2011-11-24T00:00:10 1250 lines total, including empty lines for me 2011-11-24T00:00:24 700 pure code, 270 comments, 33 code and comments, etc 2011-11-24T00:00:40 wait a moment 2011-11-24T00:00:48 that means most of my code is in a single file 2011-11-24T00:00:55 well single translation unit 2011-11-24T00:01:21 most of mine is in two files 2011-11-24T00:01:25 state.cc and bot.cc 2011-11-24T00:01:30 i don't comment much 2011-11-24T00:01:41 although i have commented more than usual so far 2011-11-24T00:01:46 just so i know wtf i have done 2011-11-24T00:01:57 I don't comment much either 2011-11-24T00:02:02 but I have a lot of commented out code 2011-11-24T00:02:29 i try to just throw out what i don't want 2011-11-24T00:02:39 and hope i can retrieve it from a backup if absolutely necessary 2011-11-24T00:02:57 one of my files got so big that I split it into two and included one in the other. It's not a real header, just a source file included in another 2011-11-24T00:03:04 Antimony: just use git :P 2011-11-24T00:03:07 *** Cyndre has joined #aichallenge 2011-11-24T00:03:07 oops 2011-11-24T00:03:12 stupid tab completion 2011-11-24T00:03:16 i did like a bfs inside a bfs inside a bfs for hill protection 2011-11-24T00:03:20 I should learn to use version control sometime 2011-11-24T00:03:21 i ended up breaking it up into two functions 2011-11-24T00:03:32 *** delt0r_ has joined #aichallenge 2011-11-24T00:03:36 so should i 2011-11-24T00:03:39 git init; git add files; git commit -a -m "commit msg" 2011-11-24T00:03:41 but, even with version control there's a lot of quick changes where it's easier to leave stuff as comments 2011-11-24T00:03:42 there, all you need to know 2011-11-24T00:03:49 i'm retarded with git, normally i just incrementally compress my bot and chuck it in a folder :P 2011-11-24T00:04:00 git push :P 2011-11-24T00:04:08 and git pull :P 2011-11-24T00:04:29 well, you just need to commit every change, even if they are half broken or 1-2 lines 2011-11-24T00:04:37 how i use git: git pull, do stuff, git add files, git commit -m "blah", git push :P 2011-11-24T00:04:44 ah 2011-11-24T00:05:34 http://ompldr.org/vYmZkNA 2011-11-24T00:05:37 :P 2011-11-24T00:07:08 i look forward to seeing what information people used at the end of the contest 2011-11-24T00:07:23 i've done what i think are some pretty sweet functions for determining information in the last couple of days 2011-11-24T00:07:58 do you change behavior based on the # of opponents you think there are? 2011-11-24T00:08:13 or know there are if you've seen them/killed them 2011-11-24T00:08:55 i'm planning to tweak how "offensive" my combat is based on #opponent and friend ants 2011-11-24T00:08:58 but i haven't done that yet 2011-11-24T00:09:01 I have another stupid question. How long has it been since everyone started their code? Just curious. 2011-11-24T00:09:08 at least you have combat 2011-11-24T00:09:12 October 23 for me 2011-11-24T00:09:16 or 24 2011-11-24T00:09:21 ie. normally i don't want to lose as many ants as the opponent 2011-11-24T00:09:31 oct 23rd 2011-11-24T00:09:33 but if i have lots i might just go right, assuming they aren't going to kill more go on the offense 2011-11-24T00:09:58 rajanaresh: i had done path finding/food collection etc. during beta months and months ago 2011-11-24T00:10:11 last week i stripped that bot down to having no strategy and have built up from there 2011-11-24T00:10:25 managed to write the foundations of my combat code in the last week 2011-11-24T00:10:41 rajanaresh: probably three weeks ago? mostly rewrote the starter kit, started out with a collaborative diffusion bot -- scrapped it as too slow, trying to make a real bot no 2011-11-24T00:10:42 w 2011-11-24T00:10:48 it was like a tiny minimax thing where i consider just like offensive and defensive moves from all ants in a battle partition 2011-11-24T00:10:48 *** AntDroid has joined #aichallenge 2011-11-24T00:11:31 do you use mixed strategies? 2011-11-24T00:11:31 then the real trouble comes with picking out the defensive/offensive moves 2011-11-24T00:11:35 no 2011-11-24T00:11:46 i don't give states a numerical value 2011-11-24T00:11:58 i evaluate them to give back an Outcome, and then Outcome has an order operation 2011-11-24T00:12:10 which i'm hoping totally orders the outcomes, i'm pretty sure it does 2011-11-24T00:12:31 well, no, it wont be reflexive 2011-11-24T00:12:45 *** LouisMartin has quit IRC (Quit: Page closed) 2011-11-24T00:12:46 but you're screwed if your order operation isn't transitive 2011-11-24T00:12:59 a1k0n: here's another bad idea -- what if you simulate ant moves as a draw from a multi-variate normal distribution? 2011-11-24T00:13:00 why isn't yours refleive? 2011-11-24T00:13:03 *** LouisMartin has joined #aichallenge 2011-11-24T00:13:07 lol 2011-11-24T00:13:24 Antimony: for priques etc. you just use > 2011-11-24T00:13:30 > isn't reflexive 2011-11-24T00:13:40 ie. a > a is not true 2011-11-24T00:13:46 i see two different bot's ants going back and forth sometimes. None of them want to die or something. How do they do that? Do they use minimax trees for that? 2011-11-24T00:14:03 no, they're just using cowardice 2011-11-24T00:14:09 rajanaresh: you could probably get that from just being defensive 2011-11-24T00:14:15 but yes minimax will also lead to that kind of behaviour 2011-11-24T00:14:30 any ants that worry about combat will probably try and dance around until they can move in for the kill 2011-11-24T00:14:41 my ants won't 2011-11-24T00:14:45 what's difficult is not letting your enemies just push you away and claim territory 2011-11-24T00:14:45 they're too stupid 2011-11-24T00:15:12 you need your ants to be offensive enough to push the "battle front" into their territory 2011-11-24T00:15:17 sort of like a reverse tug of war? 2011-11-24T00:15:31 in the future, I'd like to see less lag time between the problem formulation and contest opening. Not that I'm sour-grapes'ing 2011-11-24T00:15:40 antimatroid: if they are being pushed they start pumping in more ants...its' amazing how they do it... 2011-11-24T00:16:01 bmh: we never planned for it to take this long, it just toook this long to get stuff ready 2011-11-24T00:16:14 if you want things to happen faster hang around after the contest and help out :P 2011-11-24T00:18:00 *** TheLinker has quit IRC (Quit: Bye) 2011-11-24T00:19:36 question related to said lag: Will the early alpha/beta testers such as Xathis be allowed to update their bot even if they had waaay more time working on theirs? shouldn't they have at least, a deadline weeks earlier? 2011-11-24T00:19:42 *** kiv has quit IRC (Quit: Leaving) 2011-11-24T00:20:09 *** Raimondi has quit IRC (Quit: Get MacIrssi - http://www.sysctl.co.uk/projects/macirssi/) 2011-11-24T00:20:18 LouisMartin: everything has been open source and publicly available 2011-11-24T00:20:41 some people have more free time than others, some have more programming experience, some languages have different features speeds etc. 2011-11-24T00:20:48 in short, no :P 2011-11-24T00:21:37 lol 2011-11-24T00:21:51 antimatroid: can I get bonus points for using haskell? ;) 2011-11-24T00:22:04 bmh: you can get street cred :P 2011-11-24T00:22:14 i'm out for a bit 2011-11-24T00:23:55 where can I cash in street cred? 2011-11-24T00:24:37 *** Raimondi has joined #aichallenge 2011-11-24T00:25:27 *** Jak_o_Shadows has joined #aichallenge 2011-11-24T00:25:52 Xathis is still at V1 since the very beginning of the contest. It clearly has a lot more time than anyone who began at the start line. If he end up winning, I think it would appear more fair if he had an adjusted deadline than if he simply had more than twice the time of every other contestant. But don't get me wrong, I contributed zero to this project and so I respect your decision. 2011-11-24T00:26:49 *** Areks has joined #aichallenge 2011-11-24T00:26:56 *** JorgeB has joined #aichallenge 2011-11-24T00:27:04 I don't see any problem with Xathis having more time 2011-11-24T00:27:36 As I'm obviously not looking forward to win that contest (Top 100 would be awesome), I thinking about the contest and its reputation, not my poor, poooor, ranking lol 2011-11-24T00:27:55 well it's highly unlikely that he'll win with v1 2011-11-24T00:27:57 *** Jak_o_Shadows1 has quit IRC (Ping timeout: 245 seconds) 2011-11-24T00:28:08 I think a lot of top contenders are holding back their final versions 2011-11-24T00:28:16 So having more than twice the time isn't a unfair advantage? what would be then? 2011-11-24T00:28:20 including Xathis probably 2011-11-24T00:29:09 of course they are. And that's why that would be way more fair if their deadline was AT LEAST 1 week before those who began at the startline. 2011-11-24T00:30:13 I just don't see the problem. Development time isn't the main cosntraint here 2011-11-24T00:31:24 but it's an important one for anyone serious about winning this contest and ready to invest themselves into it 2011-11-24T00:31:50 Why not make it your goal to beat Xathis 1st version by the end of the contest? 2011-11-24T00:31:52 nvm 2011-11-24T00:32:01 Surely yo've spent more time than xathis had for his first version 2011-11-24T00:32:21 because that wouldn't be a realistic goal for me. 2011-11-24T00:32:26 exactly 2011-11-24T00:32:44 why would you think that? 2011-11-24T00:36:00 Antimony? 2011-11-24T00:36:38 what? 2011-11-24T00:36:59 [00:33] why would you think that? 2011-11-24T00:37:02 what's your username? 2011-11-24T00:37:06 think what? 2011-11-24T00:37:54 about what you said: [00:32] Surely yo've spent more time than xathis had for his first version 2011-11-24T00:38:10 [00:32] because that wouldn't be a realistic goal for me. [00:32] exactly 2011-11-24T00:38:32 how much time do you think Xathis spent on his first version? 2011-11-24T00:39:04 a thousand years? 2011-11-24T00:39:22 my username is LouisMartin. The best rank I had was around ~450 and I'm currently developing another bot. 2011-11-24T00:40:14 Antimony: Well since I have to presume, I can only estimate on what I know: He's been a beta tester and his bot his overly dominant from day one, without any version update. 2011-11-24T00:40:40 when did the beta period start? 2011-11-24T00:40:42 So, between 0 minutes and infinity, I'd tend to say a few days if not more than a full week. 2011-11-24T00:40:52 he cheats 2011-11-24T00:40:59 its a conspiracy 2011-11-24T00:41:04 LouisMartin: beta or not, a lot of smart people have been throwing the kitchen sink at this and no one has unseated him. He's clearly doing something smart. 2011-11-24T00:41:30 that's what I was trying to hint at 2011-11-24T00:41:48 clearly, the 1st version hasn't been worked on since october 2011-11-24T00:42:01 so other bots have at least a month with no luck unseating him 2011-11-24T00:42:17 Antimony: xathis says that every time he tries to improve it, it gets worse! 2011-11-24T00:42:19 of course his doing something smart, but time is an advantage when developing a smart solution 2011-11-24T00:42:59 haha I can see why. Awesome work. 2011-11-24T00:43:37 ? 2011-11-24T00:58:36 *** Antimony has quit IRC (Ping timeout: 252 seconds) 2011-11-24T01:03:56 heh, it's not like the beta was only open to a select group 2011-11-24T01:09:06 about the only part of all of this developed in secret was the final form of the game with hills and such and that was only known to about 4 people so far as I know. And that was only done and then launched a few weeks before the contest went live 2011-11-24T01:09:51 *** ThatSnail has quit IRC (Ping timeout: 248 seconds) 2011-11-24T01:12:19 janzert: where do I sign up for the steering committee for next time? :) 2011-11-24T01:12:36 hang around here after the contest ends :) 2011-11-24T01:12:45 and be active 2011-11-24T01:12:55 that's about it 2011-11-24T01:16:17 *** Micho has joined #aichallenge 2011-11-24T01:20:55 *** cyphase has quit IRC (Quit: http://www.cyphase.com/) 2011-11-24T01:21:49 *** cyphase has joined #aichallenge 2011-11-24T01:21:56 janzert: I didn't know about the final form thing. 2011-11-24T01:22:20 janzert: is there anything in the final that xathis (and other beta testers) don't know about? 2011-11-24T01:22:23 *** Micho has quit IRC (Quit: Page closed) 2011-11-24T01:22:36 hmm? 2011-11-24T01:22:45 *about the final 2011-11-24T01:23:03 after submissions close? or you mean the current form? 2011-11-24T01:23:13 *** thestinger has quit IRC (Quit: sleep) 2011-11-24T01:23:22 yes, after submission 2011-11-24T01:23:51 *** ltriant has quit IRC (Quit: Computer has gone to sleep) 2011-11-24T01:24:27 all the known at this point information about the finals is public 2011-11-24T01:24:48 will the final maps new and unknown? 2011-11-24T01:24:55 right 2011-11-24T01:25:06 from whatever set of generators is developed before then 2011-11-24T01:25:20 okay 2011-11-24T01:25:39 there is no select group of beta testers with any inside information 2011-11-24T01:25:39 can some of the old maps be part of the final and join the new maps? 2011-11-24T01:26:01 no or at least I'll oppose reusing any maps for the finals 2011-11-24T01:26:09 gah, i need to work on my bot. last git commit was november 1st 2011-11-24T01:26:13 not that I know of anyone in favor of it 2011-11-24T01:26:19 I wasn't suspicious in that way 2011-11-24T01:27:21 janzert: alright. 2011-11-24T01:27:25 *** roflmao has quit IRC (Quit: Leaving.) 2011-11-24T01:27:32 janzert: can we have a map where all the hills are visible from each other at the start? (but reachable by some round about way) 2011-11-24T01:27:40 cyphase: is your bot open source? 2011-11-24T01:28:15 I do have some plans (regarding length and such) for finals that I don't know if I'll actually be able to accomplish so haven't really publicly mentioned them but may have mentioned in passing to an admin or two, that's about the only "insider" information that exists ;) 2011-11-24T01:28:35 bmh: only if your bot is extra weak at that :) 2011-11-24T01:29:01 LouisMartin: I would discourage anyone from sharing a working bot before the end of the competition. 2011-11-24T01:29:13 bmh: I'm not really doing much of anything with map generation at this point, McLeopold or amstan are the ones to talk to 2011-11-24T01:29:36 janzert: oh.. shouldn't have said that, now I need my antiparanoid pills -_- lol 2011-11-24T01:29:40 LouisMartin: we ask that no one share bot code beyond the starter bot functionality before contest end 2011-11-24T01:29:42 If you want to use Haskell, I'd be happy to post most/all of my modifications to the starter bot 2011-11-24T01:29:54 bmh: I agree 2011-11-24T01:31:06 LouisMartin: the form of the finals is pretty well set and is planned to be the same as planetwars though 2011-11-24T01:31:08 bmh: monadifications to the starter bot? that'd be totally fun(ctional!) 2011-11-24T01:31:24 gonna love Haskell fanboy 2011-11-24T01:31:35 ..just trolling bmh ;) 2011-11-24T01:32:30 back to coding! 2011-11-24T01:33:47 janzert: didn't participate to planetwars and tron, but I think you guys are doing a great job. 2011-11-24T01:34:02 thanks :) 2011-11-24T01:34:09 LouisMartin, no. git isn't just for open source stuff :) 2011-11-24T01:34:40 cyphase: I know. I have sleeping (closed) projects there myself... 2011-11-24T01:35:17 LouisMartin: https://gist.github.com/1390763 <-- this is what my bot looks like. 2011-11-24T01:36:19 janzert: have you ever thought about charging for an entry fee 2, 3 weeks before the submission deadline? a few 5$ could make an interesting prizes. I doubt anyone would care to pay such a small amount 2011-11-24T01:36:37 LouisMartin: no. :) 2011-11-24T01:36:47 I don't think that will ever happen 2011-11-24T01:37:15 changes the whole tone of the contest and a big reason for the contest is to get new programmers writing a bot 2011-11-24T01:37:32 bmh: you would or you wouldn't? 2011-11-24T01:38:12 The contest is sponsored, it's largely about motivating people to educate themselves in a fun way 2011-11-24T01:38:42 that is my primary motivation for working on the contest, since I feel it can attract new programmers to try it out and that writing your own moderately complex program is one of the best ways to really learn programming 2011-11-24T01:39:08 *** JorgeB_ has joined #aichallenge 2011-11-24T01:39:12 *** Fandekasp has joined #aichallenge 2011-11-24T01:39:27 bmh: I like looking at Haskell code the way a child admire cool looking math formula on a blackboard. I eventually will get my hand in a functional language at some time, but right I don't even know the basic paradigms of Haskell.. 2011-11-24T01:39:38 *** JorgeB has quit IRC (Read error: Connection reset by peer) 2011-11-24T01:39:38 *** JorgeB_ is now known as JorgeB 2011-11-24T01:40:46 janzert: then what about accepting donation? if you don't put the wikipedia founder picture anywhere on that donation page, I'd probably contribute a little 2011-11-24T01:41:00 hehe 2011-11-24T01:41:07 I would like to accept donations 2011-11-24T01:41:56 Jeff is the person that handles all the money dealings and has been a bit afk this contest 2011-11-24T01:42:29 I'm speculating but I think the hang up now might be that he needs to get a non-profit formed to accept them first 2011-11-24T01:43:07 I see 2011-11-24T01:43:46 well, the establishment of a permanent foundation would be great 2011-11-24T01:44:16 yep, he has been working on that 2011-11-24T01:44:45 The AI World League of Gentlemen. :) 2011-11-24T01:45:03 %s/men/persons/g 2011-11-24T01:45:40 there's so much cool thing that could be done with enough help (..interest is clearly not the limit here) 2011-11-24T01:47:20 I may stick around from time to time after the contest. Not sure how I could contribute, but I'll evaluate that when we get there. 2011-11-24T01:47:54 "..of Gentlepersons." 2011-11-24T01:48:26 *** treeform_ has joined #aichallenge 2011-11-24T01:51:49 *** treeform has quit IRC (Ping timeout: 240 seconds) 2011-11-24T01:52:12 janzert: have you heard of "Terrarium"? an old tech demo (contest?) to show off .Net a few years ago? http://terrarium2.codeplex.com/ 2011-11-24T01:53:20 hmm, seems like I might have heard about it in passing. But no real knowledge about it no 2011-11-24T01:53:57 do you guys have already have ideas for contests in the future? 2011-11-24T01:54:54 I try to avoid thinking about any specifics for the next contest until after the current one is finished :) 2011-11-24T01:55:16 I have a few non-specific general ideas but nothing at all concrete 2011-11-24T01:55:44 and so far as I know there hasn't been any discussion by the other admins either 2011-11-24T01:56:31 There's many thing that could be said about Terrarium, but the coolest feature was that you could run the app (or have it run as a screensaver) and it would "teleport" different species (bots) and have then evolve/battle/coop in your "terrarium" 2011-11-24T01:56:48 cool 2011-11-24T01:57:18 sounds cheesy I guess, but I can imagine kids liking the idea that the creatures they code was teleporting from pc to pc and extend in this universe on a life of their own 2011-11-24T01:58:17 One thing I will say and seems like I've seen a bit of the same sentiment from other admins is I'd like to get away from any sort of rts for the next one since the last two have been sort of rts 2011-11-24T01:58:21 yea, well I too certainly would be stressed enough with the execution of the current contest without having to stress about the next one. 2011-11-24T01:58:47 *** amstan has quit IRC (Ping timeout: 240 seconds) 2011-11-24T01:58:49 *** Palmik has joined #aichallenge 2011-11-24T01:59:12 real-time s(?) ..simulation? 2011-11-24T01:59:44 *** McLeopold has joined #aichallenge 2011-11-24T01:59:44 real time strategy 2011-11-24T01:59:56 ok 2011-11-24T02:00:01 generally applied to games 2011-11-24T02:00:11 right 2011-11-24T02:00:42 what other kind of challenge would be interesting, yet not be a RTS? 2011-11-24T02:00:50 tron :) 2011-11-24T02:01:17 ah, so that's not considered like an RTS? 2011-11-24T02:01:32 I haven't really thought about it too deeply 2011-11-24T02:01:35 sounds pretty real time and strategic, no? 2011-11-24T02:01:41 nope :) 2011-11-24T02:01:51 maybe real time tactical ;) 2011-11-24T02:02:10 RTS generally involves multiple units and resource gathering of some sort 2011-11-24T02:02:27 yea, well played (and right) 2011-11-24T02:02:33 I thought it meant not turn based? 2011-11-24T02:02:44 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-11-24T02:03:07 planetwars and ants are really very basic rts at best really I'm sure there are people that would argue they don't qualify 2011-11-24T02:03:17 like me 2011-11-24T02:04:03 but still I'd like to see something different next, anyway I'll wait till the end of december to really think about it :) 2011-11-24T02:04:06 *** praveen_ has joined #aichallenge 2011-11-24T02:04:25 janzert: given any thought to the finals yet? 2011-11-24T02:04:38 hehe 2011-11-24T02:04:41 but you already had a Tron contest, you would think of doing another like the last one or are you thinking about another aspect of Tron? Pong Style? 2011-11-24T02:04:55 beyond running them the same as planetwars? 2011-11-24T02:05:12 LouisMartin: yeah it definitely won't be tron again 2011-11-24T02:05:33 probably not even another aspect of tron 2011-11-24T02:06:45 McLeopold: any ideas on narrowing (or getting rid of) the race condition on choosing the seed for matchups? 2011-11-24T02:07:00 if we had a real transactional database it'd be simple of course 2011-11-24T02:08:44 janzert: only ones that involve a rewrite 2011-11-24T02:09:02 :( 2011-11-24T02:09:02 maybe a simple 2 minute timer for no repeats or something 2011-11-24T02:09:18 McLeopold: ? 2011-11-24T02:09:31 I just went looking for a haskell library to do something I needed... then I realized I wrote such a library two years ago! 2011-11-24T02:09:35 (you are what you eat!) 2011-11-24T02:09:37 think moving the seed select into the matchup insert would help any? 2011-11-24T02:09:38 I tried using transactions but mysql never did them right, or I was doing it wrong from php 2011-11-24T02:09:55 *** pairofdice has joined #aichallenge 2011-11-24T02:10:06 I don't think the MyISAM backedn supports them 2011-11-24T02:10:12 that's my understanding at least 2011-11-24T02:10:19 McLeopold: i was sleeping and i'm going to sleep in a minute, i'll be available within some hour, but you can me ask me in the meantime 2011-11-24T02:10:25 to work* 2011-11-24T02:10:28 I think you are right, we need InnoDB 2011-11-24T02:10:45 Fluxid: oh, can you put up that crappy map on tcp? 2011-11-24T02:11:21 janzert: also, how many cell maps should I put on official? 1000? 2011-11-24T02:11:39 how about 20 to start :) 2011-11-24T02:11:47 so...few... 2011-11-24T02:11:51 :) 2011-11-24T02:12:13 also I think I'd like to weight maps toward 6 or less players 2011-11-24T02:12:42 any stats on game length per player? 2011-11-24T02:13:00 not that I've looked at no 2011-11-24T02:13:13 I'll use a beta distribution 2011-11-24T02:13:34 my reason is that generally larger player maps don't distinguish between more levels of rankings in the result anyway 2011-11-24T02:14:31 i.e. the lower player maps give more information to the rankings, but I do still want to see some of the crazy huge fights as well :) 2011-11-24T02:15:24 how about at least 100 new maps? 2011-11-24T02:15:59 do the 20 and let them run for a few days that way you can tweek the generator for any problems before adding mroe 2011-11-24T02:16:15 I've been tweaking the generator for a week 2011-11-24T02:16:18 then you can keep adding more 2011-11-24T02:16:19 *** Jak_o_Shadows has quit IRC (Remote host closed the connection) 2011-11-24T02:16:27 but haven't seen lots of games on them yet 2011-11-24T02:16:27 have you thought of running any multiple copies of a single bot in a matche? GreenTea v. xathis (x3) for example 2011-11-24T02:16:39 I'm in a hurry because we are running out of time to introduce new maps 2011-11-24T02:17:01 bmh: not really seeing the resource constraints we already have 2011-11-24T02:17:12 janzert: what do you think I would find? 2011-11-24T02:17:22 hard to say until you do it ;) 2011-11-24T02:17:33 we've pretty much "tested" a lot of maps and games 2011-11-24T02:17:48 I'd just rather keep adding them continually than a huge dump right now 2011-11-24T02:17:54 ok 2011-11-24T02:18:11 also I don't want to completely swamp out the other types 2011-11-24T02:18:14 but my definition of a "small" amount is much larger than yours 2011-11-24T02:20:08 janzert: players = max(2, int(random.betavariant(2.0, 3.0) * 10)) ? 2011-11-24T02:21:20 10 players? madness 2011-11-24T02:21:43 maybe alpha=2.5 to put the sweet spot at 4 players 2011-11-24T02:21:56 a1k0n: we have 10 players now 2011-11-24T02:22:19 sure enough 2011-11-24T02:22:50 There's not enough duels ;) 2011-11-24T02:23:03 those are like, were you lucky enough to get placed near weak opponents? 2011-11-24T02:23:05 Well okay, there are 2011-11-24T02:23:07 pairofdice: this would fix that 2011-11-24T02:23:19 3-5 players are good 2011-11-24T02:23:30 a1k0n: on the multi hill mazes, you should be placed near all of them 2011-11-24T02:24:00 dijkstra... how do I implement that again... :) 2011-11-24T02:24:10 with wikipedia 2011-11-24T02:24:12 yeah, sweet spot at 4 is probably good 2011-11-24T02:25:02 wouldn't alpha=2 put the mode at 4 players? 2011-11-24T02:25:30 or the mean rather 2011-11-24T02:26:04 I'm using wolfram's widget to get the top of the curve over the 4 2011-11-24T02:26:07 0.4 2011-11-24T02:26:11 if betavariant means "draw from a beta distribution" 2011-11-24T02:26:21 it does 2011-11-24T02:26:35 mean is alpha/(alpha+beta) 2011-11-24T02:26:46 that's a fun exercise to integrate gamma functions 2011-11-24T02:26:57 a1k0n: did your parents feed you a statistics textbook? 2011-11-24T02:27:08 all that shit cancels out and the answer turns out to be obvious 2011-11-24T02:27:25 my mom is a realtor and my dad is a radio personality 2011-11-24T02:28:26 i learned this all like, two years ago 2011-11-24T02:29:02 a1k0n: radio personality? please say your dad is rush limbaugh. i think most of us would have followup questions. 2011-11-24T02:29:07 hahah 2011-11-24T02:29:12 Oh right, beta distribution was that one 2011-11-24T02:29:18 no, mostly just a dj 2011-11-24T02:30:27 i heard an interview with the inventor of the 6502 recently. he really wanted to be a dj cuz it was the cool thing to do in those days but he wasn't cut out for it 2011-11-24T02:30:32 also wrapping it in max(2, ...) is going to produce too many 2 player maps probably 2011-11-24T02:31:37 (beta * 0.8 + 0.2) * 10? 2011-11-24T02:32:05 or at least maybe need to shift the distribution something like max(2, min(10, (random.betainvariate(2, 3) * 10) + 1)) 2011-11-24T02:32:06 if I round down, I won't get 10 player games... 2011-11-24T02:33:05 a1k0n: can you recommend any good books to eat to get smarter? 2011-11-24T02:33:37 im a huge fan of http://www.inference.phy.cam.ac.uk/itprnn/book.html 2011-11-24T02:33:45 antimatroid: I did do work on symmetry detection. I haven't really incorporated it much beyond a proof of concept 2011-11-24T02:33:59 can your overall map generator 'controller' produce all 3 types of maps at this point? 2011-11-24T02:34:11 also if that orientation detecting code is yours that might be the best obfuscation of symmetry in the whole codebase ;-) 2011-11-24T02:35:56 a1k0n: have you read Elements of Statistical Learning? (I'm told it's the anti-MacKay) 2011-11-24T02:36:16 i haven't cuz its such a ripoff 2011-11-24T02:36:54 but i don't get along with sampling theory 2011-11-24T02:38:00 janzert: no, not yet 2011-11-24T02:38:16 oh what, the pdf is free now? 2011-11-24T02:38:22 but my cell maze generator can produce all types of maps used so far :) 2011-11-24T02:38:48 its been on my wishlist for years 2011-11-24T02:39:02 the pdf? 2011-11-24T02:39:09 the book 2011-11-24T02:39:15 I want a LaTeX copy of Edwin Jaynes's book 2011-11-24T02:47:23 *** ikaros has joined #aichallenge 2011-11-24T02:47:26 *** praveen_ has quit IRC (Ping timeout: 265 seconds) 2011-11-24T02:47:52 *** Rav3nsW0rd has quit IRC (Quit: Leaving) 2011-11-24T02:48:59 *** bmh has quit IRC (Quit: Page closed) 2011-11-24T02:52:16 *** b0rder_ has joined #aichallenge 2011-11-24T02:54:13 *** Rav3nsW0rd has joined #aichallenge 2011-11-24T02:56:17 what in the world, the bad maps aren't disabled anymore 2011-11-24T02:56:49 or yeah, those were in beta :P 2011-11-24T02:58:01 *** Harpyon has joined #aichallenge 2011-11-24T03:00:23 *** b0rder_ has quit IRC (Ping timeout: 240 seconds) 2011-11-24T03:01:04 *** epicmonkey has joined #aichallenge 2011-11-24T03:03:08 *** LouisMartin has quit IRC (Quit: Page closed) 2011-11-24T03:09:41 *** replore has quit IRC (Read error: Connection reset by peer) 2011-11-24T03:12:18 aichallenge: McLeopold epsilon * rf7a4889 / (ants/mapgen/map.py ants/mapgen/cell_maze.py): add cell maze map generator - http://git.io/Y4fOTg 2011-11-24T03:12:18 aichallenge: McLeopold epsilon * rb01571e / (10 files in 6 dirs): Merge branch 'epsilon' of github.com:aichallenge/aichallenge into epsilon - http://git.io/NoVH7Q 2011-11-24T03:12:44 janzert: do you have a script than names the maps correctly? 2011-11-24T03:13:07 don't think I did 2011-11-24T03:13:12 just did it manually 2011-11-24T03:13:15 ouch 2011-11-24T03:13:23 ok 2011-11-24T03:13:25 *** praveen_ has joined #aichallenge 2011-11-24T03:13:33 then I'm really only doing 20 :) 2011-11-24T03:13:56 remember for the current set I also manually filtered the generator output for good maps 2011-11-24T03:14:22 so I probably looked at 10-20 maps for each one that made it in 2011-11-24T03:14:22 by site? 2011-11-24T03:14:28 sight 2011-11-24T03:14:35 yep 2011-11-24T03:14:57 take a look at what my code produces, I have the hill distance checks built in 2011-11-24T03:15:08 plus the 3x3 block traversal tests 2011-11-24T03:15:15 *** aerique_ has joined #aichallenge 2011-11-24T03:15:17 great 2011-11-24T03:15:30 I think I can just produce 20 and not have to look 2011-11-24T03:15:46 btw for the finals I want to be able to completely automatically produce the map set 2011-11-24T03:15:56 yeah, give it a try 2011-11-24T03:16:01 yeah, me too 2011-11-24T03:16:06 running out of time :( 2011-11-24T03:16:14 halfway there 2011-11-24T03:17:08 postpone the deadline like in business :) 2011-11-24T03:17:40 :) we'll be sticking to a hard deadline here 2011-11-24T03:17:54 'cause we're broke :) 2011-11-24T03:17:59 hehe 2011-11-24T03:18:12 damn.. i have lots of free time after christmas :) 2011-11-24T03:18:50 *** g0llum has joined #aichallenge 2011-11-24T03:20:03 janzert: what upper bound did you want on hill to hill distances? 2011-11-24T03:20:09 *** aerique_ has quit IRC (Client Quit) 2011-11-24T03:20:27 250 for now is what I was thinking 2011-11-24T03:20:34 ok, that's what I have 2011-11-24T03:20:51 might need to go lower later but I'd rather start high 2011-11-24T03:21:19 *** sigh has joined #aichallenge 2011-11-24T03:21:26 did you see the size per player limits I added to random walk before the cotnest? 2011-11-24T03:21:28 is pypy easy to install in ubuntu? 2011-11-24T03:21:34 noi 2011-11-24T03:22:07 it should really be an area limit but I just threw in a per dimension min and max 2011-11-24T03:22:36 I allow 40000 area for 2 players, and with the hill distance checks, it's probably ok 2011-11-24T03:23:04 yeah, the hill distance checks will hopefully take care of it 2011-11-24T03:23:05 *** mj41 has joined #aichallenge 2011-11-24T03:23:31 *** Anilm3 has joined #aichallenge 2011-11-24T03:23:42 fyi, it was min 10x10 max 50x50 2011-11-24T03:23:42 jeeze, first map generated failed :( 2011-11-24T03:23:49 :( 2011-11-24T03:24:00 check out the file though 2011-11-24T03:25:30 *** treeform_ has quit IRC (Remote host closed the connection) 2011-11-24T03:27:10 *** epicmonkey has quit IRC (Ping timeout: 258 seconds) 2011-11-24T03:27:20 *** Blkt has joined #aichallenge 2011-11-24T03:27:41 janzert: do we provide a zip for official maps? 2011-11-24T03:27:43 Hi guys 2011-11-24T03:27:43 *** Harpyon has quit IRC (Quit: Computer has gone to sleep.) 2011-11-24T03:28:02 I believe they are included in tools.zip 2011-11-24T03:28:22 but not a separate one no 2011-11-24T03:28:57 btw cell_maze.py is missing executable bit set 2011-11-24T03:29:32 janzert: I'm on windows, so I don't care :) 2011-11-24T03:29:57 ok, warning that you're gonna get a conflict then :) 2011-11-24T03:30:58 aichallenge: janzert epsilon * r0ef104c / ants/mapgen/cell_maze.py : Set executable bit on cell_maze generator - http://git.io/gg0DJQ 2011-11-24T03:32:09 so the multi-hill maps are gone ? 2011-11-24T03:32:30 no? 2011-11-24T03:32:52 are you putting the new maps on the contest site tonight? i guess it's tomorrow for most of you 2011-11-24T03:33:07 McLeopold: it's up 2011-11-24T03:33:18 Fluxid: the stupid hill map :D 2011-11-24T03:33:42 yeah 2011-11-24T03:33:49 I NEED REPLAYS 2011-11-24T03:33:56 what's it called? 2011-11-24T03:34:01 http://ants.fluxid.pl/map/mc_test_map_04p_01.map 2011-11-24T03:34:01 I really excited to see what happens 2011-11-24T03:34:26 oh god 2011-11-24T03:34:27 nooooooo 2011-11-24T03:34:31 did you get the one with land? 2011-11-24T03:34:43 McLeopold: no... 2011-11-24T03:34:45 uh oh 2011-11-24T03:34:48 i bet what happens is a bunch of timeouts 2011-11-24T03:34:55 meh 2011-11-24T03:35:02 meet your new submission test map :D 2011-11-24T03:35:04 McLeopold: i found this one in backscroll 2011-11-24T03:35:12 Fluxid: hold on... 2011-11-24T03:35:27 http://paste.aichallenge.org/Dbk8a/ 2011-11-24T03:35:29 Fluxid: can you add one/more of the cell mazes? if any of them are ready with starting locations... 2011-11-24T03:35:42 a1k0n: hold on... 2011-11-24T03:35:47 I'm generating an official set 2011-11-24T03:36:09 btw, I love strcat's pasted battle tests... 2011-11-24T03:37:06 McLeopold: I'm coming around to amstan's thinking that all maps should be 'landscape' i.e. cols >= rows 2011-11-24T03:37:23 yeah, please 2011-11-24T03:37:29 rotate them if necessary 2011-11-24T03:38:00 janzert: my code does that :) 2011-11-24T03:38:11 McLeopold: I've put both: http://ants.fluxid.pl/map/mc_test_map_04p_02.map 2011-11-24T03:38:14 the map.py has a make_landscape function 2011-11-24T03:38:20 ahh, good 2011-11-24T03:38:43 Fluxid: so, when do the stupid maps get games? 2011-11-24T03:38:54 McLeopold: it's random ;( 2011-11-24T03:39:05 like, totaly random? 2011-11-24T03:39:06 McLeopold: i can remove some maps to raie chances 2011-11-24T03:39:23 is maze_02p_01.map different? 2011-11-24T03:39:30 oh that was mmaze 2011-11-24T03:40:11 thank god for the visualizer 2011-11-24T03:41:05 aaand now i'll connect my bot, for fun 2011-11-24T03:44:53 i'm still waiting for a game 2011-11-24T03:45:24 Hi guys 2011-11-24T03:45:30 wups sorry 2011-11-24T03:45:54 wups sorry 2011-11-24T03:46:03 *** Anilm3 has quit IRC (Quit: leaving) 2011-11-24T03:46:44 *** Anilm3 has joined #aichallenge 2011-11-24T03:48:19 *** grwip has joined #aichallenge 2011-11-24T03:52:14 *** JorgeB has quit IRC (Quit: Computer has gone to sleep.) 2011-11-24T03:53:14 weird, Fluxid, your server is only picking duels now 2011-11-24T03:53:22 and i'm not getting any games 2011-11-24T03:53:34 (since the last one) 2011-11-24T03:56:01 there we go. 2011-11-24T03:59:14 *** ikaros has quit IRC (Remote host closed the connection) 2011-11-24T04:00:38 a1k0n: weird 2011-11-24T04:02:20 *** kire has quit IRC (Ping timeout: 252 seconds) 2011-11-24T04:02:41 ah, 3 player game finished 2011-11-24T04:03:47 McLeopold: how can i create such scenarios? http://paste.aichallenge.org/1dkIN/ 2011-11-24T04:04:44 it's just a .map 2011-11-24T04:04:49 and --scenario to playgame.py 2011-11-24T04:04:56 holy shit 2011-11-24T04:05:02 why i didn't know about this 2011-11-24T04:05:11 ./playgame.py -e --html=replay.html --nolaunch --player_seed 48 --end_wait=0.25 --verbose --log_dir game_logs --turns 75 --scenario --food none --map_file repel.map 2011-11-24T04:05:24 you probably know most of those but --scenario and --food none are important 2011-11-24T04:05:27 and to think i wanted to create "scenariobot" which would gather enough food to have X ants and take posiotions... 2011-11-24T04:05:30 (food because the map is asymmetrical) 2011-11-24T04:05:57 Fluxid: just look at how the tutorial works 2011-11-24T04:06:14 janzert: i didn't even bother looking at tutorial 2011-11-24T04:06:49 that was a stupid decision it seems 2011-11-24T04:06:57 haha with ant distribution code working perfectly it's like a plague of ants 2011-11-24T04:07:19 man what's up with all the 2-3 player matches on fluxid 2011-11-24T04:07:26 dunno :( 2011-11-24T04:07:35 13389 was the last big game 2011-11-24T04:10:55 a1k0n, guess, we have to wait for all those to reconnect 2011-11-24T04:11:41 *** Palmik has quit IRC (Ping timeout: 252 seconds) 2011-11-24T04:15:19 *** Rav3nsW0rd has quit IRC (Read error: Connection reset by peer) 2011-11-24T04:16:00 *** Rav3nsW0rd has joined #aichallenge 2011-11-24T04:16:40 *** ikaros has joined #aichallenge 2011-11-24T04:17:27 *** Fandekasp has quit IRC (Quit: Lost terminal) 2011-11-24T04:19:39 okay, let's hope I don't crash the servers... 2011-11-24T04:19:46 *** Palmik has joined #aichallenge 2011-11-24T04:20:16 *** Rav3nsW0rd has quit IRC (Read error: Connection reset by peer) 2011-11-24T04:20:30 *** Rav3nsW0rd has joined #aichallenge 2011-11-24T04:22:54 http://paste.aichallenge.org/mWzLA/ this map is cool 2011-11-24T04:23:48 agreed; add it! 2011-11-24T04:23:55 Fluxid: new maps are on official 2011-11-24T04:24:06 waiting for results... 2011-11-24T04:24:13 McLeopold: check them in? 2011-11-24T04:24:21 sure, can you do it? 2011-11-24T04:24:27 so tools.zip can pick them up too 2011-11-24T04:24:33 Map cell_maze/cell_maze_p03_03.map does not exist 2011-11-24T04:24:38 http://aichallenge.org/map.php?map=cell_maze/cell_maze_p03_03.map 2011-11-24T04:24:51 probably not as easily as you can 2011-11-24T04:24:53 :) 2011-11-24T04:24:59 uh... 2011-11-24T04:25:00 git add cell_maze 2011-11-24T04:25:04 did I screw up? 2011-11-24T04:25:12 McLeopold: is it the same as in the paste? 2011-11-24T04:25:17 no 2011-11-24T04:25:22 git commit -m "New cell map examples" 2011-11-24T04:25:46 nice, there's some games 2011-11-24T04:25:59 hmm, yeah the map links aren't working 2011-11-24T04:26:44 permissions look good 2011-11-24T04:26:58 McLeopold: but will this map from paste work? 2011-11-24T04:27:06 http://aichallenge.org/map/cell_maze/ does work, though 2011-11-24T04:27:07 oh, the map names are wrong 2011-11-24T04:27:27 cell_maze_p... instead of cell_maze_??p 2011-11-24T04:27:36 okay, 4 games have played now 2011-11-24T04:27:46 http://aichallenge.org/visualizer.php?game=122530 2011-11-24T04:28:09 how did the map addition script mess that up? 2011-11-24T04:28:20 or did you add them to the database manually? 2011-11-24T04:28:21 timeouts. nice. 2011-11-24T04:28:28 janzert: no, I used the script 2011-11-24T04:28:53 change the map names after adding? 2011-11-24T04:29:01 too weird 2011-11-24T04:29:25 *** antimatroid has quit IRC (Ping timeout: 240 seconds) 2011-11-24T04:29:47 I didn't, I swear 2011-11-24T04:30:11 I don't know how the workers are getting the right name though either 2011-11-24T04:30:34 http://aichallenge.org/visualizer.php?game=122534 2011-11-24T04:31:01 *** mj41 has quit IRC (Ping timeout: 240 seconds) 2011-11-24T04:31:17 These maps look like they could be actual anthills 2011-11-24T04:31:37 Map cell_maze/cell_maze_p06_01.map does not exist 2011-11-24T04:31:37 oops, I must have been in the wrong directory. the disk name does have the p first 2011-11-24T04:31:39 ;(((( 2011-11-24T04:31:46 ll 2011-11-24T04:31:46 janzert: the map names in the database match the ~/maps 2011-11-24T04:32:18 and permissions are the same as the others 2011-11-24T04:32:30 oh, maybe not 2011-11-24T04:33:15 did we copy them to website? 2011-11-24T04:34:13 not that I know of, it should be an alias in the apache config 2011-11-24T04:34:41 http://aichallenge.org/map/cell_maze/cell_maze_p05_02.map 2011-11-24T04:34:52 it is the visualizer? 2011-11-24T04:35:09 or map.php 2011-11-24T04:35:37 something in get_map_row 2011-11-24T04:35:54 janzert: crap, it is the memcache :( 2011-11-24T04:36:26 *** Palmik has quit IRC (Ping timeout: 252 seconds) 2011-11-24T04:36:28 ahh 2011-11-24T04:36:41 add_maps should clear out the memcache key... 2011-11-24T04:37:40 it seems to prefer cell_maze over everything else now 2011-11-24T04:38:06 is that a temporary thing to test it out? 2011-11-24T04:38:26 okay, fixed, temporarily 2011-11-24T04:38:45 a1k0n: it tries to balance the number of games played over 24 hours 2011-11-24T04:38:47 a1k0n: the logic will get those maps up to where the 24 hour average is for the others 2011-11-24T04:39:25 ahha 2011-11-24T04:39:28 clever 2011-11-24T04:39:45 http://aichallenge.org/index.php 2011-11-24T04:41:33 http://aichallenge.org/visualizer.php?game=122558 2011-11-24T04:41:34 that is an interesting map 2011-11-24T04:41:37 that map is mean 2011-11-24T04:41:57 momobot seems to be getting a shitton of games here 2011-11-24T04:42:22 and by that you mean 3? 2011-11-24T04:42:30 i guess.. 3 at once 2011-11-24T04:42:34 i've had that before too 2011-11-24T04:43:42 i'm in a game now woo 2011-11-24T04:44:26 gl 2011-11-24T04:44:41 heh, this game was a whole lot of nothing: http://aichallenge.org/visualizer.php?game=122566 2011-11-24T04:45:34 *** mj41 has joined #aichallenge 2011-11-24T04:45:52 they all did different stuff 2011-11-24T04:46:25 the maps seem a little too defensive for a lot of these recent matches tho 2011-11-24T04:46:53 ho ho ho! 2011-11-24T04:46:56 http://aichallenge.org/visualizer.php?game=122571&user=432 2011-11-24T04:47:28 map size seems to be quite a bit larger on average 2011-11-24T04:47:29 i killed xathis with extreme prejudice there 2011-11-24T04:47:34 yeah 2011-11-24T04:47:57 so now all the games will be with new maps? 2011-11-24T04:47:58 nice game 2011-11-24T04:48:16 so, in that map, you need good food gathering for a good defence 2011-11-24T04:48:33 Fluxid: only until they catch up to the number of games other maps played in the last day 2011-11-24T04:48:33 all the hills are in a line, and you have to get out to the fields 2011-11-24T04:49:10 yep, and all your foragers are vying for position 2011-11-24T04:50:23 wtf, only duels on tcp... 2011-11-24T04:50:41 *** praveen_ has quit IRC (Ping timeout: 265 seconds) 2011-11-24T04:50:50 think I'm already ready to lower the maximum hill distance ;) 2011-11-24T04:51:05 janzert: 200? 2011-11-24T04:51:42 maybe even 166 2011-11-24T04:51:48 or just shrink the max map size? 2011-11-24T04:51:48 1000 / 6 2011-11-24T04:51:55 maybe 2011-11-24T04:52:13 I was using 150x150 for the initial set btw 2011-11-24T04:52:38 but some of the maps just seem too spread out disregarding the actual size 2011-11-24T04:53:27 http://aichallenge.org/visualizer.php?game=122572 2011-11-24T04:54:06 superior food gathering made the difference 2011-11-24T04:54:36 i didn't realize the map can start with so much food on it 2011-11-24T04:54:54 .75 - 1.5% 2011-11-24T04:55:02 of land area, which is quite a lot 2011-11-24T04:55:22 -> 1.75% 2011-11-24T04:55:24 i guess so 2011-11-24T04:55:42 every 10x10 block has about 2 2011-11-24T04:55:47 ok, more people must connect to tcp for larger maps to work 2011-11-24T04:56:07 does the comments in a map give you the maximum hill distance? 2011-11-24T04:56:14 no 2011-11-24T04:56:22 I guess I could put that in 2011-11-24T04:56:42 e.g. I'm wondering what the distance is here http://aichallenge.org/map.php?map=cell_maze/cell_maze_p02_03.map 2011-11-24T04:57:00 oh hey i got a cell maze game on your server 2011-11-24T04:57:33 janzert: probably around 200 2011-11-24T04:57:45 it's just a little over manhatten 2011-11-24T04:59:30 http://aichallenge.org/visualizer.php?game=122589 2011-11-24T05:00:52 yeah these maps are huge 2011-11-24T05:02:08 how about raising turn limit for finals? 2011-11-24T05:05:13 aichallenge: janzert epsilon * r6404241 / (20 files): Initial cell maze maps - http://git.io/VD5p3A 2011-11-24T05:06:21 McLeopold: you've got local changes to cell_maze.py on the server? 2011-11-24T05:06:21 http://aichallenge.org/visualizer.php?game=122611 look at the defense structures here 2011-11-24T05:06:29 janzert: wipe them out 2011-11-24T05:07:07 I didn't add a player count switch yet, so I hard coded it to get the right count 2011-11-24T05:07:13 nice synchronized timeuts we have here 2011-11-24T05:07:31 that's a cool map, where can I find the new maps? 2011-11-24T05:08:25 Oh, damn my bot has some issues to deal with 2011-11-24T05:08:31 Anilm3: redownload the tools 2011-11-24T05:08:36 it will have the new maps 2011-11-24T05:08:42 janzert: thanks 2011-11-24T05:09:01 http://aichallenge.org/visualizer.php?game=122602 2011-11-24T05:09:05 ... >_< 2011-11-24T05:09:25 What are my ants doing 2011-11-24T05:10:17 bouncing of the walls 2011-11-24T05:11:52 I guess you coded for orthagonal mazes? 2011-11-24T05:13:49 http://aichallenge.org/map.php?map=cell_maze/cell_maze_p04_03.map I like this map 2011-11-24T05:14:09 me too 2011-11-24T05:14:26 the ones that are huge are of questionable value 2011-11-24T05:14:52 i tell me that everyone sucks and no one is going to win 2011-11-24T05:16:44 heh 2011-11-24T05:17:04 i think bots among the top 30 will be able to do something with those but the other 4000, well, not sure 2011-11-24T05:17:31 i probably won't be able to get back to top30, heh 2011-11-24T05:17:45 *** AntDroid has quit IRC (Ping timeout: 265 seconds) 2011-11-24T05:18:02 i'm in the top 10 for the first time woo 2011-11-24T05:18:15 Now I really want to update my bot 2011-11-24T05:18:17 thank you, new cell maze maps 2011-11-24T05:20:03 http://ants.fluxid.pl/replay.13529 2011-11-24T05:20:15 YEAH! 2011-11-24T05:20:33 wow, the visualizer really sucks :D 2011-11-24T05:20:51 424 point bonus ! 2011-11-24T05:21:52 lol 2011-11-24T05:22:16 not enough razing of hills there 2011-11-24T05:23:04 hahahaha 2011-11-24T05:23:27 whoa 2011-11-24T05:23:30 i can't see shit 2011-11-24T05:23:40 correction, that is all you can see 2011-11-24T05:24:14 Are the new maps in the tools package? 2011-11-24T05:24:40 yes 2011-11-24T05:24:53 but i won, i'm kind of proud, lol 2011-11-24T05:25:01 with all those bonus points, lol 2011-11-24T05:25:37 i see your ants don't like stepping on hills either 2011-11-24T05:25:49 a1k0n: yeah, penalty points 2011-11-24T05:26:06 ok i better go to bed 2011-11-24T05:28:09 me too 2011-11-24T05:28:13 *** McLeopold has left #aichallenge 2011-11-24T05:29:34 *** b0rder has quit IRC (Quit: 离开) 2011-11-24T05:40:11 *** ikaros has quit IRC (Ping timeout: 258 seconds) 2011-11-24T05:40:37 *** lericson has quit IRC (Read error: Connection reset by peer) 2011-11-24T05:42:18 *** patrisk has joined #aichallenge 2011-11-24T05:42:52 *** magikmw has joined #aichallenge 2011-11-24T05:52:17 http://ants.fluxid.pl/replay.13557 feels good man, lol 2011-11-24T05:58:31 *** praveen has joined #aichallenge 2011-11-24T06:09:01 *** Areks has quit IRC (Ping timeout: 240 seconds) 2011-11-24T06:11:31 good day everyone 2011-11-24T06:18:32 *** oprs has quit IRC (Remote host closed the connection) 2011-11-24T06:19:51 hey Blkt 2011-11-24T06:24:41 *** Inevitable has joined #aichallenge 2011-11-24T06:24:45 Hi 2011-11-24T06:25:05 *** Areks has joined #aichallenge 2011-11-24T06:25:06 Is there anyone that has the complete tutorial in Java, because we screwed something up, and have no idea where. 2011-11-24T06:25:17 of the AIchallenge 2011-11-24T06:25:20 :)? 2011-11-24T06:25:47 DERP 2011-11-24T06:26:05 tnxbai 2011-11-24T06:26:07 *** Inevitable has quit IRC (Client Quit) 2011-11-24T06:28:38 wow 8 servers running today! 2011-11-24T06:31:04 that was fast... 2011-11-24T06:31:23 Anilm3: yesterday there were twoo servers too 2011-11-24T06:37:48 I like this new map :) 2011-11-24T06:37:59 the intestines one 2011-11-24T06:38:56 *** mviel_ has joined #aichallenge 2011-11-24T06:39:55 Fluxid: things seem to be going better this week 2011-11-24T06:42:40 *** mviel__ has quit IRC (Ping timeout: 260 seconds) 2011-11-24T06:47:00 this new maps seem cool 2011-11-24T06:51:16 *** AntDroid has joined #aichallenge 2011-11-24T06:53:22 *** durarara has quit IRC (Ping timeout: 245 seconds) 2011-11-24T06:55:17 *** antimatroid has joined #aichallenge 2011-11-24T06:55:32 *** ikaros has joined #aichallenge 2011-11-24T06:55:57 those cell maze maps are awesome 2011-11-24T06:57:39 *** olexs has joined #aichallenge 2011-11-24T06:59:36 'ants not razing hills' doesn't seem very reasonable on a 2 player map with 1 hill for each player 2011-11-24T07:00:10 http://ants.fluxid.pl/replay.13611 - don't know if the official server does this 2011-11-24T07:00:33 *** olexs has quit IRC (Client Quit) 2011-11-24T07:02:18 it does 2011-11-24T07:02:57 that was perhaps cut a bit short 2011-11-24T07:03:11 McLeopold's not here 2011-11-24T07:03:34 but he also should have really taken you by now 2011-11-24T07:05:36 Well blue wasn't very aggressive 2011-11-24T07:06:20 it does, and afair cutoff parameters are the same as in the main server 2011-11-24T07:08:47 how is this cutoff working? is it just after a certain amount of some moves with no hill razed, or some ratio between blue and orange ants? If it's the latter it seems right that this game was cut off 2011-11-24T07:10:26 would be ratio of ants involved 2011-11-24T07:10:29 i don't know the specifics 2011-11-24T07:10:35 you should be able to find it in the code somewhere 2011-11-24T07:11:10 then I guess it's ok, after 50 moves with > 70 ants against < 10 ants 2011-11-24T07:11:10 *** Anilm3 has quit IRC (Quit: Lost terminal) 2011-11-24T07:11:17 *** durarara has joined #aichallenge 2011-11-24T07:13:41 hmm, 150 moves in fact in this particular game 2011-11-24T07:18:52 *** hj has joined #aichallenge 2011-11-24T07:19:48 *** hj has joined #aichallenge 2011-11-24T07:20:09 wtf is this http://ants.fluxid.pl/replay.13612 2011-11-24T07:20:22 *** hj has quit IRC (Client Quit) 2011-11-24T07:20:25 *** Conorach has joined #aichallenge 2011-11-24T07:20:45 that maps is full of hills alone 2011-11-24T07:21:17 i had a little baby bird trying to learn to fly in my backyard today 2011-11-24T07:22:22 is that expected - 20*20 maps and some >375 hills 2011-11-24T07:22:27 *** Accoun has quit IRC (Ping timeout: 244 seconds) 2011-11-24T07:23:23 lol wtf 2011-11-24T07:23:50 pretty sure McLeopold is going to implement a minimum path distance between enemy hills at some point 2011-11-24T07:24:05 then i'm not sure if we'll allow enemy hills to start in view of each other or not 2011-11-24T07:24:40 *** Rav3nsW0rd has quit IRC (Excess Flood) 2011-11-24T07:25:00 *** Rav3nsW0rd has joined #aichallenge 2011-11-24T07:25:17 that game is insane 2011-11-24T07:25:34 thats not a map even 2011-11-24T07:25:39 only hills 2011-11-24T07:25:58 that's the kind of map a crazy maps tcp server should be used for 2011-11-24T07:26:08 along with large maps and just otherwise whatever maps 2011-11-24T07:26:15 large numbers of players etc. 2011-11-24T07:26:30 max players is 10 with the new map format yeah? 2011-11-24T07:26:41 we used to use characters for ants hills but i tihnk they're using numbers now? 2011-11-24T07:51:11 *** delt0r_ has quit IRC (Ping timeout: 240 seconds) 2011-11-24T07:55:01 *** praveen has quit IRC (Quit: Page closed) 2011-11-24T07:58:04 *** TheLinker has joined #aichallenge 2011-11-24T08:03:42 *** delt0r_ has joined #aichallenge 2011-11-24T08:03:50 the crazy map i would like to see played are complete mazes with single width corridors 2011-11-24T08:04:01 the crazy part comes in when fighting can happen over water 2011-11-24T08:18:52 *** praveen has joined #aichallenge 2011-11-24T08:21:46 *** ikaros has quit IRC (Ping timeout: 244 seconds) 2011-11-24T08:23:50 *** rajanaresh has quit IRC (Ping timeout: 244 seconds) 2011-11-24T08:29:14 ok, that map was fun, time to take it down 2011-11-24T08:29:28 praveen: McLeopold decided to create and test such map 2011-11-24T08:30:16 :) yeah that was fun, but are there really going to be such crazy maps? 2011-11-24T08:34:57 *** praveen has quit IRC (Ping timeout: 265 seconds) 2011-11-24T08:35:49 *** mj41 has quit IRC (Ping timeout: 244 seconds) 2011-11-24T08:37:59 *** Palmik has joined #aichallenge 2011-11-24T08:40:59 *** sigh has quit IRC (Remote host closed the connection) 2011-11-24T08:48:15 *** ikaros has joined #aichallenge 2011-11-24T08:49:25 *** foobar_ has joined #aichallenge 2011-11-24T08:50:49 *** mj41 has joined #aichallenge 2011-11-24T09:06:38 *** phono has joined #aichallenge 2011-11-24T09:07:00 hi 2011-11-24T09:07:07 some1 herE? 2011-11-24T09:07:19 who is here? 2011-11-24T09:07:27 who is that? 2011-11-24T09:07:32 where i ame? 2011-11-24T09:08:32 *** phono has quit IRC (Client Quit) 2011-11-24T09:08:34 phono, ants can't talk, try to wave your antennae 2011-11-24T09:08:49 *** knyppeldynan has quit IRC (Quit: leaving) 2011-11-24T09:10:15 *** sigh has joined #aichallenge 2011-11-24T09:10:51 *** patrisk has quit IRC (Quit: Leaving) 2011-11-24T09:14:10 *** besh has joined #aichallenge 2011-11-24T09:18:42 *** yamal has joined #aichallenge 2011-11-24T09:18:49 \../ 2011-11-24T09:21:08 Hi, I have a question, is there a posting somewhere describing how the final tournament will be conducted? Ie, how the matchmaking will be done and what settings (maps, time constraints, #bots etc) will be used? 2011-11-24T09:21:38 *** AntDroid has quit IRC (Quit: Page closed) 2011-11-24T09:22:24 *** AntDroid has joined #aichallenge 2011-11-24T09:23:58 noone knows? 2011-11-24T09:25:42 *** avdg has joined #aichallenge 2011-11-24T09:28:04 *** sigh has quit IRC (Remote host closed the connection) 2011-11-24T09:36:09 *** g0llum has quit IRC (Read error: Connection reset by peer) 2011-11-24T09:36:44 *** foRei has joined #aichallenge 2011-11-24T09:37:49 *** Areks has quit IRC (Ping timeout: 240 seconds) 2011-11-24T09:41:34 *** dom7b5 has joined #aichallenge 2011-11-24T09:43:02 *** g0llum has joined #aichallenge 2011-11-24T09:48:23 *** HaraKiri has joined #aichallenge 2011-11-24T09:49:03 *** mcstar has joined #aichallenge 2011-11-24T09:50:26 *** besh has quit IRC (Quit: Page closed) 2011-11-24T09:50:48 *** Harpyon has joined #aichallenge 2011-11-24T09:50:54 has anyone made a map making tool? 2011-11-24T09:50:56 mcstar: ? 2011-11-24T09:51:12 hi 2011-11-24T09:51:17 like a paint thing that will then give me the map text for what i drew :P 2011-11-24T09:51:21 mcstar: hi :) 2011-11-24T09:51:36 just open the map in vim 2011-11-24T09:51:57 i guess i could make one 2011-11-24T09:52:01 but it would be qt 2011-11-24T09:52:24 wait, i don't want the map text 2011-11-24T09:52:28 i want the bot input text :P 2011-11-24T09:52:45 "then give me the map text" 2011-11-24T09:52:49 yeah my bad :P 2011-11-24T09:53:01 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-11-24T09:53:10 i want to feed the information to my bot 2011-11-24T09:53:17 atm it's a pita to set up debug scenarios 2011-11-24T09:53:23 antimatroid: are you not satisfied with scenarios? 2011-11-24T09:53:33 your thing? 2011-11-24T09:53:36 no 2011-11-24T09:53:40 mcleo's 2011-11-24T09:53:46 link? 2011-11-24T09:53:49 @scenarios 2011-11-24T09:53:50 antimatroid: Run as fast as you can and don't look back. 2011-11-24T09:53:52 you have it 2011-11-24T09:54:03 its an option to playgame.py 2011-11-24T09:54:16 i started using them >1 day ago 2011-11-24T09:54:33 *** kaemo has joined #aichallenge 2011-11-24T09:54:35 what does it do? 2011-11-24T09:54:46 i want to basically turn an ascii version of a map into bot input 2011-11-24T09:54:46 you edit the map, and you run the scenario 2011-11-24T09:54:53 maybe i should just write that 2011-11-24T09:54:55 that's easy enough 2011-11-24T09:55:07 antimatroid: i think that is exactly what you want 2011-11-24T09:55:11 yeah 2011-11-24T09:55:14 and easy to do :P 2011-11-24T09:55:15 no, a scenario 2011-11-24T09:55:20 you want that 2011-11-24T09:55:27 ok? 2011-11-24T09:55:30 just try it 2011-11-24T09:55:37 nah :P 2011-11-24T09:55:58 i hate the map file format 2011-11-24T09:56:04 it's a pita to write as well 2011-11-24T09:56:04 antimatroid: dont you want to see how does you bot react to a specific setup? 2011-11-24T09:56:12 r 2011-11-24T09:56:14 no 2011-11-24T09:56:27 i build my bot based on strategies, not how it performs with testing 2011-11-24T09:56:52 guys, can someone help me out here with antimatroid ? 2011-11-24T09:56:56 :P 2011-11-24T09:56:57 do i get him wrong? 2011-11-24T09:58:19 so you want to feed your bot the whole map at once? 2011-11-24T09:58:25 is that it? 2011-11-24T09:59:08 *** Harpyon_ has joined #aichallenge 2011-11-24T10:02:12 mcstar: my idea sucks still 2011-11-24T10:02:24 the max focus? 2011-11-24T10:02:26 what i want is to set up a scenario then feed that information to my bos 2011-11-24T10:02:30 no 2011-11-24T10:02:40 just a program to convert text to text 2011-11-24T10:02:49 i want the paint tool like your program thing to draw the scenario 2011-11-24T10:02:50 ok, i just dont see why is that good for you 2011-11-24T10:02:58 the engine does that anyway 2011-11-24T10:03:12 so is your point to do that AT ONCE? 2011-11-24T10:03:20 wo viewradius and such? 2011-11-24T10:03:23 what? 2011-11-24T10:03:31 you're off topic 2011-11-24T10:03:35 back to my tool thin 2011-11-24T10:03:37 no 2011-11-24T10:03:39 antimatroid: http://paste.aichallenge.org/Dtl76/ 2011-11-24T10:03:39 you know your tool thing? 2011-11-24T10:03:43 you want such scenarios? 2011-11-24T10:03:59 *** liberforce has joined #aichallenge 2011-11-24T10:04:08 *** deltree__ has joined #aichallenge 2011-11-24T10:04:32 antimatroid: with scenarios like what Fluxid pasted, you can start a game from an arbitrary settibng 2011-11-24T10:05:01 i found out about this today 2011-11-24T10:05:11 edit map and pass --scenario option to engine 2011-11-24T10:05:25 *** deltree_ has quit IRC (Ping timeout: 240 seconds) 2011-11-24T10:05:34 Fluxid: I want something like mcstars tool that let's you draw combat situations etc., but also with food/water/hill info, then i want to get out text like http://pastebin.com/8mjBRwYD which I can just feed straight to my bot from a file 2011-11-24T10:05:34 antimatroid: i know my tool thing, but if you want a similar thing with your bot you can just set up a scenario and voila 2011-11-24T10:05:56 ok 2011-11-24T10:05:58 i see 2011-11-24T10:06:16 and i want to be able to do larger scenarios :P 2011-11-24T10:06:16 i understand this, but i also think that a scenario is > than that 2011-11-24T10:06:56 mcstar: would it be hard for you two turn your thing into something that does this? 2011-11-24T10:07:08 it would make my life so much easier :P 2011-11-24T10:07:21 not really 2011-11-24T10:07:32 but i would have to implement map save/load 2011-11-24T10:07:36 and add more editing 2011-11-24T10:07:51 just pop up a window with the text that i could copy would be fine 2011-11-24T10:07:54 basically i wanted to jack in my bot 2011-11-24T10:08:00 or if you can just add it straight my copy thing? 2011-11-24T10:08:01 but im not doing that now 2011-11-24T10:08:49 hi there 2011-11-24T10:08:56 antimatroid: i think a scenario is better, thats why i dont pursue that idea, ie. to extend my tool 2011-11-24T10:09:04 i just saw maps have changed, was there an annouce about that ? 2011-11-24T10:09:05 *** Harpyon_ has quit IRC (Quit: Textual IRC Client: http://www.textualapp.com/) 2011-11-24T10:09:10 alright, wth is a scenario? :P 2011-11-24T10:09:20 lol 2011-11-24T10:09:29 lets you debug battles very well 2011-11-24T10:09:33 i used it yesterday 2011-11-24T10:09:49 i'm not debugging battles :P 2011-11-24T10:09:50 have you seen Fluxid 's paste? 2011-11-24T10:09:56 ok 2011-11-24T10:09:57 it's for any scenario i want to debug 2011-11-24T10:10:47 yeah? 2011-11-24T10:11:42 are people detecting if enemies have timed out? 2011-11-24T10:11:58 that would make your life much easier when killing them off i imagine 2011-11-24T10:12:23 yes, but top bots rarely time out 2011-11-24T10:12:25 no? 2011-11-24T10:12:37 but you might be in a game with another top bot with a bot that timed out 2011-11-24T10:12:46 Is it allowed to use third party libraries? And in particular can I use boost? 2011-11-24T10:12:58 if it's installed 2011-11-24T10:13:01 which i'm not sure about 2011-11-24T10:13:19 yes 2011-11-24T10:13:34 * antimatroid goes back to coding with his horribly inefficient debugging methods :) 2011-11-24T10:13:46 yamal: ask antimony when he is here, he did some thing to rename #include's 2011-11-24T10:14:16 mcstar, thanks, will do. 2011-11-24T10:14:28 fyi it's 2 14am on the east coast of australia 2011-11-24T10:14:34 pretty sure antimony lives here 2011-11-24T10:14:49 aha 2011-11-24T10:15:03 :p 2011-11-24T10:15:05 oh, why i say aha i dont mean what you do 2011-11-24T10:15:08 i was up till 7am last night 2011-11-24T10:15:17 my aha->ok, i see 2011-11-24T10:15:25 your aha->laugh, lol, haha 2011-11-24T10:15:44 why->when 2011-11-24T10:16:01 7 is a bit too much 2011-11-24T10:16:16 these days i cant go to bed before 3am 2011-11-24T10:16:21 2:30, 3 2011-11-24T10:16:43 i want to get my bot up and running 2011-11-24T10:16:50 it's still a while away yet :\ 2011-11-24T10:16:56 who doesnt? 2011-11-24T10:16:57 there's just so much to write 2011-11-24T10:17:50 antimatroid: this particular contest seems like such a resource sucker compared to the last. It's much harder just to write code that doesn't crash, as opposed to actually making decisions (like a proper AI) 2011-11-24T10:18:16 crash or time out? 2011-11-24T10:18:17 i dunno, i'm finding path finding surprisingly cheap 2011-11-24T10:18:28 mcstar: primarily time out 2011-11-24T10:18:33 and my battle code is surprisingly fast 2011-11-24T10:18:37 how did you find pathfinding cheap? 2011-11-24T10:18:43 it is 2011-11-24T10:18:50 compared to a game tree, it's nothing :P 2011-11-24T10:18:55 yep 2011-11-24T10:19:00 I can't pathfind for more than 3 squares with more than 20 ants 2011-11-24T10:19:07 :S 2011-11-24T10:19:12 you're doing it wrong :P 2011-11-24T10:19:17 of course 2011-11-24T10:19:22 antimatroid: not common wrong that is 2011-11-24T10:19:25 that sounds too slow deltree__ 2011-11-24T10:19:32 agreed 2011-11-24T10:19:43 but you don't want to be finding paths from everything to everything either 2011-11-24T10:19:44 but I'm trying a breadth first analysis 2011-11-24T10:19:56 delt0r_: most of the time i am 2011-11-24T10:20:03 and i'm doing it fairly easily 2011-11-24T10:20:09 I don't .. you don't need to 2011-11-24T10:20:13 i'm restarting the searches after moving each ant too 2011-11-24T10:20:14 I set targets, and I have to pathfind to find the distance for the nearest ant 2011-11-24T10:20:18 *** bes_ has joined #aichallenge 2011-11-24T10:20:20 you do when you do what i'm doing :P 2011-11-24T10:20:28 since most of the time you want closest pairs of things or something similar 2011-11-24T10:20:46 delt0r: are you using scenarios? 2011-11-24T10:20:51 ? 2011-11-24T10:20:53 if you do talk to antimatroid 2011-11-24T10:20:58 you dont, nvm 2011-11-24T10:20:58 guess not 2011-11-24T10:21:10 delt0r: you should :D 2011-11-24T10:21:11 what are scenarios? 2011-11-24T10:21:22 antimatroid what method are you using to pathfind 2011-11-24T10:21:30 depends on the situation 2011-11-24T10:21:33 delt0r: you can start a game from any setting 2011-11-24T10:21:33 either bfs or a* 2011-11-24T10:21:47 deltree__: you can find closest ants to food for example with a single bfs 2011-11-24T10:21:53 in what language? 2011-11-24T10:22:00 c++ 2011-11-24T10:22:07 pfff 2011-11-24T10:22:13 who uses c++ anymore? 2011-11-24T10:22:17 mcstar 2011-11-24T10:22:18 I am using java... my path finding is taking about 5ms 2011-11-24T10:22:21 c++ would be faster 2011-11-24T10:22:33 antimatroid: i dont, i use c++11 2011-11-24T10:22:35 though I heard a rumor that python is compiled to c at runtime 2011-11-24T10:22:52 I use the term "compiled" loosely there 2011-11-24T10:22:56 python is ridiculously slow 2011-11-24T10:23:02 deltree__: nothing is compiled to C at run time.. its compiled to machine code 2011-11-24T10:23:07 python is not 2011-11-24T10:23:08 java is 2011-11-24T10:23:09 I'm using c# but it's awful 2011-11-24T10:23:14 You're ridiculously slow! 2011-11-24T10:23:22 c# has JIT too 2011-11-24T10:23:30 come on ppl 2011-11-24T10:23:33 :P 2011-11-24T10:23:40 here am I with plain C :p 2011-11-24T10:23:44 his slowness is not because implementaiton 2011-11-24T10:23:48 so i would think C# should be quite fast? 2011-11-24T10:23:58 well, not cause of language implementation but becauseof his alg. 2011-11-24T10:23:59 mcstar: you mean it is? 2011-11-24T10:24:08 yeah, it's implementation not tool 2011-11-24T10:24:09 I'm tempted to write this f***er in some obscure language, like Shakespeare since it's not gonna work well anyway 2011-11-24T10:24:10 It's really a matter of coding stuff properly. 2011-11-24T10:24:11 delt0r: the algorithm is what gets you fast or slow 2011-11-24T10:24:34 liberforce: you are saying that to me like i don't know or something... 2011-11-24T10:24:39 *** ikaros has joined #aichallenge 2011-11-24T10:25:08 liberforce: the OP brought up the idea of language speed 2011-11-24T10:25:13 sorry, I missed my name completion, wanted to say that to deltree__ :p 2011-11-24T10:25:43 magikmw: liberforce do you think I joined the channel to get told "You're bad" or to hear what other people suggest to solve it. So far I've heard "lower level language" and "bfs from food to ant" 2011-11-24T10:26:11 anyway deltree__ I think its possible you have a bug or a doing something not right that is killing your perfomance 2011-11-24T10:26:23 deltree__: sit back, and think about what you really want and how are you doing in now, you will see your errors 2011-11-24T10:26:31 it* 2011-11-24T10:26:35 well, I haven't implemented my stuff yet, I'm just trying to get pathfinding working on the starter bot 2011-11-24T10:26:44 deltree__: relax and take your pills 2011-11-24T10:26:51 deltree__: you quoted specific numbers 2011-11-24T10:26:59 yup 2011-11-24T10:27:02 also finding paths from everything to everthing is also not really a win.. that is O(n^2) 2011-11-24T10:27:02 I did 2011-11-24T10:27:03 how did you get them wo having a working implementation? 2011-11-24T10:27:16 antimatroid may have performance issues yet 2011-11-24T10:27:32 ? 2011-11-24T10:27:43 mcstar: those are teh numbers with the starter bot, to be specific, it's 22 ants with a 3 - 5 point bfs 2011-11-24T10:27:45 as in bot quality or bots ability to not timeout? 2011-11-24T10:27:47 antimatroid: wasn't talking to you ;) 2011-11-24T10:28:05 deltree__: what starter bot has bfs in it? 2011-11-24T10:28:12 c++ one does 2011-11-24T10:28:21 mcstar: I'm trying to get bfs working in the starter bot 2011-11-24T10:28:24 but thats for vr2 2011-11-24T10:28:39 I followed the tutorial to get a feel for the code as it was written 2011-11-24T10:28:49 antimatroid: performance... you get more time on the "trees" if your not farting around getting all this path info you ultimately throw away 2011-11-24T10:29:07 I haven't yet started my own implementation, mostly because the DI in the starter c# project was difficult to work with 2011-11-24T10:29:10 delt0r: my combat code is actually surprisingly fast :P 2011-11-24T10:29:11 deltree__: ok, obviously you want to throw that bfs out the window and strart from sratch 2011-11-24T10:29:13 for example 2011-11-24T10:29:22 i'm breaking each battle partition down into a 2x2 game 2011-11-24T10:29:24 or something like that 2011-11-24T10:29:25 mcstar: thus the reason I'm here 2011-11-24T10:29:36 accepting even lame abuse from guys like liberforce 2011-11-24T10:29:46 deltree__: just start coding 2011-11-24T10:30:02 deltree__: c'mon 2011-11-24T10:30:06 i like to work things out on paper before coding :P 2011-11-24T10:30:07 and be sure you read the whole source of your starter bot at least once 2011-11-24T10:30:08 that's neither a suggestion nor a solution 2011-11-24T10:30:17 i just pointed the obvious 2011-11-24T10:30:41 language matter, but not as much as the algorithm 2011-11-24T10:30:41 deltree__: think your reading more into what was said that is there... tbh 2011-11-24T10:30:45 period 2011-11-24T10:31:08 no offense was intended, I don't really get astonishing results 2011-11-24T10:31:13 language matters much, but only when you maxed out your algorithm 2011-11-24T10:31:30 noone will write as fast a pathfinder in python as i can in c++ 2011-11-24T10:31:43 might be a strong opinion, but i think its true 2011-11-24T10:31:46 (not in cpython) 2011-11-24T10:31:47 mcstar: even then, it matters less than most think.. most of it is dogma 2011-11-24T10:32:10 mcstar: the language limitations arise when you already in the 100 or 200 first 2011-11-24T10:32:24 not when at 1000+ ranks 2011-11-24T10:32:48 I chose C instead of python, but had to reimplement simple collection types 2011-11-24T10:32:52 another lang consideration is not "is C faster than X" but "can i write C that is faster?" 2011-11-24T10:32:57 for example 2011-11-24T10:33:06 but didi it on thinking I could grab some performance there 2011-11-24T10:33:27 delt0r: you're right 2011-11-24T10:33:47 mcstar, speaking of this, do you find any extra few % of performance from using C++11 vs C++ or is it "just" a cleaner code w.o. performance loss? (or do you find a performance loss?? I guess optimizer is less mature for C++11 atm?) 2011-11-24T10:33:52 i think also that you are better off starting from a higher level language 2011-11-24T10:33:58 I chose C because I knew it would be easier for me to optimize it 2011-11-24T10:34:09 in python you might get a better idea(and faster) what you really supposed to write 2011-11-24T10:34:17 and then you write it in a faster language 2011-11-24T10:34:27 but a friend of mine had a completely different approach, and started learning D to crete his bot 2011-11-24T10:34:56 mcstar: Yeah, python is more for prototyping 2011-11-24T10:35:00 but that takes time too 2011-11-24T10:35:03 yamal: c++11 cant be worsely optimized than c++ 2011-11-24T10:35:16 it only adds 2011-11-24T10:35:27 but i could be wrong, im a newbie 2011-11-24T10:35:34 well it offers hashmaps 2011-11-24T10:35:41 c++11 minly adds more functinality 2011-11-24T10:35:41 guess thats what he meant 2011-11-24T10:35:42 Well, most are newbies on C++11 atm ;) 2011-11-24T10:35:44 but it's not really that hard to mess with c++ and completely rehaul code quickly it's not like these bots are thousands of lines of code 2011-11-24T10:36:03 reaching 3k soon 2011-11-24T10:36:34 (not bragging, that might be a bad thing) 2011-11-24T10:36:37 lol 2011-11-24T10:37:16 I'm around 1.8k lines of code in C, but far from finished 2011-11-24T10:37:38 mebe thousands was a bad choice 2011-11-24T10:37:38 i know i couldnt write my bot in c 2011-11-24T10:38:03 well, OOP in C is quite fun ;) 2011-11-24T10:38:15 I'm more a GTK guy 2011-11-24T10:39:23 *** yamal has quit IRC (Quit: Lämnar) 2011-11-24T10:39:23 O_o, 235 lines, nout counting ants.py 2011-11-24T10:39:32 :) 2011-11-24T10:39:43 youre nout 2011-11-24T10:40:27 pairofdice: whats in ants.py? 2011-11-24T10:40:32 why dont you count that too? 2011-11-24T10:41:12 It's what the starter set provides and it's just 140 lines 2011-11-24T10:41:32 where is that bot ranked? 2011-11-24T10:41:35 oh wait, 288 2011-11-24T10:41:44 It's ~150 2011-11-24T10:41:48 thats nice 2011-11-24T10:41:51 pretty nice 2011-11-24T10:42:11 mcstar: what's your ranking ? 2011-11-24T10:42:22 532 atm 2011-11-24T10:42:25 nice 2011-11-24T10:42:32 not really :D 2011-11-24T10:42:44 i dont have hill defense in that code 2011-11-24T10:42:44 well, nicer than me :p 2011-11-24T10:43:17 I don't have hill defense, have only poor exploration, i'm stateless, no combat strategy... 2011-11-24T10:44:37 I'm #1558, my best attempt for now (before trying an anti-object approach) was #1200 2011-11-24T10:44:46 still much work... 2011-11-24T10:45:26 i was 399 before i uploaded this one 2011-11-24T10:45:26 Are there going to be new maps for the final? I see some new ones today. 2011-11-24T10:45:29 but i'm having a hell of fun :) 2011-11-24T10:45:52 bes_: probably all the maps will be played before the finals 2011-11-24T10:46:44 new maps will be generated for the final 2011-11-24T10:47:07 the same kinds of map generators will be used though 2011-11-24T10:47:10 no surprises there 2011-11-24T10:47:24 oh 2011-11-24T10:47:33 a whole new set? 2011-11-24T10:47:43 but they still work on the map generation so things could change 2011-11-24T10:48:06 yea else you could just include all maps into your "knowledge base" 2011-11-24T10:48:23 true 2011-11-24T10:48:25 sorry 2011-11-24T10:48:32 i dont think in terms of cheating 2011-11-24T10:48:45 (and thats cheating according to my standards _D) 2011-11-24T10:48:47 : 2011-11-24T10:48:57 hehe 2011-11-24T10:49:05 yea would be pretty boring then anyways :) 2011-11-24T10:49:50 those who chose genetic programming will have a ard time... 2011-11-24T10:49:55 *hard 2011-11-24T10:50:30 The visualiser overlay is horribly slow for me 2011-11-24T10:50:55 depends on what they evolve 2011-11-24T10:51:05 but they have a hard time anyways i guess :) 2011-11-24T10:52:36 *** amstan_ has joined #aichallenge 2011-11-24T10:52:36 *** ChanServ sets mode: +o amstan_ 2011-11-24T10:55:18 *** Fandekasp has joined #aichallenge 2011-11-24T10:57:22 *** praveen has joined #aichallenge 2011-11-24T11:00:12 *** dom7b5 has quit IRC (Remote host closed the connection) 2011-11-24T11:00:52 *** amstan_ has quit IRC (Ping timeout: 245 seconds) 2011-11-24T11:02:01 *** hanshoi has quit IRC (Remote host closed the connection) 2011-11-24T11:05:54 http://aichallenge.org/profile.php?user=3358 2011-11-24T11:07:31 52 versions 2011-11-24T11:08:57 in visual basic.. must be hell 2011-11-24T11:11:11 well, we know what ai doesnt like lisp and visual basic than 2011-11-24T11:11:18 that* 2011-11-24T11:12:36 (i used to know the difference between than and then, but i no longer do, please, excuse me) 2011-11-24T11:12:37 Doesn't like lisp? 2011-11-24T11:12:47 pairofdice: apparently 2011-11-24T11:13:13 they tried to create ai in lisp for soo long, and there is none 2011-11-24T11:13:21 heh 2011-11-24T11:13:36 *** cowbandit has joined #aichallenge 2011-11-24T11:13:40 I give it another 10 years 2011-11-24T11:13:57 *** rajanaresh has joined #aichallenge 2011-11-24T11:14:50 i think ray kurtzweil will die by the time we will have a general purpose /strong ai 2011-11-24T11:16:31 has a deadline been set yet for this contest? 2011-11-24T11:16:51 I think Dec. 18 is the submission deadline 2011-11-24T11:17:15 but i'm not sure about teh final rounds if that's what you mean 2011-11-24T11:18:11 nah I meant submission deadline. Is that anywhere on the website? 2011-11-24T11:18:15 I have quite a lot of submissions too, but that's because the server's D compiler has bugs that caused my bot to compile wrongly 2011-11-24T11:18:35 *** jstrong has joined #aichallenge 2011-11-24T11:18:59 *** jstrong is now known as roflmao 2011-11-24T11:19:16 i just wanted to highlight the guy's evolutionary effort, and that he is doing pretty badly 2011-11-24T11:19:57 of course i cant be sure whether his bot is a geneticist or not 2011-11-24T11:20:01 but its possible 2011-11-24T11:20:44 mcstar: you know bocsimacko wrote the winning planet wars bot in lisp right? :P 2011-11-24T11:20:52 *** cowbandit has quit IRC (Quit: Page closed) 2011-11-24T11:20:53 yes so? 2011-11-24T11:21:01 that isnt ai 2011-11-24T11:21:06 :D 2011-11-24T11:21:17 what is ai :P 2011-11-24T11:21:26 antimatroid: he is the reason i learned lisp 2011-11-24T11:21:33 *g 2011-11-24T11:22:37 *** aerique has quit IRC (Quit: ...) 2011-11-24T11:22:46 When AI will be able to create stronger AI, that's AI ;) 2011-11-24T11:23:07 thats the singularity 2011-11-24T11:23:18 Shortly after, yes 2011-11-24T11:24:27 well, yeah, thats a brief point in time(on a human timescale even) so might be right to define the "real stuff" by its ability to reach singularity 2011-11-24T11:24:52 but there might be more than 1 singularities 2011-11-24T11:24:58 i think i might just do all searches at once sometimes 2011-11-24T11:25:01 it can make life easier 2011-11-24T11:25:22 *** patrisk has joined #aichallenge 2011-11-24T11:25:29 antimatroid: life would even be more easier if i stop anything related to this contest 2011-11-24T11:25:38 shhhh :P 2011-11-24T11:26:23 Random map generator: http://pastebin.com/WBDkKFFj 2011-11-24T11:26:42 I need to share that with one of the contest admins but I can't remember which 2011-11-24T11:26:44 bbl 2011-11-24T11:27:08 Extrarius: amstan or meclopold 2011-11-24T11:27:12 lol 2011-11-24T11:27:14 not mec 2011-11-24T11:27:16 mc 2011-11-24T11:27:25 and not lopold 2011-11-24T11:27:28 but leopold 2011-11-24T11:28:30 And 3d printers that can build better 3d printers... *starts drooling* 2011-11-24T11:29:08 pairofdice: ah, you want a 3d woman right? 2011-11-24T11:29:15 Shh! 2011-11-24T11:29:18 yep, today's printers quality are not acceptabnle 2011-11-24T11:29:32 *** tobym has joined #aichallenge 2011-11-24T11:31:04 fark 2011-11-24T11:31:12 i want to make a fairly big structural change to things 2011-11-24T11:31:50 i wanna paste a scenario 2011-11-24T11:32:18 i have to learn to use xclip 2011-11-24T11:33:34 http://paste.aichallenge.org/RtIEO/ antimatroid 2011-11-24T11:33:43 xclip is a bliss 2011-11-24T11:35:22 problem averted 2011-11-24T11:35:29 and actually cleaned stuff up a bit with solution 2011-11-24T11:35:32 rather than complicating it 2011-11-24T11:44:07 *** deltree_ has joined #aichallenge 2011-11-24T11:45:10 scratch that 2011-11-24T11:45:17 you can't easily just continue a search going 2011-11-24T11:45:23 *** praveen has quit IRC (Ping timeout: 265 seconds) 2011-11-24T11:45:25 *** deltree__ has quit IRC (Ping timeout: 240 seconds) 2011-11-24T11:47:27 *** treeform has joined #aichallenge 2011-11-24T11:49:28 *** treeform has quit IRC (Read error: Connection reset by peer) 2011-11-24T11:53:34 *** treeform has joined #aichallenge 2011-11-24T11:55:25 depends on the search 2011-11-24T11:55:42 indiana jones didnt gave up his search for the saint graal 2011-11-24T11:56:01 but he has been stopped a couple of times 2011-11-24T11:56:15 gave->give 2011-11-24T12:00:04 *** Fandekasp has quit IRC (Quit: Lost terminal) 2011-11-24T12:05:04 *** patrisk has quit IRC (Quit: Leaving) 2011-11-24T12:07:26 *** a has joined #aichallenge 2011-11-24T12:07:52 *** a is now known as Guest72106 2011-11-24T12:10:51 *** Fandekasp has joined #aichallenge 2011-11-24T12:16:56 *** Jamel has joined #aichallenge 2011-11-24T12:26:47 antimatroid: http://paste.aichallenge.org/P7HPs/ 2011-11-24T12:26:51 *** mviel__ has joined #aichallenge 2011-11-24T12:27:01 a bit better now 2011-11-24T12:28:44 *** iglo has joined #aichallenge 2011-11-24T12:30:10 *** praveen has joined #aichallenge 2011-11-24T12:30:44 *** mviel_ has quit IRC (Ping timeout: 258 seconds) 2011-11-24T12:47:51 *** mj41 has quit IRC (Ping timeout: 244 seconds) 2011-11-24T12:50:57 *** peyton has joined #aichallenge 2011-11-24T12:57:13 *** praveen has quit IRC (Quit: Page closed) 2011-11-24T12:57:36 *** amstan has joined #aichallenge 2011-11-24T12:57:36 *** ChanServ sets mode: +o amstan 2011-11-24T12:59:57 *** ThatSnail has joined #aichallenge 2011-11-24T13:02:54 *** bes_ has quit IRC (Quit: Page closed) 2011-11-24T13:06:28 *** Blkt has quit IRC (Remote host closed the connection) 2011-11-24T13:11:19 *** tass_adar has joined #aichallenge 2011-11-24T13:11:25 *** Anilm3 has joined #aichallenge 2011-11-24T13:11:34 Hi guys! 2011-11-24T13:11:55 *** tass_adar has quit IRC (Client Quit) 2011-11-24T13:13:18 *** Redgis has joined #aichallenge 2011-11-24T13:13:52 *** epicmonkey has joined #aichallenge 2011-11-24T13:15:16 *** treeform_ has joined #aichallenge 2011-11-24T13:17:31 *** buq2 has quit IRC (Read error: Connection reset by peer) 2011-11-24T13:18:56 *** treeform has quit IRC (Ping timeout: 252 seconds) 2011-11-24T13:25:03 *** treeform_ has quit IRC (Remote host closed the connection) 2011-11-24T13:25:37 *** treeform has joined #aichallenge 2011-11-24T13:29:16 there's a cheap way for everyone's score to go up: http://aichallenge.org/visualizer.php?game=122566&user=3473 2011-11-24T13:30:55 tying with GarySWest? 2011-11-24T13:31:19 several top bots stalemating 2011-11-24T13:32:32 I think my bot failed there due to distances being so big 2011-11-24T13:32:36 I'll have to get that map and try it out 2011-11-24T13:33:07 oh yeah, heh. you very effectively walled off that dead end. 2011-11-24T13:34:30 *** Antimony has joined #aichallenge 2011-11-24T13:38:47 is A* useful here? 2011-11-24T13:39:01 yes 2011-11-24T13:40:16 what about value iteration over the whole map 2011-11-24T13:40:59 I try to understand and add A* but I fear that there could be time outs 2011-11-24T13:41:34 *** mleise has joined #aichallenge 2011-11-24T13:42:13 do u calculate the path for every single ant and one target? that seems the simpliest way to do that 2011-11-24T13:44:45 *** grwip has quit IRC (Remote host closed the connection) 2011-11-24T13:44:54 iglo: I use A* and I've had no timeouts so far 2011-11-24T13:46:26 how far are ur distance? 2011-11-24T13:46:38 I try to begin with something like 5 or 6 2011-11-24T13:47:01 iglo: c++, bfs, 12000 searches of a 60 step length path in 1 sec 2011-11-24T13:47:11 thats my benchmark 2011-11-24T13:47:18 You have to really tru to get timeouts with those distances 2011-11-24T13:47:47 Well, not much... 2011-11-24T13:51:27 That's very inefficient, I need only 5 BFS's to use that second 2011-11-24T13:51:54 *** mj41 has joined #aichallenge 2011-11-24T13:52:09 bfs? 2011-11-24T13:52:22 ??? are you trying to be efficient in time killing? 2011-11-24T13:52:23 pairofdice: what are you talking about? 2011-11-24T13:52:32 :> 2011-11-24T13:52:45 I need to redo my BFS 2011-11-24T13:52:47 mcstar: c++ here too 2011-11-24T13:52:50 mleise: good to see you 2011-11-24T13:53:30 iglo: bfs: breadth first search, mostly worse performance wise than a* 2011-11-24T13:53:31 im just peeking in :) 2011-11-24T13:54:05 pairofdice: i still dont understand you, "I need only 5 BFS's to use that second" 2011-11-24T13:54:06 wtf? 2011-11-24T13:54:28 Joke. 2011-11-24T13:55:12 * mcstar still struggles with understanding real humans 2011-11-24T13:55:24 Who are you calling a real human 2011-11-24T13:55:31 Are you Data from Startreck? 2011-11-24T13:55:36 track 2011-11-24T13:55:39 ? 2011-11-24T13:55:42 treck? 2011-11-24T13:55:44 nooo 2011-11-24T13:55:47 xDD 2011-11-24T13:55:48 trek!!!! 2011-11-24T13:55:54 *** peyton has quit IRC (Quit: peyton) 2011-11-24T13:56:13 i wish i were data.... 2011-11-24T13:56:17 wow my new bot just got murderized by my old bot 2011-11-24T13:56:25 I wish too, I would have won this game already 2011-11-24T13:56:28 new bot did *exactly* what I had envisioned...and it sucked 2011-11-24T13:56:40 Anilm3: i had better things to do if i were data 2011-11-24T13:56:58 BenJackson: congrats 2011-11-24T13:57:03 You would have time for everything since you wouldn't even sleep 2011-11-24T13:57:09 *** liberforce has left #aichallenge 2011-11-24T13:58:29 0.68 seconds, sir, for an andoid, thats nearly an eternity 2011-11-24T13:58:36 r 2011-11-24T13:59:12 has anyone gotten scenario maps to work? 2011-11-24T13:59:17 yes 2011-11-24T13:59:36 Antimony: --food none 2011-11-24T13:59:42 i went crazy over this 2011-11-24T14:00:08 oh 2011-11-24T14:00:14 I tried food random 2011-11-24T14:00:26 but that doesn't work, because it still checks symmetry when spawning inital food 2011-11-24T14:00:32 food none 2011-11-24T14:01:02 thanks 2011-11-24T14:01:07 *** g0llum has quit IRC (Read error: Connection reset by peer) 2011-11-24T14:01:34 *** g0llum has joined #aichallenge 2011-11-24T14:02:19 now I need to track down the mysterious exceptions 2011-11-24T14:02:23 mleise, mcleo had some funny mapping idea this afternoon, and lured fluxid into trying it .. http://ants.fluxid.pl/replay.13612 2011-11-24T14:02:26 in my bot 2011-11-24T14:03:44 so everyone starts out with a ton of hills? 2011-11-24T14:03:51 I couldn't even figoure out what went on in that map 2011-11-24T14:04:10 abomination 2011-11-24T14:04:20 ;) 2011-11-24T14:05:13 BenJackson: good idea ... my new version is lmost the same like the old one... no major advance... 2011-11-24T14:05:18 lol g0llum 2011-11-24T14:06:20 well, it was fun for a few minutes, then vanished for the better. 2011-11-24T14:07:07 meso-scale management is hard, very hard 2011-11-24T14:07:17 hahahaha 2011-11-24T14:08:10 too bad all the bots are pretty dumb 2011-11-24T14:09:02 dumbness, like beauty, is in the eye of the beholder 2011-11-24T14:10:03 http://paste.aichallenge.org/CHRCm/ 2011-11-24T14:10:06 im debugging this 2011-11-24T14:10:15 the same code in 2 different situations 2011-11-24T14:11:11 Spartan ants 2011-11-24T14:15:25 wow, oj wasn't doing very well 2011-11-24T14:15:39 f 2011-11-24T14:15:46 i hardcoded a length 2011-11-24T14:15:54 *** Palmik has quit IRC (Read error: Connection reset by peer) 2011-11-24T14:16:48 mcstar, o noes, yuo're goig to start *actual_size* jokes.. 2011-11-24T14:16:55 *** Palmik has joined #aichallenge 2011-11-24T14:17:23 sry, have no idea about *actual_size* jokes 2011-11-24T14:20:19 g0llum: would you pls enlighten me? 2011-11-24T14:20:39 ive been programming too long, and i really dont get humans anymore 2011-11-24T14:20:43 no, stupid one, let's settle it. 2011-11-24T14:20:47 k 2011-11-24T14:26:49 *** AntDroid has quit IRC (Ping timeout: 265 seconds) 2011-11-24T14:27:43 *** Antimony has quit IRC (Ping timeout: 248 seconds) 2011-11-24T14:28:32 hmm, i thought i had a bug but i don't 2011-11-24T14:29:16 if you have a priority queue and you insert two different but "equal priority" items would you expect them to come out in the order you inserted them? 2011-11-24T14:29:40 try and tell us 2011-11-24T14:29:57 i'm just using the stl priority queue 2011-11-24T14:29:59 no 2011-11-24T14:30:06 mcstar: why not? 2011-11-24T14:30:18 i dont know 2011-11-24T14:30:19 antimatroid: that's implementation dependent, so i would expect whatever my language docs tell me to expect :) 2011-11-24T14:30:42 depends where does the algorithm inserts the new one 2011-11-24T14:30:47 -s 2011-11-24T14:31:14 ok 2011-11-24T14:31:16 before 2011-11-24T14:31:26 so no, they would come out in the same order 2011-11-24T14:31:29 thats my bet 2011-11-24T14:32:13 i'm confused? 2011-11-24T14:32:17 http://www.cplusplus.com/reference/stl/priority_queue/ 2011-11-24T14:32:21 that doesn't seem to help either? 2011-11-24T14:32:53 antimatroid: since the documentation available doesn't help too much, isn't it easier to just try it out? 2011-11-24T14:33:32 pita to set up a test case 2011-11-24T14:34:00 i need to write order functions for something weakly and not partially ordered 2011-11-24T14:34:07 then get an example and throw it all in :\ 2011-11-24T14:35:52 if you need them to be ordered differently, can't you just add an incrementally less(or more) tiny value to the priorities when you insert them? 2011-11-24T14:36:09 not ideal 2011-11-24T14:36:42 antimatroid: just consider this: it inserts the element, where the ordering is broken 2011-11-24T14:36:47 *** magikmw has quit IRC (Ping timeout: 248 seconds) 2011-11-24T14:36:51 i.e. before the first elemnt that breaks it 2011-11-24T14:36:58 an equal element doesnt break it 2011-11-24T14:37:11 so it will skip that, and inserts it after that, before the worng one 2011-11-24T14:37:36 at least thats how i understand it 2011-11-24T14:37:46 anything else wouldnt make sense to me 2011-11-24T14:37:59 it probably doesn't find the element to break with linearly though aye? 2011-11-24T14:38:12 hence it probably does make sense that they wont necessarily come out in the right order 2011-11-24T14:38:23 oh wait 2011-11-24T14:38:28 hmmm 2011-11-24T14:38:30 so it should 2011-11-24T14:38:39 I agree with mcstar 2011-11-24T14:39:36 just to be clear, i should expect them to come out in the order added and I probably have a bug somewhere? 2011-11-24T14:40:00 would it suffice a simple example with integers? 2011-11-24T14:40:12 yeah i get that 2011-11-24T14:41:21 *** Raimondi has quit IRC (Read error: Connection reset by peer) 2011-11-24T14:41:34 antimatroid: no, i think what i said, means they come out in different order 2011-11-24T14:41:50 now i'm confused 2011-11-24T14:42:13 meaning, you get out at the end the second == element, and after that the firstly inserted one 2011-11-24T14:42:26 since you only need a weak ordering function 2011-11-24T14:42:33 i'm so confused right now 2011-11-24T14:42:34 I believe it just depends which one is the first operand 2011-11-24T14:42:36 but this is a grey area 2011-11-24T14:42:45 second and element are what? 2011-11-24T14:42:50 <= and < matters now 2011-11-24T14:43:03 so I think it just depends on implementation 2011-11-24T14:43:12 *** Antimony has joined #aichallenge 2011-11-24T14:43:12 i'm ordering it with greater 2011-11-24T14:43:17 *** roflmao has quit IRC (Quit: Leaving.) 2011-11-24T14:43:17 and it's a weak total order 2011-11-24T14:43:18 *** haashd has joined #aichallenge 2011-11-24T14:43:20 antimatroid: best is that you provide the ordering function 2011-11-24T14:43:46 antimatroid: ok, with greater than i think you dont have a "bug" 2011-11-24T14:43:56 the order function is fine... http://pastebin.com/4WQ9Zrnz 2011-11-24T14:44:06 *** Raimondi has joined #aichallenge 2011-11-24T14:44:06 antimatroid: apparently it isnt 2011-11-24T14:44:25 you could test for equality and explicitly decide this case 2011-11-24T14:44:56 make it reflexive? 2011-11-24T14:45:25 omg 2011-11-24T14:45:30 *** Antvolution has joined #aichallenge 2011-11-24T14:45:40 whats wrong with: return item1.f > item2.f; 2011-11-24T14:45:56 yeah... 2011-11-24T14:46:01 mcstar: it's highly likely that i'll add extra stuff in after that later 2011-11-24T14:46:04 i do for the a star one 2011-11-24T14:46:21 ok, but it was weird seeing that 2011-11-24T14:46:25 :p 2011-11-24T14:46:34 i'm going to sleep 2011-11-24T14:46:40 nite 2011-11-24T14:46:40 i'll worry about it when i wake up 2011-11-24T14:46:54 bye 2011-11-24T14:47:55 ahh, i should be able to test it with a priority queue for pairs of ints 2011-11-24T14:47:59 i'll do that when i awake 2011-11-24T14:49:19 errr, no i can't 2011-11-24T14:49:32 antimatroid: go to bed already 2011-11-24T14:50:48 haha 2011-11-24T14:51:02 *** haashd has left #aichallenge 2011-11-24T14:51:14 if you go to sleep now, tomorrow you'll have a reason to wake up and work directly! 2011-11-24T14:51:30 [11:29] if you have a priority queue and you insert two different but "equal priority" items would you expect them to come out in the order you inserted them? 2011-11-24T14:51:32 no 2011-11-24T14:51:43 but you can add a second lesser condition on a serial number 2011-11-24T14:51:46 reading the priority_queue docs, all it says is that priority queues satisfy the heap property. how exactly they are implemented is not mentioned (which is normal for a standards reference of course). the heap properties do not guarantee a stable sorting algorithm - meaning that you probably cannot rely on anything at all when putting in same priority items and it is _completely_ implementation dependent. 2011-11-24T14:52:26 yes, we know that 2011-11-24T14:52:33 but i tried to guess it 2011-11-24T14:53:03 mcstar: trying to guess what gcc devs might have done usually ends in tears. 2011-11-24T14:53:14 you never have to guess about STL 2011-11-24T14:53:17 bqf: not my tears 2011-11-24T14:53:20 they're templates so the source is in the header files 2011-11-24T14:53:26 you can just read it if you have a question 2011-11-24T14:53:53 but bqf has it right: priority_queue is a container adapter that satisfies the heap property and heap is not stable 2011-11-24T14:54:04 http://codepad.org/CA6fAeFM 2011-11-24T14:54:05 problem solved 2011-11-24T14:54:07 no bug 2011-11-24T14:54:29 what's wrong with std::pair? 2011-11-24T14:54:43 i needed my ordering to not be partial 2011-11-24T14:55:09 eh? 2011-11-24T14:55:40 notice i'm only ordering Pair's based off their first element 2011-11-24T14:55:47 yes 2011-11-24T14:55:57 I would call that a partial order 2011-11-24T14:55:57 pair's aren't ordered like that 2011-11-24T14:56:07 right, pairs use .second as a secondary key 2011-11-24T14:56:09 errr, none are partial 2011-11-24T14:56:10 it's late 2011-11-24T14:56:21 it's the data 2011-11-24T14:56:22 oic, you don't care about the 2nd element 2011-11-24T14:56:48 my point is i wanted to see that if two "equally ordered" elements that are not actually equal are added then do they come off in the order added 2011-11-24T14:56:50 and the answer is no 2011-11-24T14:57:08 even if they *had* the answer is still no ;-) 2011-11-24T14:57:09 it needed to not be antisymmetric 2011-11-24T14:57:13 that's what i wanted to say 2011-11-24T14:57:14 i think your example only proves that you cant rely on this 2011-11-24T14:57:21 mcstar: what? 2011-11-24T14:57:22 xDDD 2011-11-24T14:57:32 my example shows that they don't necessarily come off in the order added 2011-11-24T14:57:39 yes 2011-11-24T14:57:41 which means i don't necessarily have a bug 2011-11-24T14:57:45 which is what i thought 2011-11-24T14:57:46 you cant rely on any order 2011-11-24T14:57:57 *** epicmonkey has quit IRC (Ping timeout: 245 seconds) 2011-11-24T14:58:02 yeah that's fine 2011-11-24T14:58:14 well if your code relied on them coming out in a specific order, I daresay it had a bug 2011-11-24T14:58:23 thats why i said previously that you need to extend your < 2011-11-24T14:58:27 it doesn't rely on them coming out in a different order 2011-11-24T14:58:34 i was just surprised that they weren't coming off in the order added 2011-11-24T14:58:43 i thought the other ones i expected to see first weren't added 2011-11-24T14:58:49 but clearly they just weren't searched to 2011-11-24T14:58:49 ok 2011-11-24T15:00:39 btw std::priority_queue isn't suitable for A* if that's what you're doing 2011-11-24T15:00:47 why not? 2011-11-24T15:00:57 i do do it in other places :P 2011-11-24T15:01:02 I'm pretty sure you need "reduce key" for A* and priority_queue doesn't have that 2011-11-24T15:01:19 pretty sure i don't :P 2011-11-24T15:01:34 BenJackson: define "reduce key" 2011-11-24T15:01:38 please 2011-11-24T15:01:40 what reduce key? 2011-11-24T15:01:43 s 2011-11-24T15:01:44 why would you need that, you can just remove and object and reinsert it with a different key if you need to 2011-11-24T15:01:52 s/and/an/ 2011-11-24T15:01:56 if you rediscover an expanded node by way of a lower f you need to reorder the queue for that node 2011-11-24T15:02:08 ah, yeah very true 2011-11-24T15:02:08 remove/add 2011-11-24T15:02:11 Minthos: RIght, except priority_queue doesn't have "remove element" 2011-11-24T15:02:18 oh, mine had 2011-11-24T15:02:28 priority queues are quite inefficient for that 2011-11-24T15:02:28 i did one on top of lisp lists 2011-11-24T15:02:45 BenJackson: just check elements as you take them off if you've already found a better one? 2011-11-24T15:02:51 Anilm3: In heap parlance it's one of the operations. the heap is sorted by key. in a min-heap the "reduce key" operation means to take an existing member, reduce its key and then bubble it up closer to the front 2011-11-24T15:03:26 you can insert duplicate elements, yes 2011-11-24T15:03:35 I see, thanks 2011-11-24T15:04:05 there's a table on wikipedia with the heap operations and their cost in various heaps 2011-11-24T15:04:25 BenJackson: what is the data structure behind c++'s prique? 2011-11-24T15:04:33 binary heap 2011-11-24T15:04:41 it's just build on top of std::make_heap and friends 2011-11-24T15:04:42 thats probably an overkill 2011-11-24T15:04:46 it's a "container adapter" 2011-11-24T15:04:52 binary heap is the cheapest heap 2011-11-24T15:04:58 i tried 4 different heaps 2011-11-24T15:05:07 and my list based one was the fastest 2011-11-24T15:05:14 Hi. When is should start counting 'turntime'? When I receive 'go' or 'turn X' line? 2011-11-24T15:05:38 does it say on the "how to write a starter kit" wiki? 2011-11-24T15:05:59 *** Antimony has quit IRC (Ping timeout: 240 seconds) 2011-11-24T15:07:23 BenJackson: im just saying that for heap sizes that might occur in an a* search, it might be much faster to just use an ordered list 2011-11-24T15:07:39 im wondering if anyone did similar comparisons in c++ 2011-11-24T15:07:47 what i did was in common lisp/sbcl 2011-11-24T15:07:53 my friend Amit has a survey of queue types for A* 2011-11-24T15:07:57 on his game dev page 2011-11-24T15:08:02 friend? 2011-11-24T15:08:18 ive seen his paeg 2011-11-24T15:08:19 http://www-cs-students.stanford.edu/~amitp/gameprog.html 2011-11-24T15:08:32 is he really your friend? 2011-11-24T15:08:41 yes 2011-11-24T15:09:02 he used to hang out on a mud 2011-11-24T15:09:11 I wonder if he knows about this contest, he'd probably like it 2011-11-24T15:09:13 *** kaemo has quit IRC (Ping timeout: 276 seconds) 2011-11-24T15:10:34 who doesnt 2011-11-24T15:12:15 ok, so i think Patrick Lester advocates heaps 2011-11-24T15:12:37 maybe all 4 implementations(or 3) were "bad" that i tried 2011-11-24T15:14:55 *** Accoun has joined #aichallenge 2011-11-24T15:20:47 Hi. When is should start counting 'turntime'? When I receive 'go' or 'turn X' line? 2011-11-24T15:22:33 err ... sorry ... bad window 2011-11-24T15:22:58 sure? 2011-11-24T15:23:58 yes, 'up' and 'enter' ... xchat like my console 2011-11-24T15:26:15 googling how to do SIGALRM right in perl bot ... not sure I understadn safe vs. unsafe ones 2011-11-24T15:26:33 Totally bad window 2011-11-24T15:44:00 *** Palmik has quit IRC (Remote host closed the connection) 2011-11-24T15:45:32 mj41: why do you want to use alarm? 2011-11-24T15:45:59 you want to stop all processing after x ms no matter what? 2011-11-24T15:47:38 *** Harpyon_ has joined #aichallenge 2011-11-24T15:47:59 *** Harpyon has quit IRC (Ping timeout: 240 seconds) 2011-11-24T15:47:59 *** Harpyon_ is now known as Harpyon 2011-11-24T15:50:19 *** erer has joined #aichallenge 2011-11-24T15:51:35 *** delt0r_ has quit IRC (Ping timeout: 240 seconds) 2011-11-24T15:51:47 *** ltriant has joined #aichallenge 2011-11-24T15:52:18 *** luizribeiro has quit IRC (Ping timeout: 244 seconds) 2011-11-24T15:54:00 *** ruuhkis has joined #aichallenge 2011-11-24T15:54:51 Hello! Anyone programming ants in java and willing to tell me how should i debug my code on run as it wont print my debugs to the console. 2011-11-24T15:56:16 *** luizribeiro has joined #aichallenge 2011-11-24T15:58:22 tell playgame.py to print your bot's stderr output 2011-11-24T15:58:31 Fluxid: yes 2011-11-24T15:58:49 Minthos, thanks for the tip, I'll try 2011-11-24T15:59:30 ruuhkis, look for -e and -E there 2011-11-24T16:01:43 Thanks! :3 Will give it a shot now 2011-11-24T16:01:57 I've been stuck with my A* for ages and have tried debug it thru making log file :$ 2011-11-24T16:02:11 you can try the new visualizer 2011-11-24T16:02:22 http://forums.aichallenge.org/viewtopic.php?f=25&t=1633 2011-11-24T16:02:39 it helped me a lot with my A* 2011-11-24T16:02:57 *** killian has joined #aichallenge 2011-11-24T16:03:02 *** killian is now known as kire 2011-11-24T16:04:44 Thanks Anilm3! So i'll set the defaults to "True" and it should work (print error stream)? 2011-11-24T16:04:53 *** delt0r_ has joined #aichallenge 2011-11-24T16:06:05 *** magikmw has joined #aichallenge 2011-11-24T16:07:22 *** peyton has joined #aichallenge 2011-11-24T16:10:22 Fluxid: I'm still getting 'bot 0 timed out' ... It should be easy to use ALRM :-(. 2011-11-24T16:11:13 i didn't implement any time control until yesterday 2011-11-24T16:11:27 mj41: did you profile? 2011-11-24T16:11:58 *** LouisMartin has joined #aichallenge 2011-11-24T16:12:18 What are the Cell Maze? 2011-11-24T16:12:32 *** UncleVasya has joined #aichallenge 2011-11-24T16:13:14 or rather, Why the cell maze? 2011-11-24T16:13:25 LouisMartin: why not? they're beautiful: http://aichallenge.org/map.php?map=cell_maze/cell_maze_p02_01.map 2011-11-24T16:14:25 because you keep redefining "the problems" in the middle of the contest. Not all the approchs, strategies and their algo fit every redefinition/addition you throw at them in the middle of a contest. 2011-11-24T16:14:45 LouisMartin: bots should be able to work in any enviroment 2011-11-24T16:14:48 this is not a redefinitions of the contest 2011-11-24T16:14:51 It's like when IT customer comes with new specs in the middle of a project; it brings the same problems. 2011-11-24T16:14:56 i just think they're a bit oversized 2011-11-24T16:14:56 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-11-24T16:15:00 and the spec is still the same 2011-11-24T16:15:36 i don't see problems with new maps, actually i expect finals to use totally different maps than now 2011-11-24T16:15:55 yeah, in my case it helped me found important flaws 2011-11-24T16:15:58 Fluxid: any environment defined in the final problem definition, of course. but at some point you need to work with the problem defined and stop optimizing for how it's defined 2011-11-24T16:16:21 LouisMartin: maps are not included in problem description and spec imo 2011-11-24T16:17:37 Fluxid: where in the specs are we told about maps with a different kind of geometry and thin border like the new cell maps? why changes the data we have about the problem we're trying to nail? I don't get it. 2011-11-24T16:18:14 LouisMartin: where in spec we're told that maps won't change and look this or that way? 2011-11-24T16:18:37 afair there is no info about geometry. there is only information about symmetry 2011-11-24T16:19:02 and new maps are symmetric the same way as old ones are 2011-11-24T16:19:11 Fluxid: really?? There's MAZE as we know them (squared, thick edges) and there's RANDOM WALKS... that isn't part of the problem? of course it is! if you ask me to build something that fly in the sky and run on the ground, will you really presume I'll develop something that goes into the sea without mentioning it from the start? the keyword here is _Presume_ 2011-11-24T16:20:51 That analogy is quite unfair 2011-11-24T16:20:56 it is 2011-11-24T16:20:56 Fluxid: Unless specified, we're not to code everything in case you come up with new kind of maps that comes with new challenges. Where's the limit that we should guess when the problem is a moving target? 3D maps in the final? 2011-11-24T16:21:02 LouisMartin: so someone want you to create a bot which can walk in city 2011-11-24T16:21:18 but then you get angry because your bot is only designed to walk in moskow but not paris? 2011-11-24T16:21:42 Fluxid: if my comparison is unfair, that one is too 2011-11-24T16:21:51 Fluxid: nice comparison 2011-11-24T16:22:06 Fluxid: Yes ... I'm under 98 ms since I get start of turn ('turn \d+') ... I see ALRM was called many times but then I got e.g. 'turn 161 bot 0 timed out'. 2011-11-24T16:22:41 Traditionaly for this challenge it's a very bad idea to make any assumptions. Do not hardcode, be ready for the things that not in the game specs to be changed - it is one of the main principles. 2011-11-24T16:22:52 LouisMartin: i don't see a problem. new map meet the specs. and you made a mistake optimizing bot for only one or two kinds of maps. no one said there will be no more maps 2011-11-24T16:23:37 The TYPES of maps are integral to the problem. And Yes, if you tell me from the start that we have Open fields, Downtown and SubUrban like maps, I don't see why I wouldn't optimize for those types and coding everything just in case you'd want the bot to visit Venise 2011-11-24T16:23:57 I just can't track down my problem with the pathfinder, I am getting headache from this.. Gotta start over tomorrow :< 2011-11-24T16:24:59 LouisMartin: i think you're the only unhappy one here. talk with organizers, because i'm kindo fo a fan of the new maps 2011-11-24T16:25:26 oh well, make assumptions according the guidelines and create a bugreport if an assumption isn't right :p 2011-11-24T16:25:42 I expected more maps. I even confirm this with Janzert yesterday and was glad to hear that. What I didn't expected and don't think I should have, is to expect NEW TYPES of maps. 2011-11-24T16:26:01 *** Anilm3 has quit IRC (Ping timeout: 260 seconds) 2011-11-24T16:26:08 *** sigh has joined #aichallenge 2011-11-24T16:26:20 The only thinh I can agree with is that maybe admins have to write somewhere on forums that such things as maps generation methods easily can be changed (if it isn't done yet). 2011-11-24T16:26:43 *** BenJackson has quit IRC (Quit: Page closed) 2011-11-24T16:27:33 On this channel everybody knows such specific things about this challenge, but it would be nice to make it clear for other contestants. 2011-11-24T16:28:38 *** ruuhkis has quit IRC (Ping timeout: 265 seconds) 2011-11-24T16:29:39 The only thing that was clear related to change was about variations of specific runtime-constant and new maps, not new TYPE of maps or new TYPE of runtime-constant. 2011-11-24T16:30:28 LouisMartin: Right 2 weeks before deadline for the Planet Wars (past challenge) map generation was changed hard too. Many people argued with administration here and on forums but decision wasn't canceled. 2011-11-24T16:30:29 Somebody did something map specific that broke? Good 2011-11-24T16:30:40 i can't find anything about types of maps in game spec or problem desc 2011-11-24T16:30:44 the level of adaptability required when you allow for that kind of change is realistic and hurts those who don't expect it to be that deep 2011-11-24T16:31:03 *** g0llum has quit IRC (Read error: Connection reset by peer) 2011-11-24T16:32:29 pairofdice: I agree! except I didn't. But I did optimized for the type of environment that was pretty well defined so far: Maze and Random Walks as we knew it, as they were strongly typed. e.g. the maze were pretty squared. 2011-11-24T16:33:33 *** BenJackson has joined #aichallenge 2011-11-24T16:33:34 *** HaraKiri has quit IRC () 2011-11-24T16:34:12 Fluxid: The maps (both typed by their characteristics and their naming) themselves define implicitly the problem. 2011-11-24T16:35:40 LouisMartin: you made an unfounded assumption, deal with it 2011-11-24T16:35:54 Indeed 2011-11-24T16:36:10 There are no new types of tiles, no new types of anything 2011-11-24T16:36:23 mcstar: I disagree about it being an unfounded assumption. 2011-11-24T16:36:26 And this is about how to understand the goal of maps in the starterpack. I believe maps there are only to make your life easier so you don't need to generate them by yourself. 2011-11-24T16:37:10 If such a simple thing as more organic shapes are a disadvantage to your bot, better for me 2011-11-24T16:37:11 LouisMartin: your assumption was not based on the rules, it was based on guesswork on your part - therefore it is unfounded 2011-11-24T16:37:55 They stated from the beginning that there will be new maps 2011-11-24T16:38:00 i find those maps visually pleasing 2011-11-24T16:38:01 LouisMartin: now you have 3 weeks or so to correct your bot to not be based on unfounded assumptions 2011-11-24T16:39:58 *** Anilm3 has joined #aichallenge 2011-11-24T16:40:16 UncleVasya: "I believe" .. so you have to guess such an important thing. Those kind of things decided on the level of optimization vs generality tradeoff you will go for. It's just influence the whole architecture of the software... anything that has that kind of depth should not be let to guessing and should be explicitly warned of in the form of telling the problem solver to not assume that the initial map were defining 2011-11-24T16:40:18 into. 2011-11-24T16:40:44 mcstar: they are visually pleasing 2011-11-24T16:40:52 *** Fandekasp has quit IRC (Ping timeout: 244 seconds) 2011-11-24T16:44:10 well, i knew for a fact to expect the final maps to change. i don't know where it says it on the website (if at all) but it's definitely part of the agreed-upon rules 2011-11-24T16:44:14 those are not unfounded assumptions when a problem that are defined both implicitly and explicitly doesn't defined well that is not to be interpreted as implicit definition. In other word: The scope of the Implicit spec should be at least well shaped in the explicit spec of the problem definition. 2011-11-24T16:44:57 still talking about the maps? 2011-11-24T16:45:02 another guy, Accoun, went ballistic about this exact same thing over the last three contests 2011-11-24T16:45:39 Laughable 2011-11-24T16:45:57 not sure as I barely can understand the kind of english I used in my last comment ..... maps, penguin, WWII.. who knows 2011-11-24T16:46:05 heh, so you can get in touch with him and make a rebelition :) 2011-11-24T16:46:39 here we go.. with that map problem again 2011-11-24T16:46:43 anyway perhaps the website needs some clarifications about the maps. 2011-11-24T16:46:48 They have stated their spec. Nowhere in there does it say anything about the ordering of water tiles 2011-11-24T16:47:11 oh, I'll be alright.. I'll get over it. 2011-11-24T16:47:12 a1k0n: i'm sure there's a bunch of warnings on there and on the forums 2011-11-24T16:47:37 it is up to the contestant to make test cases and check how their bot works 2011-11-24T16:47:55 also, the whole live-scoreboard thing is mostly a courtesy, as it doesn't have anything to do with the final maps or rankings (necessarily) 2011-11-24T16:48:36 amstan: no, not about the type of maps but about the maps themselves; that the maps couldn't be the one in the finals. There's an important distinction between the map instances and their types (typed characteristics) 2011-11-24T16:48:47 of course you could use that to justify changing attackradius. and actually i'd be okay with that, cuz my bot is adaptable to that, but not everyone would 2011-11-24T16:48:57 Type in this case is purely cosmetic 2011-11-24T16:49:02 LouisMartin: yep, now I agree that you had a right to understand maps in the starterpack as a defining part of the contest. 2011-11-24T16:49:38 LouisMartin: i disagree 2011-11-24T16:50:09 amstan: Since people still have a problems with making assumptions I think it's good to write in several places with big green (red) literas a warning. :) 2011-11-24T16:50:17 LouisMartin: maps can range from all empty to all flooded with water(assuming bases are connected), anything between that is fair game 2011-11-24T16:50:27 im very flexible, i expect the turntime will be lowered to 100ms in the finals 2011-11-24T16:50:36 mcstar: that will probably not happen 2011-11-24T16:50:44 oh please don't do that! 2011-11-24T16:50:45 * mcstar is flaming 2011-11-24T16:50:45 mcstar: given that the java starter pack times out during parsing 2011-11-24T16:50:49 pairofdice: find me where it's said that those type are 'cosmetic'? those typing allows for interesting optimizations which influence the architecture and the choice of algo of the bot.. this shouldn't be let to guesswork that it's only 'cosmetic' 2011-11-24T16:51:20 Optimization in this case meaning "bound to break" 2011-11-24T16:51:43 LouisMartin: what is given: you get a fair chance on any map 2011-11-24T16:51:55 i.e. positioning wise 2011-11-24T16:52:05 hills and initial food will be distributed accordinlgy 2011-11-24T16:52:20 ok? 2011-11-24T16:52:40 now number of players, and "look and feel" of water is a variable 2011-11-24T16:52:44 amstan: why should we not assume it to go down to 100ms? and yes, it's a point. Possible variations makes those contest interesting, but some types of variations should be clearly stated because of how it impact how we attack the problem. 2011-11-24T16:53:18 amstan: it would be nice some reassurance about the time :P 2011-11-24T16:54:11 that is a much more important parameter btw 2011-11-24T16:54:16 use percents :D 2011-11-24T16:54:19 some lower limit on turntime 2011-11-24T16:54:28 Much more... 2011-11-24T16:54:40 All python bots qould start timing out 2011-11-24T16:54:59 Or just sitting still 2011-11-24T16:55:27 mcstar: I like new maps because all 2 games on them my guy have won :) 2011-11-24T16:55:44 UncleVasya: me too haha 2011-11-24T16:56:10 LouisMartin: here.. it's stated: http://aichallenge.org/specification.php#Map-Format 2011-11-24T16:56:12 pairofdice: of course it implies that! do you really think you're making a point here? you need to optimize in ways that respect the problem definition in terms of possible variations so as to not over-optimize and fragilize your bot. But the problem needs to be well defined to allow for any kind of optimization. I thought it was and didn't go with absolute generalization. 2011-11-24T16:56:15 scroll down to map generator 2011-11-24T16:57:34 esult in interesting games while not allowing for people to hard code strategies. 2011-11-24T16:57:44 ive seen this line before 2011-11-24T16:58:11 LouisMartin: Please continue making your bot as fragile as you want. 2011-11-24T16:58:27 about the turn time... what's the problem? where does it says that it couldn't go down to 100ms? you guys should anticipate that by your own logic, no? what's the difference here? 2011-11-24T16:58:28 *** kaemo has joined #aichallenge 2011-11-24T16:58:36 amstan: there was a guys here who submitted a new map generator 2011-11-24T16:58:40 for got his name 2011-11-24T16:58:51 -s 2011-11-24T16:59:31 amstan: Extrarius 2011-11-24T16:59:51 LouisMartin: Althought we are all seems to oppose you in this discussion, do not think we are evil guys :) Hope you still will find this challenge enjoyable ;) 2011-11-24T17:00:22 i'm ok with 100ms turns 2011-11-24T17:00:27 :D 2011-11-24T17:00:45 really a1k0n? 2011-11-24T17:00:49 sure 2011-11-24T17:00:59 wow 2011-11-24T17:01:23 a1k0n has an algorithm that actually looks into the future, and he can anticipate moves easily, he only needs time to output his commands 2011-11-24T17:01:39 yes, it uses tachyons 2011-11-24T17:01:49 "while not allowing for people to hard code strategies." 2011-11-24T17:01:50 bingo 2011-11-24T17:01:53 pairofdice: why the arrogance? If you want to argue, we can. Bring a point, not unwarranted sarcasm over coding philosophies such as optimization vs generalization in software engineering. please, wth. 2011-11-24T17:01:55 In that case aikon will upload DoNothingBot and will win the competition because everyone else will be timeouted :) 2011-11-24T17:02:22 timed out 2011-11-24T17:02:37 thanks 2011-11-24T17:02:43 it seems i already did that. this is pretty damn embarassing: http://aichallenge.org/visualizer.php?game=124073&user=432 2011-11-24T17:03:16 a1k0n: wth happened? 2011-11-24T17:03:33 was that some early version? 2011-11-24T17:03:41 and now I'm done with this map thing. Any "fighting" atmosphere here about such thing isn't worth it for anyone. I just wanted to make my point and get some answers and that's done. thanks for the arguing as it's way better than being ignored in any case. 2011-11-24T17:03:41 no that's my current one, in 7th place 2011-11-24T17:04:02 since my moves are scored with linear combinations, there are cases where moving away from some food in order to move towards other food results in... that. 2011-11-24T17:04:10 it's pretty dumb. 2011-11-24T17:04:21 UncleVasya: hehe, yea, I will. thanks 2011-11-24T17:04:26 a1k0n: im actually surprised that it couldnt handle tha tie 2011-11-24T17:07:33 *** sigh has quit IRC (Remote host closed the connection) 2011-11-24T17:08:26 *** Jak_o_Shadows has joined #aichallenge 2011-11-24T17:10:57 *** goffrie_ has joined #aichallenge 2011-11-24T17:10:57 *** goffrie has quit IRC (Disconnected by services) 2011-11-24T17:10:58 *** goffrie_ is now known as goffrie 2011-11-24T17:12:02 the discussion was quite entertaining... 2011-11-24T17:13:02 *** peyton has quit IRC (Quit: peyton) 2011-11-24T17:14:02 *** UncleVasya has quit IRC (Ping timeout: 244 seconds) 2011-11-24T17:16:48 *** peyton has joined #aichallenge 2011-11-24T17:19:17 *** UncleVasya has joined #aichallenge 2011-11-24T17:21:17 *** foRei has quit IRC (Quit: Bye) 2011-11-24T17:22:51 lets not ridicule him 2011-11-24T17:28:26 could you mcstar? 2011-11-24T17:28:42 sorry? 2011-11-24T17:29:00 apologies accepted 2011-11-24T17:29:08 nonono 2011-11-24T17:29:13 i didnt apologize 2011-11-24T17:29:20 yes, you did 2011-11-24T17:29:21 im just asking could i what? 2011-11-24T17:29:32 sorry, too late. you did. 2011-11-24T17:29:38 sorry==excuse me 2011-11-24T17:29:48 cause i want sure what you meant 2011-11-24T17:29:55 and i dont think i was hostile to you 2011-11-24T17:30:03 but if i was, than im sorry 2011-11-24T17:30:17 then 2011-11-24T17:30:25 *** AntDroid has joined #aichallenge 2011-11-24T17:32:37 *** amstan has quit IRC (Quit: Konversation terminated!) 2011-11-24T17:33:16 ok, I'm sorry too. Here's some cats as proof of my sincerity: http://icanhascheezburger.com/ 2011-11-24T17:33:41 *** erer has quit IRC (Quit: Page closed) 2011-11-24T17:34:09 Anilm3: How was it entertaining, if I can ask? 2011-11-24T17:39:09 *** antimatroid has quit IRC (Quit: Leaving.) 2011-11-24T17:39:21 *** antimatroid has joined #aichallenge 2011-11-24T17:39:40 *** AntDroid has quit IRC (Ping timeout: 265 seconds) 2011-11-24T17:44:20 *** Mazer1010 has joined #aichallenge 2011-11-24T17:45:16 *** Mazer1010 has quit IRC (Client Quit) 2011-11-24T17:48:00 I think we'll see a new leader in few days. 2011-11-24T17:48:37 GreenTea has updated his bot. Go go Ukraine! :) 2011-11-24T17:52:32 *** Jamel has quit IRC (Quit: Page closed) 2011-11-24T17:56:44 *** peyton has quit IRC (Quit: peyton) 2011-11-24T18:03:34 *** kire has quit IRC (Remote host closed the connection) 2011-11-24T18:04:53 *** aarossig has quit IRC (Remote host closed the connection) 2011-11-24T18:13:02 *** Antvolution has quit IRC (*.net *.split) 2011-11-24T18:13:02 *** og01 has quit IRC (*.net *.split) 2011-11-24T18:13:03 *** Lithosphere has quit IRC (*.net *.split) 2011-11-24T18:13:03 *** jstemmer has quit IRC (*.net *.split) 2011-11-24T18:13:03 *** rajanaresh has quit IRC (*.net *.split) 2011-11-24T18:13:03 *** durarara has quit IRC (*.net *.split) 2011-11-24T18:13:03 *** magiik has quit IRC (*.net *.split) 2011-11-24T18:13:03 *** Ttech has quit IRC (*.net *.split) 2011-11-24T18:13:03 *** Clex has quit IRC (*.net *.split) 2011-11-24T18:13:03 *** p_l has quit IRC (*.net *.split) 2011-11-24T18:13:04 *** f4hy has quit IRC (*.net *.split) 2011-11-24T18:13:04 *** acieroid has quit IRC (*.net *.split) 2011-11-24T18:13:04 *** pgpaskar_ has quit IRC (*.net *.split) 2011-11-24T18:13:04 *** jnwhiteh has quit IRC (*.net *.split) 2011-11-24T18:13:04 *** Jak_o_Shadows has quit IRC (*.net *.split) 2011-11-24T18:13:04 *** Raimondi has quit IRC (*.net *.split) 2011-11-24T18:13:04 *** mj41 has quit IRC (*.net *.split) 2011-11-24T18:13:04 *** mleise has quit IRC (*.net *.split) 2011-11-24T18:13:04 *** iglo has quit IRC (*.net *.split) 2011-11-24T18:13:04 *** Cyndre has quit IRC (*.net *.split) 2011-11-24T18:13:04 *** Extrarius has quit IRC (*.net *.split) 2011-11-24T18:13:04 *** AlliedEnvy has quit IRC (*.net *.split) 2011-11-24T18:13:04 *** krishna has quit IRC (*.net *.split) 2011-11-24T18:13:05 *** welterde has quit IRC (*.net *.split) 2011-11-24T18:13:05 *** dr- has quit IRC (*.net *.split) 2011-11-24T18:13:05 *** Minthos has quit IRC (*.net *.split) 2011-11-24T18:13:05 *** jix has quit IRC (*.net *.split) 2011-11-24T18:13:05 *** jmcarthur has quit IRC (*.net *.split) 2011-11-24T18:13:05 *** spacebat has quit IRC (*.net *.split) 2011-11-24T18:13:05 *** Fluxid has quit IRC (*.net *.split) 2011-11-24T18:13:05 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (*.net *.split) 2011-11-24T18:13:05 *** fpcfan has quit IRC (*.net *.split) 2011-11-24T18:13:05 *** a1k0n has quit IRC (*.net *.split) 2011-11-24T18:13:05 *** rexou has quit IRC (*.net *.split) 2011-11-24T18:13:05 *** hkraal has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** sorki has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** MikeyG has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** rb_ has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** Sir_Ragnarok has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** UncleVasya has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** kaemo has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** Anilm3 has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** magikmw has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** Redgis has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** andar has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** cbad has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** antonh has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** JamesMG has quit IRC (*.net *.split) 2011-11-24T18:13:06 *** deafferret has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** mawif has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** isbric has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** nickjohnson has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** Zao has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** mviel__ has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** Conorach has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** Israfel has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** Yexo has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** bysin has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** Insti has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** jhawthorn has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** ejls has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** galdor has quit IRC (*.net *.split) 2011-11-24T18:13:07 *** timos has quit IRC (*.net *.split) 2011-11-24T18:13:08 *** artart78 has quit IRC (*.net *.split) 2011-11-24T18:13:08 *** delt0r has quit IRC (*.net *.split) 2011-11-24T18:13:08 *** mcstar has quit IRC (*.net *.split) 2011-11-24T18:13:08 *** jbroman has quit IRC (*.net *.split) 2011-11-24T18:13:08 *** cafaro has quit IRC (*.net *.split) 2011-11-24T18:13:08 *** Kingpin13 has quit IRC (*.net *.split) 2011-11-24T18:13:08 *** bugnuts2 has quit IRC (*.net *.split) 2011-11-24T18:13:08 *** j3camero has quit IRC (*.net *.split) 2011-11-24T18:13:08 *** Thinq has quit IRC (*.net *.split) 2011-11-24T18:13:09 *** Zepp has quit IRC (*.net *.split) 2011-11-24T18:13:09 *** dai-ra has quit IRC (*.net *.split) 2011-11-24T18:13:09 *** lnx has quit IRC (*.net *.split) 2011-11-24T18:13:09 *** juuso_ has quit IRC (*.net *.split) 2011-11-24T18:13:09 *** klutometis has quit IRC (*.net *.split) 2011-11-24T18:14:28 *** LouisMartin has quit IRC (Quit: Page closed) 2011-11-24T18:14:32 *** Jak_o_Shadows1 has joined #aichallenge 2011-11-24T18:14:32 *** UncleVasya has joined #aichallenge 2011-11-24T18:14:32 *** Jak_o_Shadows has joined #aichallenge 2011-11-24T18:14:32 *** kaemo has joined #aichallenge 2011-11-24T18:14:32 *** Anilm3 has joined #aichallenge 2011-11-24T18:14:32 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-11-24T18:14:32 *** magikmw has joined #aichallenge 2011-11-24T18:14:32 *** Raimondi has joined #aichallenge 2011-11-24T18:14:32 *** mj41 has joined #aichallenge 2011-11-24T18:14:32 *** mleise has joined #aichallenge 2011-11-24T18:14:32 *** Redgis has joined #aichallenge 2011-11-24T18:14:32 *** iglo has joined #aichallenge 2011-11-24T18:14:32 *** mviel__ has joined #aichallenge 2011-11-24T18:14:32 *** rajanaresh has joined #aichallenge 2011-11-24T18:14:32 *** Conorach has joined #aichallenge 2011-11-24T18:14:32 *** durarara has joined #aichallenge 2011-11-24T18:14:32 *** Cyndre has joined #aichallenge 2011-11-24T18:14:32 *** fpcfan has joined #aichallenge 2011-11-24T18:14:32 *** Extrarius has joined #aichallenge 2011-11-24T18:14:32 *** AlliedEnvy has joined #aichallenge 2011-11-24T18:14:32 *** andar has joined #aichallenge 2011-11-24T18:14:32 *** cbad has joined #aichallenge 2011-11-24T18:14:32 *** antonh has joined #aichallenge 2011-11-24T18:14:32 *** magiik has joined #aichallenge 2011-11-24T18:14:32 *** Lithosphere has joined #aichallenge 2011-11-24T18:14:32 *** JamesMG has joined #aichallenge 2011-11-24T18:14:32 *** Israfel has joined #aichallenge 2011-11-24T18:14:32 *** krishna has joined #aichallenge 2011-11-24T18:14:32 *** a1k0n has joined #aichallenge 2011-11-24T18:14:32 *** rexou has joined #aichallenge 2011-11-24T18:14:32 *** Yexo has joined #aichallenge 2011-11-24T18:14:32 *** hkraal has joined #aichallenge 2011-11-24T18:14:32 *** Ttech has joined #aichallenge 2011-11-24T18:14:32 *** welterde has joined #aichallenge 2011-11-24T18:14:32 *** bysin has joined #aichallenge 2011-11-24T18:14:32 *** Clex has joined #aichallenge 2011-11-24T18:14:32 *** sorki has joined #aichallenge 2011-11-24T18:14:32 *** deafferret has joined #aichallenge 2011-11-24T18:14:32 *** mawif has joined #aichallenge 2011-11-24T18:14:32 *** Insti has joined #aichallenge 2011-11-24T18:14:32 *** dr- has joined #aichallenge 2011-11-24T18:14:32 *** jstemmer has joined #aichallenge 2011-11-24T18:14:32 *** Minthos has joined #aichallenge 2011-11-24T18:14:32 *** isbric has joined #aichallenge 2011-11-24T18:14:32 *** MikeyG has joined #aichallenge 2011-11-24T18:14:32 *** nickjohnson has joined #aichallenge 2011-11-24T18:14:32 *** jhawthorn has joined #aichallenge 2011-11-24T18:14:32 *** rb_ has joined #aichallenge 2011-11-24T18:14:32 *** Sir_Ragnarok has joined #aichallenge 2011-11-24T18:14:32 *** p_l has joined #aichallenge 2011-11-24T18:14:32 *** ejls has joined #aichallenge 2011-11-24T18:14:32 *** jix has joined #aichallenge 2011-11-24T18:14:32 *** pgpaskar_ has joined #aichallenge 2011-11-24T18:14:32 *** acieroid has joined #aichallenge 2011-11-24T18:14:32 *** f4hy has joined #aichallenge 2011-11-24T18:14:32 *** jnwhiteh has joined #aichallenge 2011-11-24T18:14:32 *** Fluxid has joined #aichallenge 2011-11-24T18:14:32 *** spacebat has joined #aichallenge 2011-11-24T18:14:32 *** jmcarthur has joined #aichallenge 2011-11-24T18:14:32 *** Zao has joined #aichallenge 2011-11-24T18:14:32 *** galdor has joined #aichallenge 2011-11-24T18:14:32 *** timos has joined #aichallenge 2011-11-24T18:14:32 *** artart78 has joined #aichallenge 2011-11-24T18:14:32 *** delt0r has joined #aichallenge 2011-11-24T18:14:33 *** LouisMartin has joined #aichallenge 2011-11-24T18:14:33 *** Rav3nsW0rd has quit IRC (Max SendQ exceeded) 2011-11-24T18:14:33 *** Jak_o_Shadows has quit IRC (Ping timeout: 240 seconds) 2011-11-24T18:14:33 *** UncleVasya has quit IRC (Read error: Connection reset by peer) 2011-11-24T18:14:37 *** Rav3nsW0rd has joined #aichallenge 2011-11-24T18:14:49 *** clapautius has joined #aichallenge 2011-11-24T18:15:12 *** mcstar has joined #aichallenge 2011-11-24T18:15:12 *** jbroman has joined #aichallenge 2011-11-24T18:15:12 *** bugnuts2 has joined #aichallenge 2011-11-24T18:15:12 *** cafaro has joined #aichallenge 2011-11-24T18:15:12 *** Kingpin13 has joined #aichallenge 2011-11-24T18:15:12 *** j3camero has joined #aichallenge 2011-11-24T18:15:13 *** lnx has joined #aichallenge 2011-11-24T18:15:13 *** Thinq has joined #aichallenge 2011-11-24T18:15:13 *** Zepp has joined #aichallenge 2011-11-24T18:15:13 *** dai-ra has joined #aichallenge 2011-11-24T18:15:13 *** juuso_ has joined #aichallenge 2011-11-24T18:15:13 *** klutometis has joined #aichallenge 2011-11-24T18:15:14 *** HaraKiri has joined #aichallenge 2011-11-24T18:15:27 *** Antvolution has joined #aichallenge 2011-11-24T18:15:27 *** og01 has joined #aichallenge 2011-11-24T18:16:57 my python bot suffers from slowdowns at random times (turns) - each rund at a different place. what is this sorcery? some python internal mumbo jumbo? 2011-11-24T18:18:40 *** Antvolution has left #aichallenge 2011-11-24T18:19:31 *** tdubellz has quit IRC (Read error: Connection reset by peer) 2011-11-24T18:19:46 *** tdubellz has joined #aichallenge 2011-11-24T18:20:03 *** tdubellz is now known as Guest76135 2011-11-24T18:21:50 C# and VB has a similar problem and is caused - in part - by the garbage collector. Any GC in Python? 2011-11-24T18:23:07 that's what came into my mind too. but there's really no way fixing that, right? 2011-11-24T18:23:48 oh, google says you can actually control python's gc. hmm might be worth a look 2011-11-24T18:24:00 *** Anilm3 has quit IRC (Quit: Lost terminal) 2011-11-24T18:26:06 Controlling GC in your code? if that's possible in Python. Also, some types are causing more/longer garbage collection than others. ex: Struct vs Class. You may also look into that. 2011-11-24T18:27:31 well there's no structs in python. hmm disabling the gc did not help :( 2011-11-24T18:28:06 exact same symptoms (the random peaks)? 2011-11-24T18:29:17 yes, bot behaviour is identical, just time losses (up to 80ms) at random turns, and even in random positions in code (i plastered it with timers to find the cause) 2011-11-24T18:30:40 ok. sorry then, no idea. 2011-11-24T18:31:05 *** Redgis has quit IRC (Ping timeout: 244 seconds) 2011-11-24T18:33:23 *** Harpyon has quit IRC (Quit: Computer has gone to sleep.) 2011-11-24T18:34:04 can random harddisk slowdowns be a problem of that magnitude? but 80ms or so are sooo long 2011-11-24T18:37:38 I don't know. Are you writing/reading that much? 2011-11-24T18:38:29 nah, just tried disabling everything but one tiny line at end of program - didn't change anything. this is frustrating :(. i'll let it run on other computers. but thanks for your support 2011-11-24T18:39:14 *** mviel_ has joined #aichallenge 2011-11-24T18:39:34 really? then yes, running it on another computer is a good idea 2011-11-24T18:42:23 *** mcstar has quit IRC (Quit: WeeChat 0.3.6) 2011-11-24T18:42:59 *** mviel__ has quit IRC (Ping timeout: 245 seconds) 2011-11-24T18:44:04 *** besh has joined #aichallenge 2011-11-24T18:47:04 *** iglo has quit IRC (Remote host closed the connection) 2011-11-24T18:51:18 *** mj41 has quit IRC (Ping timeout: 240 seconds) 2011-11-24T18:53:40 *** tobym has quit IRC (Remote host closed the connection) 2011-11-24T18:53:58 *** McLeopold_ has joined #aichallenge 2011-11-24T18:54:19 *** besh has quit IRC (Quit: Page closed) 2011-11-24T18:55:44 @later tell LouisMartin: the game specification page specifies that the map generators are still being written and new one will be accepted. It also gives the specifications that the maps will follow. 2011-11-24T18:55:44 McLeopold_: Ready to serve. 2011-11-24T18:56:10 @later tell LouisMartin at that was specified before the opening of the contest 2011-11-24T18:56:10 McLeopold_: I come to serve. 2011-11-24T18:59:01 *** rajanaresh has quit IRC (Ping timeout: 240 seconds) 2011-11-24T18:59:12 *** dom7b5 has joined #aichallenge 2011-11-24T19:00:06 *** fpcfan has quit IRC (Ping timeout: 252 seconds) 2011-11-24T19:03:33 *** treeform has quit IRC (Read error: Operation timed out) 2011-11-24T19:05:59 *** HaraKiri has quit IRC () 2011-11-24T19:08:13 *** treeform has joined #aichallenge 2011-11-24T19:10:13 *** fpcfan has joined #aichallenge 2011-11-24T19:14:34 *** amstan has joined #aichallenge 2011-11-24T19:14:34 *** ChanServ sets mode: +o amstan 2011-11-24T19:14:57 *** Antimony has joined #aichallenge 2011-11-24T19:17:21 *** amstan has quit IRC (Remote host closed the connection) 2011-11-24T19:17:56 *** amstan has joined #aichallenge 2011-11-24T19:17:56 *** ChanServ sets mode: +o amstan 2011-11-24T19:19:45 *** ikaros has quit IRC (Remote host closed the connection) 2011-11-24T19:20:12 *** McLeopold_ has quit IRC (Ping timeout: 265 seconds) 2011-11-24T19:20:29 of course that kubuntu managed to fuck up a ton of things, why am i surprised 2011-11-24T19:21:25 *** Jak_o_Shadows1 has quit IRC (Read error: Connection reset by peer) 2011-11-24T19:24:25 *** Jak_o_Shadows has joined #aichallenge 2011-11-24T19:25:09 *** pairofdice has quit IRC (Quit: leaving) 2011-11-24T19:26:48 what's going on? 2011-11-24T19:27:20 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Read error: Connection reset by peer) 2011-11-24T19:28:42 *** amstan has quit IRC (Remote host closed the connection) 2011-11-24T19:29:06 *** amstan has joined #aichallenge 2011-11-24T19:29:06 *** ChanServ sets mode: +o amstan 2011-11-24T19:31:31 ? 2011-11-24T19:37:58 *** amstan has quit IRC (Remote host closed the connection) 2011-11-24T19:39:41 *** amstan has joined #aichallenge 2011-11-24T19:39:41 *** ChanServ sets mode: +o amstan 2011-11-24T19:42:55 *** amstan has quit IRC (Remote host closed the connection) 2011-11-24T19:48:43 *** clapautius has quit IRC (Ping timeout: 265 seconds) 2011-11-24T19:49:00 *** fpcfan has quit IRC (Ping timeout: 260 seconds) 2011-11-24T19:50:51 *** amstan has joined #aichallenge 2011-11-24T19:50:51 *** ChanServ sets mode: +o amstan 2011-11-24T19:59:18 *** fpcfan has joined #aichallenge 2011-11-24T20:03:39 *** Antimony has quit IRC (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111104165243]) 2011-11-24T20:03:53 *** Antimony has joined #aichallenge 2011-11-24T20:04:25 janzert: shouldn't i use disable account for this? http://forums.aichallenge.org/viewtopic.php?f=21&t=1948 2011-11-24T20:04:29 the account is still there 2011-11-24T20:04:57 *** fpcfan has quit IRC (Ping timeout: 244 seconds) 2011-11-24T20:05:18 *** Thinq has quit IRC (Ping timeout: 276 seconds) 2011-11-24T20:05:37 *** Thinq has joined #aichallenge 2011-11-24T20:16:06 *** fpcfan has joined #aichallenge 2011-11-24T20:16:25 is something weird going on with the channel? 2011-11-24T20:16:33 Antimony: no? 2011-11-24T20:17:03 oh 2011-11-24T20:17:10 for a while people kept randomly joining and leaving 2011-11-24T20:17:27 everything is fine :-) *waits 5 sec before the world calapses* 2011-11-24T20:17:29 well.. that's the nature of irc, kinda random 2011-11-24T20:17:58 can someone ping me? 2011-11-24T20:18:03 to test something 2011-11-24T20:18:18 I've never seen it happen before 2011-11-24T20:18:26 it was all green and yellow, no actual chatting 2011-11-24T20:18:33 ok.. or not 2011-11-24T20:18:35 @botdoc 2011-11-24T20:18:36 amstan: botdoc could be (#1) http://ubottu.com/stdin/supydocs/plugins/, or (#2) http://supybook.fealdia.org/devel/, or (#3) http://sourceforge.net/apps/mediawiki/gribble/index.php?title=Supybot_Resources. 2011-11-24T20:18:40 * avdg trusts someone is already pinging 2011-11-24T20:19:04 and now the moment of truth. Does my shiny new bot design actually not suck too much? 2011-11-24T20:19:30 hmm visible design, or behavior design? :p 2011-11-24T20:20:14 what would visible design even be? 2011-11-24T20:20:15 * avdg needs link as well 2011-11-24T20:20:24 triangles? 2011-11-24T20:20:54 *** fpcfan has quit IRC (Ping timeout: 245 seconds) 2011-11-24T20:21:36 well my bot will now occsionally do smart combat 2011-11-24T20:21:41 but it doesn't apepar to be enough to make a differenc 2011-11-24T20:22:16 It took 4th out of 7 agaisnt previous versions of my bot 2011-11-24T20:22:22 :( 2011-11-24T20:22:23 well, mine doesn't even know the definition of attacking 2011-11-24T20:22:35 and I'm rebuilding my bot from js to c++ 2011-11-24T20:22:53 I would probably mis the deadline 2011-11-24T20:22:56 *** jstrong has joined #aichallenge 2011-11-24T20:23:31 *** FrIkI has joined #aichallenge 2011-11-24T20:23:59 *** FrIkI is now known as FrIkI_aichalleng 2011-11-24T20:25:03 Hi, i just joint to the challenge. I can't find the date limit to write the final bot code :( 2011-11-24T20:25:30 hmm, isn't it mentioned on the homepage? 2011-11-24T20:25:42 at the bottom of the page 2011-11-24T20:25:52 December 18th at 11:59pm EST 2011-11-24T20:27:42 *** Antimony has quit IRC (Ping timeout: 240 seconds) 2011-11-24T20:29:29 *** Antimony has joined #aichallenge 2011-11-24T20:32:11 *** fpcfan has joined #aichallenge 2011-11-24T20:32:13 sorry , you are right 2011-11-24T20:32:34 thanks avdg , i promise i look for it a while! :-P 2011-11-24T20:36:09 *** FrIkI_aichalleng has quit IRC (Quit: Page closed) 2011-11-24T20:41:50 *** replore has joined #aichallenge 2011-11-24T20:44:23 meh, I'm now learning more about registry types than programming an ai 2011-11-24T20:57:35 and I'm busy doing webpages 2011-11-24T20:57:35 shame 2011-11-24T20:57:41 cause I really wanted to get my final bot working well with combat 2011-11-24T20:57:44 *** jstrong is now known as roflmao 2011-11-24T21:00:54 *** Conorach has quit IRC (Ping timeout: 245 seconds) 2011-11-24T21:12:30 *** b0rder has joined #aichallenge 2011-11-24T21:26:19 *** fpcfan has quit IRC (Ping timeout: 245 seconds) 2011-11-24T21:30:02 *** AntDroid has joined #aichallenge 2011-11-24T21:30:27 *** fpcfan has joined #aichallenge 2011-11-24T21:31:58 extrarius 2011-11-24T21:32:14 *** snthaoeu has joined #aichallenge 2011-11-24T21:32:50 *** Antimony has quit IRC (Ping timeout: 260 seconds) 2011-11-24T21:35:30 *** LouisMartin has quit IRC (Quit: Page closed) 2011-11-24T21:39:58 *** Antimony has joined #aichallenge 2011-11-24T22:07:30 *** Jak_o_Shadows has quit IRC (Remote host closed the connection) 2011-11-24T22:07:45 *** thestinger has joined #aichallenge 2011-11-24T22:20:06 *** snthaoeu has quit IRC (Quit: Page closed) 2011-11-24T22:26:06 *** avdg has quit IRC (Quit: Leaving.) 2011-11-24T22:27:39 *** avdg has joined #aichallenge 2011-11-24T22:29:03 *** roflmao has quit IRC (Quit: Leaving.) 2011-11-24T22:37:37 check turn 800+ http://aichallenge.org/visualizer.php?game=125185 2011-11-24T22:38:51 omg! canadiancow! 2011-11-24T22:39:54 *** Antimony_ has joined #aichallenge 2011-11-24T22:39:55 *** fpcfan has quit IRC (Ping timeout: 260 seconds) 2011-11-24T22:40:52 *** Antimony has quit IRC (Ping timeout: 258 seconds) 2011-11-24T22:41:05 *** Antimony_ is now known as Antimony 2011-11-24T22:43:34 *** fpcfan has joined #aichallenge 2011-11-24T22:46:27 *** Jak_o_Shadows has joined #aichallenge 2011-11-24T22:47:04 *** AntDroid has quit IRC (Ping timeout: 265 seconds) 2011-11-24T22:56:56 interesting 2011-11-24T22:57:06 my bot assigned 0 guards wen it was 2 ants 2011-11-24T22:57:11 but 2 guars when it has 3 ants 2011-11-24T23:12:40 @later tell amstan yeah disabling the account should be good there. 2011-11-24T23:12:40 janzert: OK 2011-11-24T23:28:40 *** jstrong has joined #aichallenge 2011-11-24T23:28:49 *** jstrong is now known as roflmao 2011-11-24T23:36:30 *** fpcfan has quit IRC (Ping timeout: 260 seconds) 2011-11-24T23:45:34 *** fpcfan has joined #aichallenge 2011-11-24T23:46:49 *** TheLinker has quit IRC (Quit: Bye) 2011-11-24T23:52:15 *** delt0r_ has quit IRC (Ping timeout: 260 seconds) 2011-11-24T23:58:58 *** ltriant has quit IRC (Read error: Connection reset by peer)