2011-06-26T00:10:10 *** okay is now known as okayzed 2011-06-26T00:11:13 *** stocha has joined #aichallenge 2011-06-26T00:16:17 *** stocha has quit IRC (Quit: Page closed) 2011-06-26T01:10:13 *** xQuasar has joined #aichallenge 2011-06-26T01:14:39 *** stocha has joined #aichallenge 2011-06-26T01:14:43 stoha02 skill = -0.30 2011-06-26T01:14:57 now that's really weak :) 2011-06-26T01:15:49 _flag: your current version seems to do rather well :) 2011-06-26T01:17:32 _flag: i think it humiliated poor sir_macelon's bot 2011-06-26T01:18:04 <_flag> stocha: Thanks, I'm hoping it takes back first soon 2011-06-26T01:18:34 _flag: i bet 2 euros, it will. (unless the ranking is broken ...) 2011-06-26T01:19:11 <_flag> Well you never know, could just be luck :) 2011-06-26T01:19:38 _flag: i'm not a ranking algorithm. I do watch games. And can spot a clear superiority. 2011-06-26T01:19:41 <_flag> Either that or it hasn't played macelon on smaller maps yet, so I'm not sure how it'll do there 2011-06-26T01:19:59 ah smaller map. i forgot about that one. 2011-06-26T01:20:30 but i guess there a suficiently few of them, that normal map superiority (90% win) will make out for them. 2011-06-26T01:20:36 even if they are loosing. 2011-06-26T01:20:48 so i guess you have first for now :p 2011-06-26T01:20:50 *** sigh has joined #aichallenge 2011-06-26T01:20:51 <_flag> One of the things macelon does wronge (IMO) is he clumps up his ants too much allowing for me to run around and capture more food until I can overwhelm him 2011-06-26T01:21:18 _flag: yeah, i was wondering how his bot feeded so well with that. 2011-06-26T01:21:36 <_flag> s/wronge/wrong 2011-06-26T01:21:41 it seems that escaping on open maps is good strategy 2011-06-26T01:22:16 it would be interesting to see how tings goes on a tiny corridor map though 2011-06-26T01:22:39 stocha: this was a complete fluke, but i really like this map http://ants.zeroviz.us/wPk4p/ 2011-06-26T01:22:42 <_flag> We'd probably both just block off the corridors and it would come down to food spawns 2011-06-26T01:23:04 i'm trying to make a faster more mazey generator 2011-06-26T01:23:19 antimatroid: did you used the big room trick ? 2011-06-26T01:23:42 antimatroid: there are a lot of maze algorithms out there, are you looking at those, or rolling your own? 2011-06-26T01:23:45 that was using voronoi diagrams to order the squares then adding water while it was still valid 2011-06-26T01:23:49 * sigh is back in australia 2011-06-26T01:23:49 a terrible way of generating maps 2011-06-26T01:24:11 antimatroid: ah okay. you haven't tryed yesterdays algorithm yet then ? 2011-06-26T01:24:20 hi sigh 2011-06-26T01:24:28 hi stocha 2011-06-26T01:24:34 sigh: i'm currently trying to do random carved paths that don't overlap itself starting inside the already carved land 2011-06-26T01:24:44 and i have looked at a few other maze algorithms 2011-06-26T01:24:46 are you the google guy ? i'm a bit lost. who was the google guy 2011-06-26T01:25:06 i have no memory for speudonyms it seems 2011-06-26T01:25:10 there are lots of google guys 2011-06-26T01:25:12 sigh is one of them yes 2011-06-26T01:25:16 I don't know about *the* google guy, but I do work at google 2011-06-26T01:25:31 but he is THE (new) google guys that can play go also. 2011-06-26T01:25:36 sigh: how was mtv overall? 2011-06-26T01:25:38 i think. if i don't mess up things a bit :p 2011-06-26T01:25:47 you have the right person 2011-06-26T01:25:58 sigh: so how is your bot doing ? :p 2011-06-26T01:26:12 antimatroid: fun, but glad I don't work there permenantly :) 2011-06-26T01:26:17 it's a bit overwhelming 2011-06-26T01:26:33 stocha: not doing a bot :P 2011-06-26T01:26:36 sigh: do you know verilog ? 2011-06-26T01:26:52 the programming language. 2011-06-26T01:27:01 haven't used it 2011-06-26T01:28:44 why the sudden interest in verilog? 2011-06-26T01:29:17 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-06-26T01:29:55 sigh i don't know. I want to learn it. I want to understand a low level language. Seem more stable(universal) than assembler beside it. 2011-06-26T01:30:16 although i'm clearly lacking the base :'( 2011-06-26T01:30:18 but verilog is a hardware description language? 2011-06-26T01:30:25 sigh: yeah. 2011-06-26T01:30:32 a low level language if there is one :) 2011-06-26T01:30:45 I never tryed to dig beyond assembly, for 20 years ! 2011-06-26T01:31:03 i'm a bit ashamed all of sudden. Especially when i feel that it answers so much. 2011-06-26T01:31:21 ah ok, sure 2011-06-26T01:31:23 hardware is inerently parallel it seems.And it has interesting primitives. 2011-06-26T01:31:31 I used VHDL in uni a bit, but I don't remember much 2011-06-26T01:31:53 I've always go like : Why are there so many useless stuff in assembly language atoms ? 2011-06-26T01:32:21 it seems that hardware is indeed a lot cleaner. 2011-06-26T01:32:36 Lastly, i do most my thinking with bit manipulation. 2011-06-26T01:32:51 (high level language thinking that is) 2011-06-26T01:32:54 depends, at the transistor level, hardware is very messy :P 2011-06-26T01:33:05 yeah ? how so ? 2011-06-26T01:33:20 you mean it's physics and chemicals etc ? 2011-06-26T01:33:41 so the transistor would be the key to straightforward "high level" language then. 2011-06-26T01:33:44 it's a step above physics, you are dealing with electronics 2011-06-26T01:33:52 i know so few about electronic. 2011-06-26T01:34:09 i don't even know where to begin. i only need whatever is usefull for verilog and such though. 2011-06-26T01:34:13 but going from transistor to logic gates is messy (from my view) 2011-06-26T01:34:22 but i simply have the level of a 8 years old boy. (or girl) 2011-06-26T01:35:00 Yeah, but logic gates seems clean. Although their seems to be some electricity concept. 2011-06-26T01:35:36 Do many is taken for granted. Because verilog is tought only to people that have high electronic background. 2011-06-26T01:35:53 I don't even know where to look to have the base. 2011-06-26T01:36:02 yeah, once you have logic gates then the rest is "easy" to reason about 2011-06-26T01:36:04 if you have a suggestion ... 2011-06-26T01:36:28 have you looked at something like this: http://mines.humanoriented.com/proposals/nand-to-tetris/ 2011-06-26T01:36:32 i think i have seen signal transmission stuff talken about. 2011-06-26T01:36:45 *** SSN has joined #aichallenge 2011-06-26T01:37:22 that book/course shows how a computer is build from the gate level 2011-06-26T01:37:58 okay. thanks for the pointer :) 2011-06-26T01:38:27 keep hanging around in the coming month (years ?) so i can ask if i have more questions :p 2011-06-26T01:39:21 I haven't read that book, I've just heard it recommended a lot 2011-06-26T01:39:36 I don't know much more than the basics 2011-06-26T01:41:18 This work will not involve Web applications or the use of Java,' and that had served to almost completely eliminate U.S. applicants.' 2011-06-26T01:41:32 hehe. i have the level of a US graduate :p 2011-06-26T01:44:04 i guess i have to buy the book now. 2011-06-26T01:57:28 all this sounds like a lot of work. But i guess you have to suffer in order to be knowlegable 2011-06-26T02:20:15 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-06-26T02:31:44 *** SSN has quit IRC (Ping timeout: 252 seconds) 2011-06-26T02:40:24 *** nux67 has joined #aichallenge 2011-06-26T02:43:50 *** stocha has quit IRC (Ping timeout: 252 seconds) 2011-06-26T03:12:36 aichallenge: janzert epsilon * r530ce6f / website/api_game_result.php : Always send a hash back to the worker if possible so it can move on to next task - http://bit.ly/ieEOKQ 2011-06-26T03:14:29 contestbot: later tell McLeopold workers were getting stuck on retrying task submission at least in the case I tracked down, because the task had already been reassigned and the server would then not return the received hash 2011-06-26T03:14:29 janzert: Yes master! 2011-06-26T03:25:21 *** Accoun has joined #aichallenge 2011-06-26T04:08:01 *** stocha has joined #aichallenge 2011-06-26T04:08:28 finally i made a bot that may proved to be stronger than sto_greed_002 in my self testings. 2011-06-26T04:08:54 i hope i can take a solid 3rd with it. 2011-06-26T04:09:06 doesn't sound too great. To hope for 3rd ... well well well 2011-06-26T04:09:59 *** UncleVasya has joined #aichallenge 2011-06-26T04:15:26 *** stocha has quit IRC (Quit: Page closed) 2011-06-26T04:59:31 bah maze maps are hard 2011-06-26T05:01:09 *** gobang has joined #aichallenge 2011-06-26T05:02:08 sigh: the following things make it slightly harder for me.. 1. I want to carve the maze with nxn blocks so that a minxmin block as specified can traverse all land squares, 2. I don't want to do this from a smaller map and then blow it up so that paths may exist between regions with a 1 width opening still, just that there exists a path around from 1, 3. I want the corridors in the maze to be noisy, not straight 2011-06-26T05:09:25 *** Palmik has joined #aichallenge 2011-06-26T05:56:17 *** mleise has joined #aichallenge 2011-06-26T06:38:55 *** mleise has quit IRC (Quit: Leaving.) 2011-06-26T07:17:41 *** onensora has joined #aichallenge 2011-06-26T07:18:49 *** foRei has joined #aichallenge 2011-06-26T07:37:54 *** davidd has quit IRC (Read error: No route to host) 2011-06-26T07:37:59 *** davidd has joined #aichallenge 2011-06-26T07:46:21 *** eashoka has joined #aichallenge 2011-06-26T07:52:18 *** stocha has joined #aichallenge 2011-06-26T07:52:25 how do one query the database ? 2011-06-26T07:52:30 game data base 2011-06-26T07:53:09 *** sYnfo has joined #aichallenge 2011-06-26T07:53:18 *** carlop has joined #aichallenge 2011-06-26T07:54:13 the ranking site doesn't seem to be responding like a charm 2011-06-26T07:56:44 my new bot is a total fiasco it seems 2011-06-26T08:01:41 antimatroid: i'm not sure how guarantee that is not possible to perfect block a part of a map, but seem that 2x2 block traversal don't work 2011-06-26T08:01:59 *** stocha has quit IRC (Quit: Page closed) 2011-06-26T08:02:12 carlop: i need to do 3x3 to get rid of it properly, the point is i can 2011-06-26T08:02:22 see this example in one of your maze map: http://ants.zeroviz.us/dDw5p/ 2011-06-26T08:02:25 i don't mind having paths between regions which are 1 width wide, provided there is a work around 2011-06-26T08:02:41 a 2x2 block can traverse every land square of that 2011-06-26T08:03:20 carlop: yeah, you need to have a 3x3 block that can traverse the map 2011-06-26T08:03:32 it was just an example map 2011-06-26T08:04:04 yes, the problem is that i have not think of a method for guarantee non-blockable map 2011-06-26T08:04:18 a 3x3 block being able to traverse all land? 2011-06-26T08:04:25 why doesn't that work? 2011-06-26T08:04:29 3x3 block traversing work 2011-06-26T08:04:34 but limit a lot the possible map 2011-06-26T08:05:21 i want to do simple random dfs walks on a stack to generate a map 2011-06-26T08:05:42 but it's hard to do it with a large block moving around in a smaller grid 2011-06-26T08:05:54 i think that few 1-corridor or 2-corridor can be a good thing, but i can not understand how to check if the map is blockable or not 2011-06-26T08:06:24 carlop: notice the "openings" in the above map which are 1 width? 2011-06-26T08:06:28 those will still be possible 2011-06-26T08:06:32 yes, that is the problem, cause i think a lot of 2x2 traversable map are non blockable, but is hard to guarantee that a map is non-blockable 2011-06-26T08:06:34 so you can block off paths 2011-06-26T08:06:50 but you can't block yourself into a region 2011-06-26T08:06:52 yes, but blocking path don't seem a problem for me 2011-06-26T08:09:52 my problem is "find afast proof that a map do not have a blockable region, without limiting too much the possible pattern" 2011-06-26T08:16:22 *** xQuasar has quit IRC () 2011-06-26T08:26:31 *** carlop has quit IRC (Ping timeout: 250 seconds) 2011-06-26T08:26:42 *** carlop has joined #aichallenge 2011-06-26T08:31:15 *** Zaphus has joined #aichallenge 2011-06-26T08:51:26 *** smiley1983 has joined #aichallenge 2011-06-26T08:52:58 UncleVasya: you might have changed it already, but if not then the calls to Sys.time can now be changed to Unix.gettimeofday and the server will compile it correctly 2011-06-26T08:53:27 locally, it will also require "ocamlbuild -lib unix" 2011-06-26T08:54:05 Ш вшвтэе срфтпу ше фтв нуыеуквфн ыукмук сщьзшдув ьн ыгиьшыышщт цуддю 2011-06-26T08:54:11 sorry 2011-06-26T08:54:44 I didn't change it and yesterday server compiled my submission well. 2011-06-26T08:56:59 smiley1983: I need to replace "ocamlc -c MyBot.ml" with "ocamlbuild -lib unix MyBot.ml"? 2011-06-26T08:59:08 *** gobang has quit IRC (Read error: Connection reset by peer) 2011-06-26T08:59:40 *** gobang has joined #aichallenge 2011-06-26T09:00:38 UncleVasya: it would always compile with Sys.time without any flags 2011-06-26T09:00:48 However, Sys.time gives the wrong sort of time 2011-06-26T09:00:58 Unix.gettimeofday gives the correct sort of time 2011-06-26T09:01:12 even on Windows? 2011-06-26T09:01:22 However, it needs to be told about the Unix library when compiling 2011-06-26T09:01:31 Yes, the unix module is implemented for Windows 2011-06-26T09:02:00 If compiling with ocamlbuild "ocamlbuild -lib unix MyBot.native" will do everything, no other steps needed 2011-06-26T09:02:30 If compiling with ocamlc, add "unix.cma" somewhere in the args... let me go and check where 2011-06-26T09:03:43 So my makefile needs only "ocamlbuild -lib unix MyBot.native" and nothing else (like Ants.ml) ? 2011-06-26T09:04:08 That's correct, if you have ocamlbuild then it will calculate what it needs to compile 2011-06-26T09:04:22 That's what the server is using 2011-06-26T09:05:19 thanks 2011-06-26T09:07:41 *** Zaphus has quit IRC (Quit: Page closed) 2011-06-26T09:07:49 When compiling with ocamlc, here is the command which links together a (slow) bytecode executable: 2011-06-26T09:08:00 ocamlc -o MyBot unix.cma Ants.cmo MyBot.cmo 2011-06-26T09:08:16 *** eashoka has quit IRC (Quit: Page closed) 2011-06-26T09:08:24 All the other compile lines work without unix.cma 2011-06-26T09:10:51 I wrote a function that works very slow and I think it isn't a correct way to do what I need. 2011-06-26T09:12:21 It gets 2 lists and adds to the 2nd list all items from 1st list except ones that already in the 2nd list. 2011-06-26T09:12:28 Here is the code: http://pastebin.com/2qCjW2dm 2011-06-26T09:13:41 For example: 1 2 4 6 and 2 4 5 --> 1 2 4 5 6 2011-06-26T09:14:03 smiley1983: Do you have a time to look on that? 2011-06-26T09:14:50 I'll take a look now... 2011-06-26T09:15:21 * 6 1 2 4 5 because there is no sorting 2011-06-26T09:18:33 Okay, two things to fix (I don't know if it'll correct the speed problem but I suspect I will help) : 2011-06-26T09:18:54 replace List.exists with List.mem item l 2011-06-26T09:19:10 smiley1983: how did you make maze maps? :) 2011-06-26T09:19:33 replace ([head] @ lst_2) with (head :: lst_2) 2011-06-26T09:20:18 smiley1983: how many types of rotational symmetry did you have? i have 5 types for 2 player maps, 3 types for 4 player maps and 1 for 8 player maps 2011-06-26T09:21:29 antimatroid: make a grid of cells. They are all closed. Pick a random cell and mark it as open. Put all of its closed neighbors into the closed list. Iteratively remove an item from the closed list, connect it to a random open neighbor, and add its closed neighbors to the list 2011-06-26T09:21:49 When there are no more closed cells, stop :) 2011-06-26T09:22:19 smiley1983: i'll play around thanks 2011-06-26T09:22:48 UncleVasya: ([head] @ lst_2) builds a new list containing one item, and then adds the items of the two lists together. head :: lst_2 just adds the item to lst_2 2011-06-26T09:22:54 i also have the problem of wanting to make them while moving say a 3x3 block around in a grid, moving only 1 n/s/e/w at a time, which complicates stuff :\ 2011-06-26T09:23:01 antimatroid: I only used one sort of rotational symmetry 2011-06-26T09:25:42 there are lots :) 2011-06-26T09:26:05 *** onensora has quit IRC (Ping timeout: 260 seconds) 2011-06-26T09:26:10 There are lots of random dungeon generators used in roguelike games which do things similar to what I've described above, but instead of using a fixed granularity, attempt to find places to add features without interfering with previously added features 2011-06-26T09:26:27 *** onensora has joined #aichallenge 2011-06-26T09:26:29 so they plan and cancel features which break certain rules 2011-06-26T09:26:56 some parts of features are allowed to be overwritten by new ones, other parts have to be left intact 2011-06-26T09:27:42 gtg, might be back later tonight, otherwise tomorrow 2011-06-26T09:27:44 *** smiley1983 has quit IRC (Quit: leaving) 2011-06-26T09:35:53 *** computerwiz_222 has joined #aichallenge 2011-06-26T09:36:43 antimatroid: where I can find a reference code for map generation ? I have few ideas to try.. this code (i think you linked it) can be a starting point ? 2011-06-26T09:36:53 antimatroid: this code = http://pastebin.com/JG3kxFzG 2011-06-26T09:37:03 carlop: let me clean the current stuff up if you want 2011-06-26T09:37:05 it's a complete mess 2011-06-26T09:37:14 i don't know python very well 2011-06-26T09:37:22 it should really be broken up into multiple classes 2011-06-26T09:40:35 if this code is working i try to look at it, i don't know python very well either, only want to try something 2011-06-26T09:41:33 carlop: feel free to look, but i think it will be a lot clearer once i clean it up 2011-06-26T09:41:46 i'm reading up how to use classes properly 2011-06-26T09:44:31 ok, i try some idea, when you have new code just tell me, just remember that Ii don't want to put time pressure on you, so do it when you prefer 2011-06-26T10:13:15 antimatroid: Is it possible to make some kind of table with pares CPU_speed---Turn_time so people cold test their bots locally in the conditions that close to the server's one? 2011-06-26T10:13:56 For example: 2.66 ghz - 1 sec, 1.8ghz - 2 sec 2011-06-26T10:14:37 UncleVasya: I don't know much about that kind of stuff sorry 2011-06-26T10:15:04 Or just show somewhere server's speed and people can calculate time for their computer manually. 2011-06-26T10:23:28 *** mleise has joined #aichallenge 2011-06-26T10:24:51 UncleVasya: McLeopold/janzert posted a link to the specs of the server a week ago, but i can't find it now 2011-06-26T10:25:31 I'll told to them, thanks. 2011-06-26T10:25:35 *talk 2011-06-26T10:43:48 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-06-26T10:57:31 *** computerwiz_222 has quit IRC (Quit: Konversation terminated!) 2011-06-26T10:57:47 *** computerwiz_222 has joined #aichallenge 2011-06-26T11:06:14 *** stocha has joined #aichallenge 2011-06-26T11:06:43 my new stocha is a bit on the shy side. But at least it represents the philosophy of the house. 2011-06-26T11:08:42 food spawning speed really affect the current winning strategy a lot (for example between flap_crapper and sir_macelon) 2011-06-26T11:11:50 *** smellyhippy has quit IRC (Quit: But wait! what are those penguins up to?) 2011-06-26T11:12:25 *** smellyhippy has joined #aichallenge 2011-06-26T11:12:35 *** computerwiz_222 has quit IRC (Quit: Konversation terminated!) 2011-06-26T11:12:56 *** stocha has quit IRC (Quit: Page closed) 2011-06-26T11:13:00 *** computerwiz_222 has joined #aichallenge 2011-06-26T11:54:10 *** sigh has quit IRC (Remote host closed the connection) 2011-06-26T11:58:36 any pythoners about? 2011-06-26T11:58:53 if i have a class in a file say grid.py, how do i use that in a file say symmetr_mapgen.py 2011-06-26T11:59:42 nevermind 2011-06-26T11:59:52 from grid import myclass 2011-06-26T12:00:21 *** jmpespxoreax has quit IRC (Ping timeout: 255 seconds) 2011-06-26T12:22:01 *** computerwiz_222 has quit IRC (Quit: Konversation terminated!) 2011-06-26T12:31:05 *** amstan has joined #aichallenge 2011-06-26T12:31:05 *** ChanServ sets mode: +o amstan 2011-06-26T12:32:29 *** smellyhippy has quit IRC (Read error: Connection reset by peer) 2011-06-26T12:33:01 *** smellyhippy has joined #aichallenge 2011-06-26T12:33:01 *** smellyhippy has joined #aichallenge 2011-06-26T12:36:21 *** zeLark has joined #aichallenge 2011-06-26T12:51:06 *** computerwiz_222 has joined #aichallenge 2011-06-26T12:51:13 Hi zeLark :) 2011-06-26T12:56:26 carlop: how are you going with my code? :P 2011-06-26T12:56:32 i have separated it into two files 2011-06-26T13:00:58 *** smiley1983 has joined #aichallenge 2011-06-26T13:02:23 UncleVasya: I've seen another optimization which can be applied to the code you posted earlier 2011-06-26T13:03:44 If lst_1 is already free of duplicates within itself, you can replace line 6 with: 2011-06-26T13:03:58 smiley1983: I found a problem in program logic.There was a possibility to deal without merging lists. 2011-06-26T13:04:16 head :: merge_unique tail lst_2 2011-06-26T13:04:33 But it always interesting to learn better way, so thanks again. 2011-06-26T13:04:43 Right, merging lists like this is always going to be costly, so is best avoided 2011-06-26T13:05:26 no problem :) Does the "head :: merge_unique tail lst_2" make sense? 2011-06-26T13:08:01 currently line 6 is merge_unique tail (lst_2 @ [head]) because head must be added to the end of lst_2. 2011-06-26T13:08:48 yes - but that means that when you check for duplicates in the next pass, you are checking against the newly added items as well 2011-06-26T13:09:52 The way I'm suggesting, you build the list of results as you go, without the double-checking, which would have been a large reason why it was slow 2011-06-26T13:10:55 since merge_unique returns a list, "head :: merge_unique ..." builds the item onto the result of the next call to merge_unique 2011-06-26T13:12:48 I think I've got it. 2011-06-26T13:13:33 it's a useful structure to understand, because it appears often 2011-06-26T13:15:09 oh, though I now realise that this adds the items to the beginning of the list. 2011-06-26T13:15:49 If they must be added to the end, you probably have to use List.rev on lst_2 before passing it in, and then List.rev again on the result - but there's usually a way to avoid this 2011-06-26T13:16:17 (when doing things the way I'm describing, that is) 2011-06-26T13:16:38 antimatroid: i wrote what i have think about symmetry generation here: http://pastebin.com/udSQiHdc, basically I'm using McLeopold representation (base point, base orientation) to summarize symmetries 2011-06-26T13:17:24 *** smiley1983 has quit IRC (Quit: leaving) 2011-06-26T13:17:52 carlop: more symmetries are possible :P 2011-06-26T13:18:46 i think so, a lot of symmetries not actualy generated are possible 2011-06-26T13:20:22 http://pastebin.com/GvdeXjdb 2011-06-26T13:20:28 that's my grid.py file 2011-06-26T13:20:57 http://pastebin.com/qkhceSgk symmetric_mapgen.py 2011-06-26T13:21:24 now i take a look of the code 2011-06-26T13:21:39 look at the second link first 2011-06-26T13:21:43 as it's much shorter 2011-06-26T13:21:56 *** GreenTea has joined #aichallenge 2011-06-26T13:22:14 Hi all 2011-06-26T13:22:25 line 34 can be changed to "self.w_locs[b]В =В self.grid.get_block(self.w_locs[b],В d)" too 2011-06-26T13:22:32 GreenTea: morning :) 2011-06-26T13:22:48 I have submited my lisp bot :) 2011-06-26T13:22:54 GreenTea: interested in playing your hand at map generation? :P 2011-06-26T13:23:32 Now maybe.. 2011-06-26T13:23:48 http://aichallengebeta.hypertriangle.com/profile.php?user=21 2011-06-26T13:24:03 Strange compilations errors.. 2011-06-26T13:24:15 I can see only style warnings 2011-06-26T13:24:33 It compiled fine on my local machine 2011-06-26T13:24:52 http://pastebin.com/GvdeXjdb grid.py 2011-06-26T13:24:52 http://pastebin.com/qkhceSgk symmetric_mapgen.py 2011-06-26T13:25:03 GreenTea: that's the current state of my map generator 2011-06-26T13:28:29 antimatroid: I think is good having 2 different file, one for generate pattern and the other for do calculation / symmetry 2011-06-26T13:28:42 carlop, GreenTea: the idea is that you guys don't really need to understand all of grid.py, just be able to use the needed functions 2011-06-26T13:29:19 it's messy, but i think it makes it easy to write general map generation functions for all symmetry types at once 2011-06-26T13:29:27 yes, currently i'm working on grid.py... i found more interesting the part about generating symmetry,,, 2011-06-26T13:30:09 but later i can look at how build random walk that seem interesting (what parameter are needed, etc) 2011-06-26T13:30:19 random walks suck 2011-06-26T13:30:24 maybe other 2011-06-26T13:30:25 maze generation is where it's at 2011-06-26T13:30:48 sory I don't understand english well, do you ask me to write map generator? ) 2011-06-26T13:31:09 GreenTea: i'm trying to convince people to play around with map generation 2011-06-26T13:31:18 ignore me if you want :) 2011-06-26T13:32:39 you mean test it? 2011-06-26T13:33:47 nah, i want more stuff like random_walk_map(self) in symmetric_mapgen.py 2011-06-26T13:33:51 i know it works :P 2011-06-26T13:34:05 it's not very fast doing the random walk maps either 2011-06-26T13:34:23 calling is_valid on grid is expensive 2011-06-26T13:35:12 GreenTea: you've seen, that your last submission got classified as LISP instead of java ? 2011-06-26T13:36:48 certainly, my submission is now lisp) 2011-06-26T13:37:10 oh :) 2011-06-26T13:38:23 I dont understand why style warnings are interpreted as compilations erros.. 2011-06-26T13:40:32 brb 2011-06-26T13:42:40 *** carlop has quit IRC (Remote host closed the connection) 2011-06-26T13:52:23 *** GreenTea has quit IRC (Quit: Page closed) 2011-06-26T13:53:54 *** GreenTea has joined #aichallenge 2011-06-26T13:56:59 One my friend has a problem. He downloaded map from server but can't launch a game on it. Here is his error: http://pastebin.com/dYAt2iD5 2011-06-26T13:58:12 Do I need ask more details from him? 2011-06-26T14:00:32 not sure, hmm 2011-06-26T14:00:38 check if his bat file is good 2011-06-26T14:12:00 His map isn't from site, maybe it's the problem. Never mind. 2011-06-26T14:18:51 *** GreenTea has quit IRC (Quit: Page closed) 2011-06-26T14:32:24 I dug deeper and found that new version of playgame.py tries to spawn food symmetrically even on asymmetric maps. 2011-06-26T14:33:24 I've launched his map with old ants repository clone without any problems. 2011-06-26T14:35:11 But with current version it says "This map does not support symmetric food placement " even if I name it asymmetric_pp.map and place into asymmetric_maps directory. 2011-06-26T14:35:44 *** carlop has joined #aichallenge 2011-06-26T15:10:48 *** nann has joined #aichallenge 2011-06-26T15:13:51 *** Apophis_ has quit IRC (Ping timeout: 255 seconds) 2011-06-26T15:20:16 *** george1914 has quit IRC (Ping timeout: 252 seconds) 2011-06-26T15:24:02 *** sYnfo has quit IRC (Quit: Page closed) 2011-06-26T15:27:48 *** computerwiz_222 has quit IRC (Quit: Konversation terminated!) 2011-06-26T15:50:49 *** Accoun has quit IRC () 2011-06-26T15:54:47 *** zeLark has quit IRC (Quit: Page closed) 2011-06-26T15:54:53 UncleVasya: pretty sure you have to tell playgame what version of food spawning to use and it just uses the symmetric spawning as the default 2011-06-26T15:55:12 Why sometimes game ends before turns limit expired even if more than 1 bots are alive? 2011-06-26T15:55:38 *** stocha has joined #aichallenge 2011-06-26T15:55:59 because of the early game cutoff (basically one player has 90% or more of ants for 100 turns) 2011-06-26T15:56:12 @later tell Zaphus Mouhaha, my overskilled combat strategy took down your bot :) http://aichallengebeta.hypertriangle.com/visualizer.php?game=39442&user=107 2011-06-26T15:56:12 stocha: As you wish. 2011-06-26T15:56:14 *** computerwiz_222 has joined #aichallenge 2011-06-26T15:56:17 unspawned food in this case also counts as a player 2011-06-26T15:57:52 _flag: i'm suprised that your bot isn't first place yet. But then you don't let it play also :) 2011-06-26T16:00:03 <_flag> stocha: I submitted again because there was another bug in my previous version. I should (I think) pass macelon this time, but we'll see 2011-06-26T16:00:14 <_flag> He's still pretty good 2011-06-26T16:05:36 *** Accoun has joined #aichallenge 2011-06-26T16:17:50 *** stocha has quit IRC (Ping timeout: 252 seconds) 2011-06-26T16:20:59 *** stocha has joined #aichallenge 2011-06-26T16:26:34 _flag: we'll see 2011-06-26T16:30:29 *** gobang has quit IRC (Read error: Connection reset by peer) 2011-06-26T16:37:56 *** UncleVasya has quit IRC (Ping timeout: 260 seconds) 2011-06-26T16:50:50 *** stocha has quit IRC (Ping timeout: 252 seconds) 2011-06-26T17:00:10 *** stocha has joined #aichallenge 2011-06-26T17:07:07 *** computerwiz_222 has quit IRC (Quit: Konversation terminated!) 2011-06-26T17:08:49 *** janzert has left #aichallenge 2011-06-26T17:09:07 *** janzert has joined #aichallenge 2011-06-26T17:12:43 *** onensora has quit IRC () 2011-06-26T17:13:10 *** Cyndre__ has quit IRC (Ping timeout: 258 seconds) 2011-06-26T17:20:51 *** stocha has quit IRC (Quit: Page closed) 2011-06-26T17:28:07 *** stocha has joined #aichallenge 2011-06-26T17:28:38 @later tell Zaphus : have you seen this http://aichallengebeta.hypertriangle.com/visualizer.php?game=39638&user=206 :) 2011-06-26T17:28:38 stocha: Ready to serve, my lord. 2011-06-26T17:33:00 *** stocha has quit IRC (Ping timeout: 252 seconds) 2011-06-26T17:36:17 antimatroid: http://ants.zeroviz.us/ttXmN/ and http://ants.zeroviz.us/eKF6y/, example of map for 8 player symmetric but non-rotational and non-tesselation 2011-06-26T17:37:01 *** zaphus has joined #aichallenge 2011-06-26T17:37:21 (i haven't work on pattern, only for finding new symmetries) 2011-06-26T17:46:58 the first link is wrong, this is a true non-rotational symmetric map: http://ants.zeroviz.us/pgEje/ 2011-06-26T17:47:56 contestbot tell stocha nice work on the combat strategy bot, looks really cool! 2011-06-26T17:47:56 zaphus: Error: I haven't seen stocha, I'll let you do the telling. 2011-06-26T17:48:16 contestbot later stocha nice work on the combat strategy bot, looks really cool! 2011-06-26T17:48:17 zaphus: You have no gotten any error messages recently, so here's a random one just to let you know that we care. 2011-06-26T17:48:36 anyone know how to use the contestbot to leave a message for someone ? 2011-06-26T17:49:10 <_flag> @later tell Zaphus : Something like this? 2011-06-26T17:49:10 _flag: Aye, aye, sir 2011-06-26T17:49:35 @later tell stocha : nice work on the combat strategy bot, looks really cool! 2011-06-26T17:49:35 zaphus: Yes master! 2011-06-26T17:49:42 _flag thanks 2011-06-26T17:49:50 *** ltriant has joined #aichallenge 2011-06-26T17:50:41 *** ltriant has joined #aichallenge 2011-06-26T17:54:03 *** sigh has joined #aichallenge 2011-06-26T17:56:24 *** computerwiz_222 has joined #aichallenge 2011-06-26T18:00:45 *** Palmik has quit IRC (Remote host closed the connection) 2011-06-26T18:05:38 *** zaphus has quit IRC (Quit: Page closed) 2011-06-26T18:10:02 *** computerwiz_222 has quit IRC (Ping timeout: 258 seconds) 2011-06-26T18:15:32 *** sigh has quit IRC (Remote host closed the connection) 2011-06-26T18:20:36 *** Apophis has joined #aichallenge 2011-06-26T18:26:37 *** Cyndre has joined #aichallenge 2011-06-26T18:28:01 *** computerwiz_222 has joined #aichallenge 2011-06-26T18:37:20 *** stocha has joined #aichallenge 2011-06-26T18:37:45 antimatroid: please, make at least 1 map as big as the Visualizer can handle during the beta .... 2011-06-26T18:38:13 I finally made a bot that can support 256x256 maps, and has so much free time left, i don't know what to do with it 2011-06-26T18:40:08 @later tell And the last version can support 256x256 maps with about no limit on the numbers of ant (unoptimized). In fact there so much time left, i don't even know what to do with it :) 2011-06-26T18:40:08 stocha: Yep. 2011-06-26T18:42:06 @later tell Zaphus: yeah thanks for the compliment :) 2011-06-26T18:42:06 stocha: Yep. 2011-06-26T18:42:37 *** nux67 has quit IRC (Quit: Page closed) 2011-06-26T18:43:02 @later tell Janzert There really should be at least one map that is the maximum size the visualizer can comfortably support. At least during the beta, so we can work out the upper bound for customers. 2011-06-26T18:43:02 stocha: Ready to serve, my lord. 2011-06-26T19:10:41 *** stocha has quit IRC (Quit: Page closed) 2011-06-26T19:51:38 *** carlop has quit IRC (Read error: Operation timed out) 2011-06-26T20:17:50 *** mleise has quit IRC (Quit: Leaving.) 2011-06-26T20:50:44 *** jmcarthur has joined #aichallenge 2011-06-26T20:50:58 woo NYC! 2011-06-26T20:51:09 jmcarthur: :) 2011-06-26T21:22:24 *** computerwiz_222 has quit IRC (Ping timeout: 255 seconds) 2011-06-26T21:39:08 *** computerwiz_222 has joined #aichallenge 2011-06-26T21:54:03 *** carlop has joined #aichallenge 2011-06-26T21:55:16 *** Chris_0076 has joined #aichallenge 2011-06-26T22:00:42 *** Chris_0076 has quit IRC (Ping timeout: 252 seconds) 2011-06-26T22:02:27 *** computerwiz_222 has quit IRC (Ping timeout: 255 seconds) 2011-06-26T22:16:36 *** carlop has quit IRC (Read error: Operation timed out) 2011-06-26T22:20:10 *** smiley1983 has joined #aichallenge 2011-06-26T22:32:04 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-06-26T22:41:01 *** foRei has quit IRC (Read error: Connection reset by peer) 2011-06-26T22:42:05 *** foRei