2011-05-16T00:27:40 *** Eruonen has joined #aichallenge 2011-05-16T00:34:45 *** computerwiz_222 has quit IRC (Remote host closed the connection) 2011-05-16T00:39:26 janzert: when using secure jails, I'm not getting output from the bots 2011-05-16T00:40:16 not sure what could cause that 2011-05-16T00:40:33 I'm running with full logging 2011-05-16T00:40:34 so play_one_game times out all the bots on the first move? 2011-05-16T00:40:53 Well, I think it's only from within the worker... 2011-05-16T00:41:06 I need to compare options then... 2011-05-16T00:41:38 within the worker for a functional test, both timeout 2011-05-16T00:42:32 I guess I'd start by checking if play_one_game works 2011-05-16T00:42:37 playgame is find 2011-05-16T00:47:00 wait, no, playgame does a timeout as well 2011-05-16T00:48:04 janzert: have you gotten a bot to reply to input inside of a jail? 2011-05-16T00:50:33 janzert: this is my current guess -- shlex.split(shell_command.replace('\\','/')) 2011-05-16T00:56:31 yes, play_one_game works fine with secure jails here 2011-05-16T00:56:48 McLeopold: works for hypertriangle too 2011-05-16T00:58:10 *** Eruonen has quit IRC () 2011-05-16T00:58:19 ['sudo', '-u', 'jailuser2', 'schroot', '-u', 'jailuser2', '-c', 'jailuser2', '-d', '/home/jailuser', 'python', 'submission_test/TestBot.py'] 2011-05-16T00:58:30 It seems like the last 2 shouldn't be split? 2011-05-16T00:59:09 why not? 2011-05-16T00:59:28 It's a guess 2011-05-16T00:59:34 they are the program to run and the argument to give it 2011-05-16T00:59:55 if that's how schroot works, the good 2011-05-16T00:59:58 btw, the shlex.split happens even when not run in a secure jail 2011-05-16T01:01:08 and taking a look at the schroot man page again, it seems like that should be correct 2011-05-16T01:02:38 okay, let me try on hypertriangle 2011-05-16T01:03:25 I've got the setup scripts running from scratch here again to test my latest changes, but I can check after it finishes as well 2011-05-16T01:15:49 aichallenge: Contest Worker epsilon * r67d7533 / (4 files in 2 dirs): added some worker debugging - http://bit.ly/mqV1pA 2011-05-16T01:16:10 amstan: any issue with pulling that as contest on hypertriangle? 2011-05-16T01:16:25 ? 2011-05-16T01:17:25 contest@aichallengebeta:!/aichallenge$ git pull (yes/no) 2011-05-16T01:18:15 probably not, why would there be? 2011-05-16T01:18:29 it's your machine, just want to be sure 2011-05-16T01:18:50 np, you can do it 2011-05-16T01:18:56 i don't have the stuff open now 2011-05-16T01:21:01 why the changes to sandbox.py, they don't seem to make any functional change? 2011-05-16T01:22:16 So I can pass a --debug switch to worker.py, and a --secure_jail switch to playgame.py 2011-05-16T01:22:47 how do the changes in sandbox.py effect that? 2011-05-16T01:23:19 ahh, nevermind I figured it out :P 2011-05-16T01:23:30 so you can pass in an explicit None and get the default 2011-05-16T01:23:51 yeah, otherwise the server_info setting takes over 2011-05-16T01:24:01 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-05-16T01:24:40 although single quote to double quote was probably a bit pedantic 2011-05-16T01:25:44 especially since my slightly preferred style there would be double quotes :P 2011-05-16T01:26:12 I avoid double quotes unless I really want a lot of single quotes in a string 2011-05-16T01:27:27 I generally default to double quotes, except for certain situations, primarily dictionary keys 2011-05-16T01:30:19 *** chris__0076 has joined #aichallenge 2011-05-16T01:31:33 *** Frontier has joined #aichallenge 2011-05-16T01:31:52 good morning 2011-05-16T01:31:57 howdy 2011-05-16T01:32:49 *** chris___0076 has quit IRC (Ping timeout: 246 seconds) 2011-05-16T01:33:10 aichallenge: Contest Worker epsilon * rb3dca5f / worker/worker.py : ignore cleanup in debug mode - http://bit.ly/iIQPn3 2011-05-16T01:36:08 aichallenge: janzert epsilon * racb8be2 / (setup/server_setup.py setup/worker_setup.py): Try not to stomp on contest file ownership in setup scripts - http://bit.ly/lLM5qU 2011-05-16T01:38:39 Contest Worker? An anonymous commit? 2011-05-16T01:38:50 janzert: what's the cwd of a jail user when running the bot command? 2011-05-16T01:39:01 Frontier: that was me using bad git info :( 2011-05-16T01:40:38 ok, I just did not know that works on git. 2011-05-16T01:40:51 *github 2011-05-16T01:41:39 I always for get 2011-05-16T01:43:38 *** Ice_Harley has joined #aichallenge 2011-05-16T01:44:42 the engine works again for me, yay! I had trouble with it trying to writing to /srv/chroot 2011-05-16T01:45:16 janzert: any status on an answer to the working dir question? 2011-05-16T01:46:26 It seems like it is /home/jailuser? 2011-05-16T01:46:31 without a number? 2011-05-16T01:46:39 that's what it appears to the sandboxed program yes 2011-05-16T01:47:10 so, what's really going on? 2011-05-16T01:48:06 on the outside it is /srv/chroot/jailuser/root/home/jailuser, which is a unionfs and any data written will actually be at /srv/chroot/jailuser/scratch/... 2011-05-16T01:48:39 it's a chroot so the /srv/chroot/jailuser/root appears to be the root of the filesystem to the programs inside it 2011-05-16T01:49:18 Where does the command run at then? I'm getting a 'cant find MyBot.py' 2011-05-16T01:49:39 in /home/jailuser 2011-05-16T01:49:49 I have a submission_test dir in there 2011-05-16T01:50:03 so, it's not in the correct location to run the MyBot.py 2011-05-16T01:50:05 the sandbox copies the working directory you give for the bot to that spot 2011-05-16T01:50:30 shouldn't it copy the contents of the dir instead? 2011-05-16T01:50:41 that's what I mean 2011-05-16T01:50:52 so, that's a bug then 2011-05-16T01:51:08 it's working here though 2011-05-16T01:51:11 but it doesn't make sense that you guys see things working 2011-05-16T01:51:41 see the Jail.prepare_with method 2011-05-16T01:51:42 Unless the worker is getting the wrong dir maybe 2011-05-16T01:52:13 does an ending slash matter? 2011-05-16T01:52:34 does it to "cp -r"? 2011-05-16T01:53:33 it is using cp -r to make the copy 2011-05-16T01:53:37 hmm things like that may depend on the distribution 2011-05-16T01:55:15 janzert: playgame passes a '.' for the working dir, worker does something different 2011-05-16T01:56:20 playgame changes to the bot directory before calling the sandbox? 2011-05-16T01:56:37 no, it just assumes where you run it is your working dir 2011-05-16T01:57:12 but that isn't the bot's working directory so should never work 2011-05-16T01:57:29 so, if I fix it, jails will break for you then :) 2011-05-16T01:58:35 to be sure, I'm talking about what ends up as the bot_cwd variable at line 40 of engine.py 2011-05-16T01:58:54 and is the first argument to the Sandbox constructor 2011-05-16T01:59:03 McLeopold: Did you add the "status" array to the replays and if so, does it use the same values as the "playerstatus" field described in the wiki? 2011-05-16T01:59:38 janzert: yes, playgame hardcodes to '.', which I am fixing 2011-05-16T02:00:03 and when I do, I think you'll see the issue I get with worker, where the dir name is copied to the jailuser 2011-05-16T02:00:07 not the contents 2011-05-16T02:00:13 Frontier: I'll check later tonight 2011-05-16T02:01:35 @#$^%!@#$%, stupid print functions 2011-05-16T02:01:36 janzert: User error -- Replace user. 2011-05-16T02:01:50 python 2 has a print statement for crying out loud 2011-05-16T02:03:03 it also has future imports 2011-05-16T02:05:26 * janzert isn't going to discuss print functions because all I'm going to do is start screaming 2011-05-16T02:06:03 I do see playgame is using the "." so that means the whole ants/ directory structure is getting copied into the sandbox 2011-05-16T02:06:35 janzert: so, you prefer the print statement over the print function? 2011-05-16T02:06:50 I prefer to stick with standard python 2 style 2011-05-16T02:07:06 if I was writing python 3 then I'd prefer print functions 2011-05-16T02:07:14 Noted. I was preparing for 3 2011-05-16T02:09:41 so the working directory passed into the sandbox should be the directory you want the commands run from 2011-05-16T02:09:59 and the commands run should assume they are run from that directory obviously :) 2011-05-16T02:10:20 yes, but I think if I actually did that, it would be broken 2011-05-16T02:10:22 the current playgame follows both those rules but in a rather wierd way :P 2011-05-16T02:11:24 basically it looks like instead of setting up the commands to run from the bots directory it sets them up to run from the current directory 2011-05-16T02:12:11 since the working directory is given as '.' and the bot command is "python dist/sample_bits/python/LeftyBot.py" for example 2011-05-16T02:12:25 with playgame, someone should be able to pass in a command, which could be "python my/dir/struct/MyBot.py" 2011-05-16T02:12:38 so, the command has the dir info inside of it 2011-05-16T02:12:59 okay, we are saying the same thing :) 2011-05-16T02:13:41 so long as the sandboxes are running non-secure or the my/dir/struct/ is under the current working directory that will work fine 2011-05-16T02:13:46 we can make playgame take 2 params for each bot, or we can try and parse it out properly 2011-05-16T02:14:04 I would just leave it the way it is 2011-05-16T02:14:04 worker is a whole different story 2011-05-16T02:14:37 it will work fine for people using it themselves since they won't be running in secure mode anyway 2011-05-16T02:15:20 but, worker passes ("../ants/submission_test/", "python TestBot.py"), and the jail has a "submission_test" dir. Shouldn't the contents be copied? 2011-05-16T02:16:08 what does "cp -r ../ants/submission_test/ some_test_destination" do? 2011-05-16T02:16:37 not the contents 2011-05-16T02:16:47 you need a splat on the end 2011-05-16T02:16:49 also using a relative path like that is going to cause the behaviour to be dependent on the scripts current working directory 2011-05-16T02:17:14 don't use * cause you won't get any .dot files 2011-05-16T02:17:18 I think I mentioned making these module dir to solve that 2011-05-16T02:17:59 I would recommend getting an absolute directory to pass to the sandbox 2011-05-16T02:18:02 *** smiley1983 has joined #aichallenge 2011-05-16T02:18:21 I think I can make one from the server_info stuff 2011-05-16T02:18:21 then you can run worker.py from anywhere and have it run correctly 2011-05-16T02:20:55 *** Mathnerd314 has quit IRC (Read error: Connection reset by peer) 2011-05-16T02:23:34 ok, I see the problem with the jail copying going on 2011-05-16T02:27:56 aichallenge: janzert epsilon * r8b7d530 / worker/sandbox.py : Ensure working directory contents are copied into sandbox - http://bit.ly/mf8SQT 2011-05-16T02:27:58 see if that does it for you 2011-05-16T02:28:03 k 2011-05-16T02:29:42 * janzert hates that copying a directory behavior based changes whether or not the target given exists 2011-05-16T02:29:54 I should know better since it has tripped me up before 2011-05-16T02:30:49 *** chris___0076 has joined #aichallenge 2011-05-16T02:31:13 Are other people finding that the visualize_locally.py script doesn't work? copy_paste.html still works. 2011-05-16T02:32:09 smiley1983: I've recently added launching the visualizer to playgame.py. Are you using that? 2011-05-16T02:33:22 *** chris__0076 has quit IRC (Ping timeout: 246 seconds) 2011-05-16T02:34:10 *** Chris_0076 has joined #aichallenge 2011-05-16T02:34:11 McLeopold: I noticed that, but on my system both playgame and visualize_locally give me a blank page in chromium. 2011-05-16T02:34:48 I ran into that bug, it's probably a bad relative path. 2011-05-16T02:35:06 wait, that was an IE issue... 2011-05-16T02:35:22 The replay.html file contains the game data, but loads a blank page. I'll go and check it in firefox... 2011-05-16T02:35:33 if you check your game_logs dir, there should be a replay.html 2011-05-16T02:35:35 *** chris___0076 has quit IRC (Ping timeout: 250 seconds) 2011-05-16T02:36:06 janzert: dir copy is better, checking some other stuff 2011-05-16T02:38:09 The replay.html is being generated in the viewer dir, and I don't seem to have a game_logs dir. Firefox also loads a blank page. 2011-05-16T02:38:28 when's the last time you pulled? 2011-05-16T02:38:41 A few minutes ago 2011-05-16T02:38:56 using play_one_game.sh? 2011-05-16T02:39:42 it has a "--log_dir game_logs" switch 2011-05-16T02:40:08 ... Ha! Thankyou! play_one_game works. I had made my own game-running scripts, I should be able to work out the problem from this. 2011-05-16T02:40:14 Sorry to have bothered you. :) 2011-05-16T02:40:26 np, I switched a lot of the switches around lately 2011-05-16T02:40:49 I found some of them and updated my scripts, obviously missed some. 2011-05-16T02:40:53 the logging is much better, you can pipe bot output anywhere now 2011-05-16T02:48:56 *** smiley1983 has quit IRC (Ping timeout: 264 seconds) 2011-05-16T02:49:07 janzert: no luck on playgame :( 2011-05-16T02:49:41 no luck in what way? 2011-05-16T02:49:51 *** smiley1983 has joined #aichallenge 2011-05-16T02:49:56 I still get timeouts with the secure jails 2011-05-16T02:49:58 it does still work here btw. I checked that before pusing 2011-05-16T02:50:00 pushing 2011-05-16T02:50:30 ok, how about running the sandbox directly does that work? 2011-05-16T02:51:04 let me get a command line for you 2011-05-16T02:53:45 *** smiley1983 has quit IRC (Client Quit) 2011-05-16T03:04:04 heh, ok so play_one_game works fine but something is going on trying to use the sandbox command line directly here as well 2011-05-16T03:04:15 okay 2011-05-16T03:04:29 I'm going to commit the playgame fix, to be consistent with the worker 2011-05-16T03:04:51 how? 2011-05-16T03:05:33 I mean how can it be consistent with the worker when we don't necessarily know the "correct" working directory 2011-05-16T03:05:50 or more specifically what are you using to get the working directory? 2011-05-16T03:06:19 aichallenge: McLeopold epsilon * r5ad7d7a / (ants/playgame.py worker/compiler.py worker/worker.py): fixed working dir of playgame - http://bit.ly/l8GfP9 2011-05-16T03:06:39 janzert: get_cmd_wd does the fix for me 2011-05-16T03:07:10 so "python test/MyBot.py" returns ("/real/path/test", "python MyBot.py") 2011-05-16T03:07:36 McLeopold: then your ants will see the outcome of all battles it's involved in 2011-05-16T03:07:38 what about "../mybot/Mybot"? 2011-05-16T03:08:17 the worker is passing the correct commands defined here: https://github.com/aichallenge/aichallenge/blob/epsilon/worker/compiler.py#L227 2011-05-16T03:08:34 janzert: look at the code, it should work 2011-05-16T03:08:44 and "../mylanguage/interpreter ../mybot/MyBot.mylanguage"? :) 2011-05-16T03:08:58 janzert: that last one may fail :) 2011-05-16T03:09:14 I guess I'm saying I think an option argument might be better way to go 2011-05-16T03:09:33 although actually there probably isn't a way to make the latter one work sanely 2011-05-16T03:10:06 antimatroid: yes 2011-05-16T03:10:41 antimatroid: if you have a lone scout ant, and it dies, it gets to see who it killed 2011-05-16T03:11:27 antimatroid: ants move -> vision calc -> battle -> birth 2011-05-16T03:12:03 antimatroid: although, I'm thinking of this as well: birth -> move -> vision -> battle 2011-05-16T03:12:40 that last one prevents the boost from spawning an ant, slightly harder to spawn under contention 2011-05-16T03:14:18 bah, on my development VM sandbox.py command line works, on my server test vm it doesn't 2011-05-16T03:15:36 and play_one_game, even with the latest playgame.py, works fine on both 2011-05-16T03:15:59 hypertriangle worker works now! 2011-05-16T03:17:49 the box where it's not working, does "./sandbox.py -o echo test" give you a "response: test" line? 2011-05-16T03:17:51 McLeopold: do you mean move->birth? 2011-05-16T03:18:00 and how about without the -o argument 2011-05-16T03:18:50 you will get more draws for food doing that? 2011-05-16T03:19:04 janzert: No more responses. Terminating. 2011-05-16T03:19:13 so, no 2011-05-16T03:19:35 antimatroid: I was specifying the order in which things are calculated in the engine 2011-05-16T03:19:50 heh, so even where it's not working correctly for me I still get that response 2011-05-16T03:20:17 antimatroid: it means you have to be next to food *before* your ant moves, so you have to wait a turn 2011-05-16T03:20:40 janzert: I guess your not in jail? Get out of jail free card? :p 2011-05-16T03:24:19 McLeopold: I don't get the difference 2011-05-16T03:24:26 except your ant spawns after the change of turn 2011-05-16T03:24:44 which you can't then control right away 2011-05-16T03:25:16 yeah okay, i think that's more confusing than anything :P 2011-05-16T03:25:28 it is... 2011-05-16T03:25:34 and it could still spawn then too 2011-05-16T03:25:37 the new ant that is 2011-05-16T03:25:57 I just don't like the boost in speed you get from walking towards food 2011-05-16T03:26:03 why not? 2011-05-16T03:26:11 I guess it can be a technique for faster growth 2011-05-16T03:26:27 everyone gets the same deal 2011-05-16T03:26:51 because, if 2 players are distance 20 from the same spot, random food shouldn't give either an advantage in speed 2011-05-16T03:28:18 ... symmetric food spawning == win 2011-05-16T03:28:25 it's too hard to make things fair asymmetrically 2011-05-16T03:28:30 Frontier: is there anyway to speed up the replays? 2011-05-16T03:28:54 antimatroid1: it's a nitpick, so it's not worth messing with 2011-05-16T03:30:09 aichallenge: McLeopold epsilon * rd62f535 / setup/server_setup.py : fixed php memcache package - http://bit.ly/mTkEwN 2011-05-16T03:31:33 McLeopold: ah well, good question. You have to convince me that is is bad to make the speed configurable through 3 settings you must enter in a text input box 2011-05-16T03:32:01 *** ltriant has quit IRC (Quit: Computer has gone to sleep) 2011-05-16T03:32:16 Frontier: I'll do anything to be able to control speed. 2011-05-16T03:32:46 *** amstan has quit IRC (Remote host closed the connection) 2011-05-16T03:33:06 It's just, the front page of the website needs to look exciting, and the ants move around like molasses. 2011-05-16T03:33:13 bad answer :p. what do you have in mind? clicking on + everytime you load a replay? 2011-05-16T03:34:03 http://hypertriangle.com:13080/ 2011-05-16T03:34:11 lol, we can add a constructor field to override settings of course. and we should do that really to avoid having people with fullscreen mode as default not seeing the front page 2011-05-16T03:34:13 Frontier: you already have a save settings... 2011-05-16T03:35:02 hypertriangle renders a crazy bunch of red and black text, horizontal lines and escape sequences 2011-05-16T03:35:32 is the visualiser on that page meant to error on me? 2011-05-16T03:35:43 McLeopold: I thought you pointed me at http://stackoverflow.com/questions/1442411/using-memcache-vs-memcached-with-php before as the reason to use php-memcached not php-memcache 2011-05-16T03:35:43 looks like it's trying to load an apache dir listing... 2011-05-16T03:36:00 janzert: that wasn't me 2011-05-16T03:36:32 it looks like ... it lives and browses the file system. it evolved, omg. reddit this!!!11oneeleven 2011-05-16T03:37:16 okay, maybe it was me, but it's a small detail 2011-05-16T03:37:42 2011-04-07T00:20:00 http://stackoverflow.com/questions/1442411/using-memcache-vs-memcached-with-php 2011-05-16T03:38:04 heh, was busy looking for it :P 2011-05-16T03:38:28 Fetching replay from: games/ <-- this is wrong obviously. 2011-05-16T03:38:30 yeah, I read the page and then thought, crap, I've seen this before :p 2011-05-16T03:38:37 Frontier: memcache issue... 2011-05-16T03:39:15 oh how does that happen? i mean how can "games/0/0.replay" become "games/" ? 2011-05-16T03:39:40 aichallenge: McLeopold epsilon * rcde2531 / website/index.php : fixed var issue with default game - http://bit.ly/jnofuy 2011-05-16T03:39:49 Frontier: btw, it's now "/games/0" 2011-05-16T03:39:58 thx to the magic of mod_rewrite 2011-05-16T03:42:42 McLeopold: currently I set the playback speed as 1 turn / second, but keep the playback time at max 1 minute. That means any game longer than 60 turns gets sped up. That doesn't work too well with a simple +/-. So I wonder if I should drop the automagic speed correction or make it configurable. 2011-05-16T03:43:27 *** chris__0076 has joined #aichallenge 2011-05-16T03:43:31 It's some fairly simple math to set it either way. I like that idea 2011-05-16T03:44:04 can't +/- buttons change based on a percentage? 2011-05-16T03:44:22 http://hypertriangle.com:13080/ looking kinda good 2011-05-16T03:44:22 say 25% faster or slower each time they're pressed 2011-05-16T03:44:23 ? 2011-05-16T03:45:15 janzert: sure, the point is I have a time limit in place: no replay runs longer than a minute no matter what the speed setting is. 2011-05-16T03:45:15 *** qacek has quit IRC (Quit: WeeChat 0.3.4) 2011-05-16T03:45:50 ahh, I thought the time limit was just for an initial speed calculation 2011-05-16T03:46:11 *** Chris_0076 has quit IRC (Ping timeout: 240 seconds) 2011-05-16T03:46:15 who would want to watch a 900 turns game at 1 or 2 turns a second? that would take 7:30 to watch 2011-05-16T03:46:30 well, for a starting speed, sure, but you want to be able to slow it down when reviewing a match 2011-05-16T03:46:47 I certainly would when wanting to check out a specific part of the battle 2011-05-16T03:47:11 ... which makes it difficult again to save that setting, because it depends on the actual replay length :p 2011-05-16T03:47:39 *** berak has joined #aichallenge 2011-05-16T03:48:08 for now, don't save it 2011-05-16T03:48:16 but at least give the controls 2011-05-16T03:48:20 can't you set a moves to play per second variable when first loading based on the replay length then just modify that variable faster or slower when requested by the buttons? 2011-05-16T03:49:05 you could even save it as a default speed as well if you want to provide a quick way for them to get back to a reasonable speed 2011-05-16T03:49:05 yeah sure, so no saving of the speed setting. 2011-05-16T03:50:24 *** delt0r___ has quit IRC (Read error: Operation timed out) 2011-05-16T03:50:55 janzert: http://hypertriangle.com:13080/games/6 2 timeouts, for sample_bots, that is worrisome 2011-05-16T03:51:22 *** delt0r___ has joined #aichallenge 2011-05-16T03:52:18 http://hypertriangle.com:13080/visualizer.php?game_id=6 also seems worrisome 2011-05-16T03:52:46 open that issue up for sigh 2011-05-16T03:53:08 I don't even know what to file ;) 2011-05-16T03:53:14 Yeah, someone else reported that as well with one of the start ants dying after 4 turns like this 2011-05-16T03:53:24 you or Frontier are almost certainly better to diagnose that 2011-05-16T03:53:29 janzert: that's a bug in the ants.py engine 2011-05-16T03:54:22 May I assume this happens when a bot times out? 2011-05-16T03:54:22 refresh the front page, game 7 ran \o/ 2011-05-16T03:54:23 that is a really old game 2011-05-16T03:54:44 *** berak has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2011-05-16T03:54:53 antimatroid1: these are games being produced right now on hypertriangle! 2011-05-16T03:54:58 submit your bots! 2011-05-16T03:55:30 *** berak has joined #aichallenge 2011-05-16T03:55:48 oh I see I have to change the colors above player number 4 again ;) 2011-05-16T03:57:10 btw I issued a "later tell sigh" about the bug in replay 6. I haven't seen him here for a while though 2011-05-16T03:57:24 @seen sigh 2011-05-16T03:57:24 Frontier: sigh was last seen in #aichallenge 3 days, 19 hours, 36 minutes, and 18 seconds ago: oh one thing with js closures... if you aren't careful with closures you can create cyclic references which aren't handled well in all browsers 2011-05-16T04:00:34 Fronter: can there be options to strip the UI from the replay, for the front page? 2011-05-16T04:05:34 arr, just make a flash video already ;). I'll do that. 2011-05-16T04:05:46 I mean... strip the ui as an option 2011-05-16T04:06:06 McLeopold: oh, didn't realise that :P 2011-05-16T04:06:26 I think compilation is failing though :( 2011-05-16T04:06:29 what should be visible besides the map? scores, ant counts, ...? 2011-05-16T04:06:32 yeah 2011-05-16T04:06:44 Frontier: just remove buttons 2011-05-16T04:06:52 links should still work 2011-05-16T04:10:57 *** smiley1983 has joined #aichallenge 2011-05-16T04:12:17 *** stocha has joined #aichallenge 2011-05-16T04:12:19 hello. 2011-05-16T04:12:33 how do one submit a bot ? 2011-05-16T04:12:42 *** chris___0076 has joined #aichallenge 2011-05-16T04:13:42 *** Naktibalda has joined #aichallenge 2011-05-16T04:14:28 *** chris__0076 has quit IRC (Ping timeout: 258 seconds) 2011-05-16T04:17:08 Will I hurt the temporary hypertriangle server if I make an account and attempt to upload the ocaml starter package? 2011-05-16T04:17:22 probably not 2011-05-16T04:18:17 I might try that then :) 2011-05-16T04:21:17 *** boegel has joined #aichallenge 2011-05-16T04:21:28 *** chris__0076 has joined #aichallenge 2011-05-16T04:21:28 *** McLeopold has left #aichallenge 2011-05-16T04:21:34 it didn't like it when i tried to create an account :\ 2011-05-16T04:22:06 as if passwords need to be 8 characters 2011-05-16T04:23:08 yeah, I'd like to see that restriction removed too 2011-05-16T04:23:17 it's not like we're a bank ;) 2011-05-16T04:24:28 it shouldn't really be necessary for a bank either 2011-05-16T04:24:42 they usually block your account for 24 hours if you screw the password up 3 times or so 2011-05-16T04:25:08 *** chris___0076 has quit IRC (Ping timeout: 276 seconds) 2011-05-16T04:29:30 *** chris___0076 has joined #aichallenge 2011-05-16T04:30:06 smiley1983, you have successfuly created an account ? 2011-05-16T04:30:22 you do know what a brute force password attack is? 2011-05-16T04:31:22 I think who ever set a minimum length of 8 had good intentions 2011-05-16T04:32:22 *** chris__0076 has quit IRC (Ping timeout: 246 seconds) 2011-05-16T04:33:04 Maybe one of the 5000 participants thinks the whole challenge is unfair and starts to crack user's passwords. :) 2011-05-16T04:33:35 *** smiley1983 has quit IRC (Ping timeout: 276 seconds) 2011-05-16T04:33:58 *** smiley1983 has joined #aichallenge 2011-05-16T04:34:37 Do you remember the HL2 source code leak? Gabe Newell, CEO of Valve had 'gaben' as his password -.-. With a minimum length of 8 this would not happen 2011-05-16T04:36:50 yeah, it'd be gabenewell :P 2011-05-16T04:36:50 would be "gabenewell" 2011-05-16T04:36:57 hehe 2011-05-16T04:36:57 ) 2011-05-16T04:37:48 Frontier, ok somebody has my pass. What can he do? 2011-05-16T04:37:50 you have to get a hold of the contest database to start cracking passwords and at that point we're screwed anyway 2011-05-16T04:38:08 .. i still haven't got a confirmation email 2011-05-16T04:38:50 i think the tmp server isn't ready to registration 2011-05-16T04:39:23 *** chris__0076 has joined #aichallenge 2011-05-16T04:39:53 antimatroid1: looks like email isn't setup 2011-05-16T04:40:17 just click the "Click Here" link at the bottom of the page after registering 2011-05-16T04:41:38 if you already closed that page then you'll need someone that has an account on hypertriangle to set the confirmation manually 2011-05-16T04:41:49 *** chris___0076 has quit IRC (Ping timeout: 246 seconds) 2011-05-16T04:43:04 with your account password guessed right, or brute forced, someone can replace your bot with an own bot of course any time you upload 2011-05-16T04:43:52 which isn't like accessing my bank account ;) 2011-05-16T04:43:59 i clicked the confirm link but still can't login :\ 2011-05-16T04:44:17 antimatroid1: hmm, it worked for me http://hypertriangle.com:13080/profile.php?user_id=29 2011-05-16T04:44:21 many ppl have the same password for several websites, that's the problem 2011-05-16T04:44:44 i am profile 28 :\ 2011-05-16T04:44:59 i posted one of my main passwords in here a few weeks ago 2011-05-16T04:45:10 dr-: and forcing them to use a password with at least length 8 won't do anything to change that 2011-05-16T04:45:11 i'm pretty sure it's now different for anything i care about 2011-05-16T04:45:20 yeah, i just throw a 1 on the end :P 2011-05-16T04:46:09 yesterday, a satire show asked someone in the pedestian prescinct to give them all their bank account information so they can look for that on the internet and remove it. the person was so aware of securing personal data that she agreed. 2011-05-16T04:46:34 how do you think fishing works? :) 2011-05-16T04:47:02 or phishing if you prefer 2011-05-16T04:47:31 I don't know, I am no phish I hope/think 2011-05-16T04:48:04 what are you.. 2011-05-16T04:49:25 well I remember a mail from my bank written with a complete lack of grammar asking me to enter my pin. i found it rather funny 2011-05-16T04:52:16 janzert, look at forum and delete spam, please 2011-05-16T04:52:39 oh no, are they back? 2011-05-16T04:53:37 one seems to slip through every so often still 2011-05-16T04:54:09 they are only active users on forum now 2011-05-16T05:34:02 *** Chris_0076 has joined #aichallenge 2011-05-16T05:36:04 *** chris__0076 has quit IRC (Ping timeout: 246 seconds) 2011-05-16T05:51:05 *** Chris_0076 has quit IRC (Ping timeout: 260 seconds) 2011-05-16T06:03:39 *** Chris_0076 has joined #aichallenge 2011-05-16T06:13:27 *** needsch has joined #aichallenge 2011-05-16T06:18:46 *** smiley1983 has quit IRC (Ping timeout: 246 seconds) 2011-05-16T06:20:58 *** smiley1983 has joined #aichallenge 2011-05-16T06:30:28 wtf is "differential viewpoints"? 2011-05-16T06:56:22 and does the stuff on hypertriangle work in IE8? 2011-05-16T06:57:41 It seems to me that since the the visualizer had been moved into a php widget the Java version doesn't get triggered any more. At least I cannot use "java=true" to override it in other browsers 2011-05-16T07:01:38 *** BtbN has joined #aichallenge 2011-05-16T07:13:37 *** Redgis has joined #aichallenge 2011-05-16T07:18:49 *** Redgis has quit IRC (Quit: ... mains libres) 2011-05-16T07:53:01 *** boegel has quit IRC (Quit: *poof!*) 2011-05-16T08:04:51 hmm, now I seem to need modrewrite or something like that to load replays? 2011-05-16T08:29:18 wtf amazon and wikipedia 2011-05-16T08:29:31 they both have the wrong season listed for "Rapture's Delight" the american dad episode 2011-05-16T08:29:58 by amazon i of course mean imdb :P 2011-05-16T08:36:51 aichallenge: janzert epsilon * r66fc2c0 / worker/sandbox.py : Correct read timeouts in the sandbox - http://bit.ly/mwrB5k 2011-05-16T08:36:52 aichallenge: janzert epsilon * r7f1a3b0 / (setup/server_setup.py website/index.php): Merge branch 'epsilon' of github.com:aichallenge/aichallenge into epsilon - http://bit.ly/lEFJNN 2011-05-16T08:43:35 aichallenge: janzert epsilon * r123607e / worker/sandbox.py : Set the correct timeout parameter - http://bit.ly/l7k8JB 2011-05-16T08:45:12 aichallenge: janzert epsilon * r4e2beb1 / worker/sandbox.py : Fix parameter name typo - http://bit.ly/l6Hilz 2011-05-16T08:45:37 *** sigh has joined #aichallenge 2011-05-16T08:53:50 *** FireFly has joined #aichallenge 2011-05-16T08:58:59 1s * 6 players * 600 turns = 1 hour for one game to complete :p 2011-05-16T09:03:44 *** Bhaskerh has joined #aichallenge 2011-05-16T09:17:18 antimatroid1, one guy on russian forum is asking about C++ bot. He doesn't have time.h file. File from VS2010 doesn't work for him 2011-05-16T09:21:18 Ice_Harley: I'll take a look 2011-05-16T09:21:23 i really need to update that bot anyway, i'll do that now 2011-05-16T09:21:50 ok 2011-05-16T09:22:41 where is it? :P 2011-05-16T09:23:46 what? 2011-05-16T09:24:04 the forum thread? 2011-05-16T09:24:12 *** sigh has quit IRC (Ping timeout: 240 seconds) 2011-05-16T09:24:16 http://www.sdchess.net/phpBB2/viewtopic.php?f=7&t=454&p=10427#p10427 2011-05-16T09:24:35 it's russian anyway :) 2011-05-16T09:24:43 wtf chrome 2011-05-16T09:24:55