2011-04-05T00:00:44 afaik we're pretty close to getting done 2011-04-05T00:02:17 except for the part where some pretty fundamental parts of the game are not specified 2011-04-05T00:03:01 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-04-05T00:06:10 what are we missing from being done 2011-04-05T00:06:20 mcleopold had a checklist 2011-04-05T00:06:23 is it just an easily locally runnable visualiser? 2011-04-05T00:12:20 *** suicidolt has quit IRC (Ping timeout: 252 seconds) 2011-04-05T00:28:08 *** ltriant has joined #aichallenge 2011-04-05T00:32:56 *** suicidolt has joined #aichallenge 2011-04-05T00:50:15 *** Mathnerd314 has quit IRC (Ping timeout: 246 seconds) 2011-04-05T00:55:54 huzzah gsoc application is in (excepting revisions) 2011-04-05T01:20:27 *** ltriant has quit IRC (Ping timeout: 240 seconds) 2011-04-05T01:20:48 *** sigh has quit IRC (Ping timeout: 240 seconds) 2011-04-05T01:21:07 *** ltriant has joined #aichallenge 2011-04-05T01:21:50 *** sigh has joined #aichallenge 2011-04-05T01:40:15 antimatroid: do you work on gsoc remotely? 2011-04-05T01:46:09 *** Kingpin13 has joined #aichallenge 2011-04-05T01:59:29 *** AlliedEnvy_ has quit IRC (Ping timeout: 246 seconds) 2011-04-05T02:03:47 sigh: yeah 2011-04-05T02:10:42 *** AlliedEnvy has joined #aichallenge 2011-04-05T02:48:08 *** kaemo has quit IRC (Ping timeout: 250 seconds) 2011-04-05T02:51:31 *** andy_ has joined #aichallenge 2011-04-05T02:58:25 *** amstan has quit IRC (Remote host closed the connection) 2011-04-05T03:01:10 *** kaemo has joined #aichallenge 2011-04-05T03:09:36 *** andy_ has quit IRC (Quit: Page closed) 2011-04-05T03:24:16 *** ltriant has quit IRC (Read error: Connection reset by peer) 2011-04-05T03:24:45 *** ltriant has joined #aichallenge 2011-04-05T03:30:26 *** aerique has joined #aichallenge 2011-04-05T03:34:30 *** ltriant has quit IRC (Ping timeout: 250 seconds) 2011-04-05T03:40:38 *** sigh has quit IRC (Ping timeout: 252 seconds) 2011-04-05T04:01:23 *** delt0r___ has joined #aichallenge 2011-04-05T04:02:47 *** delt0r_ has quit IRC (Ping timeout: 246 seconds) 2011-04-05T04:30:37 aichallenge: Nick Ham epsilon * r66c8712 / (3 files): fixed vision for the c++ starter package - http://bit.ly/gxu3Ha 2011-04-05T04:32:15 there we go 2011-04-05T04:32:20 i was getting worried, that took like 2 minutes 2011-04-05T04:37:54 *** Stocha has joined #aichallenge 2011-04-05T04:38:18 *** kaemo has quit IRC (Read error: Connection reset by peer) 2011-04-05T04:39:59 *** sigh has joined #aichallenge 2011-04-05T04:44:26 *** smellyhippy has quit IRC (*.net *.split) 2011-04-05T04:44:29 *** drhodes has quit IRC (*.net *.split) 2011-04-05T04:44:30 *** slkjcna_ has quit IRC (*.net *.split) 2011-04-05T04:44:30 *** davidd has quit IRC (*.net *.split) 2011-04-05T04:50:56 *** contestbot_ has joined #aichallenge 2011-04-05T04:50:56 *** contestbot has quit IRC (Disconnected by services) 2011-04-05T04:51:32 *** smellyhippy has joined #aichallenge 2011-04-05T04:51:32 *** drhodes has joined #aichallenge 2011-04-05T04:51:32 *** slkjcna_ has joined #aichallenge 2011-04-05T04:51:32 *** davidd has joined #aichallenge 2011-04-05T04:51:38 *** znutar has quit IRC (Excess Flood) 2011-04-05T04:52:37 *** antimatroid1 has joined #aichallenge 2011-04-05T04:52:37 *** contestbot_ is now known as contestbot 2011-04-05T04:52:38 *** ChanServ sets mode: +o contestbot 2011-04-05T04:52:59 lol, that's two times i've crashed ubuntu since installing 2011-04-05T04:53:35 Has anyone been able to build tests that support the claim that "number of ant at the end" will favor not expanding territory ? 2011-04-05T05:03:57 *** kwketh has quit IRC (*.net *.split) 2011-04-05T05:03:57 *** antimatroid has quit IRC (*.net *.split) 2011-04-05T05:03:58 *** Frontier has quit IRC (*.net *.split) 2011-04-05T05:05:23 @later tell keith i have a basic bot. You can see it there (against a lesser version of himself i think http://ns368394.ovh.net/stocha/x80Fh02_2vsFh02_r6x1) 2011-04-05T05:05:23 Stocha: Ready to serve. 2011-04-05T05:05:53 @later tell keith it has all basic features people seems to think of. Except fighting :p 2011-04-05T05:05:53 Stocha: Yep. 2011-04-05T05:06:19 @later tell keith you can message me here, and explain me what protocol you want to use in order to generation the comparatives games. 2011-04-05T05:06:19 Stocha: Job's done. 2011-04-05T05:07:02 *** znutar_ has joined #aichallenge 2011-04-05T05:07:02 *** kwketh has joined #aichallenge 2011-04-05T05:07:02 *** Frontier has joined #aichallenge 2011-04-05T05:07:11 *** znutar_ has quit IRC (Excess Flood) 2011-04-05T05:07:18 *** znutar has joined #aichallenge 2011-04-05T05:07:57 mm wrong nickname that was :'( 2011-04-05T05:08:08 @later tell keth i have a basic bot. You can see it there (against a lesser version of himself i think http://ns368394.ovh.net/stocha/x80Fh02_2vsFh02_r6x1) 2011-04-05T05:08:08 Stocha: OK 2011-04-05T05:08:21 @later tell keth you can message me here, and explain me what protocol you want to use in order to generation the comparatives games. 2011-04-05T05:08:21 Stocha: As you wish. 2011-04-05T05:14:40 *** Eruonen has joined #aichallenge 2011-04-05T05:18:44 @later tell keth i'm quite doubtfull about more than 1vs1 myself. But i disagree that it shouldn't be discussed :p Where on the forum did you see anything that supports the claim that more than 1vs1 will not work ? 2011-04-05T05:18:44 Stocha: Yep. 2011-04-05T05:19:48 Stocha: is your bot actually conforming to the specified io formats? 2011-04-05T05:20:02 keth is using that 2011-04-05T05:20:25 kwketh* 2011-04-05T05:21:00 *** antimatroid1 is now known as antimatroid 2011-04-05T05:26:18 i have an adaptor that i made from the beginner bot of McLeopold 2011-04-05T05:26:40 maybe it is old enougth not to conform anymore though. But that won't be hard to correct. 2011-04-05T05:27:06 but maybe "protocol" wasn't clear enough :p 2011-04-05T05:27:25 i meant by it : how do we exchange the exec, who runs them ? is javabot okay for you ? :p 2011-04-05T05:27:57 @later tell keth by "protocole" i mean, where do i send the exec (or do you), who runs the experiments ? how do we provide the result to the other ? 2011-04-05T05:27:57 Stocha: Job's done. 2011-04-05T05:30:53 maybe i should just provide a link to the exec so people can use it to test their bot. It's pretty simple anyway. 2011-04-05T05:31:12 i mean that isn't a bot that will win the competition. It is more an advanced starter bot :p 2011-04-05T05:32:52 Stocha: he is using c++, so if you can provide him with a runnable compiled version that'd be best 2011-04-05T05:32:57 that way neither of you see each others source 2011-04-05T05:33:00 @later tell keth okay here is what i propose : i will try to put the bot on a http link tonight (greenwich meridian) Then you can play with it, and tell me :p 2011-04-05T05:33:00 Stocha: Job's done. 2011-04-05T05:33:08 i'm using java. 2011-04-05T05:33:31 yeah, so you should be able to give a jar version of it that doesn't have the source viewable? 2011-04-05T05:33:34 i can provide a McLeopold like java bot. that's it. 2011-04-05T05:33:46 java code is always viewable. 2011-04-05T05:34:05 hmmm, okay, we need tcp :P 2011-04-05T05:34:08 you can try to obscate it. but i wouldn't count too much on that. And i don't want to put time into it :p 2011-04-05T05:34:20 obfuscate it 2011-04-05T05:34:44 you have tools for obfuscating java, but i don't want to spend the 10 minutes or so i would need to do that. 2011-04-05T05:34:53 yeah neither would i 2011-04-05T05:35:24 so if people want to read it, i just want to state that i dissallow that formally ! 2011-04-05T05:35:43 yeah. really. i may take proper lawsuit or i don't know what. 2011-04-05T05:35:52 yeah. burn them hackers. 2011-04-05T05:36:07 it's not like it is a friendly challenge or anything :p 2011-04-05T05:37:54 I've never been able to figure someone else's code myself anyway. 2011-04-05T05:45:35 Stocha: yeah, i look at others code and go ergh 2011-04-05T05:46:44 sigh: is that neighbourhood offsets stuff your fastest vision code? 2011-04-05T05:47:05 fastest vision code is in fast-vision branch 2011-04-05T05:47:23 note that that is fastest for *python* 2011-04-05T05:48:24 I wouldn't do it the same way in (for example) C++ if I was going for speed 2011-04-05T05:49:33 how would you do it in c++ for speed? 2011-04-05T05:51:32 I would profile the painting algorithm vs a modified version of the current vision algorithm where I cached only the bounds of the vision radius 2011-04-05T05:52:13 my bet would be that painting is still slower 2011-04-05T05:54:06 also note that the incremental method more difficult to implement in the bots than the engine 2011-04-05T05:54:22 so that can slow things down too 2011-04-05T05:54:39 ok, I'm going to merge fast-vision in 2011-04-05T05:54:44 i might just store all offsets for vision in my start up time then check for all ants using that 2011-04-05T05:56:27 in a language like c++ you are better of storing just (start_col_offset, end_col_offset) for each row_offset 2011-04-05T05:57:29 aichallenge: sigh epsilon * rae2c7f3 / ants/ants.py : Merge branch 'fast-vision' into epsilon - http://bit.ly/gXxHxL 2011-04-05T05:58:37 ok, now the engine has the fastest vision that I've come up with so far :) 2011-04-05T05:59:59 antimatroid: any particular reason that you are asking? 2011-04-05T06:00:21 sigh: i've finished the starter bot, i've starter setting my own "starter bot" up again 2011-04-05T06:00:29 ahh 2011-04-05T06:00:45 particularly if i get gsoc, then i will run out of time pretty early 2011-04-05T06:01:07 fair enough 2011-04-05T06:08:14 @seen keth 2011-04-05T06:08:14 Stocha: keth was last seen in #aichallenge 12 hours, 14 minutes, and 54 seconds ago: well, never mind. not really a priority 2011-04-05T06:09:27 kwkety!=keth ? 2011-04-05T06:09:34 kwketh!=keth ? 2011-04-05T06:13:21 kwketh hi 2011-04-05T06:14:39 @later tell McLeopold bug in starter_bot a dead ant square could have food spawned on it that turn, and then the food on the map will be overwritten 2011-04-05T06:14:39 sigh: Aye, aye, sir 2011-04-05T06:30:08 here is a link to what i think is a good standard bot for testing purposes 2011-04-05T06:30:09 http://ns368394.ovh.net/stocha/StandardBotFh2_2/ 2011-04-05T06:30:30 you have no right of looking at the code reusing the code, etc etc. Until the challenge is over. 2011-04-05T06:30:36 i'll post on the forums also 2011-04-05T06:31:20 @later tell kwketh here is what i want to call a standard bot :http://ns368394.ovh.net/stocha/StandardBotFh2_2/ 2011-04-05T06:31:20 Stocha: Aye, aye, sir 2011-04-05T06:32:29 *** Naktibalda has joined #aichallenge 2011-04-05T06:37:14 *** antimatroid has quit IRC (Ping timeout: 258 seconds) 2011-04-05T06:37:22 @seen kwketh 2011-04-05T06:37:22 Stocha: I have not seen kwketh. 2011-04-05T06:39:07 @seen Keth 2011-04-05T06:39:07 Frontier: Keth was last seen in #aichallenge 12 hours, 45 minutes, and 47 seconds ago: well, never mind. not really a priority 2011-04-05T06:40:29 i'm a little confused with how to deliver a message :p 2011-04-05T06:40:42 well i guess i made my part so be it for today. 2011-04-05T06:41:43 sigh: ping 2011-04-05T06:41:45 the bot wins in 164 turns against 3 hunter bots, on the first provided map. 2011-04-05T06:41:57 Frontier: pong 2011-04-05T06:42:39 I have converted the replay in github to json, take a look at it. (games/0/2.replay) 2011-04-05T06:43:45 I see no games directory 2011-04-05T06:43:50 I haven't written the spec page yet. 2011-04-05T06:44:00 ok, i found it 2011-04-05T06:44:40 the original was 1.replay. I used mostly regex replace :p 2011-04-05T06:47:11 :P 2011-04-05T06:47:47 for the record: it is 9% larger than the original 2011-04-05T06:50:52 *** antimatroid has joined #aichallenge 2011-04-05T06:52:30 Frontier: that's fine... anyway if you are worried about space the compact repr of json is about the same size 2011-04-05T06:52:49 *** p4p4p5 has joined #aichallenge 2011-04-05T06:53:38 no I'm not worried. It is not like the streaming format which is 900% bigger 2011-04-05T06:55:09 what was this food bonus you wanted to add as an additional score value? 2011-04-05T06:59:06 get antimatroid or someone to explain food bonus :) 2011-04-05T06:59:39 Frontier: the food bonus is meant to encourage the game to end early 2011-04-05T06:59:55 and help ensure the last surviving bot wins 2011-04-05T07:00:21 how does it work? 2011-04-05T07:01:36 if at the end of the game there is only one "participating player" left, they get the food bonus, which is currently turnsRemaining*spawnRate 2011-04-05T07:01:46 but that may be + currentFoodSquares 2011-04-05T07:03:52 ah this is like in battle for wesnoth, where you get the production of all towns * remaining turns as a bonus 2011-04-05T07:08:06 sigh: do you write scores for bots that already died? 2011-04-05T07:08:24 Frontier: yes 2011-04-05T07:09:43 You don't need to. I'd like to use the length of the score array to tell when a bot died/crashed/etc and I replicate the last score to the end already. 2011-04-05T07:11:04 see 's' in the replay specs 2011-04-05T07:12:51 This is a strange situation. I tell people to look at the official specs that I just made up some afternoon myself. ^^ 2011-04-05T07:22:46 *** BtbN_ is now known as BtbN 2011-04-05T07:24:55 why is it strange ? 2011-04-05T07:25:32 @seen phire 2011-04-05T07:25:32 Stocha: phire was last seen in #aichallenge 16 hours, 33 minutes, and 25 seconds ago: It spends 60% of the game just sitting there doing nothing, and it still practically draws with hunter bot. 2011-04-05T07:27:56 *** davidd has quit IRC (Read error: Operation timed out) 2011-04-05T07:30:12 *** davidd has joined #aichallenge 2011-04-05T07:40:45 Stocha: power corrupts a person. 2011-04-05T07:46:10 ... anyone got any idea on time limits we should use? 2011-04-05T07:46:26 i would like at least 500 milliseconds guaranteed per bot 2011-04-05T07:46:31 1000 would be better 2011-04-05T07:48:44 is that on a 64bits processor ? 2011-04-05T07:48:55 with a 64 bits jvm :p 2011-04-05T07:50:03 jvm? 2011-04-05T07:57:24 must be an insider about slow 64-bit jvms from sun 2011-04-05T07:58:05 http://asserttrue.blogspot.com/2008/11/why-64-bit-java-is-slow.html 2011-04-05T08:03:03 A problem I see with most programming languages is that they settled for 32-bit ints. There should really be an obvious short keyword for architecture dependent ints. 'machine_word' is a bad option for example: "for (machine_word i = 0; i < 12; i++) { ... }" 2011-04-05T08:09:26 You can't do that with a compile once run everywhere language though :p 2011-04-05T08:10:42 i use long shift. So i guess it's faster with 64 bits jvm. i would like it too at least. Anyway, i have no 32bits systems to test on. 2011-04-05T08:13:12 *** boegel has joined #aichallenge 2011-04-05T08:18:31 i have no 64-bit system 2011-04-05T08:35:07 I think I have an issue blocking the release date: Loading the huge replay antimatroid gave me takes 97 seconds and uses 2 GB of RAM in Opera. 2011-04-05T08:35:39 I know you all have 4 GB of RAM, but still it could be less :p 2011-04-05T08:36:01 what kind of huge replay is that ? 2011-04-05T08:36:08 what game settings. 2011-04-05T08:41:25 Map: 120x120, Turns: 500, Viewradius²: 96 2011-04-05T08:43:12 is that a real time spooling ? 2011-04-05T08:44:06 or standard offline text format ? 2011-04-05T08:44:10 how many players ? 2011-04-05T08:44:17 lots of food ? 2011-04-05T08:44:53 oh it's not the input file that is really bit. it's your internal representation i guess :p 2011-04-05T08:44:54 That is the offline text format (storage format) with 4 players and quite some food: http://marco-leise.homedns.org/antgui/work/visualizer.php?game_id=3 2011-04-05T08:45:32 i wouldn't want to open an ant window here :p 2011-04-05T08:45:35 Stocha: yes, noone seems to care that I want to bring animated ant pictures back :( 2011-04-05T08:45:49 I understand ^^ 2011-04-05T08:47:35 *** p4p4p5 has quit IRC (Remote host closed the connection) 2011-04-05T08:47:59 how big is the input file ? 2011-04-05T08:48:26 ∞ 2011-04-05T08:50:11 how bit is the input file in the usecase frontier is talking of :p 2011-04-05T08:50:14 big 2011-04-05T08:50:43 376 KB 2011-04-05T08:51:17 +9% (roughly) when stored in 100% json 2011-04-05T08:51:39 376 kb --> 2GB is a big ratio i guess :p 2011-04-05T08:52:58 would it be hard to use an internal serverstate frame representation and only transform it to picture when you need to ? 2011-04-05T08:53:02 on the fly. 2011-04-05T08:58:07 *** FireFly has joined #aichallenge 2011-04-05T09:01:26 Stocha: That is what I use. It is Opera's garbage collector implementation I assume, that wont run a sweep while code is executing. 2011-04-05T09:01:48 So all the little temporary variables, arrays and objects keep adding up. 2011-04-05T09:02:03 After the parsing process the memory is freed again. 2011-04-05T09:09:58 *** fgump has joined #aichallenge 2011-04-05T09:12:19 Stocha: The pictures are alway rendered from an internal representation of the objects on screen. Only the map and other static stuff is prerendered. The fog, score/ant count bars are updated once a turn. 2011-04-05T09:18:29 the solution to making it fast again is to only verify the replay on load and turn it into the internal representation on demand. 2011-04-05T09:19:15 oh and I see that I don't seem to cache replays. either that or the browsers refuse to cache files of that size 2011-04-05T09:19:56 I saw that there preg,next buttons are missing in the latest version 2011-04-05T09:20:11 is it intentionally? 2011-04-05T09:20:21 *** Guest88630 has joined #aichallenge 2011-04-05T09:21:15 I like to watch and analyze games in step by step mode, not in animation mode 2011-04-05T09:21:57 Naktibalda: only those two buttons are missing? I can hardly imagine that. 2011-04-05T09:22:26 *** sigh has quit IRC (Remote host closed the connection) 2011-04-05T09:22:30 They should all be there still. 2011-04-05T09:24:21 README is missing either :) 2011-04-05T09:24:26 could you update https://github.com/aichallenge/aichallenge/blob/epsilon/ants/viewer/README.md 2011-04-05T09:27:29 *** Guest88630 has quit IRC (Read error: Connection reset by peer) 2011-04-05T09:41:55 *** fgump has quit IRC (Quit: Leaving) 2011-04-05T09:52:08 *** kwketh has quit IRC (Read error: Connection reset by peer) 2011-04-05T09:53:59 *** Keth has joined #aichallenge 2011-04-05T09:54:14 hi keth 2011-04-05T09:54:17 are you here ? 2011-04-05T09:54:50 i posted a bot for you both here and in the "Tcp server" thread on the forum 2011-04-05T09:55:01 keth : 2011-04-05T09:56:08 *** needsch has joined #aichallenge 2011-04-05T10:01:23 *** needsch1 has joined #aichallenge 2011-04-05T10:01:35 *** aerique has quit IRC (Quit: ...) 2011-04-05T10:01:52 *** needsch has quit IRC (Ping timeout: 240 seconds) 2011-04-05T10:06:16 Naktibalda: I don't have any affiliation with the stuff in viewer 2011-04-05T10:08:52 *** kwketh has joined #aichallenge 2011-04-05T10:08:52 *** Keth has quit IRC (Read error: Connection reset by peer) 2011-04-05T10:35:27 *** Kingpin13 has quit IRC (Quit: You're a kitty!) 2011-04-05T10:48:54 *** phire has joined #aichallenge 2011-04-05T10:49:15 hi phire what's up ? 2011-04-05T10:49:33 good question 2011-04-05T10:49:55 i have made my bot available for those who want to do local testing against an average bot :p 2011-04-05T10:50:18 the old one you saw on my demos. 2011-04-05T10:51:15 it would totally own my bot in its current state 2011-04-05T10:51:27 probably But it gives a mean to test bots. 2011-04-05T10:51:47 even if you have only 20% win, that's a good idea of bot performances. 2011-04-05T10:52:02 and people will be able to try multiplayer games and see what they think of it :p 2011-04-05T10:52:04 but I can probally fix that in a few 30 min 2011-04-05T10:52:08 hopefully they will hate them :p 2011-04-05T10:52:16 ah good :) 2011-04-05T10:52:19 if I actually make it target enemies, it will do a lot better 2011-04-05T10:52:28 so you have built a strong bot, but it's broken ? 2011-04-05T10:54:12 its easier to test the scheduling system when its doing less 2011-04-05T10:54:43 ah you have built a scheduling system :) 2011-04-05T10:54:51 well the start of one 2011-04-05T10:54:53 i'm still to do that. 2011-04-05T10:55:09 i have parts of scheduling. That probably can't fit together anyway. 2011-04-05T10:55:40 I've set up my mind to do style exercise at the moment. So i'm trying to build a bot that loose with style :p 2011-04-05T10:55:51 at least i do make some progress that way. 2011-04-05T10:58:14 hmm 2011-04-05T10:58:19 its not working 2011-04-05T10:59:23 OSError: [Errno 2] No such file or directory 2011-04-05T10:59:48 mgn ? 2011-04-05T11:00:05 can you paste the link you used ? 2011-04-05T11:00:17 http://ns368394.ovh.net/stocha/StandardBotFh2_2/ 2011-04-05T11:00:34 you are the guy on ubuntu ? 2011-04-05T11:00:46 *** mceier has joined #aichallenge 2011-04-05T11:00:53 it does work for me :'( 2011-04-05T11:00:55 well, yes 2011-04-05T11:01:03 but I haven't posted on the forum 2011-04-05T11:05:02 so it does work now ? :p 2011-04-05T11:15:15 once you have stabilized a strategy, post some results (ideally after a few 100 or 1000 games :p ) 2011-04-05T11:15:42 I would love to see results for multiplayers also. Like 3 vs 1 2011-04-05T11:15:59 both 3 strong vs 1 lesser. And 3 lesser vs 1 strong. 2011-04-05T11:16:17 I never tested the provided map generators also yet. 2011-04-05T11:17:23 ok, lets build my scheduler 2011-04-05T11:19:22 then it might actually start winning 2011-04-05T11:20:03 *** Daniel350 has joined #aichallenge 2011-04-05T11:22:07 and I must work out how to get gprof running 2011-04-05T11:22:14 I'm seeing some large delays 2011-04-05T11:23:39 *** Stocha has quit IRC (Quit: Page closed) 2011-04-05T11:25:57 *** phire has quit IRC (Quit: Leaving) 2011-04-05T11:26:50 *** boegel has quit IRC (Ping timeout: 260 seconds) 2011-04-05T11:37:42 *** Daniel350 has quit IRC (Quit: Ex-Chat) 2011-04-05T12:01:54 *** delt0r_ has joined #aichallenge 2011-04-05T12:03:24 *** delt0r___ has quit IRC (Ping timeout: 246 seconds) 2011-04-05T12:17:11 *** _aj_ <_aj_!~aj@66.133.136.43> has quit IRC (Ping timeout: 250 seconds) 2011-04-05T12:17:51 *** _aj <_aj!~aj@66.133.136.43> has joined #aichallenge 2011-04-05T12:40:16 *** amstan has joined #aichallenge 2011-04-05T12:40:26 *** contestbot_ has joined #aichallenge 2011-04-05T12:40:26 *** contestbot has quit IRC (Disconnected by services) 2011-04-05T12:42:21 *** contestbot_ is now known as contestbot 2011-04-05T12:42:22 *** ChanServ sets mode: +o contestbot 2011-04-05T12:49:00 *** _jmcarthur <_jmcarthur!~jake@76.73.216.201> has joined #aichallenge 2011-04-05T12:56:57 *** jmcarthur has quit IRC (*.net *.split) 2011-04-05T13:19:23 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 3.6.16/20110319135224]) 2011-04-05T13:20:53 *** fgump has joined #aichallenge 2011-04-05T13:29:26 *** amstan has quit IRC (Remote host closed the connection) 2011-04-05T13:30:28 *** Eruonen has quit IRC (Ping timeout: 276 seconds) 2011-04-05T13:36:41 ergh, Stocha posted his code in the forums anyway :\ 2011-04-05T14:06:40 *** Eruonen has joined #aichallenge 2011-04-05T14:09:35 *** Stocha has joined #aichallenge 2011-04-05T14:10:35 antimatroid: what about it ? 2011-04-05T14:11:52 antimatroid: i think it is good for people to have something to aim for. The medium bots provides quite a challenge. 2011-04-05T14:12:08 especially if you want 100% victory on asymetric maps :^ 2011-04-05T14:14:51 Stocha: as i said earlier, if no code is shared people know how they fair against the other people who decide to compete 2011-04-05T14:15:17 if you share your code, then people can only know how they do against people who decide to also use your bot or adapt it 2011-04-05T14:15:56 a number of users expressed a preference to that for planet wars 2011-04-05T14:16:10 while i don't really mind, i see their point 2011-04-05T14:18:20 *** Cyndre_ has quit IRC (Read error: Connection reset by peer) 2011-04-05T14:18:27 *** mceier has quit IRC (Quit: leaving) 2011-04-05T14:19:06 *** Cyndre has joined #aichallenge 2011-04-05T14:27:25 the code is not shared 2011-04-05T14:27:33 it's only easy to reverse engineer 2011-04-05T14:27:42 only someone malicious would do that 2011-04-05T14:27:55 it's hundreds of times easier than it is with c 2011-04-05T14:28:08 but if you let debut info in c code, it's more than doable. 2011-04-05T14:28:18 i modified sim city rules, when i was young :p 2011-04-05T14:28:25 by reverse engineer to you mean inspecting the contents of the jar file or actually decompiling it? 2011-04-05T14:28:30 if the later, then that's alright :) 2011-04-05T14:28:33 decompiling it. 2011-04-05T14:28:43 but all debug info are in by default. 2011-04-05T14:28:44 ignore my above comments :P 2011-04-05T14:28:54 you have to do something drastic to somewhat hide them 2011-04-05T14:29:17 i can strip all my debug stuff when compiling by commenting 1 line 2011-04-05T14:29:41 and i just "bug << "this is my error stuff" << state << endl" to debug 2011-04-05T14:31:30 i once decompiled a full java application (to test risks of being hacked). The decompiled version was actuallly more readable than the source java code. 2011-04-05T14:32:06 i never tryed again, but i Know that without obfuscation giving the jar, is like giving the code. Without the comments. 2011-04-05T14:32:43 You have full access to class name, method names, propriety names, etc. 2011-04-05T14:33:25 Still, you have to get your hand on a decompilation software. 2011-04-05T14:33:48 (which you can have for free but well). And then, it is illegal in a lot of place to do that :p 2011-04-05T14:35:29 Well by debut info i meant the functions names :) 2011-04-05T14:35:32 and signature. 2011-04-05T14:36:26 anyway nobody reads the forums :p 2011-04-05T14:37:06 i doubt they will try to read my code. good for them if they try (and they'll go to jail) 2011-04-05T14:37:23 Further more, the bot is really basic. 2011-04-05T14:42:31 all the ideas implemented in it, if have seen some people discussing already :) 2011-04-05T14:49:38 *** Stocha has quit IRC (Quit: Page closed) 2011-04-05T15:01:03 *** kwketh has quit IRC () 2011-04-05T15:01:22 *** Keth has joined #aichallenge 2011-04-05T15:17:23 *** Accoun has quit IRC () 2011-04-05T15:20:52 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-04-05T15:28:21 *** boegel has joined #aichallenge 2011-04-05T15:30:40 *** Accoun has joined #aichallenge 2011-04-05T15:40:51 Anybody else have difficulty with Stocha bot? Every replay with Stocha bot gets this error: Error stack: Number of orders does not match life span. 2011-04-05T15:42:16 might be problem with his bot, will need to ask once he's on 2011-04-05T15:42:23 sounds funny. it seems to exploit a security hole in the engine, causing it to emit invalid data 2011-04-05T15:42:55 true 2011-04-05T15:42:57 you had the same thing? 2011-04-05T15:43:21 I can get the output of ants e.g. [9, 8, 38, 40] at the end 2011-04-05T15:43:35 and I can see "END" 2011-04-05T15:43:40 but the replay seems corrupted 2011-04-05T15:44:43 I once reported a bug about this symptom (number of orders ...). sigh then fixed the engine. 2011-04-05T15:45:19 Later on I heard about a bot sending invalid orders, causing a bug in the engine. It may be related 2011-04-05T15:47:16 I see 2011-04-05T15:47:46 The nice thing about working on the visualizer is that most bugs are easily spotted ^^. But I have to fight with performance issues. 2011-04-05T15:49:58 I actually don't get how his bot works 2011-04-05T15:50:10 I'm not sure if you looked at it 2011-04-05T15:52:46 I might've just fixed it 2011-04-05T15:54:04 *** dr- has quit IRC (Ping timeout: 246 seconds) 2011-04-05T15:56:21 updating to newest svn once again fixed that 2011-04-05T15:59:37 *** dr- has joined #aichallenge 2011-04-05T16:14:33 *** boegel has quit IRC (Quit: Leaving) 2011-04-05T16:15:16 *** amstan has joined #aichallenge 2011-04-05T16:15:16 *** ChanServ sets mode: +o amstan 2011-04-05T16:25:55 *** Naktibalda has joined #aichallenge 2011-04-05T16:40:58 *** delt0r___ has joined #aichallenge 2011-04-05T16:42:18 *** delt0r_ has quit IRC (Ping timeout: 246 seconds) 2011-04-05T16:54:58 *** delt0r_ has joined #aichallenge 2011-04-05T16:56:24 *** delt0r___ has quit IRC (Ping timeout: 258 seconds) 2011-04-05T17:11:30 *** Accoun has quit IRC (Ping timeout: 276 seconds) 2011-04-05T17:14:43 *** Accoun has joined #aichallenge 2011-04-05T17:20:19 *** skynetz has joined #aichallenge 2011-04-05T17:41:30 *** sigh has joined #aichallenge 2011-04-05T17:48:24 *** _jmcarthur is now known as jmcarthur 2011-04-05T18:00:36 *** ltriant has joined #aichallenge 2011-04-05T18:06:15 *** sigh has quit IRC (Remote host closed the connection) 2011-04-05T18:28:17 *** FireFly has quit IRC (Quit: swatted to death) 2011-04-05T18:33:20 *** McLeopold1 has joined #aichallenge 2011-04-05T19:12:14 *** Ttech is now known as Avanone 2011-04-05T19:12:32 *** Avanone is now known as Ttech 2011-04-05T19:22:08 *** sigh has joined #aichallenge 2011-04-05T19:23:14 *** needsch1 has quit IRC (Ping timeout: 258 seconds) 2011-04-05T19:30:38 *** needsch has joined #aichallenge 2011-04-05T19:33:35 *** fgump has quit IRC (Quit: Leaving) 2011-04-05T19:34:30 *** Eruonen has quit IRC () 2011-04-05T20:00:32 *** Stocha has joined #aichallenge 2011-04-05T20:04:23 *** dlila has joined #aichallenge 2011-04-05T20:06:45 *** Chris_0076 has joined #aichallenge 2011-04-05T20:07:01 aichallenge: McLeopold epsilon * r5fb53fd / ants/mapgen/McMaps.py : finished make_symmetric function - http://bit.ly/gPfREY 2011-04-05T20:08:22 *** McLeopold1 has left #aichallenge 2011-04-05T20:09:24 *** Mathnerd314 has joined #aichallenge 2011-04-05T20:16:03 *** skynetz has quit IRC (Quit: .) 2011-04-05T20:23:05 *** Cyndre_ has joined #aichallenge 2011-04-05T20:26:26 *** Cyndre has quit IRC (Ping timeout: 264 seconds) 2011-04-05T20:28:45 *** Keth has quit IRC (Read error: Connection reset by peer) 2011-04-05T20:29:11 *** Keth has joined #aichallenge 2011-04-05T20:46:22 *** needsch has quit IRC (Quit: Leaving.) 2011-04-05T20:46:40 *** jbroman_ has joined #aichallenge 2011-04-05T20:46:42 *** jbroman