2011-04-13T00:05:22 *** sigh has joined #aichallenge 2011-04-13T00:31:22 *** amstan_ has joined #aichallenge 2011-04-13T00:31:22 *** ChanServ sets mode: +o amstan_ 2011-04-13T00:33:22 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-04-13T01:40:24 *** Stocha has joined #aichallenge 2011-04-13T01:41:06 hi 2011-04-13T01:41:14 what about officially trying a fighting rule ? 2011-04-13T01:41:32 anything will do as long as it is official 2011-04-13T01:46:51 more and more people are starting to build bot. It would be good if the "we have no idea what rules could be accepted by jeff" could be lifted a bit. 2011-04-13T01:47:30 *** Ice_Harley has joined #aichallenge 2011-04-13T01:47:49 i think some people are even starting to build warlight bots. it's about time we got an idea of were ant is heading. 2011-04-13T02:03:40 *** chris___0076 has joined #aichallenge 2011-04-13T02:05:20 *** chris__0076 has quit IRC (Ping timeout: 246 seconds) 2011-04-13T02:12:28 *** Stocha has quit IRC (Quit: Page closed) 2011-04-13T02:36:47 *** chris__0076 has joined #aichallenge 2011-04-13T02:38:10 *** needsch has joined #aichallenge 2011-04-13T02:39:24 *** chris___0076 has quit IRC (Ping timeout: 276 seconds) 2011-04-13T02:45:06 *** amstan_ has quit IRC (Ping timeout: 260 seconds) 2011-04-13T02:45:19 *** chris___0076 has joined #aichallenge 2011-04-13T02:47:22 *** dr- is now known as lotus 2011-04-13T02:47:29 *** lotus is now known as dr- 2011-04-13T02:48:26 *** chris__0076 has quit IRC (Ping timeout: 248 seconds) 2011-04-13T02:53:19 *** Mathnerd314 has quit IRC (Read error: Connection reset by peer) 2011-04-13T03:02:00 *** needsch has quit IRC (Quit: Leaving.) 2011-04-13T03:12:46 *** antimatroid has quit IRC (Ping timeout: 240 seconds) 2011-04-13T03:13:22 *** Chris_0076 has joined #aichallenge 2011-04-13T03:15:38 *** chris___0076 has quit IRC (Ping timeout: 248 seconds) 2011-04-13T03:28:27 *** antimatroid has joined #aichallenge 2011-04-13T03:28:36 *** Ice_Harley has quit IRC (Ping timeout: 252 seconds) 2011-04-13T03:30:52 *** kaemo has joined #aichallenge 2011-04-13T03:30:58 *** ltriant has quit IRC (Quit: Computer has gone to sleep) 2011-04-13T03:39:32 *** berak_ has joined #aichallenge 2011-04-13T03:46:51 *** chris__0076 has joined #aichallenge 2011-04-13T03:48:36 *** Chris_0076 has quit IRC (Ping timeout: 246 seconds) 2011-04-13T03:55:38 *** woudshoo has joined #aichallenge 2011-04-13T04:04:35 *** mceier has joined #aichallenge 2011-04-13T04:04:40 *** boegel has joined #aichallenge 2011-04-13T04:21:39 *** smellyhippy has quit IRC (Quit: Take your stinking clause off me you damn dirty predicate!) 2011-04-13T04:24:40 *** Ice_Harley has joined #aichallenge 2011-04-13T04:30:19 *** smellyhippy has joined #aichallenge 2011-04-13T04:41:55 *** Naktibalda has joined #aichallenge 2011-04-13T05:04:34 *** ahmedhelal has quit IRC (Read error: Connection reset by peer) 2011-04-13T05:15:20 *** mceier has quit IRC (Quit: leaving) 2011-04-13T05:15:42 *** berak_ has quit IRC (Read error: Connection timed out) 2011-04-13T05:16:49 *** berak has joined #aichallenge 2011-04-13T05:36:35 *** berak has quit IRC (Read error: Connection timed out) 2011-04-13T05:37:50 *** berak has joined #aichallenge 2011-04-13T05:40:43 *** woudshoo` has joined #aichallenge 2011-04-13T05:43:51 *** woudshoo has quit IRC (Ping timeout: 260 seconds) 2011-04-13T05:56:53 dead channel is dead 2011-04-13T06:08:44 it's hibernation 2011-04-13T06:10:08 everybody are coding ) 2011-04-13T06:10:28 Ice_Harley: how're you going with food collection? :P 2011-04-13T06:10:43 i don't think too many people have actually starter yet, but i could be wrong 2011-04-13T06:11:16 *** needsch has joined #aichallenge 2011-04-13T06:11:55 I have an algorithm... in my head. Now I am rewriting starter package, because it's very bad 2011-04-13T06:12:20 what language? 2011-04-13T06:12:30 C# 2011-04-13T06:12:38 fewww 2011-04-13T06:14:34 *** Eruonen has joined #aichallenge 2011-04-13T06:16:46 Naktibalda: you use PHP? (Found your profile on ai-contest) 2011-04-13T06:20:08 *** fgump has joined #aichallenge 2011-04-13T06:20:09 yes 2011-04-13T06:23:06 *** sigh has quit IRC (Read error: Connection reset by peer) 2011-04-13T06:23:32 *** sigh has joined #aichallenge 2011-04-13T06:25:41 I wrote c++, glad you were talking about that 2011-04-13T06:25:49 although i expect any serious competitor to rewrite their starter bot 2011-04-13T06:25:54 wreen't* 2011-04-13T06:26:38 ... my current food collection is iteratively moving the ant that can make it to a food square fastest, accounting for ants that I expect to own in the future etc. 2011-04-13T06:28:01 heh, I have same algorithm 2011-04-13T06:28:25 I haven't it yet :) 2011-04-13T06:28:39 maybe everybody in top1000 will have that. Except of top10 ) 2011-04-13T06:31:39 *** mceier has joined #aichallenge 2011-04-13T06:33:12 Ice_Harley: I expect it to be quite popular 2011-04-13T06:33:17 it's fast enough too :) 2011-04-13T06:33:28 *** fgump has quit IRC (Ping timeout: 258 seconds) 2011-04-13T06:33:29 i haven't updated for avoiding "collisions" with paths though 2011-04-13T06:34:14 *** kaemo has quit IRC (Ping timeout: 258 seconds) 2011-04-13T06:34:59 unfortunately i don't think you'll be able to do it to the food squares you can reach first iteratively 2011-04-13T06:35:03 that'll be too costly 2011-04-13T06:36:34 *** fgump has joined #aichallenge 2011-04-13T06:41:21 *** berak has quit IRC (Read error: Connection timed out) 2011-04-13T06:47:46 *** berak has joined #aichallenge 2011-04-13T07:09:56 *** berak has quit IRC (Read error: Connection timed out) 2011-04-13T07:10:47 *** berak has joined #aichallenge 2011-04-13T07:13:09 *** namina has joined #aichallenge 2011-04-13T07:13:45 antimatroid: You know anything about time-invariant systems? 2011-04-13T07:14:07 not really, i can take a stab though 2011-04-13T07:14:32 what about them? 2011-04-13T07:14:41 ok, how do i wrok out if y[n]=x[n-2]-2x[n-10] is time-invariant 2011-04-13T07:17:35 hmmm, not really sure 2011-04-13T07:17:50 ok, no worries 2011-04-13T07:17:53 surely it'd be something like y[n+c] = ____ similar 2011-04-13T07:17:56 but not reallly sure 2011-04-13T07:18:35 you are familiar with invariance as a general concept yes? 2011-04-13T07:19:35 I guess you add something to input and output changes similarly? 2011-04-13T07:20:45 yeah, the idea is something is invariant under some kind of "transformation" if you can transform the input and leave the output unchanged 2011-04-13T07:20:53 or changed in an expected way 2011-04-13T07:20:58 i'm a bit iffy on it myself :) 2011-04-13T07:21:44 right, think i get it... 2011-04-13T07:22:20 want to do my assignemnt for me too :P 2011-04-13T07:22:33 what's it on? 2011-04-13T07:22:44 functional analysis 2011-04-13T07:22:48 ha 2011-04-13T07:22:53 no way Jose 2011-04-13T07:23:07 :p 2011-04-13T07:23:30 worth a shot :P 2011-04-13T07:28:49 *** FireFly has joined #aichallenge 2011-04-13T07:30:48 *** berak has quit IRC (Read error: Connection timed out) 2011-04-13T07:32:43 *** berak has joined #aichallenge 2011-04-13T07:35:22 *** kaemo has joined #aichallenge 2011-04-13T07:40:41 antimatroid: sorry, I was busy. What do you mean "avoiding "collisions" with paths"? 2011-04-13T07:41:45 Ice_Harley: say you iteratively move an ant to a food square, then you want to send another ant through that path, you don't want to think the shortest path will consist of having two ants in the same square at once 2011-04-13T07:42:27 that isn't as important for food collection as it is diverting ants to get close to battles 2011-04-13T07:44:06 i imagine i'll end up using a set of tuplets (turns, location, taken) to do that 2011-04-13T07:44:47 i have a 4-tuple with my food collection queue :\ 2011-04-13T07:44:59 antimatroid: when I say "iteratively" I mean one iteration - one food collect. It's not all ant steps to food, it's just fact - ant goes to food 2011-04-13T07:45:38 Ice_Harley: what would you method do here: 2011-04-13T07:45:39 *..a...* 2011-04-13T07:45:39 *..a...* 2011-04-13T07:45:50 i would send one ant left and one ant right 2011-04-13T07:47:55 antimatroid: first iteration: found closest food (left upper one for example). Send upper ant for it. Second Iteration: found closest food - left down one. Send upper ant for it. Third - closest food is right one. Send second ant for it - he is closer 2011-04-13T07:48:06 i can't think of a good general method to do the following optimally: 2011-04-13T07:48:07 b....*...a..* 2011-04-13T07:49:12 you risk doing stuff very sub optimally if you go for food closest to an enemy that you can reach first 2011-04-13T07:50:13 it's an option to test. Maybe it doesn't matter till my ant and enemy ant will die together 2011-04-13T07:52:44 http://pastebin.com/QKsT6ykF 2011-04-13T07:53:00 check that example, i think that gives a situation where i currently would get distances wrong 2011-04-13T07:53:03 *** berak has quit IRC (Read error: Connection timed out) 2011-04-13T07:53:10 by expecting two ants to be able to go in the same square at the same time 2011-04-13T07:53:49 ergh, except the bottom a is closer with spawn radius = sqrt(2) 2011-04-13T07:53:57 no, bottom is closer ) 2011-04-13T07:54:00 imagine spawn radius is 1 :P 2011-04-13T07:54:06 do you get what i mean though? 2011-04-13T07:54:43 *** berak has joined #aichallenge 2011-04-13T07:54:53 i'm not sure if the food collection algorithm can have ants cross paths? i think it can, maybe not with 2 ants 2011-04-13T07:54:55 to "eat" food bottom ant should go up-up-left - 3 steps 2011-04-13T07:55:03 but it's certainly something i want to worry about with other things 2011-04-13T07:55:16 yes, but what if spawn radius was 1? 2011-04-13T07:55:41 i'm trying to point out where i think it might give the wrong results, even if this example doesn't work with the current parameters 2011-04-13T07:56:20 due to moving the closest ant to a food square, it's not quite as obvious that it's breakable, but i'm pretty sure it is 2011-04-13T07:57:39 I think i understand you. 2011-04-13T08:01:25 i wonder why only water cells are sent to the bots (whenever the fog of war is lifted a little more). 2011-04-13T08:02:07 is that by purpose or may it be subject to change? 2011-04-13T08:04:11 it was to minimise io data 2011-04-13T08:04:17 it is very unlikely that it will change 2011-04-13T08:05:25 yeah, i wonder whether that'll pay off well. because every bot author has to worry about updating his local map accordingly. i.e. find out what cells can be assumed to be LAND/FLOOR because they should be visible by now, too. 2011-04-13T08:05:27 i was against it, so much information is now only implicitly given 2011-04-13T08:05:36 *** UncleVasya has joined #aichallenge 2011-04-13T08:06:16 but i cared more about other things, so i let it go :) 2011-04-13T08:06:20 any idea how much time the bots will be granted for their turns in the contest? 2011-04-13T08:06:30 originally people used x,y indexing, i didn't shut up about that until it changed 2011-04-13T08:06:40 no idea, we don't know what will be needed yet 2011-04-13T08:06:57 aye. i suggest 1s like for planet wars is not enough. 2011-04-13T08:07:03 but we'll see 2011-04-13T08:07:12 thanks for the info, I have to go now. bye bye. 2011-04-13T08:07:19 you were only actually guaranteed 0.5 seconds with planet wars in the worst case 2011-04-13T08:07:36 hmm, didnt know that 2011-04-13T08:07:44 two bots ran on the same cpu at the same time 2011-04-13T08:08:07 i didn't find that out till just before the final contest ran, hopefully it's not going to be like that this time, particularly when there's n-players 2011-04-13T08:08:09 anyway bye 2011-04-13T08:08:17 *** needsch has quit IRC (Quit: Leaving.) 2011-04-13T08:16:37 *** phire has quit IRC (Ping timeout: 240 seconds) 2011-04-13T08:16:49 *** berak has quit IRC (Read error: Connection timed out) 2011-04-13T08:19:12 *** berak has joined #aichallenge 2011-04-13T08:25:08 *** kaemo has quit IRC (Read error: Connection reset by peer) 2011-04-13T08:33:58 *** UncleVasya has quit IRC (Ping timeout: 260 seconds) 2011-04-13T08:56:54 *** ahmedhelal has joined #aichallenge 2011-04-13T08:57:55 *** needsch has joined #aichallenge 2011-04-13T09:22:36 *** fgump has quit IRC (Quit: Leaving) 2011-04-13T09:44:07 *** mceier has joined #aichallenge 2011-04-13T10:19:17 *** amstan_ has joined #aichallenge 2011-04-13T10:19:17 *** ChanServ sets mode: +o amstan_ 2011-04-13T10:20:37 *** sigh has quit IRC (Remote host closed the connection) 2011-04-13T10:21:50 stocha: +1 for that 2011-04-13T10:21:53 *** amstan_ is now known as amstan 2011-04-13T10:33:29 amstan: we need jeff 2011-04-13T10:35:41 antimatroid: have you talked to mcleopold about the issues you want to solve with jeff? 2011-04-13T10:35:44 can't he solve them? 2011-04-13T10:35:58 mcleopold can decide battle if you want 2011-04-13T10:36:09 but i was under the impression jeff was going to make that call 2011-04-13T10:36:26 ok.. what issues are we talking about? 2011-04-13T10:36:31 battle resolution 2011-04-13T10:36:33 i need a list 2011-04-13T10:36:37 i want mcleopold to have the right to make other decisions 2011-04-13T10:36:47 amstan: the reason things have been left hanging is that jeff took a fairly strong position on them last time he was around that was different than what everyone else though 2011-04-13T10:36:50 thought 2011-04-13T10:37:14 or you or janzert 2011-04-13T10:37:20 someone that is around often enough 2011-04-13T10:37:23 ok, so we have to get him then 2011-04-13T10:37:32 but i need to know everything that he needs to decide 2011-04-13T10:37:37 but battle is the big decision 2011-04-13T10:38:04 battle, i think we are fine to make the other decisions, although it would be good if jeff would okay mcleopold (or one of you two) to have the ruling decision 2011-04-13T10:39:21 does mcleopold still want to compete? 2011-04-13T10:39:32 i think so 2011-04-13T10:39:40 i was going to say it'd probably be better as one of you two 2011-04-13T10:39:55 yeah... some people might not want contestants to decide on rules 2011-04-13T10:40:14 if janzert was willing he'd possibly be the most suitable 2011-04-13T10:40:17 but like.. i don't want to do this since i'm still unclear about the choices 2011-04-13T10:40:30 janzert: wanna have such a role? 2011-04-13T10:40:36 antimatroid: i still need a list of what needs to be decided 2011-04-13T10:40:53 amstan: from jeff, it's really just battle, everything else is fine i think 2011-04-13T10:41:09 ok... i need to more know than just "battle" 2011-04-13T10:41:14 I don't mind but I'm going to be around much less for the next couple of months 2011-04-13T10:41:17 did you have a page explaining the options 2011-04-13T10:41:22 https://github.com/aichallenge/aichallenge/wiki/Ants-Battle-Resolution-Options 2011-04-13T10:41:31 probably still around some, but much less active 2011-04-13T10:41:53 janzert: yeah, it's really only to have someone we can talk to to make a decision if people are really at odds 2011-04-13T10:42:01 we can always pull jeff in if we can't find you 2011-04-13T10:42:36 sigh is another option 2011-04-13T10:42:39 finding someone who can do the main system administration is going to be critical for this as well 2011-04-13T10:42:40 i don't think he'll be competing 2011-04-13T10:42:52 but neither of them are options for system admin 2011-04-13T10:43:33 i can do some stuff around there, i could probably setup the contest too, but it might crash 2011-04-13T10:43:45 .. that would be bad :P 2011-04-13T10:43:48 yes 2011-04-13T10:44:08 there were some moments where janzert saved us, idk if i would do good in those situations 2011-04-13T10:44:31 yeah, we need an equivalent janzert 2011-04-13T10:44:51 we could ask reddit? 2011-04-13T10:44:56 jbroman is an equivalent janzert, but he probably doesn't want this 2011-04-13T10:45:03 antimatroid: it should be someone trusted by us 2011-04-13T10:45:10 idk if randomly going on reddit for this is a good idea 2011-04-13T10:45:15 pretty sure jbroman has already been offered the position 2011-04-13T10:45:25 Well, I just don't know enough about how epsilon works to know how hard it is to keep it running. 2011-04-13T10:45:50 it would be more a, "we're looking for someone that can do this", if we don't find anyone, no real loss 2011-04-13T10:46:03 I can certainly be around as much as time permits; I just can't promise that I'll know what's wrong when compile_a_bot_and_stuff.py fails. 2011-04-13T10:46:22 antimatroid: my point is that we might find some people that volunteer but are sketchy 2011-04-13T10:46:36 yeah fair call 2011-04-13T10:46:42 can already happen now, but not as likely 2011-04-13T10:46:46 antimatroid: I'm a little skeptical about arbitrarily asking "Hey, want root on our systems?" 2011-04-13T10:47:42 jbroman: you would potentially be able to bug janzert a little bit :P 2011-04-13T10:47:43 on reddit.. 2011-04-13T10:48:36 antimatroid: Could well be good enough. The "moderate number of people each with some time and expertise" approach works well for the CSC. 2011-04-13T10:50:15 Until May, I won't be in IRC by day (due to work), but once school starts again I'll at least be more present. 2011-04-13T10:50:33 i think most people are out during the day for their local time 2011-04-13T10:50:47 Ah right, I forgot that the definition of "day" varies. 2011-04-13T10:51:00 if i get accepted for gsoc, i'm going to be quite busy with that and uni 2011-04-13T10:54:10 *** namina has quit IRC (Ping timeout: 258 seconds) 2011-04-13T10:54:37 *** namina has joined #aichallenge 2011-04-13T10:58:26 ok.. news 2011-04-13T10:58:32 jeff is in florida 2011-04-13T10:58:35 antimatroid: ^ 2011-04-13T10:58:53 what does that mena? 2011-04-13T10:59:05 he might not be available to come over for a week 2011-04-13T10:59:39 hmmm.. that's really too long 2011-04-13T10:59:51 can you ask him to either read the page and make a decision or let someone else make the decision 2011-04-13T11:00:11 it would be good if he was at least told that almost everyone prefers damage to the other options 2011-04-13T11:00:18 and that people are able to understand it 2011-04-13T11:05:57 antimatroid: you here at 10pm, in 11 hours? 2011-04-13T11:06:21 he'll come over then 2011-04-13T11:06:32 i'll be at uni then 2011-04-13T11:06:41 ok.. suggest a better time 2011-04-13T11:06:44 +-2 hours? 2011-04-13T11:06:59 actually.. hmm midnight will probably not work for jeff 2011-04-13T11:07:11 it would need to be even later 2011-04-13T11:07:28 what about 8? in 9 hours 2011-04-13T11:08:15 how about if i send him an email? 2011-04-13T11:08:28 it's 1am, i start uni at 9am :P 2011-04-13T11:08:33 ok, do that then 2011-04-13T11:08:38 *** namina has quit IRC (Ping timeout: 240 seconds) 2011-04-13T11:08:40 well.. can't you come earlier? 2011-04-13T11:08:52 i need to finish this assignment and sleep :P 2011-04-13T11:08:58 ok 2011-04-13T11:09:02 he'll go by email then 2011-04-13T11:09:17 easy, i'll do it in a little bit 2011-04-13T11:09:18 *** namina has joined #aichallenge 2011-04-13T11:15:31 *** superflit has joined #aichallenge 2011-04-13T11:27:32 *** Ice_Harley has quit IRC (Ping timeout: 252 seconds) 2011-04-13T11:30:08 *** boegel has quit IRC (Ping timeout: 248 seconds) 2011-04-13T11:33:07 *** woudshoo` has quit IRC (Ping timeout: 246 seconds) 2011-04-13T11:35:01 *** kaemo has joined #aichallenge 2011-04-13T11:37:21 contestbot: topic add Meeting with Jeff on IRC @ http://www.timeanddate.com/worldclock/fixedtime.html?iso=20110413T22&p1=1181&ah=1 2011-04-13T11:37:21 *** contestbot changes topic to "Official Google AI Challenge: http://ai-contest.com/ || Channel Logs: http://contestbot.hypertriangle.com/ || Code Repo: http://github.com/aichallenge/aichallenge || Launch Checklist: https://github.com/aichallenge/aichallenge/issues/labels/launch || Meeting with Jeff on IRC @ http://www.timeanddate.com/worldclock/fixedtime.html?iso=20110413T22&p1=1181&ah=1" 2011-04-13T11:38:32 *** Eruonen has quit IRC (Ping timeout: 260 seconds) 2011-04-13T11:42:39 contestbot: later tell Mcleopold See topic for meeting, he might come over here. 2011-04-13T11:42:39 amstan: Yes master! 2011-04-13T11:44:12 *** amstan has quit IRC (Quit: Konversation terminated!) 2011-04-13T11:44:17 *** amstan_ has joined #aichallenge 2011-04-13T11:44:17 *** ChanServ sets mode: +o amstan_ 2011-04-13T11:45:19 *** Frontier has joined #aichallenge 2011-04-13T11:49:16 *** amstan_ has quit IRC (Ping timeout: 250 seconds) 2011-04-13T12:03:46 *** bhasker has joined #aichallenge 2011-04-13T12:08:01 *** bhasker has quit IRC (Ping timeout: 248 seconds) 2011-04-13T12:14:51 *** bhasker has joined #aichallenge 2011-04-13T12:17:52 *** berak has quit IRC (Read error: Connection timed out) 2011-04-13T12:18:53 *** berak has joined #aichallenge 2011-04-13T12:24:55 *** amstan_ has joined #aichallenge 2011-04-13T12:24:55 *** ChanServ sets mode: +o amstan_ 2011-04-13T12:28:29 *** namina has quit IRC (Ping timeout: 252 seconds) 2011-04-13T12:29:09 *** namina has joined #aichallenge 2011-04-13T12:34:19 *** namina has quit IRC () 2011-04-13T12:35:24 *** boegel has joined #aichallenge 2011-04-13T12:44:10 *** boegel has quit IRC (Quit: *poof!*) 2011-04-13T12:55:49 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 4.0/20110318052756]) 2011-04-13T12:55:54 *** iFire` has left #aichallenge 2011-04-13T13:20:08 *** bhasker has quit IRC (Quit: bhasker) 2011-04-13T13:26:08 *** kaemo has quit IRC (Read error: Operation timed out) 2011-04-13T13:29:25 *** smellyhippy has quit IRC (Read error: Connection reset by peer) 2011-04-13T13:29:43 *** smellyhippy has joined #aichallenge 2011-04-13T13:29:44 *** smellyhippy has joined #aichallenge 2011-04-13T13:39:36 *** phire has joined #aichallenge 2011-04-13T13:50:06 *** bhasker has joined #aichallenge 2011-04-13T13:58:34 *** fgump has joined #aichallenge 2011-04-13T14:50:00 *** amstan_ has quit IRC (Ping timeout: 250 seconds) 2011-04-13T14:50:08 *** helo_ has joined #aichallenge 2011-04-13T15:04:25 *** helo_ has quit IRC (Quit: Page closed) 2011-04-13T15:11:25 *** fgump has quit IRC (Ping timeout: 258 seconds) 2011-04-13T15:12:43 *** superflit has quit IRC (Quit: superflit) 2011-04-13T15:17:59 *** berak has quit IRC (Read error: Connection timed out) 2011-04-13T15:18:52 *** berak has joined #aichallenge 2011-04-13T15:28:22 *** fgump has joined #aichallenge 2011-04-13T15:35:25 *** Accoun has quit IRC () 2011-04-13T15:47:21 *** Accoun has joined #aichallenge 2011-04-13T15:52:46 *** amstan has joined #aichallenge 2011-04-13T15:52:46 *** ChanServ sets mode: +o amstan 2011-04-13T16:04:00 *** berak has quit IRC (Read error: Connection timed out) 2011-04-13T16:04:51 *** berak has joined #aichallenge 2011-04-13T16:13:55 *** fgump has quit IRC (Ping timeout: 258 seconds) 2011-04-13T16:13:58 *** Naktibalda has joined #aichallenge 2011-04-13T16:15:20 *** fgump has joined #aichallenge 2011-04-13T16:23:59 *** choas has joined #aichallenge 2011-04-13T16:49:42 *** FireFly has quit IRC (Quit: swatted to death) 2011-04-13T16:50:51 *** berak has quit IRC (Read error: Connection timed out) 2011-04-13T16:52:51 *** berak has joined #aichallenge 2011-04-13T17:18:51 *** berak has quit IRC (Remote host closed the connection) 2011-04-13T17:20:16 *** ahmedhelal has quit IRC (Ping timeout: 276 seconds) 2011-04-13T17:25:41 *** antimatroid has quit IRC (Ping timeout: 240 seconds) 2011-04-13T17:41:17 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 3.6.16/20110323143040]) 2011-04-13T17:47:22 *** Mathnerd314 has joined #aichallenge 2011-04-13T18:02:48 *** javagamer has quit IRC (Ping timeout: 260 seconds) 2011-04-13T18:06:48 *** ltriant has joined #aichallenge 2011-04-13T18:19:43 *** choas has quit IRC (Quit: leaving) 2011-04-13T18:25:37 *** Eruonen has joined #aichallenge 2011-04-13T18:28:46 *** antimatroid has joined #aichallenge 2011-04-13T18:44:03 *** Frontier has quit IRC (Quit: Leaving.) 2011-04-13T18:46:06 *** needsch1 has joined #aichallenge 2011-04-13T18:48:26 *** needsch has quit IRC (Ping timeout: 264 seconds) 2011-04-13T18:55:43 *** sigh has joined #aichallenge 2011-04-13T19:05:10 *** javagamer has joined #aichallenge 2011-04-13T19:18:18 *** needsch1 has quit IRC (Quit: Leaving.) 2011-04-13T19:28:17 *** bhasker has quit IRC (Ping timeout: 252 seconds) 2011-04-13T19:31:19 *** dlila has joined #aichallenge 2011-04-13T19:33:16 *** McLeopold has joined #aichallenge 2011-04-13T20:03:58 *** delt0r_ has joined #aichallenge 2011-04-13T20:05:52 *** delt0r___ has quit IRC (Ping timeout: 258 seconds) 2011-04-13T20:07:53 @later tell sigh As for the python3 in play_one_game.cmd, I was doing a test and accidentally pushed it. 2011-04-13T20:07:53 McLeopold: Ready to serve, my lord. 2011-04-13T20:10:43 aichallenge: McLeopold epsilon * r3454988 / ants/play_one_game.cmd : removed full path to python in play_one_game.cmd - http://bit.ly/goEzFz 2011-04-13T20:10:46 *** McLeopold has left #aichallenge 2011-04-13T20:27:57 *** drhodes has quit IRC (Read error: Operation timed out) 2011-04-13T20:41:04 *** McLeopold has joined #aichallenge 2011-04-13T21:08:43 *** drhodes_ has joined #aichallenge 2011-04-13T21:12:36 *** guest has joined #aichallenge 2011-04-13T21:25:42 *** Eruonen has quit IRC () 2011-04-13T21:27:18 *** Crivens has quit IRC (Read error: Connection reset by peer) 2011-04-13T21:57:55 *** ahmedhelal has joined #aichallenge 2011-04-13T21:59:43 *** drhodes_ is now known as drhodes 2011-04-13T22:15:20 McLeopold: ping 2011-04-13T22:15:25 hi 2011-04-13T22:15:27 janzert: sigh: ping 2011-04-13T22:15:32 meeting might near 2011-04-13T22:15:37 jeff just came to gtalk 2011-04-13T22:15:38 pong 2011-04-13T22:16:08 I'm around :) 2011-04-13T22:16:20 cool 2011-04-13T22:16:34 he's coming! woot 2011-04-13T22:16:50 jbroman: ping 2011-04-13T22:16:51 he's on beach wireless. lol 2011-04-13T22:16:52 might want to ping antimatroid too 2011-04-13T22:16:59 sigh: he's at school 2011-04-13T22:17:09 ahh 2011-04-13T22:17:32 *** j3camero has joined #aichallenge 2011-04-13T22:17:38 Yo yo 2011-04-13T22:18:05 welcome back :} 2011-04-13T22:18:58 Thanks! I am in Florida right now, using shaky beach wireless. I may cut in and out. 2011-04-13T22:20:35 unfortunately I think antimatroid and mcleopold are the two that have the most they want to run by you 2011-04-13T22:20:59 I think antimatroid fired off an email to you earlier though 2011-04-13T22:21:21 who's running this meeting? 2011-04-13T22:21:30 I'm only hanging around the edges since work is picking up here again 2011-04-13T22:21:34 So right now we're picking the battle resolution logic that we're going to open the next contest with. Everybody who's going to vote, quickly take a look at https://github.com/aichallenge/aichallenge/wiki/Ants-Battle-Resolution-Options and weigh in with which methods you like. You can vote for more than one, all of them, or none at all. Whichever has the most votes, we'll go with. 2011-04-13T22:21:45 damage 2011-04-13T22:21:50 damage 2011-04-13T22:22:09 antimatroid votes damage too :P 2011-04-13T22:22:19 damage 2011-04-13T22:24:16 Antimatroid and I are prepared to write something to eliminate small choke points in maps so that blocking doesn't become too bad 2011-04-13T22:24:20 Cool. So far I have 2011-04-13T22:24:25 janzert: damage 2011-04-13T22:24:25 sigh damage 2011-04-13T22:24:25 antimatroid: damage 2011-04-13T22:24:25 McLeopold: damage 2011-04-13T22:24:25 j3camero: iterative closest 2011-04-13T22:24:36 Anybody else voting before we make the decision? 2011-04-13T22:24:42 amstan? 2011-04-13T22:24:52 for score? 2011-04-13T22:25:01 amstan: battle 2011-04-13T22:25:04 no dying 2011-04-13T22:25:19 yeah, i'll go with damage 2011-04-13T22:25:56 Okay. We'll wait another few minutes in case anyone else wants to get on the record. Failing a large influx of people, it looks like we're going with the damage option! 2011-04-13T22:26:33 History will remember this as the most decisive five minutes in #aichallenge history. 2011-04-13T22:26:52 :) 2011-04-13T22:26:57 so, what's next, review of issues? 2011-04-13T22:26:58 lol 2011-04-13T22:27:32 Yes, for sure. 2011-04-13T22:27:39 damage 2011-04-13T22:27:41 is there anything else that we need j3camero for though first? 2011-04-13T22:27:47 McLeopold, I'm sure you probably have a bunch to say. 2011-04-13T22:27:52 has the scoring system been settled yet? 2011-04-13T22:28:00 sigh: it was never not settled 2011-04-13T22:28:10 not at all 2011-04-13T22:28:15 so the food bonus stuff is locked in? 2011-04-13T22:28:50 there was a poll on the forums about that: http://ai-contest.com/forum/viewtopic.php?f=21&t=1364 . people seem inclined to want points for killing ants and number of ants alive at the end 2011-04-13T22:28:51 j3camero: Actually, I don't have much to say about the current state of things. Except, we need someone doing a server build every so often to ensure we have a working set of code. 2011-04-13T22:28:54 Food bonus system is cool with me. If it ends up totally sucking, that should become clear soon after launch. Shouldn't be too hard to fix it. 2011-04-13T22:29:05 hmm, I meant it wasn't settled but that is just because I don't think we can know if it works or not without a fair number of games for people to go over 2011-04-13T22:29:35 cool, I was just clarifying to check if it was something we needed to go over now 2011-04-13T22:29:51 but seems we can leave it 2011-04-13T22:30:11 McLeopold: re server builds. If we set up a buildbot server or other continuous integration system, would that do the trick? 2011-04-13T22:30:13 if we do have a beta period that should give a good chance for it to shake out 2011-04-13T22:30:19 I think the current scoring is good as it is. We will need to wait until after beta to see if it needs to be tweaked. 2011-04-13T22:30:43 does the code actually function end-to-end at this point? 2011-04-13T22:30:59 j3camero: that sounds like a good idea. It should also create an entry for each starter pack and example bot and run games. 2011-04-13T22:31:03 I didn't know pairing was functional yet 2011-04-13T22:31:12 janzert: I know it doesn't right now 2011-04-13T22:31:37 pairing is the last big piece, and then fixing everything that broke with the schema change 2011-04-13T22:31:55 *** Stocha has joined #aichallenge 2011-04-13T22:32:02 yeah, I've been waiting for that to even attempt a full setup 2011-04-13T22:32:02 I have some website-php issues I'd like to hand off to someone. 2011-04-13T22:32:10 McLeopold: which one? 2011-04-13T22:32:35 McLeopold: do we know how the pairing algorithm is going to work yet? 2011-04-13T22:32:42 #53 2011-04-13T22:32:46 damage ! 2011-04-13T22:33:12 j3camero: yes, the pairing works right now, with a trueskill approximation for match quality 2011-04-13T22:33:12 McLeopold: that's php? seems like a cronjob task for me 2011-04-13T22:33:41 amstan: it is, but the leader board php files need to be updated to the new schema 2011-04-13T22:33:59 McLeopold: okay nice. I don;t know the algorithm details for that. Is it relatively feasible? 2011-04-13T22:34:01 pairing is going to happen real time to provide the best matches possible 2011-04-13T22:34:03 McLeopold: ok, i'll look over it 2011-04-13T22:34:38 pairing is done in an sql procedure https://github.com/aichallenge/aichallenge/blob/epsilon/sql/2_generate_matchup.sql 2011-04-13T22:34:39 we need someone with the power of making small choices i guess. 2011-04-13T22:34:53 so that matters are settled more quickly now 2011-04-13T22:35:12 we need a BDFC (Benevolent Dictator For the Contest) ;) 2011-04-13T22:35:23 So, the part left for pairing is updating the sigma and mu after a match completes 2011-04-13T22:35:54 which is pretty easy, I'm working on that tonight 2011-04-13T22:35:58 or BDFA (Benevolent Dictator For Ants) rather 2011-04-13T22:36:11 some expressed to me some concern about the trueskill precision and multiplayers games. He didn't give any detail though. 2011-04-13T22:36:18 someone 2011-04-13T22:36:41 Stocha: Trueskill was first used for multiplayer games 2011-04-13T22:37:05 It's a matter of fine tuning the beta, which we can't do until after our beta run of games 2011-04-13T22:37:28 He said he knew true skill, and that there would be problems. But didn't give any detail :p 2011-04-13T22:37:36 I think that having a dictator for technical matters is a good idea. I propose the following: if amstan, janzert, and McLeopold achieve consensus on a technical issue, then it is decided. Is that okay with everyone? 2011-04-13T22:37:47 yes 2011-04-13T22:37:49 yes 2011-04-13T22:38:01 it's okay to me also :p 2011-04-13T22:38:06 I'd prefer one person with final say, but... 2011-04-13T22:38:20 three persons consensus is more stable 2011-04-13T22:38:20 Then I'll always vote last :) 2011-04-13T22:38:44 McLeopold: assuming janzert and me disagree 2011-04-13T22:38:48 lol 2011-04-13T22:38:53 The triumvirate scheme that I propose is meant to eliminate myself as a decision bottleneck, which I know is annoying to people. 2011-04-13T22:38:53 :) 2011-04-13T22:39:11 no, it would be unanimous or we don't do it 2011-04-13T22:39:28 Yep. It's gotta be unanimous. 2011-04-13T22:39:32 i think it's good 2011-04-13T22:39:39 if things go wrong, you'll need j3camero 2011-04-13T22:39:40 we really, really need a sysadmin for ants 2011-04-13T22:39:45 unless no decision is a decision, then it's whatever wins 2011-04-13T22:39:53 j3camero: in the agenda we still have: finding an admin, launch date, money, reminder about issue assignments, pick one! 2011-04-13T22:40:44 amstan: what happened to the launch tag in github issues? 2011-04-13T22:41:06 McLeopold: it's a milestone now 2011-04-13T22:41:11 it's a milestone now? 2011-04-13T22:41:12 ah 2011-04-13T22:41:26 McLeopold: we should use those more often, they go late when we're late 2011-04-13T22:41:31 16 days!? 2011-04-13T22:41:33 red* 2011-04-13T22:41:39 McLeopold: so.. ignore that 2011-04-13T22:41:41 I think if matters are clear, you'll all 3 agrees. If you don't, it's better to refer to some hight authority 2011-04-13T22:41:47 higher 2011-04-13T22:41:48 McLeopold: it sounds like things are moving along nicely on the technical front. I really appreciiate that you have stepped into a leadership role in this matter. Now what we need in order to progress with picking a launch date is to see a fully working contest instance. When we're able to submit bots and watch them play (non-public server of course, so the contest doesn't get off to a false start), then we can pick a launch date. 2011-04-13T22:41:53 McLeopold: the date on it.. is impossible to remove, so ... yeah 2011-04-13T22:42:41 so, who wants to do a nightly build server? 2011-04-13T22:42:45 j3camero: we still need an ETA for that almost working state, otherwise we'll just take our time 2011-04-13T22:43:00 McLeopold: i have a server.. i could play with something like that 2011-04-13T22:43:19 we realy need an official tcp framework. if not a tcp server. 2011-04-13T22:43:33 Stocha: idk if that's important at the moment 2011-04-13T22:43:35 Stocha: not really, the community can provide it again 2011-04-13T22:43:36 amstan: ever set up buildbot before? I recomend it for nighly builds. Or even builds after every commit. 2011-04-13T22:43:37 Stocha: I think that may be too much to do before launch 2011-04-13T22:43:38 Stocha: let's get it launched first 2011-04-13T22:43:46 j3camero: nope, never tried CI before 2011-04-13T22:43:48 janzert: so far it's failed to do so. 2011-04-13T22:44:02 if it does, that that will become the official thing 2011-04-13T22:44:02 j3camero: buildbot seems cool though 2011-04-13T22:44:08 so far it has for the last two contests, we never have had an official one 2011-04-13T22:44:11 but officiality will make it go faster. 2011-04-13T22:44:30 i have heard complaintes about starter bots 2011-04-13T22:44:39 A TCP server will fall out naturally after the launch. It's not launch critical. 2011-04-13T22:44:42 namely i recall c# and c++ 2011-04-13T22:44:49 Stocha: what about them? 2011-04-13T22:45:43 i'm only say things from memory. My memory is quite bad :p 2011-04-13T22:46:09 I guess that covers the launch date iitem on the agenda. We can pick a launch date when we have a fully working contest instance that we can actually submit bots to and watch them play. 2011-04-13T22:46:10 i think i heard a memory leak problem on c++ (might be a library that the guy used rather than starter bot in fact) 2011-04-13T22:46:29 and somone said c# was badly written :p 2011-04-13T22:46:48 i think antimatroid answered that people would rewrite the starter bots anyway. 2011-04-13T22:46:51 Stocha: i don't see any c# starter pack: https://github.com/aichallenge/aichallenge/tree/epsilon/ants/dist/starter_bots 2011-04-13T22:47:05 then i may be confused i suppose 2011-04-13T22:47:11 Stocha: and i don't think c++ uses any libraries 2011-04-13T22:47:19 So we still have to talk about finding a sysadmin and finding money. I can briefly address the money point. 2011-04-13T22:47:47 might be the guy, he used a library. not the starter bot. well well, probably not big deal i suppose. 2011-04-13T22:48:29 amstan: https://github.com/aichallenge/aichallenge/issues/64 2011-04-13T22:48:30 When I get back from Florida (one week from now) I am going to incorporate a non-profit corporation asap. That will allow me to take donations from people without having to pay personal income tax on them. I will give our traditional sponsor, Google, right of first refusal. They are the sexiest sponsor! 2011-04-13T22:48:44 The "ant participating into a battle" things may need precision. if it's not been done yet :p 2011-04-13T22:49:28 Stocha: what? 2011-04-13T22:49:32 If not for google as a sponsor, that might be a problem 2011-04-13T22:49:34 To incorporate a non-profit, I believe I need at least two directors besides myself. They may have to be Canadian. Not sure. amstan, jbroman? 2011-04-13T22:49:36 j3camero: sounds good 2011-04-13T22:49:40 The scoring scheme 2011-04-13T22:49:44 j3camero: yep, i can do that 2011-04-13T22:49:51 j3camero: someone from microsoft got in the top 5 last time. Maybe you can contact high ranking bots of big corporations to see if we can get more sponsors? 2011-04-13T22:49:52 * amstan is a citizen 2011-04-13T22:49:56 it states that participating in killing an ant gives points. 2011-04-13T22:50:06 that needs clarification i think. 2011-04-13T22:50:22 Stocha: this was undecided yet 2011-04-13T22:50:27 *** antiuni has joined #aichallenge 2011-04-13T22:50:30 okay 2011-04-13T22:50:38 ... hi guys 2011-04-13T22:50:40 any news? 2011-04-13T22:50:47 antiuni: we're in a meeting 2011-04-13T22:51:14 antiuni : the damage thing has been agreed on. 2011-04-13T22:51:19 McLeopold: good idea. I actually already have contacts at Google, Facebook, Amazon, and Microsoft ready to go. So I am optimistic. Google and Facebook are explicitly interested. Amazon and Msft I haven't probed too much. 2011-04-13T22:51:38 wow, facebook? what's their angle? 2011-04-13T22:51:48 Stocha: :D 2011-04-13T22:51:49 McLeopold: they wanna cheat at farmville, lol 2011-04-13T22:51:53 jk 2011-04-13T22:51:55 haha 2011-04-13T22:52:23 Yeah I thought of that. If we ever do the Facebook AI Challenge, there can only be one game. 2011-04-13T22:52:37 agricola? 2011-04-13T22:52:43 to be fair, facebook has many other games 2011-04-13T22:52:48 Just kidding obvsiously. Farmville would be a terrible AI Challenge. 2011-04-13T22:53:17 * janzert wouldn't know, having avoided farmville at all costs ;) 2011-04-13T22:53:18 There should be some path that can be blocked in some maps though. Even if the algo ensure it always exists an alternative. Removing all blocking would be bad. 2011-04-13T22:53:32 Somebody will inevitably write a bot that figures out how to grow weed in Farmville and sell it at a profit in Gang Wars. 2011-04-13T22:54:01 A new update on our development is the introduction of issues 2.0 on github, we can now assign issues to people: https://github.com/aichallenge/aichallenge/issues/assigned/j3camero 2011-04-13T22:54:09 Stocha: I think our solution was to ensure each starting position has a non-blockable path to all other starting positions 2011-04-13T22:54:21 i have to go i guess. see you. Thanks for the great job :p 2011-04-13T22:54:28 cya 2011-04-13T22:54:37 *** Stocha has quit IRC (Quit: Page closed) 2011-04-13T22:55:28 so, if amstan gets a nightly build going, I'm good for now 2011-04-13T22:55:39 j3camero: oh right.. do you have the db for RPS? 2011-04-13T22:55:48 i was looking into archiving contests 2011-04-13T22:56:00 i had an idea for allowing old contests to run actually 2011-04-13T22:56:16 amstan: not sure. We can try to dig it up some other time. It's gotta be around somewhere. I think I ran RPS out of my personal CSC account. 2011-04-13T22:56:28 what if all old contests were just run on the "official tcp server", then we have some bots on the servers to play people on tcp if they need anyone 2011-04-13T22:56:33 j3camero: the website used to work, but now it doesn't anymore... 2011-04-13T22:56:42 j3camero: can you look around for the db, it might be gone.. 2011-04-13T22:56:54 antiuni: right now you can attempt to beat bocsimacko on tcp.zeroviz.us 2011-04-13T22:57:06 McLeopold: I want to play around with tron again :P 2011-04-13T22:57:28 antiuni: people have open sourced their bots.. you can probably play against them locally 2011-04-13T22:57:45 amstan: it's a lot more fun playing people on a tcp 2011-04-13T22:57:58 antiuni: but not a lot of people are on tcp 2011-04-13T22:58:01 j3camero: if you non-profit allows US citizens, I'm willing to join. Maybe it would help take in US donations? 2011-04-13T22:58:05 i'm referring to zeta with that idea, i stil think it's ideal to have old contest games runnable online 2011-04-13T22:58:35 antiuni: that might be possible, we just need to think hardly about it 2011-04-13T22:58:38 but yeah, not important now, only mentioned it from another comment 2011-04-13T22:59:25 McLeopold: do we have sandboxed compilation yet? 2011-04-13T22:59:28 antiuni: ideally a new bot would only play enough games to give it an accurate rank, otherwise games wouldn't run continuously. 2011-04-13T22:59:42 amstan: no... 2011-04-13T22:59:46 janzert: are these fixed yet? https://github.com/aichallenge/aichallenge/issues/13 2011-04-13T22:59:53 McLeopold: hmm? are we planning to? 2011-04-13T23:00:33 McLeopold: if you had it running on tcp, you wouldn't need to run games unless people wanted, and it'd be reasonably cheap 2011-04-13T23:00:34 McLeopold: sounds good. It will become clear when I begin the incorporation. It's definitely possible that we can haave US directors. I know for-profits can have US directors. 2011-04-13T23:00:49 amstan: yes, I don't think that would be to hard. Just some changes to worker.py 2011-04-13T23:02:30 contestbot: topic remove 4 2011-04-13T23:02:30 *** contestbot changes topic to "Official Google AI Challenge: http://ai-contest.com/ || Channel Logs: http://contestbot.hypertriangle.com/ || Code Repo: http://github.com/aichallenge/aichallenge || Meeting with Jeff on IRC @ http://www.timeanddate.com/worldclock/fixedtime.html?iso=20110413T22&p1=1181&ah=1" 2011-04-13T23:02:32 contestbot: topic remove 4 2011-04-13T23:02:32 *** contestbot changes topic to "Official Google AI Challenge: http://ai-contest.com/ || Channel Logs: http://contestbot.hypertriangle.com/ || Code Repo: http://github.com/aichallenge/aichallenge" 2011-04-13T23:03:19 contestbot: topic add Launch Checklist: https://github.com/aichallenge/aichallenge/issues?milestone=1 2011-04-13T23:03:19 *** contestbot changes topic to "Official Google AI Challenge: http://ai-contest.com/ || Channel Logs: http://contestbot.hypertriangle.com/ || Code Repo: http://github.com/aichallenge/aichallenge || Launch Checklist: https://github.com/aichallenge/aichallenge/issues?milestone=1" 2011-04-13T23:04:13 Okay so that's all wee can really say about money at this point. 2011-04-13T23:04:23 amstan: not that I know of 2011-04-13T23:04:26 ny other comments/questions about money? 2011-04-13T23:06:01 j3camero: will we have the money in time? 2011-04-13T23:06:15 j3camero: we should get it soon if we plan to launch in like a month 2011-04-13T23:07:44 janzert: can I assume website/api_server_setup.php is no longer needed? 2011-04-13T23:08:30 something along those lines may will probably be needed but what is actually there certainly isn't correct anymore 2011-04-13T23:08:54 did you create a worker setup script that calls it? 2011-04-13T23:09:51 the ec2 instances basically run a very short (just a few lines) script on startup and the api_server_setup was the way it bootstrapped to the full setup script 2011-04-13T23:10:28 that few line script has never been in the source repository since it hardcodes the api secret 2011-04-13T23:11:14 *** j3camero has quit IRC (Quit: Lost terminal) 2011-04-13T23:11:18 aww.. he left 2011-04-13T23:11:51 uh oh 2011-04-13T23:12:03 *** dlila has quit IRC (Quit: Leaving) 2011-04-13T23:12:25 *** j3camero has joined #aichallenge 2011-04-13T23:12:27 his wifi probably died 2011-04-13T23:12:31 j3camero: yay! 2011-04-13T23:12:41 j3camero: you should get screen 2011-04-13T23:14:09 *** j3camero has quit IRC (Client Quit) 2011-04-13T23:14:28 awww! 2011-04-13T23:15:05 *** j3camero has joined #aichallenge 2011-04-13T23:15:10 yay! 2011-04-13T23:15:11 Gah. Stupid beach internets. 2011-04-13T23:15:16 j3camero: screen? 2011-04-13T23:15:31 Yeah that would be a wise idea. 2011-04-13T23:15:35 I am on taurine anyways 2011-04-13T23:15:45 One more time and I will just start a screen. 2011-04-13T23:16:00 j3camero: will we have the money in time? 2011-04-13T23:16:05 j3camero: we should get it soon if we plan to launch in like a month 2011-04-13T23:19:21 *** fgump has quit IRC (Quit: Leaving) 2011-04-13T23:22:24 amstan: yeah for sure. I will get something incorporated asap so we can accept money. 2011-04-13T23:23:09 j3camero: ok, what about ETA for launch, i understand we must have the alpha working on a server first, but when should we aim for that? 2011-04-13T23:24:23 ASAP? 2011-04-13T23:24:33 and not a moment before 2011-04-13T23:24:43 janzert: that could mean 2 months 2011-04-13T23:24:49 we could do better 2011-04-13T23:25:05 by definition we can't ;) 2011-04-13T23:25:06 McLeopold: what do you think? you seemed very shocked when you saw that fake eta 2011-04-13T23:25:22 nah, we need something to shoot for 2011-04-13T23:25:24 janzert: 2 months because of meh...iness 2011-04-13T23:25:50 16 days sounds fine 2011-04-13T23:26:06 McLeopold: i really clicked blindingly on it 2011-04-13T23:27:10 I feel as though I lack the moral authority to impose a deadline, since I have done roughly zero percent of the actual work so far. If the hardest workers think that a goal date would help, then it must be self imposed. 2011-04-13T23:27:17 2 months as in for starting the contest? 2011-04-13T23:27:19 that's ages away 2011-04-13T23:27:25 you will lose peoples interest by then 2011-04-13T23:27:25 antiuni: yes, we don't want that 2011-04-13T23:27:47 McLeopold: ok.. 16 days it is.. changing the title to Beta Launch? 2011-04-13T23:27:55 yeah, beta 2011-04-13T23:29:27 Is 16 days for a working server feasible? Working as in, I can go to a URL, follow a tutorial, download astarter package, submit it, see my ranking, watch replays, etc? 2011-04-13T23:29:50 At least for a private beta, yes 2011-04-13T23:30:01 there will be bugs 2011-04-13T23:32:14 ok, i feel confident in this 2011-04-13T23:32:47 amstan: when you get nightly builds going, we are going to need a way to check if our lastest checkin broke anything 2011-04-13T23:32:52 Sweet. 2011-04-13T23:32:59 McLeopold: we need test scripts 2011-04-13T23:33:31 McLeopold: buildbot can do this. But it;s up to you to create the functional testing scripts that actually get run. 2011-04-13T23:33:49 I was afraid of that. 2011-04-13T23:34:18 McLeopold: we might be able to do this manually, the annoying work is actually getting it installed 2011-04-13T23:34:48 I think having a least a server we can all hit and play with will be hugh 2011-04-13T23:35:03 un, huge 2011-04-13T23:35:18 dang, irc has made me a horrible typer 2011-04-13T23:38:24 amstan: do you know where my exports are setup in ubuntu? 2011-04-13T23:38:27 j3camero: cool thing: http://ai-contest.com/forum/viewtopic.php?f=21&t=1361 2011-04-13T23:38:33 McLeopold: exports? 2011-04-13T23:39:04 yea, I want to change my TERM export 2011-04-13T23:40:13 McLeopold: i'm not that avid in bash, sorry 2011-04-13T23:40:20 jbroman might know 2011-04-13T23:40:53 j3camero: the creator has done a lot of work on it too, he seems excited, we could do it as our next contest 2011-04-13T23:41:18 j3camero: though.. the contest with survival+communication+hd storage might also be cool. 2011-04-13T23:43:09 Oh my. amstan that looks cool. 2011-04-13T23:43:34 j3camero: yeah, we'll just have to wait on that till next contest, this one should be fairly cool too 2011-04-13T23:43:51 I will fire him a quick email just to thank him for all that work. 2011-04-13T23:44:07 j3camero: now.. back to ants.. are we decided on the battle option, and what about scoring? 2011-04-13T23:44:32 Can somebody do a forum announcement to update the people on where we are? 2011-04-13T23:44:43 j3camero: yes, but where are we? 2011-04-13T23:44:51 Sounded to me like scoring was decided. Go with the food bonus idea. 2011-04-13T23:45:21 scoring is at least ok until beta and we see how it works in practice 2011-04-13T23:45:45 j3camero: as in the second last option here? http://ai-contest.com/forum/viewtopic.php?f=21&t=1364 it has 0 votes... 2011-04-13T23:46:14 that poll is very poorly worded 2011-04-13T23:46:23 janzert: lol, ok 2011-04-13T23:46:48 the food bonus is what janzert and I talked about to ensure the last bot standing has the best chance at getting the win 2011-04-13T23:47:09 and to provide no incentive to "farming" or drawing out a game 2011-04-13T23:47:49 McLeopold: it's whoever ate the most food, correct? 2011-04-13T23:47:56 one thing I think got missed from the discussion to implementation is the bonus should also include your own living ants at the end as well 2011-04-13T23:48:01 amstan: no 2011-04-13T23:48:03 no, whoever killed the most ants 2011-04-13T23:48:57 but the living ant part of it can be easily added if it becomes a problem 2011-04-13T23:49:34 janzert: I think just living ants of crashed bots. It is to simulate perfect play when there are no bots left assuming all current and future food gives you a point. 2011-04-13T23:50:06 if you don't count the winners living ants then there is incentive to not spawn ants if you think you're going to win 2011-04-13T23:50:38 thus potentially causing the game to go longer 2011-04-13T23:50:44 hmm... 2011-04-13T23:50:52 Okay, I gotta scram now. Big day tomorrow. I will try to get something out to the mailing list tomorrow that summarizes the main discussions had today. 2011-04-13T23:50:57 but that may not become too big of a problem in practice 2011-04-13T23:51:08 what mailing list? ;) 2011-04-13T23:51:19 yeah, I'm not on a list :) 2011-04-13T23:51:19 j3camero: i'm doing that on the forums now 2011-04-13T23:51:26 see you later, j3camero 2011-04-13T23:51:34 As for the scoring system, amstan I encourage you to make use of the new triumvirate system. I am probably going to be seriously awol over the next week due to vacation. 2011-04-13T23:51:49 j3camero: ok, cool 2011-04-13T23:51:54 j3camero: enjoy 2011-04-13T23:52:02 amstan: thanks dude. 2011-04-13T23:52:05 j3camero: could you make a committment to check in at known times once a week? 2011-04-13T23:52:16 that would be nice, yes 2011-04-13T23:52:27 we get so much done with this guidance 2011-04-13T23:53:10 doesn't even have to be specific time, just let amstan know at least 18 hours before hand ;) 2011-04-13T23:53:25 McLeopold: that is a good idea. I really owe you at least that, if not more. Why don't I say Monday evenings EST, barring unusualy circumstances? 2011-04-13T23:53:30 janzert: let's include living ants in the food bonus for now. It will only make a difference if the turns remaining is very small, we can check the logic after beta 2011-04-13T23:54:01 j3camero: sounds good, 7:00pm PST is best for me 2011-04-13T23:54:10 10:00PM for you? 2011-04-13T23:54:29 my monday night routine is very stable 2011-04-13T23:54:40 McLeopold: yep 10 PM. That works well, since that's when I'm usually winding down to go to sleep. 2011-04-13T23:56:08 McLeopold: including the ants in bonus sounds good to me, since I think it's strictly better 2011-04-13T23:56:34 I'm writing the issue right now... 2011-04-13T23:56:53 Anything i missed? http://ai-contest.com/forum/viewtopic.php?f=3&t=1347&p=8961#p8961