2011-04-26T00:00:05 there was a closed beta period before opening to the public 2011-04-26T00:00:13 amstan: I thought about that, but it is only a test database and so I didn't care 2011-04-26T00:00:13 _flag: umm... yes.. 2011-04-26T00:00:30 didn't notice that, i saw the first post on reddit like a week before official openiong 2011-04-26T00:00:32 _flag: can you make another one? and this time fork the repo 2011-04-26T00:00:41 amstan: plus, it's a funny password 2011-04-26T00:01:21 <_flag> amstan: another what exactly? 2011-04-26T00:01:32 _flag: another c starter 2011-04-26T00:01:40 _flag: if you want me to delete this one.. 2011-04-26T00:02:14 <_flag> What I meant is you can just ignore the example bot, the starter pack is fine 2011-04-26T00:03:02 <_flag> I'm talking about this: https://github.com/aichallenge/aichallenge/issues/41 2011-04-26T00:04:52 <_flag> amstan: ping 2011-04-26T00:05:14 oh, ok 2011-04-26T00:05:23 _flag: so i don't have to do anything then? 2011-04-26T00:05:29 _flag: i'm confused 2011-04-26T00:05:51 _flag: keep in mind that we're using an older version of your repo atm, i wasn't aware you made new changes or if i'm supposed to use them 2011-04-26T00:06:17 _flag: oh, there's a new line of input btw :P 2011-04-26T00:06:29 there is now an 'r' line 2011-04-26T00:06:41 the old bots aren't broken, but it'd be cool if the starter bot got updated 2011-04-26T00:06:48 <_flag> amstan: Originally I had a sample bot instead of the starter bot because I wasn't aware the starter bot had a spec 2011-04-26T00:07:07 basically you just stop clearing food information each turn and if you get an 'r' line, remove food information at the specified location 2011-04-26T00:07:08 <_flag> antimatroid: Is it on the wiki? I can do it tommorow 2011-04-26T00:07:12 yep 2011-04-26T00:07:16 no rush :) 2011-04-26T00:07:35 it makes it easy to track information about food out of sight and know the correct value of a visible square 2011-04-26T00:07:57 if one doesn't change the bot it just keeps working as it did 2011-04-26T00:08:06 *** sigh has quit IRC (Ping timeout: 240 seconds) 2011-04-26T00:08:37 *** sigh has joined #aichallenge 2011-04-26T00:10:01 <_flag> antimatroid: okay, I'll do it tomorrow then 2011-04-26T00:10:33 *** sigh has joined #aichallenge 2011-04-26T00:19:43 *** contestbot has joined #aichallenge 2011-04-26T00:19:51 contestbot: rankings 2011-04-26T00:19:52 amstan: Top 10 players: bocsimacko(3765), _iouri_(3565), Slin-.-(3524), _Astek_(3501), jimrogerz(3500), Accoun(3498), george(3494), GreenTea(3489), asavis(3480), bix0r4ever(3476) 2011-04-26T00:23:16 amstan: it takes awhile, it's lots of test data 2011-04-26T00:24:44 McLeopold: it's still doing it.. 2011-04-26T00:24:57 aichallenge: janzert epsilon * r7675414 / (setup/worker_setup.py worker/jail_own.c): Add small program to change ownership of certain jail files between the jail user and contest user - http://bit.ly/esljh5 2011-04-26T00:25:18 and there's still no rankings on the site 2011-04-26T00:25:47 McLeopold: i'm running it with def create_test_data(user_count=5000, map_count=100, game_count=1000000, matchup_count=10): 2011-04-26T00:25:48 if there are any C Unix programmers around, a thorough review of jail_own.c above would be really good 2011-04-26T00:25:49 is that bad? 2011-04-26T00:26:02 since it is the only setuid program of the contest 2011-04-26T00:26:25 janzert: you could try #c if nobody comes up 2011-04-26T00:26:32 i found the ppl there to know their stuff 2011-04-26T00:27:12 yeah, if no one looks at it before then as we get closer that would be a good idea 2011-04-26T00:27:32 I'll hold off for now in case I need further changes in functionality first 2011-04-26T00:28:33 at least all the pieces should now be in place for sandbox.py to do real sandboxing on a worker 2011-04-26T00:28:47 now to just make it actually do it 2011-04-26T00:32:03 janzert: owner_uid = CONTEST_UID 2011-04-26T00:32:15 janzert: is that supposed to be defined in a header file? 2011-04-26T00:33:16 no, it's defined with "gcc -DCONTEST_UID=blah -DCONTEST_GID=blag -DJAIL_GID=blar jail_own.c -o jail_own" currently 2011-04-26T00:34:00 oh, ok 2011-04-26T00:34:03 I thought about a real makefile for it but the python seemed easier :} 2011-04-26T00:34:11 McLeopold: i got bored, changed it to much lower values 2011-04-26T00:34:29 McLeopold: it finished but i still don't see much 2011-04-26T00:34:38 amstan: you only need that for load testing database queries 2011-04-26T00:34:51 use small amounts just for looking at web pages 2011-04-26T00:35:06 no.. what i wanted was a way to see a rankings table 2011-04-26T00:35:18 or is that not finished yet? 2011-04-26T00:35:25 a few users is good enough for that, isn't it? 2011-04-26T00:35:38 by table i meant html table 2011-04-26T00:35:41 no, not finished yet 2011-04-26T00:35:49 ok, lol 2011-04-26T00:35:49 yeah, I know 2011-04-26T00:47:10 *** nullkuhl has joined #aichallenge 2011-04-26T00:47:48 *** nullkuhl has quit IRC (Remote host closed the connection) 2011-04-26T00:48:12 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-04-26T00:49:32 *** nullkuhl has joined #aichallenge 2011-04-26T00:57:26 oh man.. you can type so well with this: http://duiker101.tk/hackertyper/ 2011-04-26T00:59:16 amstan: I just learned c! 2011-04-26T00:59:24 I'm soo good at it 2011-04-26T00:59:43 i just learned to type a character even when i'm releasing a key 2011-04-26T00:59:49 so i can type doubly fast 2011-04-26T01:09:09 *** nullkuhl has quit IRC (Ping timeout: 250 seconds) 2011-04-26T01:11:19 *** amstan_ has joined #aichallenge 2011-04-26T01:11:19 *** ChanServ sets mode: +o amstan_ 2011-04-26T01:13:15 *** nullkuhl has joined #aichallenge 2011-04-26T01:20:40 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-04-26T01:28:01 *** lean has joined #aichallenge 2011-04-26T01:28:18 Hi! Anyone here? 2011-04-26T01:30:34 I would like to sponsor AI projects and research. Contact me at leandror@intraway.com. Will favor latin american projects. 2011-04-26T01:31:13 *** lean has quit IRC (Client Quit) 2011-04-26T01:33:50 random 2011-04-26T01:49:40 *** Accoun has quit IRC (Ping timeout: 240 seconds) 2011-04-26T01:51:37 *** Mathnerd314 has quit IRC (Ping timeout: 250 seconds) 2011-04-26T01:54:26 amstan_, that site does not work so well for me in firefox, when I press a key it types but then it launches the quickfind box so any more typing is captured by that 2011-04-26T01:54:53 quickfind box? 2011-04-26T01:55:30 sigh: wierd, it works fine here on FF 4 2011-04-26T01:55:50 pressing / in a normal page on firefox puts you in the quick find box 2011-04-26T01:55:51 amstan_: I assume what you would get by hitting / 2011-04-26T01:55:57 on that page it does it with any key 2011-04-26T01:56:11 worked for me in firefox 2011-04-26T01:56:13 weird 2011-04-26T01:56:22 works for me in chrome 2011-04-26T01:56:23 odd 2011-04-26T01:56:58 contestbot: later tell j3camero [01:30] I would like to sponsor AI projects and research. Contact me at leandror@intraway.com. Will favor latin american projects. 2011-04-26T01:56:59 amstan_: I come to serve. 2011-04-26T01:58:55 I can still type with Esc, delete, shift, ctrl, alt, return, apple key and the arrow keys 2011-04-26T01:59:55 ahh...if I hold down ctrl or alt I can press the alphanum keys 2011-04-26T02:00:19 ok, that's enough time looking at stupid workarounds 2011-04-26T02:02:25 sigh: are you ready for more changes? 2011-04-26T02:02:50 what changes? 2011-04-26T02:02:56 I probably won't get to them today 2011-04-26T02:03:02 that's fine 2011-04-26T02:03:36 *** jaycobcoleman has joined #aichallenge 2011-04-26T02:03:46 http://pastebin.com/G429rgbe 2011-04-26T02:04:02 sigh: there's a log if you care to slog through our conversation 2011-04-26T02:04:22 sigh: In a nut shell, 1 point per ant spawn in addition to the kill points 2011-04-26T02:04:24 the idea is to give every ant spawned a point 2011-04-26T02:05:39 *crickets* 2011-04-26T02:05:39 the basic reasoning being that an ant that improves their food collection should see a noticeable boost in their rankings (get 'em hooked) 2011-04-26T02:05:54 we've had a fairly lengthy discussion about what it does/n't encourage 2011-04-26T02:06:11 that seem quite easy 2011-04-26T02:06:18 makes the scoring a bit more consistent 2011-04-26T02:06:34 yeah, now there are 2 things you can do to score 2011-04-26T02:06:41 make ants and kill ants 2011-04-26T02:06:48 the best of both worlds 2011-04-26T02:06:58 animatroid: I belive you had a similar idea to me of having a last seen array, correct? :) 2011-04-26T02:07:03 and you can think of it like this, each and is worth 1 point + whatever it makes from damaging dying enemies 2011-04-26T02:07:09 plus, we think there is still enough discouragement of campers and drawn out games 2011-04-26T02:07:29 the question is, does the first ant get a point? 2011-04-26T02:07:32 the poor little bugger 2011-04-26T02:07:35 also, I love this. creating ants should definitely score points 2011-04-26T02:07:40 antimatroid, might as well 2011-04-26T02:07:41 *** Ice_Harley has joined #aichallenge 2011-04-26T02:07:41 why not? 2011-04-26T02:07:45 yep 2011-04-26T02:07:51 make 'em feel good 2011-04-26T02:07:53 so, no one can score 0 2011-04-26T02:07:55 free point 2011-04-26T02:08:08 how about -100 if he dies 2011-04-26T02:08:25 sigh: just going over the seq branch, it looks fine to merge to epsilon at least so we can play with it more and check that sequential running works for us 2011-04-26T02:08:40 The first ant? 2011-04-26T02:08:40 or the bot? 2011-04-26T02:08:53 probably better to get it back into the mainline soon so it doesn't bitrot anymore than it has 2011-04-26T02:08:56 janzert, cool, thanks 2011-04-26T02:09:08 the first ant 2011-04-26T02:09:26 meehhhhh that seems like a bad plan 2011-04-26T02:09:27 btw, I'll be working heavily in sandbox.py starting tomorrow 2011-04-26T02:09:53 oh, it's a *really* bad plan, but would be funny 2011-04-26T02:10:09 haha yes indeed 2011-04-26T02:10:12 oh ok, I'll the merge soon 2011-04-26T02:10:19 you had me worried for a second 2011-04-26T02:10:30 I forget to type smilies :D 2011-04-26T02:10:31 but it was sitting in the back of my mind "is he joking?..." 2011-04-26T02:10:41 :) 2011-04-26T02:10:44 * janzert *still* likes inverse elimination :P 2011-04-26T02:10:56 * McLeopold wonders what janzert means 2011-04-26T02:11:02 * janzert goes afk before anyone can start arguing with him over that 2011-04-26T02:11:07 for game scoring 2011-04-26T02:11:15 * McLeopold still wonders 2011-04-26T02:11:35 was that in the wiki? 2011-04-26T02:11:38 i think inverse elimination with strongly discourage attacking 2011-04-26T02:11:52 first person eliminated is ranked last, last person surviving is ranked first 2011-04-26T02:11:54 which will lead to long drawn out games 2011-04-26T02:11:55 and I'm gone 2011-04-26T02:12:20 play galcon online, then come argue that 2011-04-26T02:12:58 :P 2011-04-26T02:13:04 * McLeopold forgot another smiley 2011-04-26T02:13:30 * antimatroid was recently alerted to the feature where you can strip the smileys and get the original text 2011-04-26T02:13:33 I think the food bonus will be large enough so that last player remaining will win anyway 2011-04-26T02:13:34 you guys must think I'm a totaly asshole half the time if you though -100 was serious? 2011-04-26T02:13:38 it's very nice when maths etc. is given 2011-04-26T02:13:58 McLeopold: who took that seriously? 2011-04-26T02:14:00 in pidgen? 2011-04-26T02:14:02 serially* 2011-04-26T02:14:07 I did, I was about to push the change :( 2011-04-26T02:14:08 yeah, pidgin <3 2011-04-26T02:14:19 me :) 2011-04-26T02:14:21 "disable smileys in selected text" 2011-04-26T02:14:23 I took me 2 month to find the smiley off switch :) 2011-04-26T02:14:34 i was told about it like last week 2011-04-26T02:14:42 wait, change your theme 2011-04-26T02:14:51 I had no smileys to begin with 2011-04-26T02:14:52 perminant off 2011-04-26T02:15:00 oh, i like having them d; 2011-04-26T02:15:15 (>^_^)> 2011-04-26T02:15:25 apparenly backwards smileys makes someone a girl 2011-04-26T02:15:37 so i've been sticking it to the stereotype of late 2011-04-26T02:15:39 just because 2011-04-26T02:15:50 yeah... that'll learn 'em 2011-04-26T02:15:51 d: is my favourite 2011-04-26T02:16:17 *** sigh has quit IRC (Quit: Leaving) 2011-04-26T02:17:25 It's shocking how you guys manage to have all the important discussions as soon as I disappear 2011-04-26T02:17:29 *** sigh has joined #aichallenge 2011-04-26T02:17:33 I must seriously hinder progress around here 2011-04-26T02:21:15 Speaking of. How did you decide points would be distributed in a battle? 2011-04-26T02:21:33 rationally 2011-04-26T02:21:47 lol 2011-04-26T02:22:12 (cause we use the Fraction class, hee hee) 2011-04-26T02:22:38 jaycobcoleman: the scoring method for battles makes perfect sense for the damage option 2011-04-26T02:23:46 1/(number of ants in battle) distributed to every ant in the battle? 2011-04-26T02:23:56 yeah 2011-04-26T02:24:06 okay, maybe not perfectly 2011-04-26T02:24:08 but close 2011-04-26T02:24:10 ants adding damage to a death at to be specific 2011-04-26T02:24:10 including dying ants 2011-04-26T02:24:25 *dead ant* 2011-04-26T02:24:39 yeah, even colliding ants get points 2011-04-26T02:24:43 and colliding ants give out points 2011-04-26T02:24:46 do they? 2011-04-26T02:24:49 yes 2011-04-26T02:24:56 if an enemy is in range 2011-04-26T02:24:57 otherwise if two enemies collided no one got points 2011-04-26T02:24:59 yep 2011-04-26T02:25:23 sigh: is that right? collided ants get points from dying battle ants and vice versa? 2011-04-26T02:25:34 yeah 2011-04-26T02:25:41 i like it like that 2011-04-26T02:26:13 and you can't collide your ants in battle range of an enemy to stop them getting points 2011-04-26T02:26:17 although you can outside range 2011-04-26T02:26:25 say you have 2 ants surrounded 2011-04-26T02:28:50 aichallenge: sigh epsilon * rf889d72 / (worker/engine.py worker/sandbox.py): Merge branch 'seq' into epsilon - http://bit.ly/fw6Z8g 2011-04-26T02:28:51 aichallenge: sigh epsilon * rc350f26 / worker/engine.py : Fix up broken merge with seq branch - http://bit.ly/eRQ4tu 2011-04-26T02:29:42 https://github.com/aichallenge/aichallenge/wiki/Ants-Game-Specification 2011-04-26T02:29:54 McLeopold: want to check the scoring section of that for me? :P 2011-04-26T02:30:17 it's right at the bottom 2011-04-26T02:30:57 why does every text block start with a space? 2011-04-26T02:31:03 it's probably perfect 2011-04-26T02:31:54 jaycobcoleman: see my comments to amstan before 2011-04-26T02:32:01 the good news is they aren't there when he converts them 2011-04-26T02:32:28 http://hypertriangle.com:13080/specification.php 2011-04-26T02:32:31 see there 2011-04-26T02:32:33 alright, I figured you'd have it covered :P 2011-04-26T02:34:52 i'm still not sure about that blue 2011-04-26T02:35:02 i wonder if a dirt coloured brown would look nicer 2011-04-26T02:35:26 yes, brown would look nicer 2011-04-26T02:36:03 animatroid: I was talking to amstan about that earier :P 2011-04-26T02:36:26 would making a dirt image from scratch be the best option? 2011-04-26T02:37:28 use repeating layered backgrounds 2011-04-26T02:38:31 i like whoever had the idea of making the background reflect different maps 2011-04-26T02:39:28 oh yes I remember that 2011-04-26T02:40:03 But to like a person for a single idea seems pretty shallow :P 2011-04-26T02:41:07 i realised that mishap 2011-04-26T02:41:10 i meant i liked the idea 2011-04-26T02:41:16 but i like the person anyway, shallow or not 2011-04-26T02:42:59 :) 2011-04-26T02:44:04 *** sigh has quit IRC (Ping timeout: 260 seconds) 2011-04-26T02:44:49 *** sigh has joined #aichallenge 2011-04-26T02:59:43 *** aerique has joined #aichallenge 2011-04-26T03:13:30 lol lame 2011-04-26T03:13:45 ? 2011-04-26T03:13:46 all four of the gambit projects are for doing representations of games etc. 2011-04-26T03:13:52 no fun algorithms projects 2011-04-26T03:14:11 so is your rejection good? 2011-04-26T03:14:38 i'm sort of glad i didn't get it, but $5k would have been nice 2011-04-26T03:15:00 *** sigh has quit IRC (Ping timeout: 260 seconds) 2011-04-26T03:15:11 but now i get to keep playing around with cooperation for my thesis instead of writing some function for a horribly written library 2011-04-26T03:15:47 *** sigh has joined #aichallenge 2011-04-26T03:17:10 Well I made an image of (supposedly) dirt from scratch. And in the process discovered that I'm terrible at editing images :P 2011-04-26T03:20:09 look at that baby: http://imgur.com/HpOp6 2011-04-26T03:20:18 I'm not sure I even know what dirt looks like 2011-04-26T03:20:47 green? 2011-04-26T03:20:58 lol 2011-04-26T03:21:03 I don't think you do :p 2011-04-26T03:21:05 i was thinking more a brown shade 2011-04-26T03:21:19 evidently I'm convinced it's green 2011-04-26T03:21:24 haha so was I 2011-04-26T03:21:35 I'm colorblind leave me alone! 2011-04-26T03:21:42 haha 2011-04-26T03:21:56 maybe you should be doing the graphics then ;p 2011-04-26T03:21:56 :P 2011-04-26T03:22:03 shouldn't* :P 2011-04-26T03:22:21 probably a good idea :) 2011-04-26T03:23:15 Would you want a texture like that though or a more realistic, inconsistent one? 2011-04-26T03:23:26 or maps? 2011-04-26T03:23:26 or blue 2011-04-26T03:23:37 everything's on the table 2011-04-26T03:30:09 *** _0x47 <_0x47!~slash@p5B25ED8F.dip.t-dialin.net> has joined #aichallenge 2011-04-26T03:32:03 http://imgur.com/rAVZq 2011-04-26T03:32:09 Is that one any more brown? 2011-04-26T03:32:18 way better 2011-04-26T03:32:39 couldn't you take the average red and green and just make sure they are similar? 2011-04-26T03:32:52 yeah, that is much more dirtish 2011-04-26T03:33:39 good. Dirtish is what I was going for 2011-04-26T03:34:26 and McLeopold, don't ask me 2011-04-26T03:34:34 I know literally nothing about colors 2011-04-26T03:34:37 about what? 2011-04-26T03:34:53 avg. red/green 2011-04-26T03:34:56 oh, a paint program should be able to tell you the values 2011-04-26T03:34:56 colour! 2011-04-26T03:35:05 haha 2011-04-26T03:35:08 America 2011-04-26T03:35:19 Am'r'ca! 2011-04-26T03:35:26 you guys don't use teh metric system. 2011-04-26T03:35:33 I do :P 2011-04-26T03:35:33 so, the aussies sided with the british for spelling? 2011-04-26T03:35:43 McLeopold: yep 2011-04-26T03:35:47 I use metric just to rebel 2011-04-26T03:36:01 Nobody every knows what I'm talking about 2011-04-26T03:36:14 that doesn't make sense 2011-04-26T03:36:23 *ever 2011-04-26T03:36:23 ^^ i like how unreadable that sentence is given what it's saying 2011-04-26T03:37:01 oh irony... 2011-04-26T03:37:14 I'm an idiot sometimes 2011-04-26T03:38:07 or my second finger was feeling bored and decided to push extra buttons 2011-04-26T03:39:45 *** woudshoo` has joined #aichallenge 2011-04-26T03:40:18 *** McLeopold has quit IRC (Ping timeout: 240 seconds) 2011-04-26T03:41:07 *** McLeopold has joined #aichallenge 2011-04-26T03:44:55 *** Accoun has joined #aichallenge 2011-04-26T03:44:57 animatroid: Well, brand new git clone. I think it's time to play with the C++ bot 2011-04-26T03:49:25 we should get some sample c++ bots out 2011-04-26T03:49:43 *** janzert has left #aichallenge 2011-04-26T03:50:17 *** janzert has joined #aichallenge 2011-04-26T04:00:35 amstan_: we have to disable fullscreen mode on the front page. People who set fullscreen as the default wont see the front page otherwise. 2011-04-26T04:00:51 *** sigh has quit IRC (Ping timeout: 240 seconds) 2011-04-26T04:02:24 *** Frontier has quit IRC (Quit: Leaving.) 2011-04-26T04:03:53 *** amstan_ has quit IRC (Ping timeout: 252 seconds) 2011-04-26T04:07:17 *** Frontier has joined #aichallenge 2011-04-26T04:11:00 *** sigh has joined #aichallenge 2011-04-26T04:11:34 *** kaemo0 has quit IRC (Ping timeout: 246 seconds) 2011-04-26T04:12:20 *** kaemo has quit IRC (Ping timeout: 258 seconds) 2011-04-26T04:14:31 *** boegel has joined #aichallenge 2011-04-26T04:16:20 sigh: Do you know any c++? 2011-04-26T04:18:40 *** Blkt has joined #aichallenge 2011-04-26T04:19:38 he does 2011-04-26T04:19:44 he's one of my goto people for help 2011-04-26T04:19:47 :p 2011-04-26T04:20:23 I think a problem he brought up with the previous python code would also apply to your c++ vision code 2011-04-26T04:20:27 this one to be exact: http://pastebin.com/gscdvw3u 2011-04-26T04:20:47 *previous python vision code that is 2011-04-26T04:21:49 there's got to be a way to pick the correct ant when you merge 2011-04-26T04:22:16 i don't think so 2011-04-26T04:22:24 https://github.com/aichallenge/aichallenge/blob/epsilon/ants/dist/starter_bots/cpp/State.cc 2011-04-26T04:22:28 line 71 on has the vision code 2011-04-26T04:22:35 it uses a new visited vector for each ant 2011-04-26T04:23:46 that it does :) 2011-04-26T04:24:02 although that does seem terribly inefficient... 2011-04-26T04:25:07 *** McLeopold has quit IRC (Read error: Connection reset by peer) 2011-04-26T04:26:49 If my memory of c++ will cooperate with me then I think I'll write a rendition of sigh's vision_by_ant for c++ 2011-04-26T04:27:52 it's a starter bot 2011-04-26T04:28:02 vision isn't terribly expensive anyway 2011-04-26T04:28:06 you only need to run it once per turn 2011-04-26T04:29:09 a harder problem is finding the set of ants that might affect one of your ants with a battle 2011-04-26T04:29:14 maybe I'm just getting too used to python where everything is expensive :P 2011-04-26T04:29:23 a%%b, a%.b, a.%b begins to highlight my problem 2011-04-26T04:29:58 and of course one would like it to be general for different battle radiuses 2011-04-26T04:30:08 good day everyone 2011-04-26T04:30:11 you mean predicting whether you'll become locked in a battle the next turn? 2011-04-26T04:30:23 whether it's possible for the ants to land in a battle 2011-04-26T04:30:35 and hence whether they should be considered in the same "game" 2011-04-26T04:30:47 I see 2011-04-26T04:31:05 *** woudshoo` has quit IRC (Ping timeout: 250 seconds) 2011-04-26T04:33:50 i think the best might be to find the subset of 5 squares you might be able to land in 2011-04-26T04:34:04 then search the extended set to see if any ants can move into a square within range of one of those 2011-04-26T04:34:41 another thing would be a.a..b 2011-04-26T04:34:45 you don't need to consider the first a there 2011-04-26T04:35:43 *** sids_aquarius has joined #aichallenge 2011-04-26T04:35:53 could you just test the outermost squares of your current battle radius and see if any enemies are within 2 squares of each of them? 2011-04-26T04:36:04 or within of course 2011-04-26T04:36:24 but what about a%.b 2011-04-26T04:36:29 i mean a%%b 2011-04-26T04:36:40 you don't want to put those two in the same game 2011-04-26T04:37:14 putting more players than necessary in a game is about the most costly thing you can do 2011-04-26T04:37:31 and not good strategy 2011-04-26T04:39:53 okay well for each ant create an array of the distances to each square and if any enemy ants are within your battle radius plus 2 and the distance they need to travel to get to you is consistent with that value then they're a part of the game 2011-04-26T04:40:27 "for each ant create an array of the distances to each square" 2011-04-26T04:40:28 lol what 2011-04-26T04:40:32 that would be very expensive 2011-04-26T04:40:43 you aren't storing it 2011-04-26T04:40:56 it isn't so terrible 2011-04-26T04:41:08 and it's not for the whole map 2011-04-26T04:41:09 it is when i want it to be quite fast 2011-04-26T04:41:16 you really only need to do it for your battle radius plus 2 2011-04-26T04:41:22 which would be quite fast 2011-04-26T04:46:00 another option would be to check each square on a direct path from you to an enemy ant to see if it is water. if so +2 to the total distance to compenstate, then check all around for more water, adding to compensate each time, until you run out of water in one direction or reach BR+2 2011-04-26T04:46:30 although that's not entirely accurate... 2011-04-26T04:46:53 *** Naktibalda has joined #aichallenge 2011-04-26T04:47:14 *** _0x471 <_0x471!~slash@p3EE0B00C.dip.t-dialin.net> has joined #aichallenge 2011-04-26T04:47:21 Theres definitely a way to determine what how to compensate for various positions of water relative to your ant 2011-04-26T04:47:26 i need it to be accurate and to work for any battle radius 2011-04-26T04:47:43 If implemented corrctly that'd do it 2011-04-26T04:47:48 *correctly 2011-04-26T04:48:09 But I can't wrap my head around the compensation at the moment 2011-04-26T04:48:20 *** _0x47 <_0x47!~slash@p5B25ED8F.dip.t-dialin.net> has quit IRC (Ping timeout: 276 seconds) 2011-04-26T04:49:59 another example is 2011-04-26T04:49:59 a%% 2011-04-26T04:50:00 %%b 2011-04-26T04:50:07 yo don't want them in the same game 2011-04-26T04:55:59 *** _0x47 <_0x47!~slash@p3EE0BDC7.dip.t-dialin.net> has joined #aichallenge 2011-04-26T04:56:33 The idea of following the water to its ends or until the value is greater than what would institute a possible battle cenario is to rule out battles where travel distance becomes greater than actual distance 2011-04-26T04:56:38 *scenario 2011-04-26T04:57:01 *** woudshoo` has joined #aichallenge 2011-04-26T04:57:07 not really what I meant, but okay 2011-04-26T04:57:38 when actual travel distance is greater than what would allow a battle 2011-04-26T04:57:57 *** _0x471 <_0x471!~slash@p3EE0B00C.dip.t-dialin.net> has quit IRC (Ping timeout: 250 seconds) 2011-04-26T04:59:50 *** mceier has joined #aichallenge 2011-04-26T05:10:54 i wish we hadn't used the euclidean metric 2011-04-26T05:10:58 it makes life tough sometimes 2011-04-26T05:12:47 makes it damn tough 2011-04-26T05:12:54 but much more realistic 2011-04-26T05:18:14 that could be fixed by adding diagonal movement. and the fact that you'd actually be moving much faster going diagonally could be accounted for by making twice as high and wide and making movement in a single direction worth two squares whereas diagonal movement would be 1x1 2011-04-26T05:20:11 *** woudshoo` has quit IRC (Ping timeout: 276 seconds) 2011-04-26T05:21:32 *** dr-` has joined #aichallenge 2011-04-26T05:21:41 *** dr- has quit IRC (Ping timeout: 276 seconds) 2011-04-26T05:23:20 *** _0x471 <_0x471!~slash@p3EE0ACED.dip.t-dialin.net> has joined #aichallenge 2011-04-26T05:23:33 *** _0x47 <_0x47!~slash@p3EE0BDC7.dip.t-dialin.net> has quit IRC (Ping timeout: 240 seconds) 2011-04-26T05:25:57 There's about a 50% chance that last, surely muddled thought was coherent. If not then I should probably be sleeping 2011-04-26T05:27:58 *** jaycobcoleman has quit IRC (Quit: Page closed) 2011-04-26T05:28:21 *** jaycobcoleman has joined #aichallenge 2011-04-26T05:37:36 *** mcstar has joined #aichallenge 2011-04-26T05:38:31 *** woudshoo` has joined #aichallenge 2011-04-26T05:39:17 *** jaycobcoleman has quit IRC (Quit: Page closed) 2011-04-26T06:04:22 *** Eruonen has joined #aichallenge 2011-04-26T06:07:45 *** nullkuhl has quit IRC (Remote host closed the connection) 2011-04-26T06:08:36 *** Eruonen has quit IRC (Ping timeout: 252 seconds) 2011-04-26T06:12:50 *** Eruonen has joined #aichallenge 2011-04-26T06:16:40 *** needsch has joined #aichallenge 2011-04-26T06:31:59 *** woudshoo` has quit IRC (Ping timeout: 250 seconds) 2011-04-26T06:38:38 *** antimatroid has quit IRC (Ping timeout: 240 seconds) 2011-04-26T06:44:30 *** sigh has quit IRC (Ping timeout: 240 seconds) 2011-04-26T06:45:55 *** sigh has joined #aichallenge 2011-04-26T06:49:14 *** Frontier1 has joined #aichallenge 2011-04-26T06:50:06 *** Frontier has quit IRC (Ping timeout: 240 seconds) 2011-04-26T06:55:24 *** antimatroid has joined #aichallenge 2011-04-26T06:55:43 *** delt0r___ has joined #aichallenge 2011-04-26T06:57:07 *** delt0r_ has quit IRC (Ping timeout: 250 seconds) 2011-04-26T06:59:05 *** woudshoo` has joined #aichallenge 2011-04-26T07:09:43 *** Ice_Harley has quit IRC (Ping timeout: 252 seconds) 2011-04-26T07:15:19 *** woudshoo` has quit IRC (Ping timeout: 246 seconds) 2011-04-26T07:29:04 *** sigh has quit IRC (Ping timeout: 248 seconds) 2011-04-26T07:30:14 *** sigh has joined #aichallenge 2011-04-26T07:38:00 lol i just made a QMatrix 2011-04-26T07:38:08