2011-05-31T00:00:53 no, don't bother. it is the crazy loops I use. 2011-05-31T00:01:57 which bot outputs "food bfs with X seeds"? 2011-05-31T00:16:23 aichallenge: Marco Leise epsilon * r2f73244 / (2 files in 2 dirs): visualier: another attempt at fixing the java visualizer (was it a compiler bug?) - http://bit.ly/ixk469 2011-05-31T00:20:26 *** olexs has quit IRC (Quit: Leaving.) 2011-05-31T00:34:29 *** Cyndre__ has quit IRC (Quit: Leaving) 2011-05-31T00:36:13 janzert: my coffeescript install fails because it needs user input :( 2011-05-31T00:36:30 Is there an ants server yet 2011-05-31T00:36:45 @betatest 2011-05-31T00:36:46 McLeopold: Run as fast as you can and don't look back. 2011-05-31T00:36:54 @beta 2011-05-31T00:36:55 McLeopold: beta could be http://aichallenge.hypertriangle.com/. 2011-05-31T00:37:02 that's wrong 2011-05-31T00:37:05 just see topic 2011-05-31T00:37:11 contestbot: forget beta 2011-05-31T00:37:11 amstan: Yep. 2011-05-31T00:40:00 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-05-31T00:50:36 janzert: a fresh worker install has permission issues on using jails 2011-05-31T01:06:47 amstan: didn't haskell get fixed? it looks commented out in the worker_setup 2011-05-31T01:15:50 *** Palmik has joined #aichallenge 2011-05-31T01:17:01 *** mcstar has joined #aichallenge 2011-05-31T01:19:53 *** Bhaskerh has quit IRC (Ping timeout: 244 seconds) 2011-05-31T01:20:33 *** BtbN_ has quit IRC (Remote host closed the connection) 2011-05-31T01:22:04 *** yasith has quit IRC (Ping timeout: 258 seconds) 2011-05-31T01:32:03 aichallenge: Marco Leise epsilon * rbaa2e43 / ants/visualizer/java/com/aicontest/visualizer/js/dom/XMLHttpRequest.java : visualizer: added gzip decompression to the Java visualizer's XmlHttpRequest wrapper - http://bit.ly/iRbys4 2011-05-31T01:34:17 *** yasith has joined #aichallenge 2011-05-31T01:35:32 Now that was a fun process: edit Java, goto console and compile jar, copy jar over to contest server via ssh, move jar in place as root on the contest server, go back to ide to start listening remote debugger, reload browser window, go back to ide and debug, repeat 2011-05-31T01:36:07 curse you for using gzip that compresses the files to a few tiny %. 2011-05-31T01:38:56 just to verify it works: http://aichallengebeta.hypertriangle.com/visualizer.php?game=7629&java=true&config={%22cpu%22:1} 2011-05-31T01:44:52 *** Bhaskerh has joined #aichallenge 2011-05-31T01:45:22 *** Bhaskerh has joined #aichallenge 2011-05-31T01:48:39 I think Cellular2 == McLeopold, what do you think? 2011-05-31T01:48:50 ? 2011-05-31T01:49:12 Is Cellular2 your bot? I have a replay where they behave almost identical 2011-05-31T01:49:34 I submitted GreedyBot, bhasker made it 2011-05-31T01:50:55 took some time, but you finally got out of south america, look at the ant count spike at the end *g*: http://aichallengebeta.hypertriangle.com/visualizer.php?game=7665 2011-05-31T01:58:12 *** UncleVasya has joined #aichallenge 2011-05-31T02:08:19 *** Ice_Harley has joined #aichallenge 2011-05-31T02:08:58 *** pvarga has joined #aichallenge 2011-05-31T02:09:49 *** aeter has joined #aichallenge 2011-05-31T02:41:02 *** amstan has quit IRC (Remote host closed the connection) 2011-05-31T02:46:39 aichallenge: McLeopold epsilon * r2813ae8 / (setup/worker_setup.py worker/sandbox.py worker/worker.py): reporting bot process errors to matchup - http://bit.ly/lvyUaz 2011-05-31T02:49:55 *** aeter has left #aichallenge 2011-05-31T02:50:13 wth 2011-05-31T02:50:21 why do i need java plugin? 2011-05-31T02:50:41 did you run the ant build for the visualizer? 2011-05-31T02:51:59 Frontier, why do i need a java plugin when i try to view the front page? 2011-05-31T02:52:12 i thought the visualizer was written in javascript 2011-05-31T02:52:58 oh i see 2011-05-31T02:53:03 java=true 2011-05-31T02:53:09 so whats that? 2011-05-31T03:16:20 *** ltriant has quit IRC (Quit: Computer has gone to sleep) 2011-05-31T03:21:45 *** sir_macelon has joined #aichallenge 2011-05-31T03:22:00 hi 2011-05-31T03:22:19 i see the web page is back again 2011-05-31T03:31:42 *** needsch has joined #aichallenge 2011-05-31T03:37:10 what's up? 2011-05-31T03:37:41 McLeopold: yes i did run the ant build 2011-05-31T03:38:01 Frontier: I was talking to mcstart with that comment 2011-05-31T03:39:06 Frontier: http://aichallengebeta.hypertriangle.com/visualizer.php?game=7770 2011-05-31T03:39:29 so, a bot timeout causes the ant data to be short? 2011-05-31T03:40:18 why the 2 allowable lengths in the first place? 2011-05-31T03:40:21 yes, in case you want the game to run that turn anyway even if a bot timed out, and you said you want it that way 2011-05-31T03:40:52 because otherwise the ants would have a dying animation in the last turn 2011-05-31T03:41:38 I could not distinguish a surviving ant from a dying ant. 2011-05-31T03:41:59 *** UncleVasya has quit IRC (Ping timeout: 260 seconds) 2011-05-31T03:44:24 maybe I wrote the idea behind it down in the wiki 2011-05-31T03:44:32 Frontier: if a bot times out, then there are no moves for that turn 2011-05-31T03:44:51 *** qacek has quit IRC (Quit: WeeChat 0.3.5) 2011-05-31T03:45:28 well, i think i have found another bug in the game engine 2011-05-31T03:45:40 do tell 2011-05-31T03:46:03 McLeopold: If you want that bot's ants to be displayed for that turn they need a '-' as their move. 2011-05-31T03:46:39 if a newly spawned ant gets a vision over a new tile where an enemy ant has died this turn then the engine sends something like 'd row col None' 2011-05-31T03:47:21 it crashed my bot and probably every bot would crash in such situation where it expects player id to parse integer value 2011-05-31T03:48:13 yeah my bot would throw an exception there, too 2011-05-31T03:48:36 only JavaScript would think this is 'None' == 0 2011-05-31T03:49:07 generally ant wasn't supposed to see this battle resolution as it has spawned after it 2011-05-31T03:49:08 but! I don't parse death lines, hehe. 2011-05-31T03:49:12 *** yasith has quit IRC (Ping timeout: 244 seconds) 2011-05-31T03:49:23 but the game engine doesn't take it into account 2011-05-31T03:49:45 every starter bot does that 2011-05-31T03:50:01 Frontier: php might survive it too... 2011-05-31T03:50:46 it appeared during my local testing 2011-05-31T03:51:22 after the sigh change to report own dead ants over the squares where the vision was lost 2011-05-31T03:52:07 post that on github for us 2011-05-31T03:52:21 ok, i'll do 2011-05-31T03:54:12 *** aerique has joined #aichallenge 2011-05-31T03:56:36 McLeopold: I thought of something yesterday... 2011-05-31T03:56:44 good for you 2011-05-31T03:56:53 the engine the tcp server uses should NOT have bots select their moves sequentially 2011-05-31T03:56:54 you must not have been reading reddit :[ 2011-05-31T03:57:00 games will take n times longer than necessary pretty much 2011-05-31T03:57:16 you guys might have already thought of that, just making sure :) 2011-05-31T03:57:22 yeah, we did 2011-05-31T03:57:25 :) 2011-05-31T03:57:46 ignore me then 2011-05-31T03:58:10 so, is food spawning on symmetric maps symmetric even for the starting food? 2011-05-31T03:58:26 yes 2011-05-31T03:58:33 with symmetric spawning turned on 2011-05-31T03:58:45 it wont be symmetric depending on where ants are located 2011-05-31T03:58:52 what if I have 5 bots and request 71 starting food squares? 2011-05-31T03:58:57 but if a deterministic bot played itself, the result would be a draw 2011-05-31T03:59:01 compile error messages look sweet now on the submission page, thanks 2011-05-31T03:59:13 aerique: glad you like it 2011-05-31T03:59:15 i think that's 71 starting food squares per bot 2011-05-31T03:59:23 oh 2011-05-31T03:59:38 i could be wrong, but i vaguely remember playing around with what I think you're talking about myself 2011-05-31T03:59:42 no, it still uses a percentage of land_area 2011-05-31T03:59:49 hmm okay 2011-05-31T03:59:59 we need to get that fixed... 2011-05-31T04:00:06 sigh implemented that stuff, so he's the best person to ask 2011-05-31T04:01:53 antimatroid1: you're right, which explains why so much food is at the start 2011-05-31T04:02:25 *** yasith has joined #aichallenge 2011-05-31T04:03:26 aerique: did you notice this? http://aichallengebeta.hypertriangle.com/visualizer.php?game=7779 2011-05-31T04:04:04 McLeopold: no, i haven't really checked any games. will take a look at it 2011-05-31T04:05:58 McLeopold: you want to reduce the amount of food? 2011-05-31T04:06:02 atm i think there's way too many ants 2011-05-31T04:06:03 *** berak has joined #aichallenge 2011-05-31T04:06:20 i would consider starting with way less and maybe upping the spawn rate 2011-05-31T04:06:21 antimatroid1: yes, I already implemented random food spawn rates 2011-05-31T04:06:26 :D 2011-05-31T04:06:29 which are lower 2011-05-31T04:06:37 yep 2011-05-31T04:06:45 I think far less ants will be better 2011-05-31T04:06:58 yeah, more interesting strategies 2011-05-31T04:07:04 and more interesting to watch 2011-05-31T04:07:07 I'm going to add a "starting_food_per_X_land_area" and start it at 200 2011-05-31T04:07:16 what does an (x) mean after someone's name in the visualizer? 2011-05-31T04:07:22 then round down to the nearest multiple of # of bots 2011-05-31T04:07:36 aerique: they were eliminated from the game 2011-05-31T04:07:37 @topic 2011-05-31T04:07:37 antimatroid1: Official Google AI Challenge: http://ai-contest.com/ || Channel Logs: http://contestbot.hypertriangle.com/ || Code Repo: http://github.com/aichallenge/aichallenge || Beta Test: http://aichallengebeta.hypertriangle.com/ (amstan) 2011-05-31T04:08:55 yeah, that crashed output is going to be really helpful to people 2011-05-31T04:09:38 should really cut down the complaints on the forums :) 2011-05-31T04:10:40 *** Meatkat has joined #aichallenge 2011-05-31T04:11:19 McLeopold: have you managed to convince the missus that you should participate in this challenge? 2011-05-31T04:11:31 less food will promote better fighting bots, which is good in my opinion 2011-05-31T04:11:42 aerique: I'm not planning to. :) 2011-05-31T04:11:44 food collecting algorithms aren't that fancy 2011-05-31T04:12:01 *** Naktibalda has joined #aichallenge 2011-05-31T04:12:06 She's more important than this is. 2011-05-31T04:12:47 McLeopold: well, thanks in advance for all the time you've put into it so far then 2011-05-31T04:13:06 sir_macelon: I'm glad you agree. I also want to prevent the massive explosion of dumb bots when the trod over a "minefield" of food. 2011-05-31T04:13:51 I might still try to write a bot in haskell, but I'm not going to try to win or keep up. 2011-05-31T04:15:12 sir_macelon: how are you collecting food out of interest? 2011-05-31T04:15:29 i agree though, food collection is fairly simply, the real game is going to be with battles and macro strategies 2011-05-31T04:15:33 will the starter bots come with extra tools for playing and visualizing locally? 2011-05-31T04:15:48 amstan put together a tools pack today 2011-05-31T04:15:55 it's a separate download 2011-05-31T04:15:57 not sure, it would be best to release a separate general tools package i think 2011-05-31T04:16:12 that way people know they don't need to submit everything and we don't need to package it with every starter bot 2011-05-31T04:16:13 I'm still excepting to put up language specific packs as well. 2011-05-31T04:16:44 animatroid: it is rather complicated now after so many upgrades, but i could try to explain :) 2011-05-31T04:16:56 ok, it's just so i know what to put in my starter packs readme. 2011-05-31T04:17:06 *antimatroid1* sorry for spelling ;) 2011-05-31T04:17:12 i have refined the explanation of mine pretty well, i'll start first if you like, then you can explain yours if it's different :P 2011-05-31T04:17:39 antimatroid1: are you even on the beta server? 2011-05-31T04:17:41 fine with me :) 2011-05-31T04:18:39 basically i take every food tile and assign every my ant to it 2011-05-31T04:19:02 sir_macelon: did you use the java starter or re-write? 2011-05-31T04:19:07 then sort it depending on the distance 2011-05-31T04:19:31 McLeopold: i nearly rewritten the whole starter package 2011-05-31T04:19:46 i didn't like it's architecture 2011-05-31T04:19:54 i start with every location containing an ant as available in 0 turns, everything else as not available, then i iteratively do a bfs from all food squares simultaneously until i find the ant that can get within the spawnradius of a food square in the smallest number of turns (based on available data, you need to search till like currentDist < bestDist + bestLocAvailability or something) then move that ant towards the f 2011-05-31T04:19:59 i also write some junit tests to test it 2011-05-31T04:20:22 sir_macelon: could you make the starter better then? 2011-05-31T04:20:54 the availability information gives you a pretty big boost on food collection ability 2011-05-31T04:20:54 i could, but i need to update some javadocs so it will be better understandeble for noobs 2011-05-31T04:21:21 I'm not the best a java, so a re-write would be okay with me. 2011-05-31T04:21:29 antimatroid1: you do to much calculation for the food 2011-05-31T04:21:44 no, it's not much more expensive, and well worth it 2011-05-31T04:22:07 consider 2011-05-31T04:22:08 *..a...* 2011-05-31T04:22:08 *..a...* 2011-05-31T04:22:11 i don't make the search path until it is really necessary 2011-05-31T04:22:15 would you send both a's left? i would send one of them right 2011-05-31T04:22:20 i call it lazy path finding 2011-05-31T04:22:40 i am basically using future ant locations and future spawned ants in my calculations for collecting foo 2011-05-31T04:22:53 so i can better make use of my current ants moves 2011-05-31T04:22:57 i also use it implcitily 2011-05-31T04:23:10 i'm not sure whether i should update to CL starter bot for the problem with game 7779, since it seems to be a game engine bug and there's an issue for it anyway (and the error message is pretty clear) 2011-05-31T04:23:24 it is a bug 2011-05-31T04:23:37 even if we didn't want to tell you the owner, it should still be an int 2011-05-31T04:24:02 I just like the lovely error message that tells you exactly what is happening 2011-05-31T04:24:33 me too :) 2011-05-31T04:24:46 antimatroid1: in your situation there will be four targets and for every target i assign my two ants 2011-05-31T04:25:05 then i sort the targets in the asc order 2011-05-31T04:25:18 two on the left will be on top 2011-05-31T04:25:20 *** Meatkat has quit IRC (Ping timeout: 240 seconds) 2011-05-31T04:25:29 then i take one of them 2011-05-31T04:25:44 and reserve my closest ant to it 2011-05-31T04:25:57 then for every other target i add two new ants 2011-05-31T04:26:08 it's a traveling salesman that spawns new salesmen 2011-05-31T04:26:14 McLeopold: would you still be interested in a little video showing you my CL workflow with using a proxy-bot? (I'd have to make it) 2011-05-31T04:26:24 very much yes ! 2011-05-31T04:26:25 but with the overhead of the distance to the food 2011-05-31T04:26:42 one will be the old ant and the second one will be the new, spawned ant 2011-05-31T04:27:09 McLeopold: okey, not sure when i will do it. maybe when the challenge has officially started 2011-05-31T04:27:14 then againg i sort the assigned ants to targets and it turns the newly spawned ant will get that food 2011-05-31T04:27:14 aerique: it would bring others into making cl bots, especially since mega1 won. 2011-05-31T04:27:37 do you track when these future targets will be available? 2011-05-31T04:27:46 otherwise your calculations might not work as effectively 2011-05-31T04:27:54 i just add a delay in turns 2011-05-31T04:28:27 in your example i assume spawnradius2=1 2011-05-31T04:28:51 then after first iteration i add to the three remaining tagets two ants 2011-05-31T04:29:11 has anyone had any ideas for battles? 2011-05-31T04:29:11 with turn delay 2 2011-05-31T04:29:25 i have pretty much worked out my food collection, i have almost no idea for battles 2011-05-31T04:30:00 then it is just a matter of sorting with pereference to the ants with higher delay 2011-05-31T04:30:14 i've pretty much given up on the idea of game trees for battles, too many ants 2011-05-31T04:30:18 too many branches 2011-05-31T04:30:25 antimatroid1: I have a great idea for battles that involves learning... 2011-05-31T04:30:32 i think i already developed a quite strong battle mechanism, that is quick and effective 2011-05-31T04:30:51 but i will keep it secret :) 2011-05-31T04:30:56 :P 2011-05-31T04:30:59 *** McLeopold has left #aichallenge 2011-05-31T04:32:09 i may only upload game replay, but not sure if there's an online viewer for replays? 2011-05-31T04:33:04 i tested it against Stocha's food hunter bot and it manages too win even in worst starting positions 2011-05-31T04:39:17 *** Meatkat has joined #aichallenge 2011-05-31T04:46:35 *** janzert has quit IRC (Ping timeout: 260 seconds) 2011-05-31T04:49:02 here's one replay where you can examine some neat battle moves :) 2011-05-31T04:49:03 http://pastebin.com/RPbeqdC1 2011-05-31T04:55:23 *** Meatkat has quit IRC (Ping timeout: 276 seconds) 2011-05-31T04:57:01 http://aichallengebeta.hypertriangle.com/visualizer.php?game=7805 // is it trying to talk ? look at the end. 2011-05-31T05:03:02 funny :) 2011-05-31T05:03:13 but a waste of development effort IMO 2011-05-31T05:08:20 *** Meatkat has joined #aichallenge 2011-05-31T05:21:27 *** Accoun has quit IRC () 2011-05-31T05:27:21 i heart induction 2011-05-31T05:27:40 sir_macelon: no such thing as wasted development effort 2011-05-31T05:27:43 novelty bots are awesome 2011-05-31T05:28:10 *** Meatkat has quit IRC (Ping timeout: 258 seconds) 2011-05-31T05:34:39 there were similar situations in previous challenge where some bots performed some weird activites just before the game end, when they were sure of win 2011-05-31T05:35:45 but i must admit it looks cool :) 2011-05-31T05:40:30 *** Meatkat has joined #aichallenge 2011-05-31T05:40:49 yeah, one doesn't submit a novelty bot 2011-05-31T05:40:52 they're best for tcp imo 2011-05-31T05:48:39 *** Meatkat has quit IRC (Ping timeout: 255 seconds) 2011-05-31T05:58:59 derp, nevermind 2011-05-31T05:59:30 i think i'll have 3 castes of ants 2011-05-31T06:03:42 *** Ice_Harley has quit IRC (Ping timeout: 252 seconds) 2011-05-31T06:06:04 *** smiley1983 has joined #aichallenge 2011-05-31T06:12:10 *** Ice_Harley has joined #aichallenge 2011-05-31T06:12:15 *** xQuasar has quit IRC (Ping timeout: 258 seconds) 2011-05-31T06:18:09 *** xQuasar has joined #aichallenge 2011-05-31T06:23:45 *** Accoun has joined #aichallenge 2011-05-31T06:33:49 *** Quasar has joined #aichallenge 2011-05-31T06:34:15 *** Quasar is now known as Guest90176 2011-05-31T06:35:14 *** smiley1983 has quit IRC (Read error: Operation timed out) 2011-05-31T06:35:52 *** xQuasar has quit IRC (Ping timeout: 252 seconds) 2011-05-31T06:36:18 *** smiley1983 has joined #aichallenge 2011-05-31T06:46:20 *** delt0r___ has joined #aichallenge 2011-05-31T06:47:42 *** delt0r_ has quit IRC (Ping timeout: 250 seconds) 2011-05-31T06:51:19 *** Eruonen has joined #aichallenge 2011-05-31T06:55:15 *** berak has quit IRC (Ping timeout: 255 seconds) 2011-05-31T07:15:27 *** smiley1983 has quit IRC (Ping timeout: 252 seconds) 2011-05-31T07:17:56 *** olexs has joined #aichallenge 2011-05-31T07:57:17 *** smiley1983 has joined #aichallenge 2011-05-31T08:12:52 regarding the conversation earlier about languages which cope with "d row col None" - OCaml's pattern matching handles this correctly for the starter pack. 2011-05-31T08:14:10 *** jmpespxoreax has quit IRC (Ping timeout: 246 seconds) 2011-05-31T08:18:12 yeah, perhaps i should've used that as well 2011-05-31T08:19:11 then again, we wouldn't have know about the bug :) 2011-05-31T08:19:16 known* 2011-05-31T08:21:39 *** Guest90176 has quit IRC (Ping timeout: 255 seconds) 2011-05-31T08:28:00 *** smiley1983 has quit IRC (Ping timeout: 240 seconds) 2011-05-31T08:29:04 *** smiley1983 has joined #aichallenge 2011-05-31T08:30:33 ok, i've just cretead a pull request for the rewritten java starter package 2011-05-31T08:30:57 it also handles the 'None' bug 2011-05-31T08:43:29 *** smiley1983 has quit IRC (Ping timeout: 252 seconds) 2011-05-31T08:45:34 *** smiley1983 has joined #aichallenge 2011-05-31T08:47:58 *** mcstar has quit IRC (Quit: Leaving) 2011-05-31T09:01:30 *** sigh has joined #aichallenge 2011-05-31T09:13:14 *** BtbN has joined #aichallenge 2011-05-31T09:16:12 *** amstan_ has joined #aichallenge 2011-05-31T09:16:12 *** ChanServ sets mode: +o amstan_ 2011-05-31T09:21:30 *** berak has joined #aichallenge 2011-05-31T09:26:57 *** tomh has joined #aichallenge 2011-05-31T09:30:36 *** arakyd has joined #aichallenge 2011-05-31T09:31:28 I know we talked a lot about the battle resolution, but I have my concerns now. 2011-05-31T09:33:20 *** amstan_ has quit IRC (Ping timeout: 240 seconds) 2011-05-31T09:35:10 antimatroid1: are you there? 2011-05-31T09:35:22 *** mcstar has joined #aichallenge 2011-05-31T09:40:38 sigh: or maybe you? 2011-05-31T09:40:49 I am there 2011-05-31T09:41:21 did you follow the discussion about battle resolution? 2011-05-31T09:41:28 I'm pretty out of touch with everything atm 2011-05-31T09:41:42 *** pvarga has quit IRC (Quit: pvarga) 2011-05-31T09:41:58 I know a bit about the battle resolution discussions since I implemented them 2011-05-31T09:42:27 ok, let's not talk this in public chat 2011-05-31T09:52:15 *** sigh has quit IRC (Remote host closed the connection) 2011-05-31T09:57:25 *** mceier has joined #aichallenge 2011-05-31T10:02:27 *** berak has quit IRC (Ping timeout: 255 seconds) 2011-05-31T10:06:19 *** amstan_ has joined #aichallenge 2011-05-31T10:06:20 *** ChanServ sets mode: +o amstan_ 2011-05-31T10:08:00 *** janzert has joined #aichallenge 2011-05-31T10:33:58 *** aerique has quit IRC (Quit: ...) 2011-05-31T10:44:00 *** aerique has joined #aichallenge 2011-05-31T10:54:42 *** amstan_ has quit IRC (Ping timeout: 250 seconds) 2011-05-31T10:57:24 *** amstan_ has joined #aichallenge 2011-05-31T10:57:24 *** ChanServ sets mode: +o amstan_ 2011-05-31T10:58:08 *** Ice_Harley has quit IRC (Ping timeout: 252 seconds) 2011-05-31T10:58:19 hello 2011-05-31T10:58:21 *** amstan_ is now known as amstan 2011-05-31T11:04:17 *** berak has joined #aichallenge 2011-05-31T11:05:00 *** amstan has quit IRC (Ping timeout: 255 seconds) 2011-05-31T11:06:24 *** McLeopold has joined #aichallenge 2011-05-31T11:06:30 *** bhasker has quit IRC (Quit: bhasker) 2011-05-31T11:09:59 contestbot: seen UncleVasya 2011-05-31T11:09:59 smiley1983: UncleVasya was last seen in #aichallenge 20 hours, 15 minutes, and 8 seconds ago: So he's going to wake up soon and will come here :) 2011-05-31T11:13:59 contestbot: later tell UncleVasya I've made a small update to the OCaml starter, you can get the important part of the change just by replacing this file: https://github.com/smiley1983/aichallenge/blob/epsilon/ants/dist/starter_bots/ocaml/Ants.ml 2011-05-31T11:13:59 smiley1983: I come to serve. 2011-05-31T11:18:32 aichallenge: Maciej Rucinski epsilon * ref042f9 / (9 files): java starter package rewritten - http://bit.ly/l3ZnXx 2011-05-31T11:23:11 *** smiley1983 has quit IRC (Ping timeout: 240 seconds) 2011-05-31T11:24:28 *** smiley1983 has joined #aichallenge 2011-05-31T11:33:21 *** berak has quit IRC (Ping timeout: 255 seconds) 2011-05-31T11:34:38 *** amstan has joined #aichallenge 2011-05-31T11:34:38 *** ChanServ sets mode: +o amstan 2011-05-31T11:36:38 contestbot: later tell hjax are you still working on testing the starter packages? Do you mind making an issue about it to track your progress? 2011-05-31T11:36:38 amstan: I think that worked... 2011-05-31T11:40:29 *** berak has joined #aichallenge 2011-05-31T11:40:43 *** sir_macelon has quit IRC (Quit: Page closed) 2011-05-31T11:43:06 *** smiley1983 has quit IRC (Ping timeout: 248 seconds) 2011-05-31T11:44:01 *** smiley1983 has joined #aichallenge 2011-05-31T11:47:10 *** Bhaskerh has quit IRC (Ping timeout: 260 seconds) 2011-05-31T12:10:53 *** smiley1983 has quit IRC (Ping timeout: 276 seconds) 2011-05-31T12:11:33 *** smiley1983 has joined #aichallenge 2011-05-31T12:15:17 aichallenge: Alexandru Stan epsilon * r47af707 / (6 files in 2 dirs): deleted old images - http://bit.ly/ll741N 2011-05-31T12:15:17 aichallenge: Alexandru Stan epsilon * r5040abe / website/using_the_tools.php : redone using the tools, it's now in the wiki - http://bit.ly/l5CzMu 2011-05-31T12:23:06 aichallenge: Alexandru Stan epsilon * rb6ca500 / website/using_the_tools.php : redone using the tools, it's now in the wiki - http://bit.ly/kQDQ1a 2011-05-31T12:24:21 can someone review this: http://aichallengebeta.hypertriangle.com/using_the_tools.php ? 2011-05-31T12:24:22 *** UncleVasya has joined #aichallenge 2011-05-31T12:27:16 *** amstan has quit IRC (Quit: lunch) 2011-05-31T12:32:39 @later tell amstan With Java, it is not better to specify the class name rather than filename? 2011-05-31T12:32:39 jbroman: I think that worked... 2011-05-31T12:37:05 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 4.0.1/20110413222027]) 2011-05-31T12:40:17 *** anagri has joined #aichallenge 2011-05-31T12:40:23 hello all 2011-05-31T12:40:38 \help 2011-05-31T12:40:58 anybody active ? UK tz ? 2011-05-31T12:44:27 *** _0x47 <_0x47!~slash@p57B3EAC0.dip.t-dialin.net> has quit IRC (Ping timeout: 255 seconds) 2011-05-31T12:45:41 anagri: What is your problem? 2011-05-31T12:48:00 aichallenge: Scott Hamilton epsilon * r511c988 / ants/dist/starter_bots/java/MyBot.java : 2011-05-31T12:48:00 aichallenge: Merge pull request #135 from sir-macelon/patch-1 2011-05-31T12:48:00 aichallenge: Small fixes - http://bit.ly/lEth87 2011-05-31T12:56:34 *** anagri has quit IRC (Ping timeout: 252 seconds) 2011-05-31T12:57:36 anagri: hello and welcome :) 2011-05-31T12:57:45 oops 2011-05-31T12:59:45 *** _0x47 <_0x47!~slash@p5B12E0E0.dip.t-dialin.net> has joined #aichallenge 2011-05-31T13:00:56 *** aeter has joined #aichallenge 2011-05-31T13:04:22 UncleVasya: contestbot probably told you, but I've updated the OCaml starter 2011-05-31T13:05:01 No, he didn't. I 2011-05-31T13:05:21 I'm going to download it now, thanx/ 2011-05-31T13:05:41 it hasn't been merged into the main repo yet 2011-05-31T13:06:30 here's the link to my fork: https://github.com/smiley1983/aichallenge/tree/epsilon/ants/dist/starter_bots/ocaml 2011-05-31T13:06:45 the only important update is in Ants.ml 2011-05-31T13:07:51 I'm hoping this should be the last change to that starter, but you never know :) 2011-05-31T13:08:32 btw, I was pawned by OCaml totally :( I can't even do the first easest thing i want to. 2011-05-31T13:09:04 aichallenge: McLeopold epsilon * r7307cd9 / ants/ants.py : fix for dead ant line with 'None'. Closes #132 - http://bit.ly/iuIKdG 2011-05-31T13:09:04 aichallenge: McLeopold epsilon * r0651c0f / ants/dist/starter_bots/java/MyBot.java : Merge branch 'epsilon' of https://github.com/aichallenge/aichallenge into epsilon - http://bit.ly/jQBEet 2011-05-31T13:10:00 it's a bit unfamiliar at first, the way it expresses some things 2011-05-31T13:10:00 I may be able to offer advice if you're stuck 2011-05-31T13:12:06 I improved tile type a little: type tile = [ `Water | `Land | `Food | `Ant | `Dead | `Unseen |`Busy];; As you see, added `Busy 2011-05-31T13:13:17 Busy means that someone of my ants already going there. 2011-05-31T13:14:01 I added it to tile_of_int and int_of_tile functions (by number 500) 2011-05-31T13:14:10 yes 2011-05-31T13:14:35 I can't write function set_tile_busy 2011-05-31T13:14:54 Or I can't call it properly. 2011-05-31T13:15:32 Here is it: 2011-05-31T13:15:51 let set_tile_busy row col = 2011-05-31T13:15:51 gstate.tmap.(row).(col) <- 2011-05-31T13:15:52 {gstate.tmap.(row).(col) with content = (int_of_tile `Busy)}; 2011-05-31T13:15:52 ;; 2011-05-31T13:16:58 I'm not sure I can past code directly here, maybe it's agains rules. Can I use private mesages for that? 2011-05-31T13:17:34 *** smiley1983 has quit IRC (Remote host closed the connection) 2011-05-31T13:17:50 it is considered a good thing to use pastebin.com(for example) instead of pasting source code directly here 2011-05-31T13:17:54 *** smiley1983 has joined #aichallenge 2011-05-31T13:19:12 smiley1983: So can I talk to you in private? 2011-05-31T13:22:39 mcstar: I need only past 4-5 strings few times. 2011-05-31T13:23:14 *** smiley1983 has quit IRC (Ping timeout: 240 seconds) 2011-05-31T13:23:30 im not an authority, and i really dont mind, im just saying its customary to do so in irc channels 2011-05-31T13:24:46 mcstar: But I can merge them and show few parts of my code in one page. Didn't thought about that until read your message. It can be even better. 2011-05-31T13:25:35 + its gives you syntax highlighting(for popular languages at least) 2011-05-31T13:25:41 -s 2011-05-31T13:27:37 im going offline 2011-05-31T13:27:46 bye 2011-05-31T13:27:59 thunderbolts are sooo loud it makes me deaf 2011-05-31T13:28:05 *** mcstar has quit IRC (Quit: WeeChat 0.3.5) 2011-05-31T13:32:41 *** Bhaskerh has joined #aichallenge 2011-05-31T13:37:18 *** Bhaskerh has quit IRC (Ping timeout: 258 seconds) 2011-05-31T13:41:54 *** bhasker has joined #aichallenge 2011-05-31T13:42:07 morning folks 2011-05-31T13:43:48 hi 2011-05-31T13:44:33 *** olexs1 has joined #aichallenge 2011-05-31T13:45:39 *** olexs has quit IRC (Ping timeout: 255 seconds) 2011-05-31T13:46:25 games per minute has risen on the beta server, but skill calculation seems broken? something with the workers? 2011-05-31T13:50:17 *** smiley1983 has joined #aichallenge 2011-05-31T13:51:04 *** yasith has quit IRC (Ping timeout: 250 seconds) 2011-05-31T13:53:47 aichallenge: McLeopold epsilon * r8614bd3 / (5 files in 4 dirs): pull game options from server in api_get_task.php - http://bit.ly/jEAx2d 2011-05-31T13:58:40 *** needsch has quit IRC (Quit: Leaving.) 2011-05-31T13:58:43 lots of people testing out the bots on beta server now, should be interesting on the day the challenge actually goes live 2011-05-31T14:00:45 beta server skill calculation seems broken 2011-05-31T14:02:55 aichallenge: McLeopold epsilon * r29dded8 / (3 files in 2 dirs): implented random starting food amount >:) - http://bit.ly/lD44qh 2011-05-31T14:04:15 *** smiley1983 has quit IRC (Ping timeout: 260 seconds) 2011-05-31T14:04:36 *** yasith has joined #aichallenge 2011-05-31T14:06:48 aichallenge: McLeopold epsilon * rbf6efe3 / ants/ants.py : fixed food rate option bugs - http://bit.ly/kK9Z94 2011-05-31T14:06:58 uploaded a new version with slightly smarter combat 2011-05-31T14:07:01 lets see how it does 2011-05-31T14:07:19 *** UncleVasya has quit IRC (Ping timeout: 260 seconds) 2011-05-31T14:13:22 hah a lot of bots oscillate in place when food is scarce including mine 2011-05-31T14:13:41 shows that most people are restricting searches to nearby visible food and exploring code isn't well written 2011-05-31T14:14:01 mcleopold trueskill broken? 2011-05-31T14:14:03 *** mcstar has joined #aichallenge 2011-05-31T14:14:08 all skill updates showing 0 2011-05-31T14:14:48 *** smiley1983 has joined #aichallenge 2011-05-31T14:15:23 *** UncleVasya has joined #aichallenge 2011-05-31T14:15:51 rage 2011-05-31T14:16:40 smiley1983: http://pastebin.com/LV5BwKrP 2011-05-31T14:17:03 McLeopold: looping over 10000 elements of the map(100x100) and accessing the visible squares takes 0.33 seconds 2011-05-31T14:17:13 does that sound reasonable? 2011-05-31T14:17:36 i wonder where is this in relation to python 2011-05-31T14:19:18 thats 292 acces per square 2011-05-31T14:19:21 s 2011-05-31T14:19:36 accesses? 2011-05-31T14:27:14 antimatroid whats with the rankings/skill updates being broken? 2011-05-31T14:27:31 there's not that much food any more in the latest games 2011-05-31T14:28:14 *** bhasker has quit IRC (Quit: bhasker) 2011-05-31T14:28:29 UncleVasya: have you deleted or hidden Ants.mli? 2011-05-31T14:28:46 If not, try doing that 2011-05-31T14:33:03 UncleVasya: if that doesn't work, pastebin the compile error and I'll look at it tomorrow - I'm off to sleep now. Best of luck 2011-05-31T14:33:34 bye 2011-05-31T14:34:09 Oh, last thing before I go: correct syntax for the method would be: 2011-05-31T14:35:15 method set_tile_busy row col = set_tile_busy state row col 2011-05-31T14:35:18 I think 2011-05-31T14:35:52 *** smiley1983 has quit IRC (Quit: leaving) 2011-05-31T14:36:16 I hope we'll see my second submission tomorrow :) 2011-05-31T14:38:36 *** InsaneMalkavian has joined #aichallenge 2011-05-31T14:41:23 aichallenge: McLeopold epsilon * r78d1cc1 / manager/manager.py : fixed bug where manager doesn't update trueskill if log file can't be written to - http://bit.ly/kSSJal 2011-05-31T14:46:23 *** delt0r_ has joined #aichallenge 2011-05-31T14:47:50 *** delt0r___ has quit IRC (Ping timeout: 258 seconds) 2011-05-31T14:52:20 McLeopold: are you there? 2011-05-31T14:53:56 *** InsaneMalkavian_ has joined #aichallenge 2011-05-31T14:57:34 *** InsaneMalkavian has quit IRC (Ping timeout: 252 seconds) 2011-05-31T15:00:19 mcstar: what's up? 2011-05-31T15:00:30 hey 2011-05-31T15:00:35 have u read my question? 2011-05-31T15:00:44 couple of lines above 2011-05-31T15:01:19 remember, some time ago i asked how did you implement the distance-checks 2011-05-31T15:01:19 I think I'd have to see the code 2011-05-31T15:01:35 *** needsch has joined #aichallenge 2011-05-31T15:01:56 its in cl, and its for me :) 2011-05-31T15:02:40 do you access all indices of the board/map for a given ant that it can see? 2011-05-31T15:02:45 sry 2011-05-31T15:02:49 all at once 2011-05-31T15:03:11 *** yasith has quit IRC (Ping timeout: 240 seconds) 2011-05-31T15:03:12 how fast is that part, can you tell me? 2011-05-31T15:03:42 i have a better estimate(the prev. value i gave doesnt include the real array accesses) 2011-05-31T15:04:24 for a 9072 sized map it takes 0.43 seconds 2011-05-31T15:05:41 I haven't been in the bot code for awhile, sry. 2011-05-31T15:05:53 and do you precompute the indices to be accessed, or by an inequality on-the-fly? 2011-05-31T15:05:57 oh 2011-05-31T15:06:06 so you dont have an estimate?: 2011-05-31T15:06:08 I actually haven't tried to write a real bot yet. 2011-05-31T15:06:14 No, not really. 2011-05-31T15:06:21 im asking about the engine 2011-05-31T15:06:31 the bot doesnt need to do this 2011-05-31T15:06:36 sigh did the benchmarks on that 2011-05-31T15:06:52 and wrote the faster code 2011-05-31T15:07:06 ok ill try him 2011-05-31T15:07:58 mcstar: don't know if it will help you find what you want, but there is a script in the util directory under ants that compares various vision methods 2011-05-31T15:08:01 but does the engine occupy the full cpu btw? 2011-05-31T15:08:26 mcstar: not while the bots are doing moves 2011-05-31T15:08:51 the timing doesn't start until the engine waits for bot input, and we do a sleep to ensure we don't take time away from bots 2011-05-31T15:09:07 janzert: i wasnt going to look at the code, its easier to ask, and i wouldnt really benefit from seeing that code, its more like curiosity 2011-05-31T15:09:35 *** FireFly has joined #aichallenge 2011-05-31T15:09:49 mcstar: you're going to need every trick you can think of to speed up vision code if you need it 2011-05-31T15:10:08 This whole game is going to be too much to process in 500ms 2011-05-31T15:10:55 well, i was a bit concerned, and thats why i asked too, how does mine compare to Your production code :) 2011-05-31T15:11:32 for the moment this seems fast to me 2011-05-31T15:11:58 and you want to learn cl right? 2011-05-31T15:12:10 maybe when my engine is ready ill show it to you 2011-05-31T15:12:32 (maybe as in if youre interested) 2011-05-31T15:13:45 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-05-31T15:16:40 *** bhasker has joined #aichallenge 2011-05-31T15:20:28 the fastest python method for vision using a block of 10 x 10 ants spaced in a grid with 10 spaces in between each ant and a squared visionradius of 96 takes 5.63 milliseconds, but the engine doesn't calculate vision from scratch it uses incremental updates 2011-05-31T15:26:22 ok, i look up 272 coordinates for a 108x84(9072) map in 0.43 seconds 2011-05-31T15:26:53 that means 272 lookups for every square of the map 2011-05-31T15:28:03 *** bhasker has quit IRC (Quit: bhasker) 2011-05-31T15:29:21 *** Redgis has joined #aichallenge 2011-05-31T15:29:47 roughly 4.7 ms vs. 5.63 if i understand correctly 2011-05-31T15:30:53 janzert: did you run this? 2011-05-31T15:31:05 i guess your pc is faster than mine though 2011-05-31T15:34:04 run "vision_compare.py 10 10" in the util directory to get the timing youself 2011-05-31T15:34:20 janzert: did you see my issue with coffeescript? 2011-05-31T15:34:48 McLeopold: not that I recall offhand 2011-05-31T15:35:13 the first command to curl an install script fails because the script excepts user input 2011-05-31T15:36:04 so we can't use that script? 2011-05-31T15:36:28 I thought maybe redirecting 'yes' to the script would work, but it didn't 2011-05-31T15:36:32 I'm not sure 2011-05-31T15:38:52 aichallenge: janzert epsilon * r41edecc / (worker/compiler.py worker/sandbox.py): Have submission compilation use a sandbox - http://bit.ly/ive60E 2011-05-31T15:40:02 *** Accoun has quit IRC () 2011-05-31T15:40:07 aichallenge: McLeopold epsilon * r9c7d843 / ants/ants.py : fixed kill_player mistake - http://bit.ly/jamLcT 2011-05-31T15:41:10 aichallenge: McLeopold epsilon * r84301d5 / ants/ants.py : fixed print - http://bit.ly/kEyPGW 2011-05-31T15:42:31 *** bhasker has joined #aichallenge 2011-05-31T15:42:47 janzert: also, didn't amstan say haskell got fixed? 2011-05-31T15:43:23 he did say that, but then when I went to try it I don't see it being fixed and he couldn't find it as fixed eitehr 2011-05-31T15:43:25 either 2011-05-31T15:43:38 so not sure how it worked for him the one time 2011-05-31T15:44:32 oh, I do remember now what happened 2011-05-31T15:44:44 he only installed GHC not the full haskell-platform 2011-05-31T15:45:00 and it's the platform package that is broken 2011-05-31T15:45:29 *** hjax has joined #aichallenge 2011-05-31T15:45:39 there was some debate whether just having ghc and not the platform would be good enough for the contest 2011-05-31T15:45:53 I never heard whether that was true or not 2011-05-31T15:47:07 contestbot: later tell amstan I had school today, im starting testing now, ill make an issue for easier testing 2011-05-31T15:47:07 hjax: Ready to serve. 2011-05-31T15:47:12 11.29 ms for me the fastest 2011-05-31T15:47:35 janzert: can we get by with just ghc? 2011-05-31T15:48:29 that's what I don't know 2011-05-31T15:48:49 jmcarthur: ping 2011-05-31T15:48:50 I assume jmcarthur is best to answer that 2011-05-31T15:48:56 :) 2011-05-31T15:50:26 janzert: what kind of cpu u got? 2011-05-31T15:51:26 i5 750 2011-05-31T15:51:46 http://aichallengebeta.hypertriangle.com/visualizer.php?game=8468 // what happened to the food spawning ? 2011-05-31T15:51:57 although those timings above were also run inside a vm so maybe slower than expected from the cpu I have 2011-05-31T15:52:11 ok, nice one, i ride an old amd x2 2011-05-31T15:53:46 berak: for one it became symmetric it seems which isnt half bad 2011-05-31T15:56:25 i'd even agree on sym vs asym, but most of the time, there's neither food, or opponents visible to each player in this game 2011-05-31T15:57:40 there was talk that the option specified for spawing food controlled the initial amount by ANTs, it was misunderstood before, and now its correctly lower, also, migh be too low, nevertheless correct me if im wrong 2011-05-31T15:58:25 McLeopold, janzert: if you can get ghc and cabal-install to work, then you can use cabal-install to install the rest of the platform. it requires building the libraries from source, though. most of them don't take too long, as long as you avoid the opengl stuff 2011-05-31T15:58:42 cabal-install handles the downloading and building. it's like a package manager for haskell stuff 2011-05-31T16:01:03 McLeopold you here? 2011-05-31T16:05:13 *** Accoun has joined #aichallenge 2011-05-31T16:08:33 *** hjax has quit IRC (Quit: Page closed) 2011-05-31T16:13:34 well, since there's less food, increase the viewradius ? 2011-05-31T16:15:52 *** _0x47 <_0x47!~slash@p5B12E0E0.dip.t-dialin.net> has quit IRC (Remote host closed the connection) 2011-05-31T16:16:04 berak: try exploring 2011-05-31T16:16:14 *** McLeopold has left #aichallenge 2011-05-31T16:16:14 *** _0x47 <_0x47!~slash@p5B12E0E0.dip.t-dialin.net> has joined #aichallenge 2011-05-31T16:16:18 tia, ok. 2011-05-31T16:17:15 *** McLeopold has joined #aichallenge 2011-05-31T16:27:31 *** berak has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2011-05-31T16:28:03 *** aeter has left #aichallenge 2011-05-31T16:30:22 aichallenge: janzert epsilon * r17f9db8 / setup/worker_setup.py : Fix coffeescript install - http://bit.ly/mBaDFu 2011-05-31T16:31:54 *** UncleVasya has quit IRC () 2011-05-31T16:40:50 *** Apophis_ has joined #aichallenge 2011-05-31T16:49:56 aichallenge: Scott Hamilton epsilon * rd3e2a31 / (2 files): 2011-05-31T16:49:56 aichallenge: Merge pull request #134 from smiley1983/epsilon 2011-05-31T16:49:56 aichallenge: OCaml starter: fixed warnings, fixed bad input processing, updated comments - http://bit.ly/lvCp5D 2011-05-31T16:54:04 *** UncleVasya has joined #aichallenge 2011-05-31T16:54:43 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-05-31T16:58:35 *** hjax has joined #aichallenge 2011-05-31T16:59:15 *** mceier has quit IRC (Quit: leaving) 2011-05-31T16:59:21 whoa i just booted into ubuntu and playgame wont run at all 2011-05-31T16:59:42 hjax: have a pastebin? 2011-05-31T17:00:48 here http://pastebin.com/BP8WhuF7 2011-05-31T17:02:31 hjax: that looks pretty bad, probably my fault 2011-05-31T17:02:58 *** InsaneMalkavian_ has quit IRC (Ping timeout: 252 seconds) 2011-05-31T17:03:13 hjax: pull the latest repo and see if it still happens 2011-05-31T17:03:20 i downloaded that a few days ago though, im going to get a fresh copy 2011-05-31T17:04:05 but it was running on windows just fine :/, i copied it onto my flash drive and brought it to ubuntu 2011-05-31T17:08:12 hjax didnt u have this problem a couple of days ago with python 2.6, and 2.7 solved it? 2011-05-31T17:09:19 *** UncleVasya has quit IRC (Ping timeout: 260 seconds) 2011-05-31T17:10:28 *** Palmik has quit IRC (Read error: Connection reset by peer) 2011-05-31T17:12:25 i had this problem with 2.5 2011-05-31T17:12:36 this code should work on 2.6 and 2.7 2011-05-31T17:14:14 oh yes that was 2.6ű5 2011-05-31T17:14:16 shat 2011-05-31T17:14:18 2.5 2011-05-31T17:14:31 contestbot: later tell amstan couldnt start testing, running games is broken for python 2.6, which is version for ubuntu, i cant run games 2011-05-31T17:14:31 hjax: Ready to serve. 2011-05-31T17:14:52 *** hjax has quit IRC (Quit: Page closed) 2011-05-31T17:16:19 *** InsaneMalkavian has joined #aichallenge 2011-05-31T17:18:06 *** jmpespxoreax has joined #aichallenge 2011-05-31T17:20:43 *** UncleVasya has joined #aichallenge 2011-05-31T17:21:03 hjax: play_one_game.cmd works with python 2.6.6 on windows 2011-05-31T17:21:10 *** UncleVasya has joined #aichallenge 2011-05-31T17:22:31 *** bhasker_ has joined #aichallenge 2011-05-31T17:22:33 *** bhasker has quit IRC (Ping timeout: 255 seconds) 2011-05-31T17:32:44 aichallenge: janzert epsilon * r48de724 / setup/worker_setup.py : Use PPAs where we can - http://bit.ly/is2Eoe 2011-05-31T17:36:11 aichallenge: janzert epsilon * r31faac7 / setup/worker_setup.py : Correct function docstring - http://bit.ly/ji0TQd 2011-05-31T17:50:41 *** UncleVasya has quit IRC (Read error: Connection reset by peer) 2011-05-31T17:58:57 *** amstan has joined #aichallenge 2011-05-31T17:58:57 *** ChanServ sets mode: +o amstan 2011-05-31T17:59:53 contestbot: later tell hjax submit them to the site instead, we should only need to test locally if they don't work there 2011-05-31T17:59:53 amstan: Ready to serve, my lord. 2011-05-31T18:06:58 *** ltriant has joined #aichallenge 2011-05-31T18:10:05 *** hjax has joined #aichallenge 2011-05-31T18:10:26 ok im here 2011-05-31T18:10:31 amstan, you here? 2011-05-31T18:13:45 *** bhasker_ has quit IRC (Quit: bhasker_) 2011-05-31T18:14:46 *** bhasker has joined #aichallenge 2011-05-31T18:23:02 *** Redgis has quit IRC (Quit: ... mains libres) 2011-05-31T18:23:57 afternoon 2011-05-31T18:25:57 evening 2011-05-31T18:27:26 how goes it 2011-05-31T18:27:34 looks like skill updates are broken again 2011-05-31T18:27:48 those are weird, it's like they lag behind the games 2011-05-31T18:27:58 by 2 minutes 2011-05-31T18:28:01 ah 2011-05-31T18:28:02 after a while finished games get skill points 2011-05-31T18:28:08 ah, ok :) 2011-05-31T18:28:45 however, I now have some games much older than 2 minutes, where I still have 0 skill entered 2011-05-31T18:29:24 *** hjax has quit IRC (Quit: Page closed) 2011-05-31T18:32:03 yup 2011-05-31T18:32:05 same here 2011-05-31T18:32:30 last 1 hr 2011-05-31T18:32:32 no updates 2011-05-31T18:32:40 have like 20 + games with 0.00 2011-05-31T18:33:14 now it will be nice if my bot would just beat sir_macelon consistently 2011-05-31T18:33:26 though it seems to be doing quite well otherwise 2011-05-31T18:41:53 *** hjax has joined #aichallenge 2011-05-31T18:42:02 amstan? 2011-05-31T18:42:17 hjax: yes? 2011-05-31T18:42:41 you you want me to make a separate account for every starter package? 2011-05-31T18:46:34 mcleopold any ideas why skill updates are not happening? 2011-05-31T18:47:05 yeah, the workers are throwing errors 2011-05-31T18:47:18 oh k 2011-05-31T18:47:46 i wonder if all the starter package bots will score exactly the same :/ 2011-05-31T18:49:16 some may time out:p 2011-05-31T18:50:44 exactly lets see what language is truly the fastest 2011-05-31T18:50:57 assemly 2011-05-31T18:51:05 with a b in there somewhere 2011-05-31T18:52:17 assembly? lol 2011-05-31T18:52:38 sure, i dare you to make an assembly starter pack 2011-05-31T18:57:29 bhasker: ready for a major skill update? 2011-05-31T18:58:34 sure 2011-05-31T18:58:51 dammit still rank 2:p 2011-05-31T18:59:07 and greentea's ghost is higher :( 2011-05-31T18:59:25 hmm his ghost and mine have same trueskill but ghost is higher in list:( 2011-05-31T18:59:53 that's cause his was submitted first 2011-05-31T18:59:58 ah 2011-05-31T19:00:17 any reason for no C or haskell bots in git? 2011-05-31T19:00:23 hmm, actually, yours should be higher 2011-05-31T19:01:21 hjax: they're git submodules, you have to pull them in using "git submodule" commands 2011-05-31T19:01:30 ? 2011-05-31T19:01:44 the folders are empty on github, 2011-05-31T19:01:46 e.g git submodule init; git submodule update 2011-05-31T19:02:24 they show fine in the web interface for me 2011-05-31T19:02:41 not sure how you would get to an empty directory for them from the website 2011-05-31T19:03:33 *** FireFly has quit IRC (Quit: swatted to death) 2011-05-31T19:03:34 hmm weird my skill dropped on ranking page but nothing changed on my profile page 2011-05-31T19:03:35 weird 2011-05-31T19:03:43 i downloaded the whole thing as a zip files, no C or haskell 2011-05-31T19:03:57 hjax: sure 2011-05-31T19:04:02 yeah, they aren't in the base repo 2011-05-31T19:04:05 they are submodules 2011-05-31T19:04:22 hjax: yes, you'll have to use git 2011-05-31T19:07:07 bhasker: you were seeing the numbers rounded, his skill was 0.0017 higher 2011-05-31T19:07:22 ah k 2011-05-31T19:08:00 *** hjax has quit IRC (Ping timeout: 252 seconds) 2011-05-31T19:11:21 *** needsch has quit IRC (Quit: Leaving.) 2011-05-31T19:12:41 *** mcstar has quit IRC (Quit: WeeChat 0.3.5) 2011-05-31T19:19:39 there has gotta be a better way to get the java plugin stuff 2011-05-31T19:21:08 watching my web server setup pull in x11, firefox, etc. just seems so wrong 2011-05-31T19:27:12 McLeopold: server_setup.py is broken because of an unfilled {api_url} variable in server_info.php.template 2011-05-31T19:27:35 aichallenge: McLeopold epsilon * red330cb / setup/worker_setup.py : Merge branch 'epsilon' of https://github.com/aichallenge/aichallenge into epsilon - http://bit.ly/k2Hra3 2011-05-31T19:27:35 aichallenge: McLeopold epsilon * r2acb8fd / (ants/ants.py manager/manager.py): fixed issue with failed bot crashing engine - http://bit.ly/lZfiq5 2011-05-31T19:27:51 janzert: oh yeah, I did do that. 2011-05-31T19:28:28 *** computerwiz_222 has joined #aichallenge 2011-05-31T19:29:37 *** tomh has quit IRC (Quit: tomh) 2011-05-31T19:31:33 *** hjax has joined #aichallenge 2011-05-31T19:34:37 McLeopold: from that last commit, there should never be anything put before a docstring 2011-05-31T19:35:08 aichallenge: McLeopold epsilon * ra126e8b / setup/server_setup.py : fixed server_info.php generation - http://bit.ly/iKm6wt 2011-05-31T19:36:10 aichallenge: McLeopold epsilon * r3edec0e / ants/ants.py : removed debug statements - http://bit.ly/iPiWGF 2011-05-31T19:37:29 *** hjax has quit IRC (Quit: Page closed) 2011-05-31T19:41:14 aichallenge: janzert epsilon * rdec403f / (ants/ants.py setup/server_setup.py setup/worker_setup.py): Move code after docstring and revert executable bit on setup scripts - http://bit.ly/kjAl6W 2011-05-31T19:41:39 oh, that docstring, sry 2011-05-31T19:41:50 :) no problem 2011-05-31T19:42:12 saw the executable bit got cleared on the setup scripts somewhere along the way too 2011-05-31T19:42:15 I never think about it because I never use them 2011-05-31T19:42:27 McLeopold: yes, i noticed 2011-05-31T19:42:38 all the scripts i have to exec i have to type python in front 2011-05-31T19:43:22 I never think about executable bits cause I'm usually in windows 2011-05-31T19:43:59 we have a special 7 layer dip of extension to program mappings stored in the registry sewage 2011-05-31T19:45:08 amstan: did you get a worker going yet? 2011-05-31T19:53:48 *** olexs1 has quit IRC (Quit: Leaving.) 2011-05-31T19:54:41 *** _0x47 <_0x47!~slash@p5B12E0E0.dip.t-dialin.net> has quit IRC (Ping timeout: 240 seconds) 2011-05-31T19:55:00 *** JamesMG_ has quit IRC (Ping timeout: 240 seconds) 2011-05-31T19:55:22 *** Frontier has quit IRC (Ping timeout: 244 seconds) 2011-05-31T19:55:41 *** olexs has joined #aichallenge 2011-05-31T19:56:11 *** JamesMG has joined #aichallenge 2011-05-31T19:56:42 *** olexs has quit IRC (Client Quit) 2011-05-31T19:57:31 *** olexs has joined #aichallenge 2011-05-31T19:58:06 *** Frontier has joined #aichallenge 2011-05-31T19:58:43 *** _0x47 <_0x47!~slash@p5B12E0E0.dip.t-dialin.net> has joined #aichallenge 2011-05-31T19:59:28 McLeopold: no, i'm not at home at the moment 2011-05-31T19:59:36 and idk if janzert is available 2011-05-31T20:00:24 working on the automatic worker deployment stuff at the moment :) 2011-05-31T20:01:07 ok, should i join you then? 2011-05-31T20:01:33 not necessarily 2011-05-31T20:02:06 if you have something you need an answer on I can probably pause this and try to answer it though 2011-05-31T20:02:20 janzert: do you have to go soon? 2011-05-31T20:02:34 not that I know of right now 2011-05-31T20:02:39 ok, cool 2011-05-31T20:02:46 so then i could join you in 30 min 2011-05-31T20:09:49 aichallenge: janzert epsilon * r5ade526 / (3 files in 2 dirs): Work for automated worker deployment - http://bit.ly/kYCfjV 2011-05-31T20:11:43 *** olexs has quit IRC (Quit: Leaving.) 2011-05-31T20:11:43 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-05-31T20:29:14 *** Eruonen has quit IRC () 2011-05-31T20:31:16 *** amstan has joined #aichallenge 2011-05-31T20:31:16 *** ChanServ sets mode: +o amstan 2011-05-31T20:33:00 McLeopold: related to https://github.com/aichallenge/aichallenge/issues/47#comment_1270056 : 2011-05-31T20:33:10 i get a 404 from http://aichallengebeta.hypertriangle.com/game_list_json.php&page=1 2011-05-31T20:35:52 *** computerwiz_222 has quit IRC (Quit: Konversation terminated!) 2011-05-31T20:35:55 yeah, use a ? 2011-05-31T20:36:00 I was typing fast 2011-05-31T20:36:11 I can edit, right? 2011-05-31T20:36:14 *** computerwiz_222 has joined #aichallenge 2011-05-31T20:37:52 McLeopold: is there a json way to get info about only one game? 2011-05-31T20:37:55 or.. nvm 2011-05-31T20:37:59 game_info.php 2011-05-31T20:38:04 * amstan facepalms 2011-05-31T20:38:11 /game/# 2011-05-31T20:38:17 yeah 2011-05-31T20:38:43 I thought the same thing and almost wrote something until *I* facepalmed :) 2011-05-31T20:38:54 janzert: k, so what state is the worker installer in now? can i install? or is there some quick commits you have to do? 2011-05-31T20:39:24 the goal is 4 gpm... 2011-05-31T20:39:30 the automatic installation is humming along fine at the moment, I do know it will have a problem at the end though 2011-05-31T20:39:33 we've got 3 workers running right now 2011-05-31T20:39:43 McLeopold: we do? 2011-05-31T20:39:48 but no jails 2011-05-31T20:39:59 I still need to rewrite the worker startup script 2011-05-31T20:40:02 McLeopold: is there a way to get separate worker info in here? http://aichallengebeta.hypertriangle.com/server_stats.php 2011-05-31T20:40:07 the worker table is not what it used to be 2011-05-31T20:40:24 it's not that different 2011-05-31T20:40:35 janzert: well.. i can give you a shell account to my htpc and you can try there if you want 2011-05-31T20:40:50 I'm not familiar with what it was 2011-05-31T20:40:54 McLeopold: we need to be able to get individual worker information at least 2011-05-31T20:40:59 McLeopold: i had an error on every query for that table, i just decided to remove them then 2011-05-31T20:41:09 i.e. per worker games and errors per minute 2011-05-31T20:41:20 I'll look at the queries in the history 2011-05-31T20:41:39 we use that information to monitor the health of a worker to see if it needs a reboot 2011-05-31T20:42:00 *** computerwiz_222 has quit IRC (Ping timeout: 240 seconds) 2011-05-31T20:42:10 McLeopold: https://github.com/aichallenge/aichallenge/blob/pw/planet_wars/www/game_stats.php 2011-05-31T20:43:08 McLeopold: oh yes, what do you think? http://aichallengebeta.hypertriangle.com/using_the_tools.php 2011-05-31T20:43:38 actually the monitor script used api_worker_query.php to get the information 2011-05-31T20:43:40 i removed stuff from the old one, some parts seemed too weird for older programmers 2011-05-31T20:43:56 so having it on the human readable page isn't required I suppose but probably still nice 2011-05-31T20:44:07 include? 2011-05-31T20:44:19 so, api_worker_query is probably broken then? 2011-05-31T20:44:46 most likely since the database structure has changed so much 2011-05-31T20:45:00 amstan: it's a good start, but needs a re-write 2011-05-31T20:45:09 IMHO 2011-05-31T20:45:23 i'm not a good writer 2011-05-31T20:46:04 basically, I'd do a how-to with lists and examples rather than an essay in paragraph form 2011-05-31T20:46:45 well, here's my list: https://github.com/aichallenge/aichallenge/issues/128 2011-05-31T20:46:53 * janzert is watching the final shuttle move out of the vehicle assembly building to the launchpad 2011-05-31T20:47:05 janzert: but... worker! 2011-05-31T20:47:24 setup is running in the background :) 2011-05-31T20:54:07 and it died at the non-existent worker start script :/ 2011-05-31T20:56:12 McLeopold: dude.. you have to stop chmod -x ing worker and manager.py 2011-05-31T20:56:28 *** hjax has joined #aichallenge 2011-05-31T20:58:13 why does it say modified: ants/dist/starter_bots/c (new commits) 2011-05-31T21:00:18 amstan ok im creating an account for each starter pack now, been busy :/ 2011-05-31T21:00:27 hjax: cool, thanks 2011-05-31T21:00:47 all of them are in zip files ready to go 2011-05-31T21:01:44 aichallenge: Alexandru Stan epsilon * r1c98687 / (manager/manager.py worker/worker.py): fixed exec permissions on main files - http://bit.ly/iKSfe3 2011-05-31T21:01:52 hjax: yes, you type make in the folder 2011-05-31T21:01:54 it zips them 2011-05-31T21:02:22 i have them already, now for the longest part making the accounts :/ 2011-05-31T21:02:51 janzert: do you have anything in mind for the api worker setup php page? 2011-05-31T21:03:01 janzert: is it supposed to contain some cool convenience methods? 2011-05-31T21:03:13 s/methods/commands/ 2011-05-31T21:03:40 basically it feeds the worker an initilization script 2011-05-31T21:03:47 it should be all set at this point 2011-05-31T21:03:53 do we have such a script? 2011-05-31T21:04:01 for the new contest? 2011-05-31T21:04:01 worker_init.py 2011-05-31T21:04:21 that's what I finished setting up a little earlier 2011-05-31T21:04:37 in setup/? 2011-05-31T21:04:51 I ended up moving it to the website 2011-05-31T21:05:12 kind of icky but that's where we need it from 2011-05-31T21:05:16 ln -s? 2011-05-31T21:05:27 there's really no need for it in setup 2011-05-31T21:05:41 if i git clone, i'd want that 2011-05-31T21:05:44 it's only for automated installs 2011-05-31T21:05:58 wouldn't git pull be useful for updating automated installs? 2011-05-31T21:05:58 nah, I doubt it 2011-05-31T21:06:21 ok, can i just call that directly then now? 2011-05-31T21:06:31 call what? 2011-05-31T21:06:37 worker_init 2011-05-31T21:06:40 api_server_setup.php? 2011-05-31T21:07:02 bah, sorry firecall be back in a half hour or so 2011-05-31T21:07:08 aww 2011-05-31T21:07:59 McLeopold: 2011-05-31T21:08:01 error: Your local changes to the following files would be overwritten by merge: manager/manager.py 2011-05-31T21:08:03 McLeopold: can you pull pls? 2011-05-31T21:10:15 janzert: hmm, we can use this: https://github.com/aichallenge/aichallenge/raw/epsilon/website/worker_init.py 2011-05-31T21:12:34 well.. 2011-05-31T21:12:36 * amstan is blocked 2011-05-31T21:16:47 *** bhasker has quit IRC (Quit: bhasker) 2011-05-31T21:21:15 wow are the rankings really that unstable :/ 2011-05-31T21:21:30 go starter bot is doing good :/ 2011-05-31T21:23:51 *** computerwiz_222 has joined #aichallenge 2011-05-31T21:24:44 amstan who wrote that go starterbot? 2011-05-31T21:24:53 idk, check the last commit 2011-05-31T21:24:56 for that folder 2011-05-31T21:25:36 amstan, its not behaving like a starter bot, its intelligent and about to take my bot down on the leaderboards 2011-05-31T21:26:19 i'm not sure what antimatroid's plan is, he didn't really make it clear that they must follow a strategy 2011-05-31T21:26:27 http://aichallengebeta.hypertriangle.com/visualizer.php?game=9085 2011-05-31T21:26:43 we its better than my bot, and its free :/ 2011-05-31T21:27:03 they must at least be equal, and weak :/ 2011-05-31T21:27:12 gostarterbot is neither 2011-05-31T21:27:36 amstan, scalastarterbot froze the worker 2011-05-31T21:27:44 what? 2011-05-31T21:27:47 how? 2011-05-31T21:28:09 its happened with coffeescript before 2011-05-31T21:28:26 what do you mean worker? it seems to work just fine still 2011-05-31T21:28:30 this is the first scala bot, and the worker cant compile it 2011-05-31T21:28:59 so.. it can't compile? or did it crash the whole worker(as in there's no more games being played on the server) 2011-05-31T21:29:23 i cant tell yet 2011-05-31T21:29:37 why did you say "froze the worker"? 2011-05-31T21:30:13 hmmm either the worker is hanging or crashed 2011-05-31T21:30:44 it's executing java bots.. 2011-05-31T21:30:46 doing just fine 2011-05-31T21:30:53 i have no idea what you mean by that 2011-05-31T21:31:02 really? scalabot is still on compiling and running tests 2011-05-31T21:31:17 no waiting to compile 2011-05-31T21:31:42 it'll probably play a game in between 2011-05-31T21:32:14 it shouldnt update that status until its actually started :/ 2011-05-31T21:32:25 no idea 2011-05-31T21:32:27 McLeopold: ping 2011-05-31T21:33:00 i gtg, ill do the last couple bots tomorrow morning 2011-05-31T21:33:13 k, keep updating your issue entry 2011-05-31T21:33:18 thanks, btw 2011-05-31T21:33:42 no problem, ill watch a few battles of each bot before i add it to the issue 2011-05-31T21:34:17 i know exactly how python bot preforms, so its the only one there atm 2011-05-31T21:35:10 amstan: its been over 10 min: http://aichallengebeta.hypertriangle.com/profile.php?user=59 2011-05-31T21:35:41 *** yasith has joined #aichallenge 2011-05-31T21:35:58 hjax: yeah, ok 2011-05-31T21:36:01 i gtg, have contestbot send me any new info 2011-05-31T21:36:10 ok, cool 2011-05-31T21:36:11 cya 2011-05-31T21:36:15 *** hjax has quit IRC (Quit: bye) 2011-05-31T21:39:28 contestbot: later tell hjax try submitting again 2011-05-31T21:39:28 amstan: Yes master! 2011-05-31T21:42:13 *** yasith has quit IRC (Ping timeout: 258 seconds) 2011-05-31T21:42:19 *** yasith_ has joined #aichallenge 2011-05-31T21:42:43 janzert: when is that worker-src.zip generated? 2011-05-31T21:43:41 *** yasith_ has quit IRC (Client Quit) 2011-05-31T21:44:09 *** yasith has joined #aichallenge 2011-05-31T21:46:11 janzert: hmm.. i don't like this, so worker-src.tgz is now 7 days old on the aichallengebeta server 2011-05-31T21:46:25 any fixed that we make to workers won't be updated in there 2011-05-31T21:48:49 what's wrong with pulling from git? 2011-05-31T22:00:13 amstan: pong 2011-05-31T22:00:28 McLeopold: have you seen my wall? 2011-05-31T22:00:41 i needed to pull on the main server 2011-05-31T22:01:11 ok 2011-05-31T22:01:20 i stashed your changes 2011-05-31T22:01:40 McLeopold: also, any idea what's going on with this: http://aichallengebeta.hypertriangle.com/profile.php?user=59 2011-05-31T22:01:52 he's been stuck in that state for a while now 2011-05-31T22:02:19 the worker assigned the compile task may have crashed? 2011-05-31T22:02:49 my changes on the server were temporary, I had already committed the fixes to the repo 2011-05-31T22:03:57 what about debug? 2011-05-31T22:04:25 my worker is continually downloaded and failing to report on the scala bot 2011-05-31T22:13:13 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-05-31T22:14:44 amstan: what is your wall? 2011-05-31T22:14:56 check a terminal that you were logged into 2011-05-31T22:14:59 at aichallengebeta 2011-05-31T22:15:02 you should have seen it 2011-05-31T22:15:06 yeah I did 2011-05-31T22:15:15 what's the question about debug? 2011-05-31T22:15:43 contest@aichallengebeta:~/aichallenge$ ls debug.txt 2011-05-31T22:16:54 it's gone now 2011-05-31T22:17:42 And I'm not actively chmod -x ing the files. I blame git :) 2011-05-31T22:18:10 *** smiley1983 has joined #aichallenge 2011-05-31T22:18:14 *** antimatroid1 has quit IRC (Ping timeout: 252 seconds) 2011-05-31T22:19:34 aichallenge: Alexandru Stan epsilon * r5bacc3d / setup/worker_setup.py : fixed dmd on 32 bit - http://bit.ly/mShEYo 2011-05-31T22:20:15 *** antimatroid has joined #aichallenge 2011-05-31T22:20:18 McLeopold: just saw your question about which bot was saying "Food bfs with n seeds" - that was mine, I forgot to disable debugging output before uploading in that version. The new version should be silent. 2011-05-31T22:20:33 oshit.. 2011-05-31T22:20:51 McLeopold: i did a mistake 2011-05-31T22:21:07 McLeopold: sudo rm -Rf /srv/chroot on the server 2011-05-31T22:21:13 McLeopold: i think i'll have to reinstall the worker 2011-05-31T22:21:26 McLeopold: i tought it was my local machine, they have the same hostname 2011-05-31T22:21:31 smiley1983: the only downside is if you fill up your 1024 byte quota, you may not see a good error if you do crash 2011-05-31T22:21:44 amstan: we're not using jails anyway 2011-05-31T22:21:49 oh! 2011-05-31T22:21:52 that's right 2011-05-31T22:21:55 omg! i am relieved 2011-05-31T22:22:12 * amstan continues to rm -rf 2011-05-31T22:22:27 It would be nice if *nix had a command line trash that you could recover from 2011-05-31T22:22:52 I'm sure someone has written an alias for it at some point 2011-05-31T22:23:24 McLeopold: I'm glad to hear that there's a limit, and it wasn't just filling up the server's whole drive with my garbage :) 2011-05-31T22:23:38 no, just my debug console :) 2011-05-31T22:23:39 McLeopold: you could easily do it by mving to .trash 2011-05-31T22:23:48 McLeopold: it takes less time that mv as well 2011-05-31T22:24:00 amstan: There are edge cases here. 2011-05-31T22:24:15 jbroman: yes 2011-05-31T22:24:20 jbroman: you're probably lolling at me right now 2011-05-31T22:24:20 amstan: For example, the case where I have write access to a directory but cannot read the file. I may delete it, but I would not be able to move it to another filesystem. 2011-05-31T22:24:41 if i were to write a wrapper i would mv it to the same fs 2011-05-31T22:24:49 unless it's a flash drive 2011-05-31T22:25:26 I'm not even sure whether you could mv it to the same filesystem if it's owned by somebody edge and is 700 or similar. Probably not. 2011-05-31T22:25:36 Many GUI file managers do provide Windows/Mac-like Trash. But especially for the command-line, it's weird when you try to define certain odd situations. 2011-05-31T22:26:00 Another solution is snapshots (à la ZFS). 2011-05-31T22:26:18 yeah, but doesn't that mean more overhead when doing things? 2011-05-31T22:26:33 amstan: ZFS snapshots are, from what I understand, rather low in overhead. 2011-05-31T22:27:00 (Copy-on-write etc) 2011-05-31T22:27:19 decent, would you use it to replace ext4? 2011-05-31T22:27:42 I don't think it's stable on Linux. I hear btrfs is nice in its own right, but I haven't even looked that closely at what's new in ext4. 2011-05-31T22:27:54 Frankly, most of my important stuff is code which sits in source control anyways. :P 2011-05-31T22:28:54 *** bhasker has joined #aichallenge 2011-05-31T22:29:00 yeah, i guess i woudln't care if my desktop hd died 2011-05-31T22:29:25 i would lose recent torrents at most 2011-05-31T22:29:48 but for servers this is another story 2011-05-31T22:33:08 contestbot: later tell janzert notes: http://pastebin.com/GDa5teRx 2011-05-31T22:33:08 amstan: I come to serve. 2011-05-31T22:33:30 well.. that should have been a gist, i have more, lol 2011-05-31T22:43:50 McLeopold: so.. how much do you think we have left till launch? 2011-05-31T22:43:55 i think it's time we get jeff 2011-05-31T22:45:28 amstan: There's quite a few issues with displaying data on the website, but nobody has noticed or opened issues. 2011-05-31T22:45:41 That said, we could probably launch real soon. 2011-05-31T22:45:51 yeah, let me write an email 2011-05-31T22:46:15 Do we have server to launch on? 2011-05-31T22:46:21 *** delt0r___ has joined #aichallenge 2011-05-31T22:46:22 that's up to jeff 2011-05-31T22:48:40 *** delt0r_ has quit IRC (Ping timeout: 260 seconds) 2011-05-31T22:50:54 *** qacek has joined #aichallenge 2011-05-31T22:52:29 finally back 2011-05-31T22:52:36 janzert: yay 2011-05-31T22:54:14 anytime you want the code that is installed on workers to be updated just run setup/create_worker_archive.py again 2011-05-31T22:54:35 and this way we can exactly control when and what that is 2011-05-31T22:55:02 it creates it from the local server repo of course 2011-05-31T22:55:55 once you run create_worker_archive, does it update the workers or something? 2011-05-31T22:56:24 no, it updates the tar ball that worker_init pulls from 2011-05-31T22:56:42 worker_init is only ever used once for the initial setup of a worker 2011-05-31T22:56:59 and really only for automated worker installs on ec2 2011-05-31T22:57:09 aichallenge: McLeopold epsilon * rbd2495c / (4 files): fixed issue with new language not being created - http://bit.ly/lzW8Yd 2011-05-31T22:57:09 aichallenge: McLeopold epsilon * r496bdcb / (8 files in 5 dirs): Merge branch 'epsilon' of https://github.com/aichallenge/aichallenge into epsilon - http://bit.ly/lOPV1C 2011-05-31T22:57:46 whenever we want to update workers before it was easiest to just kill them and bring up new ones 2011-05-31T22:59:11 amstan: http://aichallengebeta.hypertriangle.com/profile.php?user=59 2011-05-31T22:59:25 McLeopold: relay it to hjax 2011-05-31T22:59:44 janzert: oh yes, i fixed dmd 2011-05-31T22:59:47 on 32 bit 2011-05-31T23:00:00 @later tell hjax scalabot "compiled" correctly, which means the errors were reported 2011-05-31T23:00:00 McLeopold: Ready to serve. 2011-05-31T23:00:24 amstan: ok :} 2011-05-31T23:01:41 *** arakyd has quit IRC (Quit: Page closed) 2011-05-31T23:05:47 aichallenge: Alexandru Stan epsilon * re66bfb8 / (website/.gitignore website/starter_packages/.gitignore): changed gitignore to be able to include the starter_packages folder in git - http://bit.ly/kovowV 2011-05-31T23:05:47 aichallenge: Alexandru Stan epsilon * r362e6a9 / (ants/dist/starter_bots/Makefile ants/dist/tools/Makefile): fixed tools installation, made the clean in the starter_pack installation better - http://bit.ly/l3OLyu 2011-05-31T23:05:48 aichallenge: Alexandru Stan epsilon * r307de97 / (4 files): Merge branch 'epsilon' of github.com:aichallenge/aichallenge into epsilon - http://bit.ly/m5JuNC 2011-05-31T23:07:45 janzert: ugh, nvm: NameError: global name 'opts' is not defined 2011-05-31T23:07:52 shouldn't that be defined? 2011-05-31T23:08:04 where? 2011-05-31T23:08:13 most places it's options 2011-05-31T23:08:13 everywhere in worker_setup 2011-05-31T23:08:15 as a global 2011-05-31T23:08:18 no 2011-05-31T23:08:24 why is it an argument? 2011-05-31T23:08:32 it is passed into functions as a paramater 2011-05-31T23:08:52 if i want to pass it to install_dmd i have to pass it to install_all_languages as well 2011-05-31T23:08:58 because globals are bad? :) 2011-05-31T23:09:10 yep, generally 2011-05-31T23:10:32 janzert: can i just do _, arch = get_ubuntu_release_info() in there? 2011-05-31T23:10:38 instead of using opts? 2011-05-31T23:11:07 did get_ubuntu_release_info even work inside a chroot? 2011-05-31T23:11:13 I thought it broke 2011-05-31T23:11:31 aichallenge: Alexandru Stan really-fix-dmd-32 * rdab82e8 / setup/worker_setup.py : ugh.. opts - http://bit.ly/jYvOsE 2011-05-31T23:11:34 like that ^ 2011-05-31T23:11:41 janzert: i fixed it 2011-05-31T23:11:59 janzert: i didn't use the command anymore, just relied on the /etc/ file 2011-05-31T23:12:11 did we have 5 workers last contest? 2011-05-31T23:12:13 * janzert would really rather not paramaterize everything for architecture actually 2011-05-31T23:12:26 McLeopold: 3 or 4 most of the time 2011-05-31T23:12:29 janzert: otherwise the worker will always crash on my system 2011-05-31T23:12:37 because i have 32 bit 2011-05-31T23:12:38 but at about 1/4th the cost too 2011-05-31T23:12:38 *** smiley1983 has quit IRC (Quit: leaving) 2011-05-31T23:12:45 amstan: yes, I know 2011-05-31T23:12:56 well.. what do you propose? 2011-05-31T23:12:58 :( 2011-05-31T23:13:16 only making it work on the setup we're using :/ 2011-05-31T23:13:23 I don't have a good solution 2011-05-31T23:13:45 yes, but if it doesn't work on 32(and that's the only reason why) i cannot test 2011-05-31T23:14:02 are you ok with me not participating in the multiple worker+sandbox tests? 2011-05-31T23:15:54 I won't stand in the way of making it architecture portable, I just don't really like the extra additions we are having to add for it 2011-05-31T23:20:40 well.. feel free to fix some things that you don't like while keeping the functionality 2011-05-31T23:20:49 anyone know of a good way to get all active process ids in python? 2011-05-31T23:23:26 aichallenge: Alexandru Stan epsilon * rbfbcf90 / setup/worker_setup.py : 2011-05-31T23:23:26 aichallenge: Revert "fixed dmd on 32 bit" 2011-05-31T23:23:26 aichallenge: This reverts commit 5bacc3d42c93c1ec72e757898ec8e715c70faeb9. 2011-05-31T23:32:49 McLeopold: do we still need ants/game_logs ? 2011-05-31T23:33:00 is it in the repo? 2011-05-31T23:33:05 yes 2011-05-31T23:33:14 and my tools makefile includes it in the zip 2011-05-31T23:33:16 no, that is just a temp dir, it shouldn't have been commited 2011-05-31T23:33:28 probably my fault 2011-05-31T23:33:29 what about play_one_game.sh? 2011-05-31T23:33:32 --log_dir game_logs 2011-05-31T23:33:38 will it fail if the dir doesn't exist? 2011-05-31T23:33:41 playgame should create it 2011-05-31T23:33:44 k 2011-05-31T23:34:12 playgame, line 234 2011-05-31T23:34:27 nvm, it's gitignored 2011-05-31T23:34:34 i guess i have to clean it in my makefile 2011-05-31T23:34:52 McLeopold: http://aichallengebeta.hypertriangle.com/tools.zip 2011-05-31T23:34:58 see what's extra in there 2011-05-31T23:35:05 i know the visualiser stuff is extra 2011-05-31T23:35:12 but i have to run ant to clean it 2011-05-31T23:35:17 and ant doesn't want to run on that server 2011-05-31T23:36:07 the .gitignore files are extra 2011-05-31T23:36:45 website-source looks like extra 2011-05-31T23:37:05 don't include compiled files like *.pyc 2011-05-31T23:37:29 but I would include jar files 2011-05-31T23:37:52 do you have a way to do this cleanly? 2011-05-31T23:37:59 maybe another git checkout? 2011-05-31T23:38:06 no, that's what your doing? 2011-05-31T23:38:11 no 2011-05-31T23:38:14 i'm cping 2011-05-31T23:38:26 but it feels really dirty 2011-05-31T23:38:28 I thought you were writting a makefiles? 2011-05-31T23:38:46 yes: https://github.com/aichallenge/aichallenge/blob/epsilon/ants/dist/tools/Makefile 2011-05-31T23:40:01 I think we should leave out the asymmetric map get 2011-05-31T23:40:03 gen 2011-05-31T23:40:50 why do we even have that? 2011-05-31T23:40:58 just git rm it 2011-05-31T23:41:27 it's okay to leave code in the repo, just the tools download should be really clean 2011-05-31T23:41:41 tools will get everything from ants/ right now 2011-05-31T23:41:45 except the starter_packages 2011-05-31T23:41:51 so we might as well clean it 2011-05-31T23:42:24 I'm not sure about the stuff in util. Probably leave it out. 2011-05-31T23:43:26 amstan: how would a git checkout work? 2011-05-31T23:43:35 like a branch with only what we want? 2011-05-31T23:43:59 we could do that, a tools branch 2011-05-31T23:44:03 and merge regularly 2011-05-31T23:44:15 so, it could pull only what we want? 2011-05-31T23:44:30 i was just thinking pulling the repo again in a subfolder then only keeping the ants folder 2011-05-31T23:44:35 that way everything is clean 2011-05-31T23:44:45 no pyc files, log files or compiled visualizer 2011-05-31T23:45:12 if you make a branch, I can update it, that might be better 2011-05-31T23:45:36 you can update epsilon fine as well 2011-05-31T23:46:09 this isn't going to be a one-to-one pull of a repo, and it shouldn't be 2011-05-31T23:46:34 let's just stick with the makefile and I can modify it where needed 2011-05-31T23:47:09 yes, i'm just saying, to clone the repo instead of using rsync to copy ants 2011-05-31T23:47:18 that way anything in gitignore actually gets ignored 2011-05-31T23:47:25 aichallenge: janzert epsilon * r1b20454 / worker/release_stale_jails.py : Add script to release stale jail locks - http://bit.ly/jms2Ii 2011-05-31T23:47:39 omg! 2011-05-31T23:47:43 my worked finished installing 2011-05-31T23:47:46 I'd rather we list each file individually in the makefile 2011-05-31T23:47:46 and it did it properly 2011-05-31T23:47:54 McLeopold: that's a lot of files 2011-05-31T23:48:03 I'm okay with that 2011-05-31T23:48:28 There's too many exceptions, and trying to make the repo into a tools download doesn't make sense right now. 2011-05-31T23:50:50 McLeopold: normal workers should run worker.py as "worker.py -t -n 0"? 2011-05-31T23:50:59 ye 2011-05-31T23:51:00 s 2011-05-31T23:51:00 *** yasith has quit IRC (Ping timeout: 240 seconds) 2011-05-31T23:54:26 aichallenge: janzert epsilon * r62d9dd0 / worker/release_all_jail_users.py : Remove old script - http://bit.ly/iHce1o 2011-05-31T23:59:15 *** amstan has quit IRC (Ping timeout: 260 seconds)