r_: that is, it's parallelizable to the granularity of individual squares 2011-01-16T13:45:53 *that i'm right 2011-01-16T13:46:02 how do i use the wrong word like that ?! 2011-01-16T13:46:37 I think you may be right... with the addition of the rule that a player cannot send two units to the same square... 2011-01-16T13:46:45 for 2 player games anyways 2011-01-16T13:46:56 yeah, it falls apart with more than 2 players 2011-01-16T13:47:26 *** Blkt has quit IRC (Read error: Connection reset by peer) 2011-01-16T13:47:56 i intentionally wrote my code as a comonad so that i would be forced to solve this problem cleanly ;) 2011-01-16T13:48:32 so the new question i have is 2011-01-16T13:48:52 a->b->c->d-> 2011-01-16T13:48:58 e^ 2011-01-16T13:49:07 does e fight c or b 2011-01-16T13:49:09 wait, where is e going? 2011-01-16T13:49:14 up 2011-01-16T13:49:34 onto the square previously held by c 2011-01-16T13:49:38 okay that's what i meant 2011-01-16T13:49:45 it would fight b 2011-01-16T13:49:50 since c moves at the same time as e 2011-01-16T13:49:53 *** Blkt has joined #aichallenge 2011-01-16T13:50:02 b and e end up on the same square, so they fight 2011-01-16T13:50:11 right 2011-01-16T13:50:23 we have fights with ones that swap 2011-01-16T13:50:33 then fights with ones that end up on the same square 2011-01-16T13:50:43 right, and movement technically happens in between 2011-01-16T13:50:49 yes 2011-01-16T13:53:23 the simultaneous movement combines the uncertainty of stratego with some micromanagement, obviously 2011-01-16T13:53:27 still think i would prefer a 1 unit per turn... game... but code wise its easy enough to do both 2011-01-16T14:04:22 just tried online stratego... what a crap web site... 2011-01-16T14:26:25 *** Blkt has quit IRC (Ping timeout: 276 seconds) 2011-01-16T15:17:25 *** Accoun has quit IRC () 2011-01-16T15:18:04 *** delt0r___ has joined #aichallenge 2011-01-16T15:19:53 *** delt0r_ has quit IRC (Ping timeout: 276 seconds) 2011-01-16T15:40:34 *** sigh has joined #aichallenge 2011-01-16T16:04:28 *** sigh has quit IRC (Remote host closed the connection) 2011-01-16T16:15:12 *** Naktibalda has quit IRC (Remote host closed the connection) 2011-01-16T16:20:09 *** Obeleh has quit IRC (Quit: Ik ga weg) 2011-01-16T16:23:16 lol 2011-01-16T16:25:01 *** dr- has quit IRC (Read error: Operation timed out) 2011-01-16T16:41:07 *** choas has quit IRC (Quit: leaving) 2011-01-16T16:51:05 lol? 2011-01-16T16:51:17 LOL 2011-01-16T16:51:22 LOL? 2011-01-16T16:51:34 well i didn't want to be left out 2011-01-16T17:01:07 *** mceier has joined #aichallenge 2011-01-16T17:10:51 *** mega1 has quit IRC (Ping timeout: 240 seconds) 2011-01-16T17:11:00 *** sigh has joined #aichallenge 2011-01-16T17:23:41 *** chris_0076 has quit IRC (Remote host closed the connection) 2011-01-16T17:25:28 *** tapwater has joined #aichallenge 2011-01-16T17:29:40 *** nexion- has joined #aichallenge 2011-01-16T17:34:33 *** Chris_0076 has joined #aichallenge 2011-01-16T17:46:01 *** chl has quit IRC (Ping timeout: 255 seconds) 2011-01-16T18:11:51 *** dlila has joined #aichallenge 2011-01-16T18:12:53 *** Meatkat has joined #aichallenge 2011-01-16T18:13:51 *** iFire has quit IRC (Read error: Connection reset by peer) 2011-01-16T18:13:52 *** iFire` has joined #aichallenge 2011-01-16T18:27:09 *** dlila has quit IRC (Quit: Leaving) 2011-01-16T18:27:39 *** dlila has joined #aichallenge 2011-01-16T19:05:48 *** dlila has quit IRC (Quit: Leaving) 2011-01-16T19:06:07 *** dlila has joined #aichallenge 2011-01-16T19:30:09 *** Frontier has left #aichallenge 2011-01-16T19:42:20 *** Migi32 has quit IRC (Quit: BOINC) 2011-01-16T19:55:41 *** sigh has quit IRC (Read error: Connection reset by peer) 2011-01-16T19:56:11 *** sigh has joined #aichallenge 2011-01-16T20:23:27 *** mceier has quit IRC (Quit: leaving) 2011-01-16T20:31:57 poo crap, is there a way to include a .txt file with source files to read from and actually know it'll be accessible when the program runs? 2011-01-16T20:32:18 *** Cold-Phoenix has joined #aichallenge 2011-01-16T20:37:44 no 2011-01-16T20:37:50 unless you use java 2011-01-16T20:37:52 ;) 2011-01-16T20:40:03 oh well, i think i only need it for testing anyway 2011-01-16T20:40:29 it's just so i can tell it what to do the first few moves, for testing the game mechanics 2011-01-16T20:40:57 there are "hacky" ways... 2011-01-16T20:41:13 the proper way... is to have the file as part of the installation... 2011-01-16T20:41:52 one has to submit source files, can a source file include a text file? 2011-01-16T20:42:38 theoretically 2011-01-16T20:43:40 but there is no expectation that you will have read access in the execution env 2011-01-16T20:43:49 basically its needs to be "supported" 2011-01-16T20:46:52 *** Accoun has joined #aichallenge 2011-01-16T20:50:53 *** Sylph has quit IRC (Read error: Connection reset by peer) 2011-01-16T20:59:26 i just realised i've never seen someone write u'r, while I hate that kind of writing, that would make me chuckle 2011-01-16T21:00:45 *** Accoun has quit IRC (Ping timeout: 260 seconds) 2011-01-16T21:12:17 *** Accoun has joined #aichallenge 2011-01-16T21:20:48 antimatroid1: wikipedia claims that u're can be used in the place you're although I would be suspicious as they do not cite any sources (http://en.wikipedia.org/wiki/SMS_language#Combinations_of_the_above_can_shorten_a_single_or_multiple_words) 2011-01-16T21:22:48 I would argue that if you are optimising over the number of characters, then u'r is just redundant and context should suffice to resolve any ambiguity between u'r and ur as is the case in spoken language 2011-01-16T21:23:41 *** Accoun has quit IRC (Ping timeout: 240 seconds) 2011-01-16T21:23:47 if I was to be so shit as to write like that, I would use u'r 2011-01-16T21:23:55 but i'm just difficult 2011-01-16T21:24:04 *** Accoun has joined #aichallenge 2011-01-16T21:24:59 If I may continue this absurd analysis, I would content that you are being ungrammatical as u'r would imply that it is a contraction, but if "you" is written as "u" and "are" is written as "r" then "you are" should be written as "ur" without requiring any contraction 2011-01-16T21:25:35 * I would contend 2011-01-16T21:25:38 my life revolves around analysing silly things :) 2011-01-16T21:26:17 so youare should be good? (my english is terrible) 2011-01-16T21:26:22 or u'? 2011-01-16T21:26:32 for ur? 2011-01-16T21:28:01 u' is beside the point, I was arguing that u'r is not a contraction, not that any contraction is valid 2011-01-16T21:29:59 and in english adjoining two words is not without precendent, many new words are formed this way... so writing ur does not violate any fundamental rules 2011-01-16T21:31:18 youare is not valid by convention only, it would easily be a word in it's own right 2011-01-16T21:31:46 hmmm, i think i'll stick to maths 2011-01-16T21:32:34 youare might even take on a slightly different meaning (cf "can not" vs "cannot") 2011-01-16T21:34:07 youare is gross :P 2011-01-16T21:36:56 *** Accoun has quit IRC (Ping timeout: 276 seconds) 2011-01-16T21:38:18 indeed, youare's lack of aesthetic appeal would hinder its acceptance into the english lexicon. I'd happily wager that any such venture would be doomed to failure. I was only considering its hypothetical existence to advance argument against u'r, and you are mistaken if you think I really do think that youare should be regarded as cromulent :P 2011-01-16T21:39:06 are you sure you weren't an english major? 2011-01-16T21:39:51 how do you know I'm not? :P 2011-01-16T21:41:03 *** Accoun has joined #aichallenge 2011-01-16T21:44:29 anyway, I would say the discussion falls more under the scope of linguistics than anything an english major would study 2011-01-16T21:52:41 there is a lot of work in making this snowballs bot 2011-01-16T21:57:20 think of the kids in africa who don't even have snowballs, not get back to work and finish your bot 2011-01-16T21:57:29 *now 2011-01-16T22:03:47 haha 2011-01-16T22:03:51 i don't have snowballs either :( 2011-01-16T22:04:03 does that mean I am comparable to an African kid? 2011-01-16T22:05:06 depending on how you implement comparable 2011-01-16T22:05:22 lack of snowballs gives full equivalence 2011-01-16T22:05:47 i feel their pain 2011-01-16T22:05:50 well, you've answered your own question then 2011-01-16T22:07:29 *** amstan has joined #aichallenge 2011-01-16T22:07:29 *** ChanServ sets mode: +o amstan 2011-01-16T22:32:51 *** Vogons has quit IRC (Ping timeout: 265 seconds) 2011-01-16T22:40:50 *** davidjliu has joined #aichallenge 2011-01-16T22:58:53 *** davidjliu has quit IRC (Quit: Page closed) 2011-01-16T23:13:36 *** dlila has quit IRC (Quit: Leaving) 2011-01-16T23:18:41 *** delt0r_ has joined #aichallenge 2011-01-16T23:20:06 *** delt0r___ has quit IRC (Ping timeout: 240 seconds) 2011-01-16T23:32:36 *** davidjliu has joined #aichallenge 2011-01-16T23:47:02 *** Accoun has quit IRC (Ping timeout: 260 seconds) 2011-01-16T23:59:07 *** tapwater has quit IRC (Quit: tapwater)