2011-03-22T00:09:50 you can block 2 width corridors with sqrt(6) 2011-03-22T00:11:46 https://github.com/aichallenge/aichallenge/wiki/Ants-Battle-Resolution-Options 2011-03-22T00:11:49 see example 4 2011-03-22T00:14:08 * amstan uses 'git describe --tags' for including the version inside the starter packages 2011-03-22T00:15:51 *** j3camero has quit IRC (Read error: Connection reset by peer) 2011-03-22T00:16:12 *** j3camero has joined #aichallenge 2011-03-22T00:17:08 *** phirenz has quit IRC (Ping timeout: 240 seconds) 2011-03-22T00:21:53 more examples are there 2011-03-22T00:22:21 janzert: there's quite a few examples where power and damage differ 2011-03-22T00:22:44 yes 2011-03-22T00:23:37 I wonder does it require three teams to make them differ? 2011-03-22T00:23:47 my first guess is it does 2011-03-22T00:24:09 and then I see example 10 :P 2011-03-22T00:24:13 so nevermind 2011-03-22T00:24:23 :p 2011-03-22T00:25:23 is 1 damage where you live if you have less damage than one of your enemies? 2011-03-22T00:25:40 is power damage where* 2011-03-22T00:26:13 i think that might be right 2011-03-22T00:26:20 I think so 2011-03-22T00:26:24 *** phirenz has joined #aichallenge 2011-03-22T00:26:36 *** dlila has quit IRC (Quit: Leaving) 2011-03-22T00:26:39 thatn all of your enemies** 2011-03-22T00:26:55 have you tried blocking 3 or 4 wide corridors with sqrt(6)? 2011-03-22T00:27:01 right 2011-03-22T00:27:05 yeah, pretty sure you can't block 3 fully 2011-03-22T00:27:09 but you can do a good job 2011-03-22T00:27:14 did i give those examples? 2011-03-22T00:27:16 or "than each of" 2011-03-22T00:27:44 6 is one 3 wide example 2011-03-22T00:27:52 5 is repeated 4 2011-03-22T00:27:53 my bad 2011-03-22T00:29:14 check example 5 now 2011-03-22T00:29:38 it's blockable with support 2011-03-22T00:29:49 ahh, so you can trade 3 for 1 2011-03-22T00:30:04 what? 2011-03-22T00:30:07 oh yeah 2011-03-22T00:30:08 yeah 2011-03-22T00:30:21 then if you come in right after and they run out, you can diminish them 2011-03-22T00:30:29 but the can also set up a nice block temporarily 2011-03-22T00:30:38 i don't mind that, i just don't want full blocking 2011-03-22T00:30:49 yeah, it's probably ok 2011-03-22T00:31:17 too bad we have to keep all corridors 3 wide 2011-03-22T00:31:48 2* 2011-03-22T00:31:57 oh err 2011-03-22T00:32:06 we could change the attack radius back to 2 2011-03-22T00:34:33 hmm, I think I'd like it better but why was it increased? 2011-03-22T00:36:07 mcleopold wanted a larger one 2011-03-22T00:38:39 i think i should restate power/occupied as damage 1 and damage 2 2011-03-22T00:38:49 sigh: ^^? 2011-03-22T00:39:06 antimatroid: no 2011-03-22T00:39:11 but they're so close 2011-03-22T00:39:15 power is power, not damage 2011-03-22T00:39:34 it's damage with you live if you have less damage than any of your enemies in range 2011-03-22T00:39:55 meh, I don't think it makes sense, but I'm too busy to argue 2011-03-22T00:40:09 i'll leave it for the moment 2011-03-22T00:40:49 chris___0076: ping 2011-03-22T00:40:54 I don't really care a lot since I think the new damage one has about the same advantages as the old power option and is much more intuitive to explain 2011-03-22T00:41:11 amstan:pong 2011-03-22T00:41:28 hmm? 2011-03-22T00:41:31 *** chris___0076 is now known as chris_0076 2011-03-22T00:41:41 chris___0076: can you delete your fork, then refork. You still appear in the aichallenge network with that one commit, even though i pulled from you 2011-03-22T00:42:02 hmm ok ^^ 2011-03-22T00:42:06 antimatroid: also it's easier to talk about damage and power than damage1 and damage2 2011-03-22T00:42:06 chris_0076: thanks 2011-03-22T00:42:12 let me figure out how to do that ^^ 2011-03-22T00:42:22 antimatroid: who remembers which is which 2011-03-22T00:42:38 i do :P 2011-03-22T00:42:44 hmm okay 2011-03-22T00:42:46 good for you 2011-03-22T00:42:58 chris_0076: https://github.com/chris0076/aichallenge > admin > delete 2011-03-22T00:43:05 hmm ok ^^ 2011-03-22T00:44:25 ok deleted ^^ 2011-03-22T00:45:08 chris_0076: you can fork it again if you want to, it was just getting annoying to have that commit teasing me, like i didn't pull properly or something 2011-03-22T00:45:09 lol 2011-03-22T00:45:18 lol ok ^^ 2011-03-22T00:45:28 *** avocado has quit IRC (Ping timeout: 240 seconds) 2011-03-22T00:45:39 amstan: can i get the makefile to delete the .o files when it's done? 2011-03-22T00:45:46 it looks like its meant to but doesn't :\ 2011-03-22T00:46:04 antimatroid: where? it should be doing it... 2011-03-22T00:46:22 antimatroid: works for me... which makefile? 2011-03-22T00:46:25 so the repo from here?: https://github.com/aichallenge/aichallenge 2011-03-22T00:46:41 https://github.com/aichallenge/aichallenge/raw/bd27c56cd9e391dba1476a6e2eca313ceb3ddc99/ants/dist/starter_bots/cpp/Makefile 2011-03-22T00:46:42 that one 2011-03-22T00:46:52 *** avocado_ has joined #aichallenge 2011-03-22T00:48:06 chris_0076: yep 2011-03-22T00:48:12 ok forked ^^ 2011-03-22T00:48:22 antimatroid: -rm -f ... ${OBJECTS} ... 2011-03-22T00:48:38 antimatroid: it's in there 2011-03-22T00:48:41 hmm, so even with attack radius of 2, 2 wide corridors can still be blocked? 2011-03-22T00:49:45 battle_sim.py 4 <<<"%bb%|a..a|.aa." 2011-03-22T00:50:31 so reducing it doesn't help 2011-03-22T00:51:01 hmmm yeah, you're right 2011-03-22T00:51:24 so if blocking is a problem we need to carve maps with 3x3 blocks? 2011-03-22T00:51:56 yes or at least ensure all areas are accessible with 3 wide corridors 2011-03-22T00:52:08 can do that 2011-03-22T00:52:19 just do the isvalid function with a 3x3 block 2011-03-22T00:52:33 you could make "shortcuts" with narrower corridors that could then be blocked 2011-03-22T00:52:53 that'd be harder to do right 2011-03-22T00:53:01 yep :) 2011-03-22T00:53:50 so... who doesn't prefer damage to the rest? 2011-03-22T01:01:32 you guys realize that water squares don't affect the battle? 2011-03-22T01:01:48 you keep making examples with water, which makes it hard to see the players :P 2011-03-22T01:02:35 ^^ antimatroid, janzert 2011-03-22T01:02:56 yes, I know they don't directly 2011-03-22T01:02:59 sigh: the point is to show the water blocking places an ant could be 2011-03-22T01:03:15 but I'm looking at what is possible in a corridor so... :) 2011-03-22T01:03:18 :) 2011-03-22T01:03:53 i didn't see a single comment on reddit about battle/spawning happening over water which is good 2011-03-22T01:04:07 there were a few about battle 2011-03-22T01:07:40 when I get home I'm going to change occupied to power and make damage the default 2011-03-22T01:07:49 :) 2011-03-22T01:08:12 i've not come across someone that didn't understand damage yet 2011-03-22T01:14:24 *** amstan_ has joined #aichallenge 2011-03-22T01:14:24 *** ChanServ sets mode: +o amstan_ 2011-03-22T01:36:44 *** amstan_ has quit IRC (Ping timeout: 255 seconds) 2011-03-22T01:55:05 *** phirenz has quit IRC (Ping timeout: 264 seconds) 2011-03-22T02:05:10 sigh: current spawn is sqrt(2) yeah? 2011-03-22T02:05:16 i'm updating the specs page 2011-03-22T02:05:21 yes 2011-03-22T02:05:38 just look at playgame.py for defaults 2011-03-22T02:31:50 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-03-22T02:33:34 *** ltriant has quit IRC (Quit: Computer has gone to sleep) 2011-03-22T02:36:45 *** Stocha has joined #aichallenge 2011-03-22T02:38:24 *** sigh has quit IRC (Ping timeout: 252 seconds) 2011-03-22T02:43:23 *** Stocha has quit IRC (Ping timeout: 252 seconds) 2011-03-22T03:20:41 *** avocado_ is now known as avocado 2011-03-22T03:32:39 *** sigh has joined #aichallenge 2011-03-22T03:45:27 *** Stocha has joined #aichallenge 2011-03-22T03:45:50 i'm beginning to believe bunching ant is really more fun in the end. 2011-03-22T03:46:56 If they was only two player per game, then one blockable corridor would be okay. As long as existis a path to either doors of the blockable corridor. 2011-03-22T03:47:25 that would allow to block for slowing down the opponent 2011-03-22T03:47:34 *** chris__0076 has joined #aichallenge 2011-03-22T03:47:51 anyway with more than two player it's too messy :p 2011-03-22T03:48:06 *** Stocha has quit IRC (Client Quit) 2011-03-22T03:49:03 *** chris_0076 has quit IRC (Ping timeout: 255 seconds) 2011-03-22T04:10:33 *** amstan has quit IRC (Ping timeout: 250 seconds) 2011-03-22T04:18:31 *** delt0r_ has joined #aichallenge 2011-03-22T04:20:06 *** delt0r___ has quit IRC (Ping timeout: 255 seconds) 2011-03-22T04:22:10 *** andy has joined #aichallenge 2011-03-22T04:24:46 *** Eruonen has joined #aichallenge 2011-03-22T04:25:41 *** Kingpin13 has joined #aichallenge 2011-03-22T04:28:51 *** boegel has joined #aichallenge 2011-03-22T04:30:04 *** aerique has joined #aichallenge 2011-03-22T04:49:36 *** adnzzzzZ has joined #aichallenge 2011-03-22T05:06:19 *** adnzzzzZ has quit IRC (Read error: Connection reset by peer) 2011-03-22T05:14:32 *** num1 is now known as num1_ 2011-03-22T05:24:25 *** Stocha has joined #aichallenge 2011-03-22T05:38:01 *** sda has joined #aichallenge 2011-03-22T05:40:21 *** kaemo has joined #aichallenge 2011-03-22T05:56:31 *** Stocha has quit IRC (Quit: Page closed) 2011-03-22T06:03:07 *** sda has quit IRC () 2011-03-22T06:18:09 *** Whalefreezer has joined #aichallenge 2011-03-22T06:27:49 *** needsch has joined #aichallenge 2011-03-22T06:28:52 *** Whalefreezer has quit IRC (Quit: Leaving.) 2011-03-22T06:40:43 *** kaemo has quit IRC (Remote host closed the connection) 2011-03-22T06:46:29 *** mceier has joined #aichallenge 2011-03-22T06:47:58 *** kaemo has joined #aichallenge 2011-03-22T06:58:59 *** Kingpin13 has quit IRC (Read error: Connection reset by peer) 2011-03-22T06:59:38 *** Kingpin13 has joined #aichallenge 2011-03-22T07:06:32 *** boegel has quit IRC (Quit: *poof!*) 2011-03-22T07:10:00 *** pyro- has joined #aichallenge 2011-03-22T07:13:12 *** pyro- has left #aichallenge 2011-03-22T07:13:31 *** Naktibalda has joined #aichallenge 2011-03-22T07:18:25 *** Stocha has joined #aichallenge 2011-03-22T07:30:02 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-03-22T07:30:13 aichallenge: sigh epsilon * r3dd217e / (ants/play_one_game.cmd ants/play_one_game.sh): Update play_one_game files to point to new bot locations (+5 more commits...) - http://bit.ly/ejXmKl 2011-03-22T07:44:58 sigh: did you update the resolution names? 2011-03-22T07:45:05 antimatroid: yes 2011-03-22T07:45:12 :) 2011-03-22T07:45:14 and damage is now default 2011-03-22T07:45:25 that was my implied question :P 2011-03-22T07:45:42 I also made a few minor changes to the wiki page 2011-03-22T07:46:10 specs page? 2011-03-22T07:46:17 i went through and cleaned it up a bit earlire 2011-03-22T07:46:28 the battle resolution page 2011-03-22T07:46:32 ah yep 2011-03-22T07:49:14 *** Stocha has quit IRC (Ping timeout: 252 seconds) 2011-03-22T07:49:45 *** avocado has quit IRC (Read error: Operation timed out) 2011-03-22T07:50:27 sigh: what do you think of the overview at the top of the specs page now? 2011-03-22T07:50:34 i tried to make it "simpler" 2011-03-22T07:50:51 i think "turn-based" might be potentially confusing to some people 2011-03-22T07:57:07 let me see 2011-03-22T07:59:38 why is turn-based is ambiguous 2011-03-22T07:59:51 turn-based is ambiguous 2011-03-22T07:59:58 I wasn't asking a question :P 2011-03-22T08:00:46 remove "turn-based" it doesn't add much to the description 2011-03-22T08:00:51 err.. because it might mean players make moves in turns? :P 2011-03-22T08:01:51 I say remove it 2011-03-22T08:02:06 done 2011-03-22T08:02:10 also remove the stuff about how to get started, that just cluters up the specs 2011-03-22T08:05:14 and done 2011-03-22T08:05:20 also, can we have a toc on the game? 2011-03-22T08:05:25 ^_^ 2011-03-22T08:05:39 *on the page 2011-03-22T08:06:39 not sure how easily 2011-03-22T08:07:05 i wanted one in the past then gave up when i couldn't find an easy way to add one like in a latex document 2011-03-22T08:07:39 ah, just googled... it seems it is a requested feature :P 2011-03-22T08:08:33 i imagine it'll happen at some point, it's a pretty common want 2011-03-22T08:10:09 *** Stocha has joined #aichallenge 2011-03-22T08:11:34 ah, i like dimkadimons most recent post in the forums 2011-03-22T08:12:06 he went through and did his own ranking of the outcome methods in each example based on the outcome 2011-03-22T08:13:13 aichallenge: sigh epsilon * r3527d43 / (ants/ants.py ants/play_one_game.sh): Refactor how ants/food are manipulated (second try) - http://bit.ly/hxVO8M 2011-03-22T08:15:08 i see 2011-03-22T08:16:19 pretty hard to make anything of that score though since it is just random examples with arbitrary scoring :P 2011-03-22T08:17:35 he's got two examples where he favors power over damage and not by much 2011-03-22T08:19:34 he has a point with the one i saw 2011-03-22T08:19:49 ab... 2011-03-22T08:19:49 ..c..a 2011-03-22T08:19:56 damage leaves a and b next to each other 2011-03-22T08:21:42 I tried what you just pasted and it just keeps the bottom right a alive 2011-03-22T08:23:00 it shouldn't 2011-03-22T08:23:15 i think that b should have died 2011-03-22T08:24:02 b does die when I run it 2011-03-22T08:24:20 *** dlila has joined #aichallenge 2011-03-22T08:24:40 ./battle_sim.py <<<'ab...|..c..a' 2011-03-22T08:25:17 sigh: it's not outputting that for me :\ 2011-03-22T08:25:19 i just tried again 2011-03-22T08:25:31 i get both a's and b still alive 2011-03-22T08:25:42 hmmm... paste what you are running? 2011-03-22T08:25:54 ab... 2011-03-22T08:25:54 ..c.a 2011-03-22T08:25:58 it makes sense though 2011-03-22T08:26:23 that is correct though 2011-03-22T08:27:02 a1(1/2 + 1/3), b(1/2 + 1/3), c(1/2 + 1/3 + 1), a2(1/3) 2011-03-22T08:27:21 only c dies from damage 2011-03-22T08:27:37 ok, I was typing it wrong 2011-03-22T08:27:44 ah yeah, just noticed :P sorry 2011-03-22T08:27:51 i did it 2011-03-22T08:27:53 my bad :P 2011-03-22T08:28:05 i keep doing those ..'s 2011-03-22T08:28:13 o_O 2011-03-22T08:28:19 but yeah, i think he's right that power method resolves better there 2011-03-22T08:28:48 i added a part to the comments about understandability 2011-03-22T08:29:17 i ranked support as easiest to understand, put damage and iterated closest together, then noted pretty much everyone struggles to grasp power on their first go 2011-03-22T08:30:13 hmm... I think I screwed something up on git 2011-03-22T08:30:51 ? 2011-03-22T08:33:06 ok, just a small mistake 2011-03-22T08:33:10 aichallenge: sigh epsilon * r4fbd476 / ants/battle_sim.py : Make battle_sim aware that occupied is now power - http://bit.ly/hTMyI4 2011-03-22T08:34:43 I don't think that example is that bad... next round they will kill each other off 2011-03-22T08:35:11 I don't think it hurts the strategy any 2011-03-22T08:35:21 in fact it makes it easier to cooperate :P 2011-03-22T08:35:33 the main issue should be blocking 2011-03-22T08:39:45 yeah 2011-03-22T08:39:54 i don't think blocking is that bad 2011-03-22T08:40:10 if it becomes an issue, we just make sure a 3x3 block can reach every square inside the grid, problem solved 2011-03-22T08:40:36 yeah 2011-03-22T08:41:17 antimatroid: funny thing, I explained the methods to a guy at work and his first thought was that power was the most intuitive :P 2011-03-22T08:41:28 or made the most sense 2011-03-22T08:41:32 :P 2011-03-22T08:41:54 but was that for intuitive way of explaining fighting in general and he understood them all anyway? 2011-03-22T08:42:06 i refer to intuitive in that fashion, but i suspect some others aren't 2011-03-22T08:42:48 I just explained the methods and asked him what he thought made the most sense 2011-03-22T08:43:24 sigh: i disagree with him on the last example 2011-03-22T08:43:59 i think the damage outcome is better 2011-03-22T08:44:15 disagree with dimkadimons? 2011-03-22T08:44:19 yeah 2011-03-22T08:48:55 *** robbs has joined #aichallenge 2011-03-22T08:58:56 How do the value threshold for how much damage is need to kill an ant affect the function ? 2011-03-22T08:59:21 i mean ant killed if damage >=1 is pretty arbitrary. Maybe 1 is not the best value. 2011-03-22T09:00:23 *** sigh has quit IRC (Remote host closed the connection) 2011-03-22T09:00:52 *** dlila has quit IRC (Quit: Leaving) 2011-03-22T09:01:15 *** dlila has joined #aichallenge 2011-03-22T09:02:35 *** kaemo has quit IRC (Ping timeout: 276 seconds) 2011-03-22T09:06:43 it's the obvious choice so that two ants coming across each other die 2011-03-22T09:06:51 anything else would be less intuitive i think 2011-03-22T09:08:06 Stocha: the power method can be restated in terms of damage where an ant lives if it has less damage than all of its enemies within range 2011-03-22T09:08:38 than each* might be a bit clearer 2011-03-22T09:14:40 *** Stocha has quit IRC (Ping timeout: 252 seconds) 2011-03-22T09:15:24 *** phirenz has joined #aichallenge 2011-03-22T09:29:14 aichallenge: aerique epsilon * r6e575e1 / ants/dist/starter_bots/common_lisp/bin/download-and-install-3rd-party-libs.sh : Add script for download and install of 3rd-party libs. - http://bit.ly/esKPqL 2011-03-22T09:29:18 aichallenge: aerique epsilon * r4065f37 / (2 files in 2 dirs): 2011-03-22T09:29:19 aichallenge: Add downloading 3rd party CL libs to Makefile. 2011-03-22T09:29:19 aichallenge: Add removing of downloaded and installed archive to script. - http://bit.ly/hd9x7u 2011-03-22T09:35:25 aichallenge: aerique epsilon * r63c31f8 / ants/dist/starter_bots/common_lisp/bin/download-and-install-3rd-party-libs.sh : Actually use URL variable. - http://bit.ly/fXZOSQ 2011-03-22T09:50:12 *** tvorryn has quit IRC (Quit: Page closed) 2011-03-22T09:54:41 *** FireFly has joined #aichallenge 2011-03-22T10:27:11 *** kaemo has joined #aichallenge 2011-03-22T10:27:20 *** McLeopold has joined #aichallenge 2011-03-22T10:29:46 antimatroid: you there? 2011-03-22T10:29:57 i am 2011-03-22T10:29:58 what's up? 2011-03-22T10:31:04 what if the damage method used 2 for each ant? 2011-03-22T10:31:19 then you might avoid blocking more 2011-03-22T10:31:37 and, we are using both sides of a 1:1 battle ratio 2011-03-22T10:32:06 1 ant can kill 2, but up against 3, the 3 all survive 2011-03-22T10:32:29 hmmm, that might be a good idea 2011-03-22T10:32:42 i will make some examples with sigh's thing 2011-03-22T10:32:58 McLeopold: have you seen the new battle resolutions page? 2011-03-22T10:33:02 yes 2011-03-22T10:33:04 feel free to add anything to that 2011-03-22T10:33:21 how easy is it to use picture instead of the text? 2011-03-22T10:33:51 oh, also, I think the attack radius should be reduced to 2 for the damage option 2011-03-22T10:33:53 i have no idea how to convert text to image 2011-03-22T10:34:10 McLeopold: so did janzert and i, until we realised you can still block 2 passages 2011-03-22T10:34:13 I'm thinking like what you did in the forums awhile back 2011-03-22T10:34:25 yes, but not as easily 2011-03-22T10:34:48 so spawn = sqrt(2) and attack = sqrt(4)? 2011-03-22T10:34:52 And it's nice to be able to know that you can do a focused attack at distance 2 orthogonally 2011-03-22T10:35:08 it's easier to see 2011-03-22T10:35:22 yeah i was originally for 2 2011-03-22T10:35:26 but don't mind too much either way 2011-03-22T10:35:31 if we do damage=2, then I'd go back up to 5 2011-03-22T10:35:46 well leave it at 5 for right now 2011-03-22T10:35:53 i'll play around with damage = 2 :P 2011-03-22T10:36:42 i find it hard to see damage being beaten though 2011-03-22T10:37:04 support sucks, power is too confusing and closest is less interesting 2011-03-22T10:37:32 if deadlocks become a problem, we can fix it by making sure a 3x3 block can traverse the map 2011-03-22T10:39:57 yes, let's make damage=2 the default right now 2011-03-22T10:44:18 it doesn't work when i change line 464 to 2011-03-22T10:44:18 "damage_per_enemy = Fraction(2, len(enemies))" :\ 2011-03-22T10:44:29 aab just kills the right most a and b 2011-03-22T10:46:15 that's weird 2011-03-22T10:46:31 it is 2011-03-22T10:46:36 i can't work out what's causing it 2011-03-22T10:46:40 is that happening for you? 2011-03-22T10:46:52 I'll have to try 2011-03-22T10:47:27 oh, is his overloading of score an issue? 2011-03-22T10:47:38 he has score for self and score inside the function 2011-03-22T10:47:43 and refers to both 2011-03-22T10:48:50 McLeopold: nevermind 2011-03-22T10:50:02 i had tried making my own second function, but i need to change more if i want to run both rules alongside each other like that 2011-03-22T10:50:59 works for me 2011-03-22T10:51:38 both a's die? 2011-03-22T10:52:04 yeah, mines working now 2011-03-22T10:53:39 that eliminates blocking 2 width passages 2011-03-22T10:53:58 you can still block 1 width, but only for a REALLY blocked off 1 width passage 2011-03-22T10:54:08 it should also make winning large battles without losses more difficult 2011-03-22T10:54:18 yes 2011-03-22T10:54:46 should i update the wiki to say that's what the damage option is? 2011-03-22T10:55:27 maybe another option, damage=2 and damage=1 2011-03-22T10:57:26 hmmm, i don't think sigh would like that 2011-03-22T10:57:34 i could rename it to be more general 2011-03-22T10:57:48 an ant has d damage to distribute each turn 2011-03-22T11:02:12 actually, if we take it back to 2 2011-03-22T11:02:15 there is no blocking? 2011-03-22T11:02:23 attack radius that is 2011-03-22T11:03:07 or move it up to 3 2011-03-22T11:03:54 McLeopold: a.b.c kills all three again for either of the alternate damage options though 2011-03-22T11:11:17 *** anders__ has quit IRC (Ping timeout: 252 seconds) 2011-03-22T11:13:35 aichallenge: Marco Leise epsilon * r608e6b6 / (6 files in 6 dirs): 2011-03-22T11:13:36 aichallenge: visualizer: (1) fixed additional '0' after score and ant counts when using the applet version in Chrome / (2) fixed bug reported by 2011-03-22T11:13:36 aichallenge: Stocha (buttons were created before the images are loaded) / (3) Makefile generates .jar file - http://bit.ly/f84niI 2011-03-22T11:37:15 *** kaemo has quit IRC (Ping timeout: 246 seconds) 2011-03-22T11:39:36 *** Stocha has joined #aichallenge 2011-03-22T11:41:38 *** bhasker has joined #aichallenge 2011-03-22T11:44:50 i have set up some experiment (with same setting i used before). you can view the applet, or/and the result. I posted links in the combat fighting option thread. 2011-03-22T11:45:32 To see how the X value influence the game. Though with bots not trying to get the most out of the setted X value, it's kinda pointless :p 2011-03-22T11:46:09 I have been toying with the idea, of making the X value part of the random map generation also. Then you WOULD have an AI challenge for sure i think :p 2011-03-22T11:46:33 *** Stocha has quit IRC (Quit: Page closed) 2011-03-22T11:48:07 no idea what the X value is 2011-03-22T11:48:44 antimatroid: I worry that damage=2 might make it too easy to take out 2 enemy ants with 1 of your own 2011-03-22T11:56:44 possibly 2011-03-22T11:56:55 it has flaws the more one thinks about it 2011-03-22T11:57:03 a.b.c i would really like to have just b dying htere 2011-03-22T11:57:11 janzert: can i ask a c++ question? 2011-03-22T11:57:25 you can always ask 2011-03-22T11:57:34 but I may or may not be able to answer :) 2011-03-22T11:58:42 also I probably gotta take off any minute here, as soon as someone shows up 2011-03-22T11:58:54 I have different types of games (NGame, EGame, etc.) and for each type of game pure strategies (NStrategy, EStrategy, etc.), I want to write functions for games generally (using templates, or a game class that they all inherit whatever) without passing the type of strategy to be used to the function too 2011-03-22T11:59:00 i think i solved my own question 2011-03-22T11:59:10 *** hrelo has joined #aichallenge 2011-03-22T11:59:22 good :) 2011-03-22T11:59:30 if i have all strategies inherit strategy, then i can just have a function inside game to return a new strategy? 2011-03-22T12:00:03 strategies can be thought of a lot like iterators, although i haven't actually got them defined inside NGame etc. 2011-03-22T12:00:25 yep, as long as the generic strategy contains all the information you function needs 2011-03-22T12:00:33 yep, that's the plan 2011-03-22T12:00:35 cool cheers 2011-03-22T12:00:38 :P 2011-03-22T12:01:11 btw, not uncommon to realize the answer for a programming problem just from stating it "out loud"/written out 2011-03-22T12:01:19 yeah i know 2011-03-22T12:01:25 i.e. see rubber ducky programming :) 2011-03-22T12:01:32 although i'm unsure again, how will my newstrategy function know what type to send back 2011-03-22T12:01:49 i can't have a template Game and call Game::Strategy, c++ hates it 2011-03-22T12:02:00 *** amstan has joined #aichallenge 2011-03-22T12:02:00 *** ChanServ sets mode: +o amstan 2011-03-22T12:02:31 the function has to return the same specific type as it received? 2011-03-22T12:02:50 it needs to take say an NGame and return an NStrategy 2011-03-22T12:03:40 I'd probably use a virtual factory function on game class 2011-03-22T12:04:00 what's that? :P 2011-03-22T12:04:38 basically Game.get_strategy(blah, blah, blah) that returns the correct strategy for that game 2011-03-22T12:04:50 or create_strategy() if you prefer 2011-03-22T12:05:37 sorry gotta run, see you 2011-03-22T12:05:43 cheers, bye 2011-03-22T12:10:45 *** teamsc has joined #aichallenge 2011-03-22T12:14:26 *** hrelo has quit IRC (Quit: Page closed) 2011-03-22T12:17:23 *** aerique has quit IRC (Quit: ...) 2011-03-22T12:19:07 *** delt0r___ has joined #aichallenge 2011-03-22T12:20:46 *** delt0r_ has quit IRC (Ping timeout: 255 seconds) 2011-03-22T12:21:51 aichallenge: Alexandru Stan epsilon * rbe5633c / (28 files in 6 dirs): Merged scala starter package (synesso/epsilon) - http://bit.ly/h11QmT 2011-03-22T12:32:11 *** amstan_ has joined #aichallenge 2011-03-22T12:32:11 *** ChanServ sets mode: +o amstan_ 2011-03-22T12:37:09 *** Eruonen has quit IRC () 2011-03-22T12:38:18 *** Stocha has joined #aichallenge 2011-03-22T12:38:30 im all for damage rule now. 2011-03-22T12:38:51 but i want the X value to be random for each game, and hidden from the bots :p 2011-03-22T12:39:09 Also maybe we need a bigger fighting radius 2011-03-22T12:39:28 (than radius2=5) 2011-03-22T12:39:53 I want the X=0.01 be really devastating for stacking ants :p 2011-03-22T12:41:22 lol, a hidden x from the bots 2011-03-22T12:41:24 is that a joke? 2011-03-22T12:41:53 it would suck if you can't correctly work out the outcome of potential moves 2011-03-22T12:42:46 you can 2011-03-22T12:42:54 only you have to explore a bit before. 2011-03-22T12:43:04 aichallenge: Marco Leise epsilon * rdc10fb8 / (7 files in 2 dirs): added Makefile for JavaScript part of visualizer - http://bit.ly/h0gq0q 2011-03-22T12:43:06 aichallenge: Marco Leise epsilon * r01c3dfd / (28 files in 6 dirs): Merge branch 'epsilon' of github.com:aichallenge/aichallenge into epsilon - http://bit.ly/fuMx77 2011-03-22T12:43:10 that's also what the dark fog of ware is all about. 2011-03-22T12:43:22 the fact that the map is unknown at the start. 2011-03-22T12:43:51 And, it doesn't make any changes needes. And is a real challenge to get a bot do things right in as much games as possible. 2011-03-22T12:44:58 It also eliminates most of the blocking trouble. Some game will have blocking in, and you can try to use that to your advantage. But you will need to take risk or time to find out that it may work 2011-03-22T12:45:14