2010-11-09T00:00:12 nope just locks again 2010-11-09T00:00:25 :( 2010-11-09T00:01:41 I gotta go-- I will think about this-- trying things out here I think you may indeed be reading before it's ready and getting stuck or something 2010-11-09T00:01:48 *** sigh has quit IRC (Ping timeout: 272 seconds) 2010-11-09T00:01:51 yea 2010-11-09T00:02:00 technically it should *block* the caller 2010-11-09T00:02:16 but i came across the same issue - my fixed my problem though 2010-11-09T00:02:19 mine* 2010-11-09T00:02:24 that's what i thought but i am totally not getting blocked it just hiccups and dies 2010-11-09T00:03:06 you can even see stderr gets jacked 2010-11-09T00:03:10 it doesn't finish HI lol 2010-11-09T00:03:25 you can fflush stderr? 2010-11-09T00:03:42 well that's not my issue 2010-11-09T00:03:48 fflush(stdin) just locks it too 2010-11-09T00:04:01 how did you have it working before? 2010-11-09T00:04:07 i just tried fgets with a fprintf after it it didn't pritn anything 2010-11-09T00:04:14 i haven't ever used stdin/out - pure socket code to tcp 2010-11-09T00:04:16 demerzel_: can you release just your framework in c so I can try and test it? 2010-11-09T00:04:28 just a bot that does nothing 2010-11-09T00:04:33 yea sure 2010-11-09T00:06:42 McLeopold: one night i made changes, incrementally testing on an old bot the whole time 2010-11-09T00:06:47 i broke so many things 2010-11-09T00:06:54 :) 2010-11-09T00:07:04 I'm sooo glad I've got scm. 2010-11-09T00:07:13 heh 2010-11-09T00:07:23 I would have lost some crucial pieces. 2010-11-09T00:07:25 my backups are just a big folder of zips aha 2010-11-09T00:07:35 I forgot about the whole idea I just restored. 2010-11-09T00:08:05 i just reached freedom from exams a few hours ago, i will be creating a new bot soon enough 2010-11-09T00:08:28 I also use eclipse with svn, so it is super easy to compare and merge. 2010-11-09T00:09:32 help me solve this one: http://72.44.46.68/canvas?game_id=917105 2010-11-09T00:10:42 you expanded a second time 2010-11-09T00:10:44 you shouldn't have 2010-11-09T00:11:01 *** davidjliu has joined #aichallenge 2010-11-09T00:11:03 yeah, but the real question is why did I think it was okay? 2010-11-09T00:11:38 *** davidjliu has quit IRC (Client Quit) 2010-11-09T00:11:40 I would have one it 2 or 3 turns if I had waited. :( 2010-11-09T00:11:48 *** davidjliu has joined #aichallenge 2010-11-09T00:12:04 i already had it pretty well isolated so i could plug in tcp 2010-11-09T00:12:05 http://www.mediafire.com/?6g75858qr4g16v7 2010-11-09T00:12:21 sorry if there's some undefined structs or something just hacked it open with notepad++ 2010-11-09T00:13:15 anyway if you find something email me themendios@gmail.com i need to sleep before i hang myself 2010-11-09T00:13:28 thanks for all your help guys :) it is bizarre 2010-11-09T00:16:01 Top 10 players: bocsimacko(4228), bix0r4ever(4060), george(4039), shangas(4013), rsergio(4003), GreenTea(3989), Accoun(3964), luksian(3960), Xrillo(3953), dmj111(3937) 2010-11-09T00:18:10 *** demerzel_ has quit IRC (Ping timeout: 265 seconds) 2010-11-09T00:22:07 Mcleopold: are you looking at what's going to happen once you take a planet? 2010-11-09T00:22:19 it seems as though your bot just saw that it could take it 2010-11-09T00:22:27 but didn't bother to check it could sustain that claim 2010-11-09T00:22:44 it's things like that which have left me doing moves to a planet at a time into the futuer 2010-11-09T00:22:44 yep 2010-11-09T00:23:25 or, more critically, it didn't notice that it couldn't sustain the first planet. 2010-11-09T00:23:25 not at a time*, i just make moves into the future assuming i need to counter them, but only send fleets that i have to at time 0 2010-11-09T00:23:47 why not? 2010-11-09T00:24:23 well, I should have left a few planets back just incase the 5 on the enemy home planet got sent, which they did. 2010-11-09T00:24:40 which means I didn't have the 18 to take the second planet 2010-11-09T00:25:11 oh, i see what you mean, i thought you meant when you claimed it 2010-11-09T00:25:43 i wonder if it's worth waiting on expanding at all if you're ahead 2010-11-09T00:25:54 absolutely 2010-11-09T00:26:04 i think it'd be fairly rare where that might hurt you, and even rarer that the other bot exploited that 2010-11-09T00:26:05 let me find an example... 2010-11-09T00:28:26 heh, my new bot seems to like the defensive strategy of allowing the enemy to capture my planet for one turn... 2010-11-09T00:28:30 http://72.44.46.68/canvas?game_id=918941 2010-11-09T00:28:41 http://72.44.46.68/canvas?game_id=916927 2010-11-09T00:29:05 *** virdo has quit IRC (Ping timeout: 265 seconds) 2010-11-09T00:29:29 I wonder if that helps? 2010-11-09T00:29:38 is that a different game? 2010-11-09T00:30:22 better example: http://72.44.46.68/canvas?game_id=917329 2010-11-09T00:31:19 hmmm okay, more interesting would be examples where a bot waited while ahead and shouldn't have 2010-11-09T00:31:29 okay 2010-11-09T00:31:30 and even better the other bot exploited the opportunity 2010-11-09T00:33:17 http://72.44.46.68/canvas?game_id=916956 2010-11-09T00:34:38 what am i looking for? 2010-11-09T00:35:10 cccf waited to expand and shouldn't have, but wasn't ahead 2010-11-09T00:36:19 yeah, i'm more curious about how unlikely it is that you should expand even though you're ahead to avoid a certain loss 2010-11-09T00:36:52 http://72.44.46.68/canvas?game_id=917370 2010-11-09T00:36:55 better? 2010-11-09T00:37:35 hmmm yeah i guess 2010-11-09T00:37:44 *** amstan_ is now known as amstan 2010-11-09T00:37:46 he didn't really have ships left to expand, but he also didn't try to leave any to do so 2010-11-09T00:42:49 *** davidjliu has quit IRC (Ping timeout: 265 seconds) 2010-11-09T00:49:04 *** phxtrack has quit IRC (Read error: Connection reset by peer) 2010-11-09T00:50:37 *** phxtrack has joined #aichallenge 2010-11-09T00:52:07 *** McLeopold has quit IRC (Quit: Leaving) 2010-11-09T00:52:07 *** bhasker has quit IRC (Read error: Connection reset by peer) 2010-11-09T00:52:12 *** bhasker has joined #aichallenge 2010-11-09T00:53:15 *** sigh has joined #aichallenge 2010-11-09T00:55:40 *** Appleman1234 has quit IRC (Remote host closed the connection) 2010-11-09T01:00:10 *** xxxss has joined #aichallenge 2010-11-09T01:06:57 *** sigh has quit IRC (Read error: Connection reset by peer) 2010-11-09T01:07:08 i am giving on my more accurate game state based bot 2010-11-09T01:07:15 *** sigh has joined #aichallenge 2010-11-09T01:07:19 the more accurate i make it , the more idiotic its getting 2010-11-09T01:07:38 *** tapwater has quit IRC (Quit: tapwater) 2010-11-09T01:08:29 *** bhasker has quit IRC (Quit: bhasker) 2010-11-09T01:10:20 i love source code formatters 2010-11-09T01:13:43 what's with the tokenizing crap when processing gamestate? 2010-11-09T01:13:55 would it be more efficient for me to just use a string stream? 2010-11-09T01:14:16 it'd be a hell of a lot neater 2010-11-09T01:16:01 Top 10 players: bocsimacko(4214), bix0r4ever(4048), george(4028), rsergio(3996), shangas(3976), GreenTea(3966), luksian(3957), Accoun(3946), Xrillo(3940), dmj111(3921) 2010-11-09T01:19:35 *** boegel has joined #aichallenge 2010-11-09T01:30:11 *** boegel has quit IRC (Ping timeout: 252 seconds) 2010-11-09T01:30:32 *** sigh has quit IRC (Remote host closed the connection) 2010-11-09T01:32:41 *** Utkarsh has quit IRC (Read error: Connection reset by peer) 2010-11-09T01:36:12 *** ratatata has quit IRC (Quit: Page closed) 2010-11-09T01:39:59 *** Utkarsh has joined #aichallenge 2010-11-09T01:42:51 *** xxxss has quit IRC (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org) 2010-11-09T01:46:45 *** xxxss has joined #aichallenge 2010-11-09T01:57:41 i forgot how painful it was to rewrite from scratch 2010-11-09T01:57:49 this is going to take a while 2010-11-09T02:05:55 *** Ice_Harley has joined #aichallenge 2010-11-09T02:06:00 *** hellman has quit IRC (Ping timeout: 276 seconds) 2010-11-09T02:08:17 *** Utkarsh has quit IRC (Ping timeout: 250 seconds) 2010-11-09T02:10:09 *** hellman has joined #aichallenge 2010-11-09T02:13:38 *** Utkarsh has joined #aichallenge 2010-11-09T02:16:01 Top 10 players: bocsimacko(4217), george(4051), bix0r4ever(4037), rsergio(3994), shangas(3975), GreenTea(3974), Xrillo(3946), Accoun(3931), dmj111(3918), medrimonia(3916) 2010-11-09T02:21:53 *** yasith has quit IRC (Ping timeout: 240 seconds) 2010-11-09T02:26:41 *** hellman has quit IRC (Remote host closed the connection) 2010-11-09T02:32:45 *** sigh has joined #aichallenge 2010-11-09T02:33:32 *** yasith has joined #aichallenge 2010-11-09T02:33:32 woot... back in the top 50 2010-11-09T02:34:40 sigh: :) 2010-11-09T02:34:55 i'm in the middle of rewriting the starter package from scratch again 2010-11-09T02:35:01 yeah, how is that coming along? 2010-11-09T02:35:05 slowly 2010-11-09T02:35:14 i'm rewriting the whole token thing 2010-11-09T02:35:16 that shit is horrible 2010-11-09T02:35:19 i left that there last time 2010-11-09T02:35:22 token? 2010-11-09T02:35:28 tokenize etc. 2010-11-09T02:35:55 what for? 2010-11-09T02:36:37 when passing teh game state string 2010-11-09T02:36:53 although my life was just made harder by the fact that they have comments in the strings 2010-11-09T02:37:55 do you know if i'm guaranteed at least that a new fleet/planet will start on a new line? 2010-11-09T02:38:01 i can work with that 2010-11-09T02:38:20 I think so 2010-11-09T02:38:35 I still don't understand why you need to rewrite the basic protocol code 2010-11-09T02:38:41 cause it's rank 2010-11-09T02:38:45 :P 2010-11-09T02:38:47 haha 2010-11-09T02:38:54 btw, how was your exam? 2010-11-09T02:39:05 surprisingly good, seeing as i didn't sleep last night 2010-11-09T02:39:17 :P 2010-11-09T02:39:25 i've been too slack studying for these exams, but i think i pulled through alright 2010-11-09T02:39:34 wont be quite up to the normal standards, but meh 2010-11-09T02:39:55 heh 2010-11-09T02:41:17 I remember that feeling :) 2010-11-09T02:41:23 stringstream ss (stringstream::in | stringstream::out); 2010-11-09T02:41:27 i always have to look up that line 2010-11-09T02:51:53 *** Appleman1234 has joined #aichallenge 2010-11-09T02:53:46 i sort of wish the number of planets was specified before actually having to pass the first game state 2010-11-09T02:54:06 i do like that about codejam problems 2010-11-09T02:56:27 sigh: i almost declared x and y as ints again :P 2010-11-09T02:57:23 :P 2010-11-09T02:59:38 *** sigh has quit IRC (Remote host closed the connection) 2010-11-09T02:59:52 *** Utkarsh has quit IRC (Read error: Connection reset by peer) 2010-11-09T03:06:14 huzzah i think it worked 2010-11-09T03:07:12 *** Utkarsh has joined #aichallenge 2010-11-09T03:08:31 *** perror has joined #aichallenge 2010-11-09T03:09:13 without adding changes specific to my bot, how much nicer is this 2010-11-09T03:09:13 http://codepad.org/btTfDUqg 2010-11-09T03:09:54 compared to http://codepad.org/mzwJs1vr 2010-11-09T03:12:30 sigh: would that make it clearer why i wanted to change it? :P 2010-11-09T03:13:01 *** Utkarsh has quit IRC (Read error: Connection reset by peer) 2010-11-09T03:13:20 *** Utkarsh has joined #aichallenge 2010-11-09T03:15:15 which is better? I don't understand 2010-11-09T03:16:03 Top 10 players: bocsimacko(4222), george(4056), bix0r4ever(4044), rsergio(3992), shangas(3988), GreenTea(3968), Xrillo(3951), Accoun(3934), medrimonia(3930), Raschi(3923) 2010-11-09T03:16:13 the second one is longer, messier, unreadable, and does a hell of a lot of operations on strings 2010-11-09T03:16:24 although the operations don't really cost much, i much prefer my new one 2010-11-09T03:19:50 antimatroid1: +1 2010-11-09T03:19:55 that looks like my original c++ one 2010-11-09T03:20:25 *** Utkarsh has quit IRC (Ping timeout: 265 seconds) 2010-11-09T03:20:51 antimatroid1: yea it would make it a lot easier on allocation 2010-11-09T03:21:49 how i handle it is that i run two passes 2010-11-09T03:22:07 two passes for what? 2010-11-09T03:22:09 one to parse the data into a fixed size array and then i allocate all my planets etc. 2010-11-09T03:22:16 i'm about to fix it up so i'm not recreating objects every move 2010-11-09T03:22:20 and i don't really store fleets like that 2010-11-09T03:22:25 i'm in the middle of doing a full rewrite 2010-11-09T03:22:31 ah ok ^^ 2010-11-09T03:22:46 my code wouldn't need a rewrite at all (but i design things like that) 2010-11-09T03:22:50 my other code has been going since the moment the contest opened 2010-11-09T03:22:54 so is rather messy 2010-11-09T03:23:18 and i've been nonstop just trying strategies all over the place 2010-11-09T03:23:22 thats why it takes me too long... i think of all the potential possibilities and make my base code as generic as possible 2010-11-09T03:23:23 except the last week cause of exams D: 2010-11-09T03:23:43 ^^ 2010-11-09T03:23:50 my other code has stuff for decision trees, game trees (with minimax) and zero sum games 2010-11-09T03:23:51 do you think you did well in the exams? 2010-11-09T03:24:08 not particularly well, i have spent too much time on ai this semester 2010-11-09T03:24:15 but don't really care, undergrad is now finished 2010-11-09T03:24:37 :P ^^ 2010-11-09T03:24:47 should still get mostly hds and dn's so i can live with that 2010-11-09T03:24:58 :D 2010-11-09T03:25:13 thats good , i assume hd = high distinctions and dn = distinctions 2010-11-09T03:25:41 not sure i'll pull that off with graphics, i didn't go to class since week one, did each of the two assignments the day they were due, studied for the exam on the day and then when i went to clean out that section of my folder, all i had was the unit outline :P 2010-11-09T03:25:45 yep 2010-11-09T03:26:16 *** Utkarsh has joined #aichallenge 2010-11-09T03:27:05 lol 2010-11-09T03:27:53 i work my arse off on my uni work (a good thing too) 2010-11-09T03:27:54 lol 2010-11-09T03:28:35 i have a couple of other projects that need attention from me :/ 2010-11-09T03:28:53 haven't had time because when i've been coding - i've been coding this instead 2010-11-09T03:29:08 yeah, i normally work my ass off 2010-11-09T03:29:16 my other 4 units were all maths and got more attention 2010-11-09T03:29:24 but i've been distracted with ai :P 2010-11-09T03:29:28 lol 2010-11-09T03:29:40 thankfully this contest only runs till 29th 2010-11-09T03:29:47 i pulled an all nighter last night so i didn't fail pdes 2010-11-09T03:29:49 my exams are in january 2010-11-09T03:30:11 i have nothing to do till the end of the contest but write my bot though :P 2010-11-09T03:30:18 :) 2010-11-09T03:31:36 oh god, even the main method is painful 2010-11-09T03:31:39 time to rewrite that 2010-11-09T03:32:47 *** aerique has joined #aichallenge 2010-11-09T03:39:47 xD 2010-11-09T03:39:54 damn it i need to get a good bot working 2010-11-09T03:40:01 then i can leave it for a week and get some other work done 2010-11-09T03:40:23 lol, that's what i kept telling myself 2010-11-09T03:40:36 i'm going to take my time with this one 2010-11-09T03:40:40 * raedwulf is naming various strategies after characters in norse mythology 2010-11-09T03:40:55 so far, Odin and Thialfi 2010-11-09T03:41:03 i was originally going to have a robot unicorn theme with my code, i'm so glad i abandoned that early 2010-11-09T03:41:09 haha 2010-11-09T03:41:22 my username on the main site is "youwillfail" though 2010-11-09T03:41:26 but no bot is up there atm 2010-11-09T03:42:16 ah ok 2010-11-09T03:42:26 my rage bot is ranked about 2300 2010-11-09T03:43:15 plenty of room for improvement :P 2010-11-09T03:43:19 :) 2010-11-09T03:43:21 yep 2010-11-09T03:43:32 i'm chasing top 10 2010-11-09T03:43:41 but my heuristic bot doesn't have a defense strategy and my other bot doesn't have an offensive strategy 2010-11-09T03:43:42 lol 2010-11-09T03:43:45 i had it robbed from me in tron 2010-11-09T03:43:54 so by themselves, they are pretty useless 2010-11-09T03:52:04 i've been awake for 38 hours :|, sleep is going to be really good tonight 2010-11-09T03:53:51 :) 2010-11-09T03:55:25 Yeah! I am first time in top100! 2010-11-09T03:58:24 congrats 2010-11-09T03:58:37 *** narnach has joined #aichallenge 2010-11-09T03:58:37 *** narnach is now known as narnach_mb 2010-11-09T03:59:24 congrats! 2010-11-09T03:59:34 *** narnach_mb has quit IRC (Client Quit) 2010-11-09T03:59:43 *** narnach has joined #aichallenge 2010-11-09T03:59:44 *** narnach is now known as narnach_mb 2010-11-09T04:00:13 The question is how long i will be there ) 2010-11-09T04:02:54 HOLY COW that was a huge bug 2010-11-09T04:03:06 lol 2010-11-09T04:03:09 my eval function was always evaluating turn 60 2010-11-09T04:03:14 even if that was in the past 2010-11-09T04:03:21 now you've squashed it - there's bug gooo everywhere 2010-11-09T04:03:36 O.o 2010-11-09T04:03:39 :P 2010-11-09T04:05:20 *** Utkarsh has quit IRC (Read error: Connection reset by peer) 2010-11-09T04:06:14 lavalamp_: :P 2010-11-09T04:07:41 *** sigh has joined #aichallenge 2010-11-09T04:13:29 *** Utkarsh has joined #aichallenge 2010-11-09T04:13:31 *** phreeza has joined #aichallenge 2010-11-09T04:16:03 Top 10 players: bocsimacko(4244), george(4076), bix0r4ever(4054), rsergio(4015), shangas(4014), GreenTea(3986), Xrillo(3962), Accoun(3953), Raschi(3944), medrimonia(3941) 2010-11-09T04:28:21 http://www.daniweb.com/forums/thread262589.html 2010-11-09T04:28:24 dudes, remember this 2010-11-09T04:29:30 damn 2010-11-09T04:29:48 this completely changes my perception of C 2010-11-09T04:29:54 no! no function calls for you! 2010-11-09T04:29:55 the implications... 2010-11-09T04:30:04 awesome :D 2010-11-09T04:31:26 *** Ice_Harley has quit IRC (Ping timeout: 265 seconds) 2010-11-09T04:31:28 well, that certainly eliminates the need for a lot of my goto's... 2010-11-09T04:32:27 haha 2010-11-09T04:32:47 i don't think that person quite understoof what if else was for aha 2010-11-09T04:33:43 why do you think that? 2010-11-09T04:33:59 if else - to break up flow of a program 2010-11-09T04:34:12 question "can i use if else to break up the flow of my program" 2010-11-09T04:34:17 hence they don't understand :P 2010-11-09T04:35:25 "is it possible to put an if-else statement inside a function" 2010-11-09T04:35:56 "so i can if-else while if-elseing" 2010-11-09T04:36:43 that's dangerous 2010-11-09T04:36:57 i am 2010-11-09T04:37:16 *** Appleman1234 has quit IRC (Read error: Connection reset by peer) 2010-11-09T04:37:31 *** Utkarsh has quit IRC (Ping timeout: 265 seconds) 2010-11-09T04:39:39 *** acieroid` has joined #aichallenge 2010-11-09T04:41:29 there was a tiny bug with that parsegamestate function i posted earlier 2010-11-09T04:41:46 stop posting bugs >=( 2010-11-09T04:41:52 *** Zepp_ has joined #aichallenge 2010-11-09T04:41:58 it's my specialty :P 2010-11-09T04:42:09 wow my eval function sucks 2010-11-09T04:42:21 but at least tuning it now accomplishes something 2010-11-09T04:43:11 eval functions are ridiculously hard to get right 2010-11-09T04:43:19 they are 2010-11-09T04:43:24 i sort of have mine worked out 2010-11-09T04:43:35 *** Utkarsh has joined #aichallenge 2010-11-09T04:43:43 well actually, my eval function is pretty straight forward, it's move selection that i find hardest 2010-11-09T04:43:43 it helps to evaluate the correct turn 2010-11-09T04:43:59 I'm actually working on move selection atm too 2010-11-09T04:44:11 mine too... kind of by force, everytime i try to it change it, it gets worse 2010-11-09T04:44:14 i've been working on move selection for over a month :P 2010-11-09T04:44:27 it's hard 2010-11-09T04:44:44 *** Appleman1234 has joined #aichallenge 2010-11-09T04:44:54 planet selection is slightly easier 2010-11-09T04:45:01 I work from there 2010-11-09T04:45:14 antimatroid1: yea i noticed 2010-11-09T04:45:17 i know what i want to write 2010-11-09T04:45:43 but its tricky to write an allocation system for it 2010-11-09T04:45:44 what i think is hardest is resolving battles with both players making future moves 2010-11-09T04:46:03 my hypothetical enemy can't make moves 2010-11-09T04:46:04 in a desirable way anyway 2010-11-09T04:46:06 *** Aaaarg has quit IRC (Read error: Connection reset by peer) 2010-11-09T04:46:06 he's stupid 2010-11-09T04:46:42 modelling the other player allows you to just lump a whole lot onto the tree to work out 2010-11-09T04:46:53 easier said than done 2010-11-09T04:47:13 *** acieroid has quit IRC (*.net *.split) 2010-11-09T04:47:14 *** Zepp has quit IRC (*.net *.split) 2010-11-09T04:47:18 i know, tron was really easy 2010-11-09T04:47:23 btw, what are you doing in your rewrite? minimax? 2010-11-09T04:47:24 this is so damn hard 2010-11-09T04:47:33 yeah going with minimax 2010-11-09T04:47:39 although i may do decision tree first 2010-11-09T04:47:55 i'm doing the entire thing from scratch though 2010-11-09T04:48:41 minimax is the devil which tempts you to the dark side 2010-11-09T04:48:59 minimax works wondefully when the calculations are correct 2010-11-09T04:49:08 lol 2010-11-09T04:49:11 and you give it smart choices 2010-11-09T04:49:35 if I could give it smart choices, I wouldn't need minimax :P 2010-11-09T04:50:04 they need to be less smart for the same intelligence imo with minimax 2010-11-09T04:50:13 :) 2010-11-09T04:52:12 *** Rubicon-|-Cross has joined #aichallenge 2010-11-09T04:52:41 *** xxxss has quit IRC (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org) 2010-11-09T04:58:20 meh not sure if i have enough time for this any more 2010-11-09T04:58:22 :( 2010-11-09T04:58:40 at least i learnt something ^^ 2010-11-09T04:59:00 * raedwulf will revisit #aichallenge in a week 2010-11-09T04:59:16 i need to get some stuff done for my other personal projects 2010-11-09T04:59:22 you'll be back 2010-11-09T04:59:23 been keeping them waiting for too long 2010-11-09T04:59:38 antimatroid1: not sure - i'm part of the urbanterror development team ^^ 2010-11-09T04:59:56 but yea i put a lot of effort into the code 2010-11-09T04:59:58 this contest is worse than crack 2010-11-09T05:00:01 haha 2010-11-09T05:00:17 i think i might wait till the next contest - because at least then i will be looking out for it 2010-11-09T05:00:21 rather than starting it mid-way 2010-11-09T05:00:34 then i would have loads more time 2010-11-09T05:01:26 e.g. if i knew it started in september, i would have at least 1 month before term starts :P 2010-11-09T05:01:34 anywas good luck everyone 2010-11-09T05:01:43 the last one was in february this year 2010-11-09T05:02:58 february is ok too i guess 2010-11-09T05:03:19 we do necessarily get the planets in the same order each turn yes? 2010-11-09T05:03:39 tbh, at the rate im going with all my work, ill get something decent working by the 29th, but it'll not be groundbreaking 2010-11-09T05:03:49 antimatroid1: yes, but i don't depend on that 2010-11-09T05:04:11 also i have an assignment tomorrow 2010-11-09T05:04:26 where i have to write a literature review... thats going to EAT time up 2010-11-09T05:04:28 :P 2010-11-09T05:16:03 Top 10 players: bocsimacko(4237), george(4068), Raschi(4066), bix0r4ever(4050), shangas(4027), rsergio(4005), GreenTea(3988), Accoun(3955), Xrillo(3953), medrimonia(3942) 2010-11-09T05:22:00 *** Ice_Harley has joined #aichallenge 2010-11-09T05:22:16 *** boegel has joined #aichallenge 2010-11-09T05:26:41 *** p4p4 has joined #aichallenge 2010-11-09T05:31:51 *** Tobu has joined #aichallenge 2010-11-09T06:04:12 *** magiczvn has joined #aichallenge 2010-11-09T06:04:50 a 2010-11-09T06:06:10 *** magiczvn has left #aichallenge 2010-11-09T06:06:11 b 2010-11-09T06:09:26 c 2010-11-09T06:15:22 d 2010-11-09T06:16:04 Top 10 players: bocsimacko(4236), george(4064), Raschi(4060), bix0r4ever(4058), shangas(4027), rsergio(4012), GreenTea(3980), Accoun(3972), luksian(3968), Xrillo(3939) 2010-11-09T06:21:52 e 2010-11-09T06:26:26 o 2010-11-09T06:27:28 *** ulope has joined #aichallenge 2010-11-09T06:47:21 Error323: back to school! 2010-11-09T06:49:27 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-11-09T06:54:54 *** Utkarsh has joined #aichallenge 2010-11-09T06:59:44 *** hellman has joined #aichallenge 2010-11-09T07:01:48 *** Utkarsh has quit IRC (Ping timeout: 264 seconds) 2010-11-09T07:02:47 phreeza: :P 2010-11-09T07:03:20 Error is an apropriate nick ;) 2010-11-09T07:07:36 *** eregon has quit IRC (Ping timeout: 245 seconds) 2010-11-09T07:08:07 *** Utkarsh has joined #aichallenge 2010-11-09T07:12:02 *** narnach has joined #aichallenge 2010-11-09T07:12:34 *** Appleman1234 has quit IRC (Ping timeout: 245 seconds) 2010-11-09T07:15:54 *** narnach_mb has quit IRC (Ping timeout: 240 seconds) 2010-11-09T07:15:55 juuup 2010-11-09T07:16:03 Top 10 players: bocsimacko(4216), george(4039), bix0r4ever(4039), luksian(4003), shangas(3999), Raschi(3998), rsergio(3992), GreenTea(3973), Accoun(3958), Xrillo(3940) 2010-11-09T07:19:04 I wanna be in there >_> 2010-11-09T07:20:51 *** eregon has joined #aichallenge 2010-11-09T07:23:59 I doubt I'll make it there hehe... my bots so far are more proof of concept than competitive 2010-11-09T07:24:26 can't be bothered with all the fine tuning that would go into making a bot competitive 2010-11-09T07:27:39 *** Utkarsh has quit IRC (Ping timeout: 250 seconds) 2010-11-09T07:33:24 *** Utkarsh has joined #aichallenge 2010-11-09T07:36:52 phreeza: how did UCT turn out? 2010-11-09T07:36:58 are you still using it? 2010-11-09T07:37:12 yea i have a bot running it 2010-11-09T07:37:20 it can win games :) 2010-11-09T07:37:22 still dimwit? :P 2010-11-09T07:37:26 but not enough 2010-11-09T07:37:28 yea 2010-11-09T07:38:23 the major flaw is that it fails to see that it can be sniped... 2010-11-09T07:38:31 dimishwit? is that you? 2010-11-09T07:38:59 http://72.44.46.68/getplayer?player=dimwit_0.4.3 2010-11-09T07:39:23 yeah, I see that one... but there is another bot: dimishwit2.5_pretrial 2010-11-09T07:39:32 oh 2010-11-09T07:39:36 i guess that one isn't you 2010-11-09T07:39:41 thats me too 2010-11-09T07:39:48 but just a single game right? 2010-11-09T07:39:49 haha... ok :) 2010-11-09T07:40:01 4 games, all losses 2010-11-09T07:40:03 that was just debugging ;) 2010-11-09T07:40:06 :P 2010-11-09T07:40:12 it's got a higher rank :D 2010-11-09T07:40:20 haha 2010-11-09T07:42:19 i might do another version before the contest, more efficient and with something like forced moves for sniping etc... 2010-11-09T07:42:21 *** Utkarsh has quit IRC (Ping timeout: 255 seconds) 2010-11-09T07:42:33 yeah, fair enough 2010-11-09T07:42:36 but for now i'm working on a case based resoning thing 2010-11-09T07:42:51 to see if that can be done ;) 2010-11-09T07:43:06 haven't heard anyone else trying that 2010-11-09T07:43:22 it's hard 2010-11-09T07:43:25 :P 2010-11-09T07:43:38 indeed :) 2010-11-09T07:44:13 the problem is that you probably don't understand the game as well as you think you do :) 2010-11-09T07:44:45 hey I have a pretty low opnion of how well I understand this game 2010-11-09T07:45:07 :P 2010-11-09T07:45:13 :D 2010-11-09T07:45:27 hmm... I really need to look into UCT sometime... I don't understand it is effective, especially in games like Go 2010-11-09T07:46:07 uct? 2010-11-09T07:46:27 UCT is a guided monte-carlo tree search algorithm 2010-11-09T07:47:14 *I don't understand how it is effective 2010-11-09T07:47:51 UCT = "Upper Confidence Bounds Applied to Trees" 2010-11-09T07:48:21 *** Utkarsh has joined #aichallenge 2010-11-09T07:48:22 i think ther main advantage over minimax is it gives the best possible results even if you abort early 2010-11-09T07:48:41 you can do that in minimax with iterative deepening 2010-11-09T07:49:06 oh 2010-11-09T07:49:07 kinda, but uct is smarter about it 2010-11-09T07:49:20 UCT/monte carlo's advantage over minimax is that it can deal with enormous branching factors 2010-11-09T07:49:32 yeah, that's my understanding 2010-11-09T07:49:41 *** Frontier has left #aichallenge 2010-11-09T07:49:59 it tries to balance exploring the most promising branches and exploring branches that haven't been looked into yet 2010-11-09T07:50:12 yes 2010-11-09T07:50:52 *** Frontier has joined #aichallenge 2010-11-09T07:51:24 a problem i ran into is that according to the pure uct algorithm every node in ply 1 gets played at least once to begin with 2010-11-09T07:51:28 you can do that in go 2010-11-09T07:51:40 but not in PW 2010-11-09T07:51:45 ah, planet was has too many moves 2010-11-09T07:51:56 owned by branching factor :D 2010-11-09T07:52:00 so i had to artificially limit the layer width 2010-11-09T07:52:28 yea the branching factor is many orders of magnitude larger than in go 2010-11-09T07:53:27 but i randomly sampled the tree, like an additional monte carlo layer ;) 2010-11-09T07:53:50 :) 2010-11-09T07:54:43 whats the action set in these type of methods 2010-11-09T07:54:56 because that seems to be the biggest issue 2010-11-09T07:55:30 yea mine is super naive... i tried o improve it but that only made it worse 2010-11-09T07:56:21 yeah 2010-11-09T07:56:26 the assumptions work both ways 2010-11-09T07:56:31 basically from a random sampling 30% of your planets, send a random amount to a random planet 2010-11-09T07:56:35 i also made a very simple version 2010-11-09T07:56:41 take 1 planet per turn 2010-11-09T07:56:47 which didn't even rank lowest ^_^ 2010-11-09T07:56:50 at tcp 2010-11-09T07:56:54 -300 or sth 2010-11-09T07:57:10 hehe 2010-11-09T07:57:13 with minimax? 2010-11-09T07:57:16 ye 2010-11-09T07:57:41 you should try it with UCT 2010-11-09T07:57:48 not a complicated algorithm 2010-11-09T07:58:05 well using alphabeta atm 2010-11-09T07:58:13 but a massive actionset 2010-11-09T07:58:20 sometimes 30 sec per turn locally 2010-11-09T07:58:23 Hostname: jarvis - OS: Linux 2.6.35-gentoo-r5/x86_64 - CPU: 4 x Intel(R) Core(TM)2 Quad (2666.923 MHz) - Processes: 191 - Uptime: 9d 4h 11m - Load Average: 1.88 - Memory Usage: 829.48MB/2008.39MB (41.30%) - Disk Usage: 657.31GB/1768.99GB (37.16%) 2010-11-09T07:58:27 not good ^_^ 2010-11-09T07:58:30 Error323: what brancing factor 2010-11-09T07:58:31 hehe 2010-11-09T07:58:39 andbut does it play well? 2010-11-09T07:58:44 on average? 2010-11-09T07:58:44 *but 2010-11-09T07:58:54 we should set up a tcp server for those bots 2010-11-09T07:58:55 it does win though 2010-11-09T07:59:02 mega1: -300 elo 2010-11-09T07:59:02 with a 60s time limit 2010-11-09T07:59:12 sigh: no that s the testversion 2010-11-09T07:59:16 oh, ok 2010-11-09T07:59:22 with my heuristics it beats my old 2010-11-09T07:59:26 but ugh 2010-11-09T07:59:29 ah 2010-11-09T07:59:38 takes all freakin night 2010-11-09T07:59:51 :) 2010-11-09T08:00:04 http://72.44.46.68/getplayer?player=XYZ 2010-11-09T08:00:07 so, what sort of branching factor do you have? 2010-11-09T08:00:10 this was my test alphabeta =) 2010-11-09T08:00:27 currently its massive, around 2000 :P 2010-11-09T08:00:35 20000 sometimes 2010-11-09T08:00:37 I think with 60s to compute i'd run out of memory before i run out of time 2010-11-09T08:00:41 yeah, that's a tad high :) 2010-11-09T08:00:59 http://72.44.46.68/canvas?game_id=913971 2010-11-09T08:01:10 i love how it was able to kick some ass here and there :P 2010-11-09T08:01:44 *** Naktibalda has joined #aichallenge 2010-11-09T08:02:08 ah so it punds a given planet until it gets it and then moves on to the next? 2010-11-09T08:02:39 branching factor is 23 (nr of planets) 2010-11-09T08:02:48 and it has my staging algo 2010-11-09T08:02:49 that's it 2010-11-09T08:02:56 depth is 2 2010-11-09T08:03:10 ehh 4 2010-11-09T08:03:13 ah with 23 why don't you go deeper? 2010-11-09T08:03:29 doesn't make sens 2010-11-09T08:03:36 this was a test anyway 2010-11-09T08:04:24 http://72.44.46.68/canvas?game_id=913814 2010-11-09T08:04:31 why does that game terminate 2010-11-09T08:04:54 doesn't say timeout or disconnect 2010-11-09T08:05:06 no idea 2010-11-09T08:05:10 and XYZ is taking a 0 growth -_- 2010-11-09T08:05:42 yep, not good :P 2010-11-09T08:07:50 defining a good heuristic evaluation is also quite tricky 2010-11-09T08:08:51 mega1: care to share? :P 2010-11-09T08:09:57 it's too hairy to explain :-) 2010-11-09T08:10:09 hehe 2010-11-09T08:10:20 currently i simulate to the end of the game 2010-11-09T08:10:24 but it rlly su 2010-11-09T08:10:25 sux 2010-11-09T08:10:41 mega1: are you going to open source your code after the tournament? 2010-11-09T08:10:47 yes, yes 2010-11-09T08:10:53 cool :) 2010-11-09T08:11:02 mega1: a hint? 2010-11-09T08:11:06 ^^ 2010-11-09T08:12:03 simulate to the end? assuming that no one makes any more moves? 2010-11-09T08:12:09 indd 2010-11-09T08:12:26 yeah, that over-values planet growth rates 2010-11-09T08:12:30 I couldn't make alpha-beta work well with more than 1 play 2010-11-09T08:12:41 mega1: heh i had the same problem 2010-11-09T08:12:47 but got it working 2010-11-09T08:13:05 *ply 2010-11-09T08:13:32 that is, alphabeta gives same amount of turns as minimax in half the cpu time. I guess that says it works 2010-11-09T08:13:48 same endresult that is 2010-11-09T08:14:30 wanna trade ? :P 2010-11-09T08:15:50 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-11-09T08:16:04 Top 10 players: bocsimacko(4241), bix0r4ever(4065), george(4058), rsergio(4020), Raschi(4019), shangas(4019), luksian(3995), GreenTea(3987), Accoun(3972), Xrillo(3958) 2010-11-09T08:16:52 oh, dmj no longer in the top 10... 2010-11-09T08:17:01 I wanna be in there >_< 2010-11-09T08:17:06 so close 2010-11-09T08:17:13 well next heuristicbot is doing quite a bit better 2010-11-09T08:17:16 so maybe... 2010-11-09T08:17:28 301 elo @ tcp \o/ 2010-11-09T08:17:29 :P 2010-11-09T08:17:36 :) 2010-11-09T08:19:12 oh, shangas dropped to 6th 2010-11-09T08:19:13 :( 2010-11-09T08:19:32 time to sleep, night everyone 2010-11-09T08:19:40 *** sigh has quit IRC (Quit: Leaving...) 2010-11-09T08:21:16 *** Utkarsh has joined #aichallenge 2010-11-09T08:24:35 *** bhasker has joined #aichallenge 2010-11-09T08:24:44 bhasker: hoooowdy 2010-11-09T08:24:55 morning! 2010-11-09T08:25:18 well, 14:20 but ok ^_^ 2010-11-09T08:26:38 *** bhasker has quit IRC (Read error: Connection reset by peer) 2010-11-09T08:26:56 *** bhasker has joined #aichallenge 2010-11-09T08:30:45 i really need to reduce my actionspace 2010-11-09T08:31:54 http://72.44.46.68/getplayer?player=E323-17.0-1-gf87222e :D 2010-11-09T08:32:12 mega1: care to put your almighty bot back in tcp? 2010-11-09T08:32:45 I will, later. My laptop is busy :-( 2010-11-09T08:32:53 kk 2010-11-09T08:52:13 i have a question about minmax mega1 2010-11-09T08:52:31 i got it working too, alphabeta that is :) 2010-11-09T08:52:31 does your bot basically do a max version only? 2010-11-09T08:52:38 oh nice 2010-11-09T08:52:54 but the actionspace is @)#$(%)#($ and my heuristic value function sux 2010-11-09T08:53:19 heh so how are you doing it , what is your move generator function like 2010-11-09T08:53:21 it does both max and min 2010-11-09T08:53:24 does it like pick a target planet 2010-11-09T08:53:28 and generate moves to capture it 2010-11-09T08:53:32 in the next X turns 2010-11-09T08:53:40 and you do the same for all the planets? 2010-11-09T08:53:43 mm no 2010-11-09T08:54:19 bhasker: yes, the max only version is the best performing one so far 2010-11-09T08:54:20 i only use a selection for attacking (those that get fed through staging) and compute all combinations against enemyplantes 2010-11-09T08:54:52 ehhh what do you guys mean by max only? 2010-11-09T08:55:08 Error I am not sure i understand what you mean by "selection for attacking" 2010-11-09T08:55:25 selection of my planets that are allowed to attack 2010-11-09T08:55:39 hmm k 2010-11-09T08:55:45 but its failing hard here xD 2010-11-09T08:55:48 heh 2010-11-09T08:56:01 though my alphabeta is working properly 2010-11-09T08:56:03 and how do you compute available ships, i am having trouble with that 2010-11-09T08:56:17 not implemented min max yet 2010-11-09T08:56:22 when i try to account for future shis 2010-11-09T08:56:27 my bot just gets too conservative:-\ 2010-11-09T08:56:30 and doesn't do crap 2010-11-09T08:56:46 available ships? 2010-11-09T08:57:02 how do you mean 2010-11-09T08:57:13 as in how do you compute how many ships you can send from a given planet 2010-11-09T08:57:17 without losing it 2010-11-09T08:57:29 i don't :P 2010-11-09T08:57:34 whenever i try to get that accurate it grinds to a halt 2010-11-09T08:57:39 indeed 2010-11-09T08:57:50 so you assume that you can always send all your ships? 2010-11-09T08:57:55 yes 2010-11-09T08:57:56 as long as its not under attack? 2010-11-09T08:58:10 yes 2010-11-09T08:58:19 hmm 2010-11-09T08:58:57 even your frontline planets? 2010-11-09T08:59:17 especially those 2010-11-09T08:59:34 hmm don't you like lose planets that way when you send out to the wrong target 2010-11-09T08:59:35 from front 2010-11-09T08:59:53 i have quite a neat target selector :> 2010-11-09T09:00:01 hmm but if its wrong 2010-11-09T09:00:03 if i don't use that 2010-11-09T09:00:07 it goes horribly wrong 2010-11-09T09:00:10 hmm k 2010-11-09T09:00:44 well 'neat'... its just simple but very effective 2010-11-09T09:01:08 but yes sometimes it doesn't work 2010-11-09T09:01:18 thats why i lose from e.g. shampoo sometimes :P 2010-11-09T09:03:12 heh 2010-11-09T09:03:28 i can't stand it 2010-11-09T09:04:01 got a very beatiful piece of code that does my alphabeta, but i can't get a good heuristic function and actionset 2010-11-09T09:04:04 >_< 2010-11-09T09:04:11 fail 2010-11-09T09:04:16 :( 2010-11-09T09:05:55 mega1: bhasker what is this max only? 2010-11-09T09:06:11 you just look at moves that maximize your value 2010-11-09T09:06:15 and same for enemies 2010-11-09T09:07:01 Error323: one ply minimax: enemy moves are not considered, the evaluation function is supposed to take care of that 2010-11-09T09:07:24 mega1 you don't even look at enemy's move is it? 2010-11-09T09:07:30 yes 2010-11-09T09:07:31 oh 2010-11-09T09:07:32 just generate all possible moves for you 2010-11-09T09:07:38 and maximize the value 2010-11-09T09:07:42 some :-) 2010-11-09T09:07:46 *** Utkarsh has quit IRC (Ping timeout: 252 seconds) 2010-11-09T09:07:49 yea some possible moves 2010-11-09T09:07:54 hmm i have been considering that 2010-11-09T09:08:16 right now i assume my priority queue will do that,since its ordered by planet value 2010-11-09T09:08:22 but that doesn't work out right 2010-11-09T09:08:27 sometimes i pick moves to attack far off planets 2010-11-09T09:08:36 which are far worse than sending those ships at nearer targets 2010-11-09T09:09:07 and mega1 what do you evaluate?, the value at next turn? 2010-11-09T09:09:15 or the value at the end of executing the curent turn 2010-11-09T09:09:31 the value at some future turn 2010-11-09T09:09:42 so thats your look ahead? 2010-11-09T09:09:54 or "horizon" 2010-11-09T09:09:56 hmm 2010-11-09T09:11:14 interesting 2010-11-09T09:14:36 yeah this is all about defining a good heuristic eval function 2010-11-09T09:14:43 which i'm failing at :P 2010-11-09T09:14:52 *** Utkarsh has joined #aichallenge 2010-11-09T09:16:05 Top 10 players: bocsimacko(4242), bix0r4ever(4063), george(4055), shangas(4020), rsergio(4020), luksian(4001), GreenTea(3996), Accoun(3988), Raschi(3967), dmj111(3965) 2010-11-09T09:20:58 40 sec for a turn sometimes 2010-11-09T09:20:59 epic 2010-11-09T09:21:20 heh 2010-11-09T09:23:20 VISITED: 19659SCORE: 802419452 2010-11-09T09:23:24 lol 2010-11-09T09:23:28 20K nodes visited ^_^ 2010-11-09T09:23:53 it does 10K nodes in 1.2 seconds it seems 2010-11-09T09:25:02 VISITED: 111135 SCORE: 89548777 TIME: 24.7882s *ouch* 2010-11-09T09:25:13 100K nodes 2010-11-09T09:25:24 nice :) 2010-11-09T09:25:41 for a 12% score increase? 2010-11-09T09:28:12 *** mega1 has quit IRC (Ping timeout: 264 seconds) 2010-11-09T09:28:17 *** mega1 has joined #aichallenge 2010-11-09T09:30:11 no thats what alpha returns 2010-11-09T09:30:21 different turn 2010-11-09T09:33:43 ah i see 2010-11-09T09:35:28 i have a weird idea in my head... somehow mapping out future probabilities instead of concrete ship distributions. not sure if that makes any sense though :D 2010-11-09T09:36:08 i tried that a bit 2010-11-09T09:36:19 but my implementation sucked hard :P 2010-11-09T09:36:34 heh 2010-11-09T09:38:40 *** fawek has joined #aichallenge 2010-11-09T09:38:46 *** Azrathud has quit IRC (Ping timeout: 272 seconds) 2010-11-09T09:42:23 *** mega1 has quit IRC (Ping timeout: 260 seconds) 2010-11-09T09:43:48 *** mega1 has joined #aichallenge 2010-11-09T09:56:30 lol timeout @ over 100 sec 2010-11-09T09:56:33 and it times out 2010-11-09T09:57:47 *** demerzel_ has joined #aichallenge 2010-11-09T10:01:57 trying to upload just my io code 2010-11-09T10:03:11 nope plain crash 2010-11-09T10:03:13 wtf 2010-11-09T10:03:40 C#? 2010-11-09T10:03:48 C 2010-11-09T10:04:26 I have 2 last games lost in first turn... http://ai-contest.com/profile.php?user_id=6494 2010-11-09T10:04:45 like before they fix that bug 2010-11-09T10:05:00 oh my old bot was C# i know that problem hehe 2010-11-09T10:05:39 but that was already fixed 2010-11-09T10:06:13 nope i was hoping McLeopold or lavalamp_ would have something to say, had the problem since yesterday 2010-11-09T10:06:19 after that fix i go to near 380 position from near 1000 without changing bot ) 2010-11-09T10:06:41 what bug is that 2010-11-09T10:07:20 first turn loses. Especially for C#, but other langs too 2010-11-09T10:07:50 no mine is very specific 2010-11-09T10:07:57 C stdin is choking for some reason 2010-11-09T10:08:09 spent hours on it yesterday 2010-11-09T10:08:21 it's only when stdin is piped/redirected/whatever, works fine typed to 2010-11-09T10:09:06 *** JamesMG has joined #aichallenge 2010-11-09T10:15:21 mega1: is your heuristic some linear function or is it like you said really more hairy :P 2010-11-09T10:16:03 sorry, the evaluation function is top secret 2010-11-09T10:16:06 Top 10 players: bocsimacko(4241), luksian(4071), george(4059), bix0r4ever(4056), shangas(4025), rsergio(4015), GreenTea(4002), Accoun(3981), Xrillo(3965), Raschi(3965) 2010-11-09T10:17:00 mega1: yea makes sense =) 2010-11-09T10:18:05 jeez what does everybody use locally 2010-11-09T10:18:16 mm? 2010-11-09T10:18:18 playgame is bugged i gotta compile albertz iouri's i can't find the thread 2010-11-09T10:18:27 jbot doesn't spit out debug 2010-11-09T10:18:31 ceridan's just plain doesn't work right 2010-11-09T10:18:42 is playgame bugged? 2010-11-09T10:18:46 o_O 2010-11-09T10:19:17 i haven't tested locally in forever but i need something that isn't bugged to test stdin with 2010-11-09T10:19:30 it'd be nice if it spit out debug so i can see if it's getting anything at least 2010-11-09T10:19:43 demerzel_: i use playgame and jbot for viewing 2010-11-09T10:19:48 such a small problems but it's taken me like 6 hours and no solution 2010-11-09T10:20:36 yea i guess i will use playgame - it's useful to have more than just 'didn't work' 2010-11-09T10:21:19 demerzel_: are you using an ancient playgame.jar and not an updated one?: http://ai-contest.com/forum/viewtopic.php?f=18&t=534&sid=4c5b683c631ac16a0481cf006ec15861 2010-11-09T10:21:26 nevermind the sid 2010-11-09T10:21:56 yea i'm looking at that now 2010-11-09T10:22:00 where is iouri's 2010-11-09T10:22:15 i can't remember what the hell the thread was called 2010-11-09T10:22:25 been foreeever since i tested locally 2010-11-09T10:24:17 a1k0n: is iouri? 2010-11-09T10:24:42 *** superflit has quit IRC (Quit: superflit) 2010-11-09T10:25:49 i have no idea 2010-11-09T10:25:53 i am so goddamn frustrated right now 2010-11-09T10:26:03 6 hours of googling this shit and i still have no idea 2010-11-09T10:26:15 raedwulf says he has code that works fine, i look at it and it's the same as mine 2010-11-09T10:26:33 i stripped everything but the io code and 'go\n' and uploaded it and it's still busted 2010-11-09T10:39:59 *** wh1teside has joined #aichallenge 2010-11-09T10:42:10 *** bhasker has quit IRC (Read error: Connection reset by peer) 2010-11-09T10:48:49 *** Utkarsh has quit IRC (Ping timeout: 245 seconds) 2010-11-09T10:53:17 *** siconize has joined #aichallenge 2010-11-09T10:54:34 *** Utkarsh has joined #aichallenge 2010-11-09T11:02:16 *** superflit has joined #aichallenge 2010-11-09T11:07:41 *** boegel has quit IRC (Ping timeout: 252 seconds) 2010-11-09T11:08:01 *** bhasker has joined #aichallenge 2010-11-09T11:09:28 *** lavalamp_ has quit IRC (Quit: Page closed) 2010-11-09T11:11:09 *** mceier has joined #aichallenge 2010-11-09T11:15:09 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-11-09T11:16:07 Top 10 players: bocsimacko(4250), luksian(4085), bix0r4ever(4072), george(4057), shangas(4040), rsergio(4026), GreenTea(4011), Raschi(4010), Accoun(3999), Xrillo(3993) 2010-11-09T11:18:29 *** tapwater has joined #aichallenge 2010-11-09T11:21:52 *** Utkarsh has joined #aichallenge 2010-11-09T11:27:06 *** Ice_Harley has quit IRC (Ping timeout: 265 seconds) 2010-11-09T11:28:26 demerzel_: still having problems with io? If it helps I will test run it under my server code and maybe find something. 2010-11-09T11:28:33 yea i am 2010-11-09T11:29:22 completely stripped down i/o code: http://www.mediafire.com/?6g75858qr4g16v7 2010-11-09T11:29:28 no state no nothing still broken 2010-11-09T11:31:03 #endif without #if? wth? 2010-11-09T11:31:18 oh yea i chopped some crap out there's an extra endif in mybot.h sorry 2010-11-09T11:33:48 so there is a duplicate io_lastP 2010-11-09T11:34:07 yea there are a couple random bits leftover from my hackjob 2010-11-09T11:34:18 assume that nothing insanely obvious like that is the problem ;) 2010-11-09T11:34:30 or rather - anything that won't compile 2010-11-09T11:34:56 i can process it for you if you want, i grabbed it this morning and been playng with it, only had like 2 things iirc 2010-11-09T11:35:23 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-11-09T11:35:25 *** iFire has quit IRC (Read error: Connection reset by peer) 2010-11-09T11:35:43 *** iFire has joined #aichallenge 2010-11-09T11:37:18 what does printf with only one parameter do except for spitting out a warning? 2010-11-09T11:39:10 oh it says hi, very nice of it. 2010-11-09T11:41:55 i use 'hi' as a 'yes i got here' a lot haha 2010-11-09T11:45:30 Strange, I recieve the stderr, but stdout remains empty. 2010-11-09T11:45:45 i fprintf(stderr 2010-11-09T11:45:53 the go\n never happens it dies around p18 2010-11-09T11:48:17 It says: HI go, so it gets to planet 23 2010-11-09T11:49:11 how is 2010-11-09T11:49:11 void io_sendInput(const char *newInput){ printf("%s", newInput); } 2010-11-09T11:49:11 supposed to look like? 2010-11-09T11:51:13 just print it out 2010-11-09T11:51:30 what are you using to redirect to it 2010-11-09T11:52:00 problem is the server/game engines don't pipe out continously it's hitting some wall and breaking on main server and here on windows 2010-11-09T11:52:09 if you pipe the file to it it does fine 2010-11-09T11:54:16 oh it says HI go? did you tyype go at it or give it a file? it should be saying HI 2010-11-09T11:54:26 per line 2010-11-09T11:55:38 i gave it a map through my server 2010-11-09T11:56:39 *** bhasker has quit IRC (Quit: bhasker) 2010-11-09T12:00:38 *** mega1 has quit IRC (Ping timeout: 260 seconds) 2010-11-09T12:00:54 part of the problem is it's also a DOS issue 2010-11-09T12:00:58 but it's broken on the main server too 2010-11-09T12:01:13 it's frustrating because i can't get it working here so i can't focus on making it work on main 2010-11-09T12:04:16 lol, what did you DO? why does it not pipe correctly? 2010-11-09T12:04:25 i have no goddamn idea to either 2010-11-09T12:04:33 it should work, it does work in all situations BUT the one i need 2010-11-09T12:04:48 works here piped to works typed to doesn't work when it's a child process 2010-11-09T12:05:27 apparently same for server because i uploaded a barebones io like that to the server and it's still broke 2010-11-09T12:07:47 the hardest part of thsi is i have no way to really check it as a child process - it just does/doesn't do whatever 2010-11-09T12:08:01 testing is impossible i just make a change, copy exe and pray 2010-11-09T12:10:01 *** mega1 has joined #aichallenge 2010-11-09T12:10:21 *** lavalamp_ has joined #aichallenge 2010-11-09T12:10:25 oh wait! 2010-11-09T12:10:48 *** Bobng has joined #aichallenge 2010-11-09T12:11:39 ? 2010-11-09T12:11:40 problem solved, do you pay me? 2010-11-09T12:11:55 you won't have to hear me bitch about it any more lol 2010-11-09T12:12:35 lol, send me a virtual beer. anyway, you didn't flush after the "go\n". 2010-11-09T12:12:50 flush out? 2010-11-09T12:13:04 i don't even get that far 2010-11-09T12:13:22 in c++ "cout << "go" << endl;" does the same as "... fflush(stdout);" 2010-11-09T12:13:33 yea i know - it doesn't get that far 2010-11-09T12:13:39 doesn't even finish reading the state 2010-11-09T12:13:48 so it never sends that 2010-11-09T12:13:53 but if it did, it would help ^^ 2010-11-09T12:14:05 well if it did and then froze i would hvae a diff problem ;) 2010-11-09T12:14:41 how can i make your bot fail while reading the state? 2010-11-09T12:15:05 upload it tothe server lol 2010-11-09T12:15:14 or use iouri's planetwarrior 2010-11-09T12:15:26 i had it set up to fgetc instead of fgets and spit out each char 2010-11-09T12:15:38 this was the output, dead at p18: http://pastebin.com/rQq5Kt5g 2010-11-09T12:16:07 Top 10 players: bocsimacko(4273), bix0r4ever(4107), luksian(4074), george(4072), shangas(4069), rsergio(4047), GreenTea(4035), Accoun(4026), Xrillo(4012), Raschi(4005) 2010-11-09T12:16:13 it dies with playgame too 2010-11-09T12:16:18 but i don't know how to make it output there 2010-11-09T12:16:54 *** bhasker has joined #aichallenge 2010-11-09T12:16:55 *** aerique has quit IRC (Quit: ...) 2010-11-09T12:18:01 *** Bobng has quit IRC (Quit: Leaving) 2010-11-09T12:19:17 how can i know you flushed stderr in this example ;) 2010-11-09T12:19:40 i did 2010-11-09T12:19:41 :P 2010-11-09T12:20:04 And you are sure you can read whole lines from stdin? 2010-11-09T12:20:09 no i am not 2010-11-09T12:20:14 i don't think i can 2010-11-09T12:20:22 then dont ^^ 2010-11-09T12:20:23 you can see i have a hook after fgets 2010-11-09T12:20:30 it never has output anything 2010-11-09T12:20:33 that's why i tried single char 2010-11-09T12:20:52 man you don't understand how many different methods i've tried lol 2010-11-09T12:20:57 that is like 2 of probably 20 2010-11-09T12:21:48 *** yasith has quit IRC (Ping timeout: 240 seconds) 2010-11-09T12:23:51 now what if the server sends you 1k of map data, the flushes and sends you the rest of the line in the next go. first of all: your buffer only reads 128 bytes and expects the data to be a single line, then the next read starts in the middle of a line which is also bad. try to buffer what you read from stdin and extract the lines manually. that's what i do anyway 2010-11-09T12:25:36 yea i'm working on that angle ;) 2010-11-09T12:34:53 *** yasith has joined #aichallenge 2010-11-09T12:39:36 *** siconize has quit IRC (Ping timeout: 265 seconds) 2010-11-09T12:50:31 *** AlliedEnvy has quit IRC (Ping timeout: 240 seconds) 2010-11-09T12:54:36 *** Cold-Phoenix has joined #aichallenge 2010-11-09T12:55:01 *** acieroid` is now known as acieroid 2010-11-09T12:59:15 *** Cold-Phoenix has quit IRC (Client Quit) 2010-11-09T13:16:08 Top 10 players: bocsimacko(4259), bix0r4ever(4104), luksian(4077), shangas(4068), george(4065), rsergio(4056), GreenTea(4051), Accoun(4042), Raschi(4009), dmj111(4007) 2010-11-09T13:17:48 *** amin_ has joined #aichallenge 2010-11-09T13:18:21 hello 2010-11-09T13:18:27 any one here ? 2010-11-09T13:19:12 *** fawek has quit IRC (Ping timeout: 264 seconds) 2010-11-09T13:19:18 *** boegel has joined #aichallenge 2010-11-09T13:19:55 no 2010-11-09T13:20:25 i need some help 2010-11-09T13:20:38 i'm new ! 2010-11-09T13:20:46 please... please.... 2010-11-09T13:20:47 :( 2010-11-09T13:21:04 *** irchs has joined #aichallenge 2010-11-09T13:24:24 what? 2010-11-09T13:25:45 just ask your question 2010-11-09T13:28:54 mega1: what is your branching factor if i may ask? 2010-11-09T13:31:05 *** narnach has quit IRC (Ping timeout: 240 seconds) 2010-11-09T13:32:03 goddamn - making a little progress but it's windows specific lol 2010-11-09T13:32:14 got it finally reading the entire input but it's using direct windows definitions :( 2010-11-09T13:35:31 Error323: someone has asked this question today. I think it's about a 100. 2010-11-09T13:35:41 ah ok 2010-11-09T13:35:42 ty 2010-11-09T13:35:47 Can be made smaller if needed. 2010-11-09T13:35:51 that was me that asked ;) 2010-11-09T13:36:07 then i spent the next 24 hours trying to debug c stdin redirection problems lol 2010-11-09T13:36:14 but candidate move generation is very expensive 2010-11-09T13:36:34 better than fighting cygwin I guess 2010-11-09T13:36:43 it's been a nightmare honestly 2010-11-09T13:36:53 i /think/ STD_INPUT_HANDLE is workable in linux 2010-11-09T13:36:55 have you solved it? 2010-11-09T13:36:59 at least i'm finding examples of it 2010-11-09T13:37:18 i solved it by basically defining my own stdin/stdout via GetStdHandle(STD_INPUT_HANDLE) 2010-11-09T13:37:51 *** yasith has quit IRC (Ping timeout: 240 seconds) 2010-11-09T13:37:58 i still have no idea what was wrong with the standard stdin/out but i'm happy i'm reading the entire input now 2010-11-09T13:38:33 yeah its extremely expensive 2010-11-09T13:38:35 :/ 2010-11-09T13:38:42 what is 2010-11-09T13:40:52 i'm not sure how to make it portable tho cuz STD_INPUT_HANDLE is in winbase 2010-11-09T13:40:57 " but candidate move generation is very expensive" 2010-11-09T13:41:01 ohoh 2010-11-09T13:41:10 ^ makes me think I'm on the right track 2010-11-09T13:41:44 *** amin_ has quit IRC (Quit: Page closed) 2010-11-09T13:46:47 |A| = 160000 xD that's a bit too much 2010-11-09T13:49:29 *** perror has quit IRC (Quit: Bye all !) 2010-11-09T13:49:56 *** Migi32 has joined #aichallenge 2010-11-09T13:50:40 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86 [Firefox 3.6.8/20100722155716]) 2010-11-09T13:50:56 *** yasith has joined #aichallenge 2010-11-09T13:51:50 anybody know what GetStdHandle is in gnu c lol 2010-11-09T13:54:21 damn, my fancy schmancy method with voronoi (inspired by the tron contest) don't lead to any decent result... 2010-11-09T13:54:26 time for heuristics, I guess 2010-11-09T14:01:57 amin_ did you ever ask your question? 2010-11-09T14:06:11 *** hellman has quit IRC (Remote host closed the connection) 2010-11-09T14:09:09 fuck this i guess it's virtualbox time 2010-11-09T14:09:41 i have no idea what i'm doing with ubuntu tho - it is a gui or command line? i have run centos vps so command line linux is relatively familiar but never installed the fancy ones 2010-11-09T14:10:32 *** justin_pdx has joined #aichallenge 2010-11-09T14:11:09 goddamn this is ridiculous i swear 2010-11-09T14:11:09 demerzel_, if you install the same one the game server is on it's command line only 2010-11-09T14:11:17 oh good 2010-11-09T14:11:25 they're running heron or something 2010-11-09T14:11:27 right 2010-11-09T14:11:32 8.04 LTS I think 2010-11-09T14:11:39 server edition or something 2010-11-09T14:11:51 if I were at home I could tell you ^^ 2010-11-09T14:12:35 I had to search for it, the .iso is not linked from their home page 2010-11-09T14:16:08 Top 10 players: bocsimacko(4263), bix0r4ever(4128), luksian(4117), george(4084), shangas(4077), rsergio(4074), GreenTea(4072), Accoun(4057), dmj111(4029), Xrillo(4011) 2010-11-09T14:19:04 *** boegel has quit IRC (Quit: *poof!*) 2010-11-09T14:22:51 *** amin10043 has joined #aichallenge 2010-11-09T14:22:57 hello 2010-11-09T14:23:08 i am a simple question 2010-11-09T14:23:20 can i ask ? 2010-11-09T14:23:31 *** boegel has joined #aichallenge 2010-11-09T14:23:32 hey dude ! 2010-11-09T14:23:52 hey! 2010-11-09T14:24:10 nobody here ? 2010-11-09T14:24:42 amin10043 just ask your question 2010-11-09T14:24:49 okey 2010-11-09T14:24:57 there are people here but not all watching the window every instant 2010-11-09T14:25:28 how can use IssueOrder function in loop 2010-11-09T14:25:31 can i ? 2010-11-09T14:25:41 if you want to, sure 2010-11-09T14:26:03 *** tapwater_ has joined #aichallenge 2010-11-09T14:26:05 but you have to make sure you don't send more ships than you have 2010-11-09T14:26:06 but when i use it there is no moving in my planet 2010-11-09T14:26:34 uhhhh 2010-11-09T14:26:57 do you ever call FinalizeOrder (or whatever it's called) 2010-11-09T14:27:07 i think i can just one move in every time that call DoTurn function isn't ? 2010-11-09T14:27:29 no, you can issue as many orders as you have ships for 2010-11-09T14:27:31 haaah ? what is it ? :D 2010-11-09T14:28:50 *** tapwater has quit IRC (Ping timeout: 272 seconds) 2010-11-09T14:28:50 *** tapwater_ is now known as tapwater 2010-11-09T14:28:56 ah, FinishTurn() 2010-11-09T14:29:11 IssueOrder() as many times as you want to, then FinishTurn() 2010-11-09T14:29:16 lavalamp_ are you c 2010-11-09T14:29:18 owwwww, yes. 2010-11-09T14:29:18 or c++ 2010-11-09T14:29:32 thank you 2010-11-09T14:29:35 demerzel_, C++ but I have dealt with C in the past 2010-11-09T14:29:51 ah well i think i'll just go c++ for cin this is ridiculous 2010-11-09T14:30:05 it's a stupid stupid problem that i've spent wayyy too much time on 2010-11-09T14:30:10 yeah I really don't know why it is so difficult :/ 2010-11-09T14:30:48 aww what happened to http://www.ai-contest.com/language_profile.php?lang=C 2010-11-09T14:30:51 nobody? lol 2010-11-09T14:31:09 that page is flaky, sometimes it shows 4 gobots and sometimes just me 2010-11-09T14:31:22 oh i forgot bout "leaderboard" 2010-11-09T14:32:32 damn there really aren't many tho - none in irc :( 2010-11-09T14:33:09 my submitted bot is 25 Elo on tcp and 3473 on main server >.< 2010-11-09T14:34:27 oh good, that means mine should still be rising on main 2010-11-09T14:36:02 JamesMG: mine was stuck at place 250 for a week and then suddenly went up to 150ish... 2010-11-09T14:36:22 mine has climbed from 2000th over the weekend 2010-11-09T14:36:32 it's starting to lose a few at rank 375ish though 2010-11-09T14:36:58 hey dude, it doesn't work ! 2010-11-09T14:38:02 i put in a plaintive cry for help in C but i doubt any of hte <20 plain C folks read it haha 2010-11-09T14:38:11 i cannot believe out of all the shit i've done that is what is going to push me to c++ 2010-11-09T14:38:24 amin10043 if you can describe your problem you will get more responses 2010-11-09T14:38:33 ok, sure 2010-11-09T14:38:53 i want to call IssueOrder in a loop 2010-11-09T14:39:21 after call IssueOrder , call FinishTurn immediately 2010-11-09T14:39:30 demerzel_, once in college I got totally stuck on an assignment because I didn't understand why I needed copy constructors... so much frustration for such a simple problem... :( 2010-11-09T14:39:37 but it's not work 2010-11-09T14:40:00 amin10043, no, you can only call FinishTurn when you're all done issuing orders for your turn 2010-11-09T14:40:42 ok, so how to move from all my planet in one time 2010-11-09T14:41:10 amin10043: you call IssueOrder N times. You call FinishTurn 1 time. 2010-11-09T14:41:19 IssueOrder();IssueOrder();FinishTurn(); 2010-11-09T14:41:21 i need multi threading 2010-11-09T14:41:40 amin10043: no you don't 2010-11-09T14:41:51 and it's not allowed, anyway 2010-11-09T14:42:03 i don't know how ?! 2010-11-09T14:42:22 amin10043 maybe you should pastebin your code because we clearly aren't communicating at all 2010-11-09T14:42:27 pastebin.com 2010-11-09T14:42:41 when i use IssueOrder in loop , no moving happen in runtime 2010-11-09T14:43:00 does your loop ever end? 2010-11-09T14:43:32 yeeeees 2010-11-09T14:43:43 for (Planet p : pw.MyPlanets()) 2010-11-09T14:43:48 this is my loop 2010-11-09T14:44:29 amin10043 there are any number of things you could be doing wrong 2010-11-09T14:44:47 what ? 2010-11-09T14:44:52 without a better description and/or seeing some of your code it is really hard for us to help 2010-11-09T14:44:57 lavalamp_ i've never used copy constructors but shallow copies are usually fine by me haha 2010-11-09T14:45:39 demerzel_ not if you have private, allocated data and don't realize you accidentally made some temporary copies... 2010-11-09T14:45:49 oic 2010-11-09T14:45:49 hey ... 2010-11-09T14:45:56 yes? 2010-11-09T14:46:23 i just want to use IssueOrder for all my planet 2010-11-09T14:46:27 *** irchs has quit IRC (Quit: irchs) 2010-11-09T14:46:38 how can i do it ? 2010-11-09T14:46:39 amin10043, you could be: timing out, crashing, sending more ships than you have, sending ships to an invalid planet... the list goes on 2010-11-09T14:47:05 no one ! 2010-11-09T14:47:30 demerzel_, why undergo all this pain ? can't you just use the c++ input, compile as c++ and still do everything in c from there on ? 2010-11-09T14:47:47 amin10043, if you post your code on pastebin, I will look at it. Otherwise I don't think this is a very productive conversation. 2010-11-09T14:47:51 hey , can i describe ? 2010-11-09T14:47:55 of course -- i could just use python too 2010-11-09T14:48:02 i am actually bout to do that tbh tho 2010-11-09T14:48:02 okey 2010-11-09T14:48:32 deep copies are not necessary when you restrict yourself to immutable data :) 2010-11-09T14:48:39 i found it in an old tron bot 2010-11-09T14:48:41 if (!((r = fcntl(STDIN_FILENO, F_GETFL)) & O_NONBLOCK)) fcntl(STDIN_FILENO, r | O_NONBLOCK); 2010-11-09T14:48:47 that should do me 2010-11-09T14:49:14 demerzel_ good thinking 2010-11-09T14:49:26 i just needed to know what STDIN_FILENO was - in windows its STD_INPUT_HANDLE 2010-11-09T14:49:36 hey , this is my code : http://pastebin.com/YSN9fmfp 2010-11-09T14:49:42 can you help me ? 2010-11-09T14:49:49 *** irchs has joined #aichallenge 2010-11-09T14:50:08 amin, give me a sec to look at it 2010-11-09T14:50:28 sure 2010-11-09T14:50:36 amin10043: verify that dest is not null in getNearest 2010-11-09T14:51:03 or if it returns null do not issue an order with source == null 2010-11-09T14:51:09 * jmcarthur cries a little every time he sees this: "catch (Exception e) { /* owned */ }" 2010-11-09T14:51:33 haha 2010-11-09T14:52:12 jmcarthur: same 2010-11-09T14:52:18 it's one thing to have a catch-all at the top level. it's another to not even report it 2010-11-09T14:52:21 first time i saw that i thought 'yea by your shitty code' 2010-11-09T14:52:40 catchall at top level is good 2010-11-09T14:52:45 yeah 2010-11-09T14:52:46 just make it say 'go\n' lol 2010-11-09T14:52:55 amin10043: can you successfully run the starter bot? 2010-11-09T14:52:56 when it actually tells you when things go wrong... 2010-11-09T14:53:37 no , it's successful 2010-11-09T14:53:51 why it's seems wrong ? 2010-11-09T14:54:34 amin10043: your code is not obviously wrong. Your problem, whatever it is, has nothing to do with calling issueorder in a loop. 2010-11-09T14:54:42 http://pastebin.com/Y5DEN2Vz // demerzel_ 2010-11-09T14:54:50 you will have to debug it to see what is going wrong. 2010-11-09T14:55:17 it's another problem, :) how can i debug java code ?! 2010-11-09T14:55:28 p4p4 ?? 2010-11-09T14:55:51 i use netbeans , 2010-11-09T14:56:01 i cry a little every time i see # std::vector Tokenize(const std::string& s, const std::string& delimiters) 2010-11-09T14:56:03 how can debug it ? 2010-11-09T14:56:06 amin10043 I'm not a java coder so you're on your own for that 2010-11-09T14:56:08 c++ makes thing SO MUCH EASIER 2010-11-09T14:56:53 aha, okey my firend, thanks a lot 2010-11-09T14:57:22 *** RainCT has joined #aichallenge 2010-11-09T14:58:42 "c++ makes thing SO MUCH EASIER" seeing that makes ME cry 2010-11-09T14:59:01 you must be deep in some hairy faeces to make a statement like that :D 2010-11-09T14:59:03 haha 2010-11-09T14:59:10 its all relative 2010-11-09T14:59:13 :D 2010-11-09T14:59:15 Agreed 2010-11-09T14:59:25 i'm sorry i need to put my std::ears on my std::head include to ears << hear << you :: 2010-11-09T14:59:35 omg you must get all the gurlz 2010-11-09T14:59:36 :P 2010-11-09T14:59:48 if you want java program in java :P 2010-11-09T14:59:50 (or boys, nothing heteronormative about the google ai challenge) 2010-11-09T15:00:16 *** irchs is now known as janstaunton 2010-11-09T15:00:16 google ai challenge is nerdonormative 2010-11-09T15:00:19 lol 2010-11-09T15:00:24 heheh 2010-11-09T15:00:40 quite 2010-11-09T15:00:53 c++ was cool in 95 now there are better languages 2010-11-09T15:01:02 for that kind of programming at least 2010-11-09T15:01:33 i LIKE pedantic languages for a reason ;) 2010-11-09T15:01:35 for what kind of programming? 2010-11-09T15:01:45 painful? 2010-11-09T15:01:55 no because it either works or it doesn't 2010-11-09T15:01:59 not hidden away in some implementation 2010-11-09T15:02:00 :D 2010-11-09T15:02:07 c++ is pretty pedantic 2010-11-09T15:02:11 imho 2010-11-09T15:02:17 compared to python maybe 2010-11-09T15:02:23 yes :) 2010-11-09T15:02:41 but then again every language qualifies there 2010-11-09T15:02:59 except maybe vbscript ;) 2010-11-09T15:04:28 ruby, perl... don't they all have duck typing? 2010-11-09T15:04:52 oh and then theres Javascript, that tries to match {} if you happen to forget one :D 2010-11-09T15:05:33 js guesses about ; too, I think 2010-11-09T15:05:44 yes 2010-11-09T15:05:54 its actually defined in the standard i think 2010-11-09T15:05:55 ECMA 2010-11-09T15:05:57 *** fawek has joined #aichallenge 2010-11-09T15:06:22 Planet IDs remain static between turns right? 2010-11-09T15:07:42 yeah 2010-11-09T15:07:49 they freaking better 2010-11-09T15:11:57 hey , when timeout happens ? 2010-11-09T15:12:17 i don't know when and why ? 2010-11-09T15:13:58 *** Zepp_ is now known as Zepp 2010-11-09T15:14:04 no, it's too late, you're already out, program got terminated already 2010-11-09T15:15:01 amin10043: I just noticed a problem in your code 2010-11-09T15:15:50 You need a line "if (p == planet) {continue}" 2010-11-09T15:16:03 inside the getNearest function 2010-11-09T15:16:08 Top 10 players: bocsimacko(4263), bix0r4ever(4133), luksian(4117), george(4084), shangas(4077), rsergio(4075), GreenTea(4064), Accoun(4055), dmj111(4030), Xrillo(4002) 2010-11-09T15:16:15 as it is a planet is 0 distance from itself 2010-11-09T15:16:33 that is why your code does nothing, you're sending ships to the planet they start on 2010-11-09T15:17:29 amin10043 did you get all that? 2010-11-09T15:17:35 yeah, i do it. 2010-11-09T15:18:14 yes, my problem solved ! 2010-11-09T15:18:28 lavalamp_ thanks 2010-11-09T15:18:38 np 2010-11-09T15:18:50 where are you from ? :) 2010-11-09T15:19:15 when you can beat my bot I'll answer that :) 2010-11-09T15:19:43 http://ai-contest.com/profile.php?user_id=10821 2010-11-09T15:19:45 sure 2010-11-09T15:19:55 well, that tells you country, I guess, haha 2010-11-09T15:19:59 *** Accoun has quit IRC () 2010-11-09T15:20:02 you are a good boy, thanks 2010-11-09T15:22:43 * phreeza pats lavalamp_ on the head 2010-11-09T15:22:45 good boy 2010-11-09T15:23:01 * lavalamp_ wags tail 2010-11-09T15:23:16 i wish my dog could debug my code for me 2010-11-09T15:23:40 my cat would like to help if I would just let her on the keyboard 2010-11-09T15:27:25 sigh i can't even make a bot that beats my current bot 2010-11-09T15:27:37 my so called new improved bot gets thrashed by my current bot ./cry 2010-11-09T15:28:54 my best bot so far doesn't deal well at all with my other bots but gets rated better than them anyway 2010-11-09T15:28:56 *** mk has joined #aichallenge 2010-11-09T15:28:56 Android devs get help from monkeys sometimes i think 2010-11-09T15:28:59 http://developer.android.com/reference/android/app/ActivityManager.html#isUserAMonkey() 2010-11-09T15:29:24 *** mk is now known as Guest82843 2010-11-09T15:29:39 hmm 2010-11-09T15:31:53 *** Frontier has left #aichallenge 2010-11-09T15:32:32 *** Accoun has joined #aichallenge 2010-11-09T15:38:42 well i still don't have a bot that can read stdin lol 2010-11-09T15:56:18 :D i think i found a good actionset 2010-11-09T15:56:45 goddamned 2010-11-09T15:56:56 my bot wins against all example bots, on all 100 example maps 2010-11-09T15:57:05 but it loses it's first three games on the server :( 2010-11-09T15:57:53 can the maps being used at TCP be downloaded somewhere? 2010-11-09T15:59:01 well, the examplebots are a farcry from whats on the tcp server 2010-11-09T15:59:43 tcp server is pretty tough 2010-11-09T16:00:08 you have to have a top 200 bot just to get a positive ranking 2010-11-09T16:00:25 I meant the official server :( 2010-11-09T16:00:31 tcp isn't the happy time funhouse it was a month ago 2010-11-09T16:00:37 oh lol 2010-11-09T16:00:50 all my adversaries were top 500 though 2010-11-09T16:00:52 but still 2010-11-09T16:01:01 well you asked for the maps on teh tcp server XP 2010-11-09T16:01:12 and yes, those can be downloaded I think 2010-11-09T16:01:40 boegel: http://72.44.46.68/maps/map851.txt 2010-11-09T16:01:44 repeat for all 999 2010-11-09T16:01:52 JamesMG: ah, thx! 2010-11-09T16:02:09 not sure if there is a more official/easier way 2010-11-09T16:02:52 http://72.44.46.68/maps.tar.gz 2010-11-09T16:03:01 there is 2010-11-09T16:03:35 ah :) 2010-11-09T16:04:17 coolm thx guys! 2010-11-09T16:04:27 my laptop will know what to do tonight :P 2010-11-09T16:04:48 I think it was mega1 who pointed out that some of those maps are dups 2010-11-09T16:05:04 I tihnk all the ones in the 200's IIRC 2010-11-09T16:06:32 awesome, I'm timing out again 2010-11-09T16:09:01 JamesMG: your fault or server's? 2010-11-09T16:09:21 looking like mine 2010-11-09T16:09:48 *** mceier has quit IRC (Quit: leaving) 2010-11-09T16:09:55 wow, I wonder what loop I screwed up to get it to take a second 2010-11-09T16:10:04 seeing as it was running in .1s last friday 2010-11-09T16:10:38 O.o 2010-11-09T16:12:44 *** amin10043 has quit IRC (Quit: Page closed) 2010-11-09T16:16:09 Top 10 players: bocsimacko(4263), bix0r4ever(4140), luksian(4136), george(4089), rsergio(4082), shangas(4064), Accoun(4061), GreenTea(4060), dmj111(4021), medrimonia(4015) 2010-11-09T16:17:55 *** boegel has quit IRC (Quit: Leaving) 2010-11-09T16:18:53 *** Cold-Phoenix has joined #aichallenge 2010-11-09T16:25:09 I really feel bad for the neutral citizens in all this carnage, they didn't ask for this unjust war in space! 2010-11-09T16:26:40 The real question is why don't they make some ships to defend themselves with 2010-11-09T16:26:51 yeah they should see it coming 2010-11-09T16:26:57 They would easily dominate the aggressors 2010-11-09T16:27:07 I can't tell if it's mono or hardware, but on my laptop the gamestate that timed me out on official takes .25s :( 2010-11-09T16:27:11 I wish I was playing as neutral :( 2010-11-09T16:27:16 linux desktop takes a second though 2010-11-09T16:27:51 brb roasting some monkey nuts 2010-11-09T16:28:03 the monkeys must hate you 2010-11-09T16:29:04 fried tastes just as good and takes less time :P 2010-11-09T16:29:21 Seriously? You can fry monkey nuts? :) 2010-11-09T16:30:42 can't you fry anything? 2010-11-09T16:30:53 I suppose with some caramelised sugar you are sorted! 2010-11-09T16:36:40 yuck. 2010-11-09T16:37:05 fried icecream is pretty decent 2010-11-09T16:37:20 perhaps heartstoppingly so 2010-11-09T16:37:26 *** rebelxt has joined #aichallenge 2010-11-09T16:38:45 *** rebelxt has left #aichallenge 2010-11-09T16:42:31 deep fried? 2010-11-09T16:42:34 in batter? 2010-11-09T16:42:40 or how does that work? 2010-11-09T16:42:53 thats the way they fry snickers 2010-11-09T16:44:16 How will the final rankings be calculated? 2010-11-09T16:44:40 Surely on the last day there will be loads of last minute submissions - is it going to be left to settle for a while or what? 2010-11-09T16:44:50 *** Mathnerd314 has quit IRC (Ping timeout: 252 seconds) 2010-11-09T16:45:42 hrm, seems the final rankings are going to be released only a few days after submission closes 2010-11-09T16:46:20 wait there's an actual product called monkey nuts? 2010-11-09T16:47:12 d0ugal I think in addition to that the top 100 will play a round robin but who knows, it's hard to get a straight answer 2010-11-09T16:47:18 lavalamp_: A product of nature even. 2010-11-09T16:47:29 O.O 2010-11-09T16:47:39 http://en.wikipedia.org/wiki/Peanut - Peanuts are known by many other local names, like earthnuts, ground nuts, goober peas, monkey nuts, pygmy nuts and pig nuts. 2010-11-09T16:47:59 I certainly didn't know most of them, but we call 'em monkey nuts here. 2010-11-09T16:48:13 lavalamp_: It'll be interesting to see the final 100 battle it out 2010-11-09T16:48:14 ah thanks, lol 2010-11-09T16:48:36 I hope some people will release code and/or articles about their bots. I'd love to learn and see how other people are doin' it 2010-11-09T16:48:37 we should make a betting pool on how many bocsimacko will lose if they do that 2010-11-09T16:49:44 mmmmm deese nuts are nice 2010-11-09T16:49:47 Need some seasoning tho 2010-11-09T16:51:00 * d0ugal is going to much on some Pistachio nuts 2010-11-09T16:51:09 s/much/munch/ 2010-11-09T16:52:38 *** Mathnerd314 has joined #aichallenge 2010-11-09T17:04:30 *** sigh has joined #aichallenge 2010-11-09T17:10:07 bhasker: ugh i hate those last to games >_< 2010-11-09T17:10:25 really should've gained more neutrals 2010-11-09T17:10:28 but nooo 2010-11-09T17:10:38 hehe:) 2010-11-09T17:10:42 this is my latest submission bot 2010-11-09T17:10:52 my new version sucks 2010-11-09T17:10:54 you win a lot like that =) 2010-11-09T17:10:58 oh 2010-11-09T17:11:05 i jsut fixed a couple of bugs in the last submission 2010-11-09T17:11:08 and resubmitted 2010-11-09T17:14:04 i see 2010-11-09T17:14:22 i'm working out a good action set on paper 2010-11-09T17:14:26 for alphabeta 2010-11-09T17:14:32 i think i got it... 2010-11-09T17:14:44 its gonna be quite expensive though =) 2010-11-09T17:14:51 but nice nontheless 2010-11-09T17:15:05 branchingfactor should not exceed 200 2010-11-09T17:15:11 or sth 2010-11-09T17:16:09 Top 10 players: bocsimacko(4253), luksian(4127), bix0r4ever(4122), george(4082), rsergio(4079), Accoun(4060), GreenTea(4051), shangas(4051), Raschi(4009), Xrillo(4008) 2010-11-09T17:17:56 the thing is 2010-11-09T17:18:03 i got a good idea now 2010-11-09T17:18:11 but the implementation will take some time 2010-11-09T17:18:24 s/some/a lot 2010-11-09T17:18:27 and i hate that :P 2010-11-09T17:19:24 you gonna hate this one too error323 2010-11-09T17:19:40 lol 2010-11-09T17:19:43 >_> 2010-11-09T17:19:51 another 200 turn game coming up 2010-11-09T17:19:58 lovely 2010-11-09T17:20:04 *** sigh has quit IRC (Read error: Connection reset by peer) 2010-11-09T17:20:27 *** sigh has joined #aichallenge 2010-11-09T17:21:24 the next one is yours 2010-11-09T17:21:27 my bot made a blunder 2010-11-09T17:21:32 sent like 220 ships to a far neutral:-\ 2010-11-09T17:23:09 how are you watching that live? 2010-11-09T17:24:27 @jbotmanager 2010-11-09T17:24:28 lavalamp_: jbotmanager = http://www.mabs.at/ewing/JBotManager.jar 2010-11-09T17:24:53 warning, watching tcp games live will kill your productivity 2010-11-09T17:24:59 lol 2010-11-09T17:26:06 :D 2010-11-09T17:26:09 already does for me 2010-11-09T17:26:45 E323 http://72.44.46.68/canvas?game_id=931100 2010-11-09T17:27:21 ahw 2010-11-09T17:27:23 got sniped 2010-11-09T17:27:38 oh man.. epic plans for next contest... we'll have hack nights for coding 2010-11-09T17:28:31 good heavens please don't do a game that interests me 2010-11-09T17:28:37 indeed 2010-11-09T17:28:45 this stuff is way to addictive ;-) 2010-11-09T17:28:53 can you announce what the game will be yet? 2010-11-09T17:28:58 or my wife may try to kill one of us 2010-11-09T17:29:03 all i do is check my ratings or think about improvements :P 2010-11-09T17:31:22 *** p4p4 has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2010-11-09T17:31:25 damn this is gona be nice 2010-11-09T17:31:31 mega1: here i come :P 2010-11-09T17:31:39 :> 2010-11-09T17:32:01 Error323: much improved? 2010-11-09T17:32:39 mega1: well the idea is on paper 2010-11-09T17:32:43 and the framework is there 2010-11-09T17:32:58 Now lets hope it also works in practice ^_^ 2010-11-09T17:33:15 * mega1 was already checking tcp 2010-11-09T17:33:27 soon =) 2010-11-09T17:35:15 *** Bobng has joined #aichallenge 2010-11-09T17:35:21 lavalamp same problem here :) 2010-11-09T17:35:29 my wife is already losing it with me spending so much time on the bot 2010-11-09T17:36:25 how old are you bhasker if i may ask 2010-11-09T17:36:42 <-- 27 2010-11-09T17:36:43 *** RainCT has quit IRC (Remote host closed the connection) 2010-11-09T17:36:55 <--32 2010-11-09T17:37:05 = 2010-11-09T17:37:06 0 2010-11-09T17:37:08 = 2010-11-09T17:37:10 ugh 2010-11-09T17:37:14 i mean =) 2010-11-09T17:37:20 :) 2010-11-09T17:37:23 <-- 28 2010-11-09T17:39:01 me 2010-11-09T17:39:03 men 2010-11-09T17:39:07 *** Bobng has quit IRC (Client Quit) 2010-11-09T17:39:08 i'm gonna be coding all night 2010-11-09T17:39:09 lol 2010-11-09T17:39:10 well that was a mighty nice sleep 2010-11-09T17:39:26 after spending 40 hours awake aha 2010-11-09T17:42:07 *** Prillicy has joined #aichallenge 2010-11-09T17:42:10 antimatroid1: so when can we expect to see you on the leaderboard? 2010-11-09T17:42:37 I am not doing well playing with my bot on the TCP server, keeps timing out for some reason 2010-11-09T17:43:06 Prillicy: first move? or later? 2010-11-09T17:43:09 works fine on the main challenge 2010-11-09T17:43:11 first move 2010-11-09T17:43:34 :( 2010-11-09T17:43:50 I suspect my port isn't open? 2010-11-09T17:43:58 try telnet manually 2010-11-09T17:44:11 um...k 2010-11-09T17:44:17 amstan: no clues about the next contest? 2010-11-09T17:44:24 well, that will tell you if you can actually connect 2010-11-09T17:44:36 I want a clue! 2010-11-09T17:44:37 don't know how 2010-11-09T17:44:38 phreeza: what? as in game? no 2010-11-09T17:44:43 phreeza: suggestions are open though 2010-11-09T17:45:03 Prillicy: what os? 2010-11-09T17:45:13 ah ok it sounded like you already had an idea 2010-11-09T17:45:14 windows 7 2010-11-09T17:46:59 I dunno how to do much on windows 7 :( 2010-11-09T17:47:06 lol 2010-11-09T17:47:16 too bad we can't have competitions on IRC 2010-11-09T17:47:17 sgh: a few days at least 2010-11-09T17:47:21 I should write a bot for that 2010-11-09T17:47:25 * Prillicy ponders. 2010-11-09T17:47:26 :) 2010-11-09T17:47:31 i'm about to ride to the uni to give a friend a vote in some election 2010-11-09T17:47:39 then i'll get back to work, but i don't want to rush it 2010-11-09T17:47:40 *** demerzel_ has quit IRC (Quit: Page closed) 2010-11-09T17:47:44 yeah, fair enough 2010-11-09T17:48:08 who's j3camero and aren't we all developers? 2010-11-09T17:48:19 j3camero started all this 2010-11-09T17:48:22 j3camero is the mythical jeff 2010-11-09T17:48:34 thought it was started by google 2010-11-09T17:48:35 he was more present last comp, he started a new job when this started 2010-11-09T17:48:38 actually I have seen him say something now 2010-11-09T17:48:46 @seen j3camero 2010-11-09T17:48:46 JamesMG: j3camero was last seen in #aichallenge 4 days, 19 hours, 50 minutes, and 48 seconds ago: I shall stay out of this debate. Anybody who brings up strong typing obviously has an extreme opinion on the subject. They're usually from Haskell-land, where the trolls lurk. 2010-11-09T17:48:47 no, googles name is there because it looks good for google and the competition 2010-11-09T17:48:50 wow 2010-11-09T17:48:52 nothing more, nothing less 2010-11-09T17:48:56 so I guess I finally have to admit he exists 2010-11-09T17:49:06 *** Azrathud has joined #aichallenge 2010-11-09T17:49:14 :p 2010-11-09T17:49:23 so google doesn't get involved at all? 2010-11-09T17:49:27 no 2010-11-09T17:49:28 haha, I remember that conversation about strong typing 2010-11-09T17:49:34 so do i :P 2010-11-09T17:50:18 but it says "sponsored by" 2010-11-09T17:50:23 hmm 2010-11-09T17:50:32 "here, have our name" 2010-11-09T17:50:55 Prillicy: like i said, it looks good for both parties 2010-11-09T17:51:03 yay, party 2010-11-09T17:51:11 heh 2010-11-09T17:51:19 who knows, if it keeps blowing out, they might have an interest to invest some time/money into the competitions 2010-11-09T17:51:30 hey, anyone who was in the tron contest... how many entries were therein the end? 2010-11-09T17:51:43 around 750 at the start of final contest 2010-11-09T17:51:52 a few, like myself, weren't there at the end 2010-11-09T17:52:00 wow... this contest is much larger 2010-11-09T17:52:16 http://csclub.uwaterloo.ca/contest/rankings.php?page=8 2010-11-09T17:52:19 708 ended 2010-11-09T17:52:28 j3camero in 704th place aha 2010-11-09T17:53:00 like i said, "if it keeps blowing out" :P 2010-11-09T17:53:03 bocsimacko is 11th :) 2010-11-09T17:53:20 4000+ entries is pretty big already 2010-11-09T17:53:21 yeah i noticed that the other night, he also has a blog 2010-11-09T17:53:26 link? 2010-11-09T17:53:35 look on his profile 2010-11-09T17:53:48 there may/may not have been posts following tron *whistles* 2010-11-09T17:53:58 although nothing really new 2010-11-09T17:54:09 anyway, i'll be back later 2010-11-09T17:55:39 ok cool, he's tried UCT as well 2010-11-09T17:55:41 for tron 2010-11-09T17:55:58 and gave up on it 2010-11-09T17:55:59 *** nann has joined #aichallenge 2010-11-09T17:56:15 it really wasn't hard to get a good minimax bot going for tron 2010-11-09T17:56:25 i pretty much half learnt to program properly with that contest 2010-11-09T17:56:32 :) 2010-11-09T17:57:11 did you see some of the games of my tron bot? :P 2010-11-09T17:57:26 looking at my user page the other night and those of the top 5 bots was a little bit of a bummer 2010-11-09T17:57:35 #rankings 2010-11-09T17:58:03 http://csclub.uwaterloo.ca/contest/profile_games.php?user_id=2788 :( 2010-11-09T17:59:06 http://csclub.uwaterloo.ca/contest/visualizer.php?game_id=4115976 he would take them out as soon as he could haa 2010-11-09T17:59:27 *** Apophis has quit IRC (Ping timeout: 240 seconds) 2010-11-09T18:01:18 http://csclub.uwaterloo.ca/contest/visualizer.php?game_id=4103894 the stupid bot could have won that too 2010-11-09T18:04:11 *** Apophis has joined #aichallenge 2010-11-09T18:05:24 *** Utkarsh has quit IRC (Ping timeout: 265 seconds) 2010-11-09T18:05:50 hmm 2010-11-09T18:06:20 @rankings 2010-11-09T18:06:21 Prillicy: Top 10 players: bocsimacko(4273), bix0r4ever(4133), luksian(4128), george(4115), rsergio(4091), Accoun(4085), GreenTea(4070), shangas(4060), medrimonia(4042), Raschi(4041) 2010-11-09T18:06:58 *** nann has quit IRC (Ping timeout: 260 seconds) 2010-11-09T18:09:11 *** superflit has quit IRC (Quit: superflit) 2010-11-09T18:11:12 *** Utkarsh has joined #aichallenge 2010-11-09T18:11:27 *** Prillicy has quit IRC (Quit: I be bangin') 2010-11-09T18:12:04 i thoroughly enjoyed mega1's post on UCT for tron. i share those sentiments regarding public discussion of strategies / algorithms. there's a sense of panic within whenever i read someone in here or on the forums skirt around ideas I want to try. :) 2010-11-09T18:13:09 *** fawek has quit IRC (Read error: Operation timed out) 2010-11-09T18:13:26 as often in life, ideas are a dime a dozen, implementing well is what matters most, I think 2010-11-09T18:13:40 and picking the good ideas, maybe 2010-11-09T18:13:49 but mostly execution 2010-11-09T18:14:20 that's true, of course. 2010-11-09T18:14:23 i just branched my code off to throw away the previous version and write a min max bot 2010-11-09T18:14:32 back to -1000 elo rating soon 2010-11-09T18:15:21 i do remember instances of competitions where only a handful of people stumbled across an idea outrageous enough that not everyone was considering it. 2010-11-09T18:16:09 Top 10 players: bocsimacko(4275), bix0r4ever(4135), luksian(4131), george(4118), rsergio(4093), Accoun(4087), GreenTea(4072), shangas(4062), medrimonia(4044), Raschi(4043) 2010-11-09T18:16:18 *** Utkarsh has quit IRC (Ping timeout: 276 seconds) 2010-11-09T18:17:06 *** JamesMG has quit IRC (Quit: Leaving.) 2010-11-09T18:18:18 i guess the transition between genuine ideas and implementation details is fluid, too... 2010-11-09T18:18:50 *** Migi32 has quit IRC (Remote host closed the connection) 2010-11-09T18:19:28 really looking forward to mega1s postmortem :) 2010-11-09T18:21:46 *** mceier has joined #aichallenge 2010-11-09T18:28:00 *** yasith has quit IRC (Ping timeout: 276 seconds) 2010-11-09T18:28:12 I'm secretly hoping he got UCT working because that would be a really interesting read :) 2010-11-09T18:29:01 but I don't think that's likely 2010-11-09T18:29:20 hm 2010-11-09T18:29:34 yea 2010-11-09T18:29:52 phreeza: do you play Go? 2010-11-09T18:30:23 no... i mean i have dabbled, but i've never won a game 2010-11-09T18:30:29 haha :) 2010-11-09T18:30:36 *** Mathnerd314 has quit IRC (Ping timeout: 264 seconds) 2010-11-09T18:30:47 *** Utkarsh has joined #aichallenge 2010-11-09T18:31:01 a guy came into a lab i used to work at a week before i left 2010-11-09T18:31:10 he aparently got everyone there hooked 2010-11-09T18:31:24 so if i had stayed around... but no 2010-11-09T18:31:25 :) 2010-11-09T18:31:28 :) 2010-11-09T18:31:39 do you? 2010-11-09T18:31:46 yeah, it's a lot easier and more fun to learn if you have other people starting with you 2010-11-09T18:31:56 yeah, not so much recently though 2010-11-09T18:31:58 i bet 2010-11-09T18:32:31 whats your handicap? i forget the fancy japanese word :D 2010-11-09T18:33:01 I used to be able to hold a 4d on KGS 2010-11-09T18:33:07 4d rank 2010-11-09T18:33:31 aha so if you play against a total beginner you give him 4 stones? 2010-11-09T18:34:14 no... ranks go from ~30 kyu to 1k then 1 dan to ~7 d 2010-11-09T18:34:33 well, more like ~9 usually 2010-11-09T18:34:48 ah wow 2010-11-09T18:34:52 so thats pretty good 2010-11-09T18:35:48 it's not bad... but improvement in the dan ranks is hard :) 2010-11-09T18:35:58 Advanced amateur according to wikipedia 2010-11-09T18:35:58 so there is a *very* large difference from where I am to the top 2010-11-09T18:36:17 ok 2010-11-09T18:37:29 *** Utkarsh has quit IRC (Ping timeout: 240 seconds) 2010-11-09T18:37:38 as a comparison... best go bots are around 2d 2010-11-09T18:39:57 *** yasith has joined #aichallenge 2010-11-09T18:39:57 nice so you can beat them 2010-11-09T18:40:12 *** Mathnerd314 has joined #aichallenge 2010-11-09T18:40:37 yeah... well, also I know how to exploit them to some extent 2010-11-09T18:40:46 so it's not exactly fair :P 2010-11-09T18:41:09 haha 2010-11-09T18:41:39 is there something like computer/human team go? 2010-11-09T18:41:52 hmm... I haven't seen anyone do that 2010-11-09T18:41:58 thats what i find really interesting 2010-11-09T18:42:03 in chess too 2010-11-09T18:42:16 eliminate the chance for tactical blunders 2010-11-09T18:42:30 I suspect for go the human would find it frustrating, computer style is VERY different to human style 2010-11-09T18:42:55 tactical blunders are easy in go, computers make them too :P 2010-11-09T18:42:57 *** Utkarsh has joined #aichallenge 2010-11-09T18:43:05 oh :) 2010-11-09T18:45:46 *** jaspervdj has quit IRC (Quit: NEVER GONNA GIVE YOU UP NEVER GONNA LET YOU DOWN) 2010-11-09T18:46:18 even locally, searching the variations is not trivial and humans bring together a lot of heuristics and pattern matching to be able to do it effectivly 2010-11-09T18:47:06 and in go it is easy to win the battle, but lose the war... capture a group you were aiming for but come out behind overall 2010-11-09T18:47:52 *** Utkarsh has quit IRC (Ping timeout: 272 seconds) 2010-11-09T18:49:42 *** justin_pdx has quit IRC (Quit: justin_pdx) 2010-11-09T18:50:32 yea really need to get into it more... 2010-11-09T18:50:39 so many things to do 2010-11-09T18:50:41 so little time 2010-11-09T18:50:51 anyone in here need jeff? 2010-11-09T18:50:54 yeah, it can suck up a lot of time :) 2010-11-09T18:51:05 speaking of which... i should get off irc 2010-11-09T18:51:13 i can summon him! 2010-11-09T18:51:26 :D 2010-11-09T18:51:49 janzert: ping 2010-11-09T18:51:55 thats j3camero? 2010-11-09T18:52:02 no 2010-11-09T18:52:32 hell get a load of messages from contestbot if he shows up here i think :) 2010-11-09T18:53:47 *** Utkarsh has joined #aichallenge 2010-11-09T18:54:33 *** Guest82843 has quit IRC (Quit: Page closed) 2010-11-09T19:03:57 damnit 2010-11-09T19:04:14 i forgot the name of this symbol in math: "x" 2010-11-09T19:04:23 f : S x A 2010-11-09T19:04:29 what's the "x" again 2010-11-09T19:04:33 omg 2010-11-09T19:04:52 *** janstaunton has quit IRC (Quit: janstaunton) 2010-11-09T19:05:04 *** Utkarsh has quit IRC (Ping timeout: 245 seconds) 2010-11-09T19:06:11 depends :D 2010-11-09T19:06:16 cross product 2010-11-09T19:06:29 outer product 2010-11-09T19:06:30 i thought so too, but thats in linear algebra 2010-11-09T19:06:55 ahh 2010-11-09T19:06:59 the cartesian product 2010-11-09T19:07:00 :D 2010-11-09T19:07:19 karthesian product 2010-11-09T19:07:21 ah yea 2010-11-09T19:07:25 carthesian 2010-11-09T19:07:30 :D 2010-11-09T19:10:15 *** bhasker_ has joined #aichallenge 2010-11-09T19:12:15 *** bhasker__ has joined #aichallenge 2010-11-09T19:13:12 *** bhasker has quit IRC (Ping timeout: 272 seconds) 2010-11-09T19:14:35 *reads back* sigh is being modest... 4d KGS is pretty good 2010-11-09T19:14:39 *** bhasker_ has quit IRC (Ping timeout: 245 seconds) 2010-11-09T19:14:54 thought so 2010-11-09T19:15:28 It's probably at least 6d AGA 2010-11-09T19:15:31 er, 5d 2010-11-09T19:15:49 where AGA = usgo.org 2010-11-09T19:16:10 Top 10 players: bocsimacko(4280), bix0r4ever(4150), luksian(4139), george(4116), Accoun(4102), rsergio(4092), shangas(4080), GreenTea(4079), medrimonia(4049), Raschi(4048) 2010-11-09T19:25:03 *** JensTi has joined #aichallenge 2010-11-09T19:58:56 *** AlliedEnvy has joined #aichallenge 2010-11-09T20:00:03 *** tapwater has quit IRC (Remote host closed the connection) 2010-11-09T20:00:23 *** tapwater has joined #aichallenge 2010-11-09T20:07:44 *** tapwater has quit IRC (Remote host closed the connection) 2010-11-09T20:08:12 *** tapwater has joined #aichallenge 2010-11-09T20:16:16 Top 10 players: bocsimacko(4275), luksian(4182), bix0r4ever(4155), george(4114), Accoun(4112), rsergio(4094), GreenTea(4091), shangas(4080), Raschi(4075), Hazard(4052) 2010-11-09T20:23:57 *** mceier has quit IRC (Quit: leaving) 2010-11-09T20:27:33 *** JensTi has quit IRC (Quit: Leaving.) 2010-11-09T20:37:27 *** xxxss has joined #aichallenge 2010-11-09T20:42:53 *** demerzel has joined #aichallenge 2010-11-09T20:47:47 *** lavalamp__ has joined #aichallenge 2010-11-09T20:50:10 *** Sylph has joined #aichallenge 2010-11-09T20:50:42 holy fucking crap 2010-11-09T20:50:45 its working 2010-11-09T20:51:00 i created awesomeness 2010-11-09T20:51:02 ^_^ 2010-11-09T20:56:50 i am screwed 2010-11-09T20:56:58 i must solve this stdin problem now i don't even care how stupid it is 2010-11-09T20:57:00 *** virdo has joined #aichallenge 2010-11-09T20:57:04 gonna waste a week on it haha 2010-11-09T20:57:15 MUST SOLVE 2010-11-09T21:09:47 *** Mathnerd314 has quit IRC (Quit: ChatZilla 0.9.86-rdmsoft [XULRunner 1.9.2.12/20101026210630]) 2010-11-09T21:12:08 *** phreeza_ has joined #aichallenge 2010-11-09T21:15:26 *** phreeza has quit IRC (Ping timeout: 252 seconds) 2010-11-09T21:16:01 does anyone know where i can find the format of the strings the engine gives to me each turn? 2010-11-09T21:16:06 am i going to have to just debug it 2010-11-09T21:16:12 Top 10 players: bocsimacko(4266), luksian(4184), bix0r4ever(4153), george(4124), Accoun(4119), rsergio(4107), GreenTea(4094), Raschi(4092), shangas(4072), Hazard(4054) 2010-11-09T21:20:23 the format? 2010-11-09T21:20:26 what do you mean 2010-11-09T21:20:43 i know you've been around too long to just give you a link to the specs lol 2010-11-09T21:21:00 ...lol I was about to do that http://ai-contest.com/specification.php 2010-11-09T21:21:58 lavalamp_ figure out my stdin problem please 2010-11-09T21:22:02 i can't let it go now 2010-11-09T21:22:07 lol 2010-11-09T21:22:34 i actually got it working in windows with the HANDLE thing but obv that isn't gonna work in linux 2010-11-09T21:25:48 lol 2010-11-09T21:25:54 hrm 2010-11-09T21:26:29 in the middle of coding furiously, I will look at that again when I get stuck 2010-11-09T21:28:12 well here's what actually works in windows - using the STD_INPUT_HANDLE which is a DWORD -12 and freading it 2010-11-09T21:28:20 but that doesn't apply in linux obv 2010-11-09T21:30:01 *** Mathnerd314 has joined #aichallenge 2010-11-09T21:30:49 maybe the ansi functions are buggy on windows (won't be the first time) and you have a differente problem in linux 2010-11-09T21:31:14 yes more than likely 2010-11-09T21:31:18 but i have no idea how to identify the problem in linux 2010-11-09T21:32:53 have you tried searching for a program in C that just echos stdin to stdout? 2010-11-09T21:38:28 nope i will do that tho 2010-11-09T21:43:14 yeah, no i mean what the engine is passing me each turn 2010-11-09T21:43:17 it has "go" in it 2010-11-09T21:43:48 i'm rewriting my entire thing, and the starter pack from scratch 2010-11-09T21:44:20 i'm going to just debug it to file, i'll share the results 2010-11-09T21:44:25 antimatroid1, it passes you planets, then fleets, then the word "go" 2010-11-09T21:44:38 the spec page spells it out pretty clearly, I thought 2010-11-09T21:44:39 oh shit, so it is just like that 2010-11-09T21:44:51 ^^ 2010-11-09T21:44:51 yeah, it does, sorry 2010-11-09T21:45:11 okay, so just read line till i receive go then bam 2010-11-09T21:48:16 i just look for 'g' 2010-11-09T21:48:26 it's either P F or g 2010-11-09T21:48:29 or I if you're on tcp 2010-11-09T21:48:34 yeah i just changed that 2010-11-09T21:48:56 i like to be able to look at something and know what it's doing 2010-11-09T21:49:07 i removed all those tokenize functions / classes last night 2010-11-09T21:49:09 they were horrible 2010-11-09T21:51:51 yea it's amazingly easy in C 2010-11-09T21:51:58 i fscanf(stdin line 2010-11-09T21:52:10 1 even 2010-11-09T21:52:25 same with any C really though i dno't know they went with tokenize 2010-11-09T21:55:45 this is just hilarious http://72.44.46.68/canvas?game_id=934189 2010-11-09T21:55:50 talk about comeback 2010-11-09T21:56:01 it's amazingly easy... says the guy who's spent hours trying to get it to work >.< 2010-11-09T21:57:39 yea no shit 2010-11-09T21:57:47 Error323 that's impossible 2010-11-09T21:57:49 i am totally baffled at how goddamn ridiculously hard this is 2010-11-09T21:57:57 more so that there isn't a quick google solution 2010-11-09T21:58:21 yeah.... 2010-11-09T21:59:05 i've been coding in different languages for over a decade and i've never hit a brick wall like this 2010-11-09T21:59:18 it'd be different if i had no idea what stdin was or how files worked in C 2010-11-09T21:59:36 now it's turned into a MISSION 2010-11-09T22:00:42 i wish dhartmei was around i'd ask him about his c code from tron 2010-11-09T22:01:08 *** Rubicon-|-Cross has joined #aichallenge 2010-11-09T22:01:32 isn't his code open source? 2010-11-09T22:01:42 tron was a while ago ^^ 2010-11-09T22:02:01 yea i have the link to it 2010-11-09T22:02:05 unfortunately i don't udnerstand it 2010-11-09T22:02:13 link me? 2010-11-09T22:02:19 he's got some kind of fcntl fd_set thing going on and part of it has async stuff 2010-11-09T22:02:37 http://www.benzedrine.cx/tron/tronbot/MyTronBot.cc 2010-11-09T22:04:45 I don't think that's async in the sense of multiple threads, he's just (potentially?) waiting on more than one file handle 2010-11-09T22:05:05 i have no idea 2010-11-09T22:05:14 i could copy it word for word but it wouldn't work in windows 2010-11-09T22:06:00 i want to know why raedwulf's code works and mine doesn't 2010-11-09T22:06:06 i copied his code directly lol 2010-11-09T22:07:22 hahaha that sucks 2010-11-09T22:07:39 his is literally just a looping fgets 2010-11-09T22:07:50 he said he 'had problems with it going too fast' so parsed right then and there 2010-11-09T22:07:58 and that was literally it 2010-11-09T22:08:21 i already parse and on top of that assign values to structs so idk wtf 2010-11-09T22:08:34 *** tapwater_ has joined #aichallenge 2010-11-09T22:08:51 look up the documentation for select? I don't see where dhartmei is actually reading the value out... :/ 2010-11-09T22:10:12 yea me neither 2010-11-09T22:10:24 i once had a game where our ships count graph was a near perfect polynomial 2010-11-09T22:10:35 like, way less noise than that one 2010-11-09T22:10:42 *** tapwater has quit IRC (Ping timeout: 265 seconds) 2010-11-09T22:10:43 *** tapwater_ is now known as tapwater 2010-11-09T22:10:52 he's treating stdin like a socket somehow lol 2010-11-09T22:11:01 so you can compile raed's code and it works for you but not when you copy it? :/ 2010-11-09T22:11:16 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-11-09T22:11:19 well his doesn't work for me for a variety of reasons - i don't know what some of his decs are 2010-11-09T22:11:27 it's very basic though 2010-11-09T22:11:30 i just completely changed my code for handling game state strings without hitch, i have no idea how you're having this much trouble 2010-11-09T22:11:31 http://pastebin.ca/1986003 & 4 2010-11-09T22:11:35 *** Rubicon-|-Cross has joined #aichallenge 2010-11-09T22:11:42 i can't read stdin anti 2010-11-09T22:11:46 on unix each file (socket, pipe-- all the same) gets a number, fdset is a bitarray of them 2010-11-09T22:11:50 are you vanilla C? i'd like to see your stdin 2010-11-09T22:12:27 it complicates the issue that what might fix it in windows won't in linux and vice versa and i have no way to test in linux other than just blindly uploading 2010-11-09T22:12:49 never got that virtualbox setup? 2010-11-09T22:13:07 mines in c++ and i'm using windows 2010-11-09T22:13:19 i'm still cleaning it up a bit, then i'll post what i did for you 2010-11-09T22:14:00 well if its cin << std::string then don't worry about it but if it's using C stuff i'd like to see it 2010-11-09T22:14:25 lavalamp_ - there is no way it would be like a good 4-5 hour learning experience plus i have no idea how i would setu p an engine etc etc 2010-11-09T22:15:03 i'm just using getline and passing cin as an input stream 2010-11-09T22:15:31 that's to construct the map_data string, for passgamestate i'm using string streams 2010-11-09T22:16:13 Top 10 players: bocsimacko(4262), luksian(4157), bix0r4ever(4157), george(4123), Accoun(4123), GreenTea(4096), rsergio(4096), Raschi(4090), shangas(4063), Hazard(4043) 2010-11-09T22:16:22 we did our tron bot trying to use mostly just c, but it ended up being way too messy to be worthwhile imo 2010-11-09T22:19:11 well i've got full-on tcp socket code, nodes, parsing etc already setu p in C 2010-11-09T22:19:19 http://pastebin.com/75TWrjbG 2010-11-09T22:19:29 demerzel, does that work for you? 2010-11-09T22:19:40 no fgets just freezes 2010-11-09T22:20:03 i mean i'll try it 2010-11-09T22:20:19 but it never works - i would say a windows thing but fgets was in the version i uploaded to the main server 2010-11-09T22:20:43 hm that works fine for me, typing and piping 2010-11-09T22:21:14 maybe you have some problem other than reading? 2010-11-09T22:21:28 is *everyone* rewriting the io code? 2010-11-09T22:21:53 sigh, all the cool kids are doing it 2010-11-09T22:21:59 sigh: imo it really doesn't matter, i just did it for aesthetic purposes 2010-11-09T22:22:09 I rewrote it only because I did the go starter package... 2010-11-09T22:22:23 yeah, well that actually makes sense :P 2010-11-09T22:22:41 what i do makes sense in the fact that what i do hardly ever makes sense 2010-11-09T22:22:41 what got you interested in go lang, anyway? 2010-11-09T22:23:02 i rewrote it in all the other ones thsi is the only one that makes any problem lol 2010-11-09T22:23:43 well the name at first, obviously ^^ 2010-11-09T22:24:09 :P 2010-11-09T22:24:11 but then I decided to learn it (in the process made dailyjoseki.com) and it turned out I really like it 2010-11-09T22:24:40 *** Azrathud has quit IRC (Read error: Connection reset by peer) 2010-11-09T22:25:11 i expect i'll really like haskell when i get around to learning it 2010-11-09T22:25:15 *** Azrathud has joined #aichallenge 2010-11-09T22:25:20 cool :) 2010-11-09T22:25:23 *** Azrathud has quit IRC (Read error: Connection reset by peer) 2010-11-09T22:26:26 *** Azrathud has joined #aichallenge 2010-11-09T22:31:44 lavalamp_ i c/p'd your code and uploaded it to the server 2010-11-09T22:31:53 only thing i added was a strcpy from line to io_ lol 2010-11-09T22:32:17 *** Azrathud has quit IRC (Read error: Connection reset by peer) 2010-11-09T22:32:27 welp: http://www.ai-contest.com/visualizer.php?game_id=6888066 2010-11-09T22:32:28 crashed 2010-11-09T22:32:40 well more like didn't respodn 2010-11-09T22:40:31 huh well that code will probably issue invalid orders... >.< 2010-11-09T22:42:09 it doesn't do jack except go 2010-11-09T22:42:23 anyway i just plugged in c++ for one line and it works now 2010-11-09T22:42:24 how frustrating 2010-11-09T22:42:37 i am tired of dicking with it 2010-11-09T22:42:52 but it annoys me to death so i will probably pick at it forever :( 2010-11-09T22:43:34 http://pastebin.ca/1986979 2010-11-09T22:43:41 is this what you did? 2010-11-09T22:45:00 wait, that doesn't work locally 2010-11-09T22:45:37 ....weird 2010-11-09T22:46:07 it works when I cat file | it, and when I type at it, but not from jbotman 2010-11-09T22:46:46 no i don't use strcmp just check first char for 10 2010-11-09T22:46:48 '\n' 2010-11-09T22:47:47 haha and now it works 2010-11-09T22:47:49 how fucking stupid 2010-11-09T22:48:10 whatever i think i managed to burn myself out lo 2010-11-09T22:49:08 wait that works for you? 2010-11-09T22:49:12 lol now it doesn't work for me 2010-11-09T22:49:16 no not that 2010-11-09T22:49:27 i plugged in C++ for one line - cin.getline(io_input) 2010-11-09T22:49:29 and it works 2010-11-09T22:49:49 i knew it would i just have no idea why lol 2010-11-09T22:49:56 *** Azrathud has joined #aichallenge 2010-11-09T22:51:40 *** Sylph has quit IRC () 2010-11-09T22:53:25 well now my experimental c PLUS PLUS bot works on the site yay 2010-11-09T22:53:32 man that is going to bug me forever :( 2010-11-09T22:56:05 *** demerzel has quit IRC (Quit: Page closed) 2010-11-09T22:57:21 is there any difference to cin.getline(string) and getline(cin, string)? 2010-11-09T23:10:09 are peoples bots able to handle starting with fleets already in motion 2010-11-09T23:10:12 and was it messy? 2010-11-09T23:12:38 my heuristic bot on the server is stateless 2010-11-09T23:12:59 antimatroid1: it's only messy if your bot has state 2010-11-09T23:13:01 can throw whatever situation i like at it 2010-11-09T23:13:26 *** Rubicon-|-Cross has quit IRC (Quit: Leaving) 2010-11-09T23:13:27 geez... 5:10 am here, what am i doing in here 2010-11-09T23:13:35 and in this case, move number counts as state 2010-11-09T23:14:28 i have move number 2010-11-09T23:15:07 i've worked out how to do it anyway, the problem arises in needing to construct future states to add fleets, but i construct those based off the max distance between planets 2010-11-09T23:15:36 i'm just gonna add fleets to a queue when inputting them then add them after 2010-11-09T23:16:13 Top 10 players: bocsimacko(4259), luksian(4153), bix0r4ever(4151), george(4121), Accoun(4114), Raschi(4105), rsergio(4099), GreenTea(4099), shangas(4059), Hazard(4034) 2010-11-09T23:26:17 *** Appleman1234 has joined #aichallenge 2010-11-09T23:35:00 *** Sylph has joined #aichallenge 2010-11-09T23:39:13 *** ChanServ sets mode: +o sigh 2010-11-09T23:57:45 *** TheMacca has joined #aichallenge 2010-11-09T23:58:22 amstan: still need me? 2010-11-09T23:59:28 janzert: nah... he's probably gone