2011-04-07T00:01:46 *** dlila has quit IRC (Quit: Leaving) 2011-04-07T00:08:21 hmm :( 2011-04-07T00:08:47 It will be faster. It's the choosing of the map that is slowing me down, which isn't that critical. 2011-04-07T00:09:02 even 1.5 seconds will be workable 2011-04-07T00:09:04 I'll be happy if we get it to 500ms 2011-04-07T00:09:10 just would be nice to be faster 2011-04-07T00:09:12 yeah 2011-04-07T00:09:45 This time around, I don't think the database will be to much of an issue. I plan to get memcache working for most php pages. 2011-04-07T00:10:06 great 2011-04-07T00:10:07 So it should be free to churn on matchups and ranking. 2011-04-07T00:10:28 btw, can you add memcache to the install scripts? 2011-04-07T00:10:33 what I can do will be very limited from here on out, work is picking up into our busy season now 2011-04-07T00:10:36 I can create an issue if you want 2011-04-07T00:10:52 just using the apt package? 2011-04-07T00:11:10 yeah, that's what I'm using locally. 2011-04-07T00:11:44 I've never really looking into things like NoSQL, CouchDB or MongoDB 2011-04-07T00:12:23 just need to add the package name to the list at line 27 in install_website_packages() of server_setup.py then 2011-04-07T00:12:58 And something to get it in the startup scripts 2011-04-07T00:13:16 apt won't have it start automatically already? 2011-04-07T00:13:39 I'm still new to ubuntu 2011-04-07T00:13:57 I'm used to gentoo, you always roll your own. :) 2011-04-07T00:14:05 ahh, I'd be rather surprised if apt doesn't configure it to start 2011-04-07T00:14:58 we may need to add some manual configuration changes though but the defaults are probably fine to start with 2011-04-07T00:15:32 from the little I've used memcache it seems like the only major knob to twist is how much memory to use 2011-04-07T00:15:38 am I remembering right? 2011-04-07T00:15:50 yep 2011-04-07T00:16:10 plus telling it of other server to help with the load, but we would only have one 2011-04-07T00:16:39 in which case I wouldn't bother worrying with that for now since I've no idea what the final server capability will look like 2011-04-07T00:16:55 I'm don't know if I'm using php-memcache or php-memcached :/ 2011-04-07T00:17:04 hehe 2011-04-07T00:17:48 oh, php-memcached is newer and better 2011-04-07T00:18:39 and looks like php-memcache is just the php interface not the actual memcached daemon 2011-04-07T00:19:09 both php-* are just clients 2011-04-07T00:19:46 http://packages.ubuntu.com/natty/php5-memcached 2011-04-07T00:20:00 http://stackoverflow.com/questions/1442411/using-memcache-vs-memcached-with-php 2011-04-07T00:20:36 oops, yeah 2011-04-07T00:20:43 *** sigh has quit IRC (Remote host closed the connection) 2011-04-07T00:20:57 I saw the libmemcached5 dependency and thought it was daemon itself 2011-04-07T00:21:11 janzert: do you think you could do the chroot stuff before you can't work on this anymore? 2011-04-07T00:21:30 was jbroman not working on it? 2011-04-07T00:21:37 I don't know 2011-04-07T00:21:40 I could try 2011-04-07T00:21:50 It's been really quite the last 2 weeks 2011-04-07T00:21:57 I haven't really done much myself 2011-04-07T00:22:09 although mainly I was hoping once you had the pieces in place that everything worked end to end I would finish off the setup scripts 2011-04-07T00:22:33 I'm more worried about chroot than install scripts 2011-04-07T00:22:45 it shouldn't take much to do that so should be able to squeeze it in :} 2011-04-07T00:22:58 yeah, it's rather a bigger benefit 2011-04-07T00:23:06 more work too I think ;) 2011-04-07T00:23:51 although even more important is getting the sigstop stuff working (and seeing if it will really work) 2011-04-07T00:24:06 I'm hoping sigh will look at that. 2011-04-07T00:24:18 Also, Frontier is worried about buffer stuff with the engine. 2011-04-07T00:24:43 was just typing about that :) 2011-04-07T00:25:09 the easiest portable way to do it is have a thread for each bots input and output 2011-04-07T00:25:42 kinda like dmj setup? 2011-04-07T00:25:46 so long as we're only going to have 10's and not hundreds of bots in a game it should work just fine 2011-04-07T00:26:04 not sure what he did, I've never looked at his tcp server code at all 2011-04-07T00:26:19 I'm pretty sure you did 2011-04-07T00:26:29 I wouldn't be surprised though since it is a common technique 2011-04-07T00:26:32 there was a comment about mixing threads and processes... 2011-04-07T00:26:57 maybe back in the tron timeframe I might have but I don't remember doing it :) 2011-04-07T00:27:51 you can look at my arimaa AEI interface for an example of doing it too, although it's not quite what you want so probably not as good of an example if it's also in the tcp server 2011-04-07T00:29:15 actually though now that I think about it more I'd be a bit surprised if he is using it in the tcp server 2011-04-07T00:29:38 non-blocking socket IO is pretty easy and portable 2011-04-07T00:29:59 that's where I got the example that you looked at 2011-04-07T00:30:31 ok, I don't remember the example now though 2011-04-07T00:31:47 basic idea in python is kick of a thread that just reads or writes to the bot and then communicate to it from the main thread using a queue 2011-04-07T00:32:08 I think the current sandbox does that 2011-04-07T00:32:50 I think the old sandbox did for reads, but not writes 2011-04-07T00:33:01 not sure if the current one still does 2011-04-07T00:33:40 yeah, your right, not for writes 2011-04-07T00:35:09 yep, looks like 2011-04-07T00:36:24 anything else offhand? I've got a headache and looking at code isn't doing it any good 2011-04-07T00:36:37 going to take off unless you need something 2011-04-07T00:36:40 just chroot 2011-04-07T00:37:14 yeah, I'll try and look at getting that setup next time I get some time 2011-04-07T00:37:30 thx 2011-04-07T00:38:09 see you around 2011-04-07T00:38:17 later 2011-04-07T01:06:03 *** chris__0076 has quit IRC (Quit: Leaving) 2011-04-07T01:23:40 *** fgump has joined #aichallenge 2011-04-07T01:24:58 *** McLeopold has left #aichallenge 2011-04-07T01:46:05 *** Daniel350 has joined #aichallenge 2011-04-07T02:26:26 aichallenge: McLeopold epsilon * r7fca3fb / sql/2_generate_matchup.sql : updated generate_matchups sp, added some sql for setting positions (in progress) - http://bit.ly/hFFsmP 2011-04-07T02:27:41 *** sigh has joined #aichallenge 2011-04-07T02:37:44 *** lundgren has joined #aichallenge 2011-04-07T03:13:56 *** Stocha has joined #aichallenge 2011-04-07T03:37:37 *** amstan has quit IRC (Remote host closed the connection) 2011-04-07T03:37:43 *** Stocha has quit IRC (Quit: Page closed) 2011-04-07T03:50:45 *** antimatroid has quit IRC (Ping timeout: 252 seconds) 2011-04-07T03:58:45 *** boegel has joined #aichallenge 2011-04-07T04:00:14 *** stocha has joined #aichallenge 2011-04-07T04:02:55 *** stocha has quit IRC (Client Quit) 2011-04-07T04:06:00 *** antimatroid has joined #aichallenge 2011-04-07T04:07:11 *** aerique has joined #aichallenge 2011-04-07T04:10:17 aichallenge: sigh epsilon * r326515f / worker/engine.py : Ensure that bots can't DoS engine with a large stream of input without "go" - http://bit.ly/fBDeEu 2011-04-07T04:10:25 *** antimatroid has quit IRC (Ping timeout: 260 seconds) 2011-04-07T04:28:53 *** antimatroid has joined #aichallenge 2011-04-07T04:29:37 *** fgump has quit IRC (Quit: Leaving) 2011-04-07T04:30:22 *** lundgren has quit IRC (Quit: Page closed) 2011-04-07T04:58:09 Frontier: http://pastebin.com/nkMvs2cV - bugged replay (perhaps too long? it went up to 1000 rounds) 2011-04-07T05:01:03 Frontier: http://pastebin.com/3cxRPTiU - one more, much smaller 2011-04-07T05:03:12 *** Stocha has joined #aichallenge 2011-04-07T05:39:22 *** Naktibalda has joined #aichallenge 2011-04-07T06:08:12 *** Keth has quit IRC (Read error: Connection reset by peer) 2011-04-07T06:08:20 *** Keth has joined #aichallenge 2011-04-07T06:32:05 *** p4p4p5 has joined #aichallenge 2011-04-07T06:42:57 *** delt0r_ has quit IRC (Ping timeout: 246 seconds) 2011-04-07T06:43:19 *** Eruonen has joined #aichallenge 2011-04-07T06:49:21 *** needsch has joined #aichallenge 2011-04-07T07:29:03 *** delt0r has joined #aichallenge 2011-04-07T07:30:24 *** Daniel350 has quit IRC (Ping timeout: 252 seconds) 2011-04-07T07:54:47 *** Keth has quit IRC (Read error: Connection reset by peer) 2011-04-07T07:55:09 *** Keth has joined #aichallenge 2011-04-07T07:57:01 *** Daniel350 has joined #aichallenge 2011-04-07T08:03:36 Keth: I get a "Number of orders does not match life span." with that replay. 2011-04-07T08:04:59 but I have an idea. it must be another regression from the json switch 2011-04-07T08:05:22 Frontier: no idea either 2011-04-07T08:08:42 ok lets do the math... the line says 'end turn' = 756 and 'conversion turn' = 70 2011-04-07T08:09:29 that makes 686 turns that must be filled with orders (unless this ant survived the game, but you said it went over 1000 turns) 2011-04-07T08:11:32 you telling me it's my bot? 2011-04-07T08:11:34 blame the negine 2011-04-07T08:11:35 engine 2011-04-07T08:11:36 :3 2011-04-07T08:11:42 I see 684 commands in the command string. But even if the ant survived that would only allow it to have one less move order than life span 2011-04-07T08:11:48 the thing is 2011-04-07T08:11:50 I gave you two 2011-04-07T08:11:55 the second one has only like 80 turns 2011-04-07T08:12:02 and gives same error 2011-04-07T08:12:19 see if you can figure out the shorter one 2011-04-07T08:13:16 ok command string has 42 chars 2011-04-07T08:13:18 we'll need some regression tests for then engine somehow 2011-04-07T08:13:35 *** Stocha has quit IRC (Quit: Page closed) 2011-04-07T08:13:41 ant is owned by player from 0 - 44 2011-04-07T08:14:36 same story. somehow the ant lives longer than the engine outputs orders for it 2011-04-07T08:15:43 i guess report to whoever does the engine since I honestly have no clue how the repaly information is stored 2011-04-07T08:16:03 I just thought I would let you know I often get these with my bot and stocha's 2011-04-07T08:16:51 I know it might be bot fault but engine shouldn't output the corrupted replay in first place 2011-04-07T08:18:21 sigh can probably tell you what is going wrong. But I can imagine it happens with other bots, too. 2011-04-07T08:19:00 i see 2011-04-07T08:19:20 so I talk to sigh and I tell him what? 2011-04-07T08:19:21 Any bot that survives the game has one less order than it should. Or maybe the other way round: The engine tries to artificially prolong the replay. 2011-04-07T08:19:30 what's happening? 2011-04-07T08:19:35 =D 2011-04-07T08:19:35 :D 2011-04-07T08:19:46 one replay 2011-04-07T08:19:48 well 2011-04-07T08:19:51 many replays are corrupted 2011-04-07T08:20:05 This chat turns out to be a sketch comedy 2011-04-07T08:20:13 hmm... those replay's that you pasted above? 2011-04-07T08:20:21 let me paste again, there is one short 2011-04-07T08:20:21 why are they the old style replays? 2011-04-07T08:20:42 that's true 2011-04-07T08:20:49 people like 'em better that way 2011-04-07T08:21:07 more to the point, the latest engine can't generate them 2011-04-07T08:21:16 yeah, i realized ujst now 2011-04-07T08:21:30 sketch comedy over 2011-04-07T08:31:34 *** FireFly has joined #aichallenge 2011-04-07T08:33:35 sigh: Is it the same bug that you already solved a few days ago? 2011-04-07T08:34:05 Frontier: no idea... probably 2011-04-07T08:34:41 I thought it was maybe another attempt at adding a 'bonus round'. 2011-04-07T08:36:10 *shrug* there's been a few bugs fixed... I'm not going to worry about errors in old versions 2011-04-07T08:37:25 sigh: I thought it was a new bug since it looked like the bug with the scores having an additional entry. 2011-04-07T08:37:51 Its good to know that I can reject any non-json replay now :) 2011-04-07T08:38:02 ...I mean in bug reports 2011-04-07T08:38:39 Frontier: to be honest it sounds more like a bug I had ages ago in the conflict resolution 2011-04-07T08:55:45 *** p4p4p5 has quit IRC (Remote host closed the connection) 2011-04-07T08:56:13 *** smellyhippy has quit IRC (Ping timeout: 276 seconds) 2011-04-07T08:58:09 sigh: the leap seconds issue looks like an april fool's joke ^^ 2011-04-07T08:59:05 I didn't know about it though. It was an interesting read. May come handy one day at a table quiz 2011-04-07T09:00:03 yeah, I don't know why he put that in there 2011-04-07T09:22:17 *** amstan has joined #aichallenge 2011-04-07T09:22:18 *** ChanServ sets mode: +o amstan 2011-04-07T09:29:15 *** sigh has quit IRC (Remote host closed the connection) 2011-04-07T09:33:27 *** amstan has quit IRC (Remote host closed the connection) 2011-04-07T09:54:47 why do we have compile.py and compile_anything.py and compile.py doesn't run because it tries to call a function with 0 arguments that takes 1 argument. *puzzled* 2011-04-07T10:12:25 *** helo_ has joined #aichallenge 2011-04-07T10:17:34 *** Daniel350 has quit IRC (Quit: Ex-Chat) 2011-04-07T10:33:04 aichallenge: Marco Leise epsilon * rd92ba5f / (worker/server_info.py.sample worker/worker_setup.py): added D language .deb to worker setup script / added missing comma in server_info.py.sample - http://bit.ly/f1J8mM 2011-04-07T10:37:30 @later tell McLeopold: Please review your commit on compiler.py from the 15th of March. It looks as if the old compile_anything.py and the new compiler.py got mixed up. Also the mysql import may no longer be required. 2011-04-07T10:37:30 Frontier: As you wish. 2011-04-07T10:42:47 aichallenge: aerique epsilon * rf8b0c60 / ants/dist/starter_bots/common_lisp/src/game-state.lisp : Prevent dead ants from overwriting food tiles. - http://bit.ly/gRxYGn 2011-04-07T10:51:08 *** boegel has quit IRC (Ping timeout: 246 seconds) 2011-04-07T10:57:50 *** aerique has quit IRC (Quit: ...) 2011-04-07T11:01:25 *** contestbot_ has joined #aichallenge 2011-04-07T11:01:26 *** contestbot has quit IRC (Disconnected by services) 2011-04-07T11:03:44 *** contestbot_ is now known as contestbot 2011-04-07T11:03:46 *** ChanServ sets mode: +o contestbot 2011-04-07T11:07:11 *** contestbot_ has joined #aichallenge 2011-04-07T11:07:12 *** contestbot has quit IRC (Disconnected by services) 2011-04-07T11:09:06 *** contestbot_ is now known as contestbot 2011-04-07T11:09:08 *** ChanServ sets mode: +o contestbot 2011-04-07T11:16:18 *** needsch has quit IRC (Quit: Leaving.) 2011-04-07T11:16:35 Since it just came out yesterday I'm guessing the leap second issue was a reaction to http://queue.acm.org/detail.cfm?id=1967009 2011-04-07T11:17:31 but I think for us it's clearly YAGNI and adds way more complexity than it's worth to handle it 2011-04-07T11:23:19 *** needsch has joined #aichallenge 2011-04-07T11:29:29 *** McLeopold has joined #aichallenge 2011-04-07T11:33:35 Frontier: you around? 2011-04-07T11:34:15 hi 2011-04-07T11:36:58 McLeopold: Yeah, I found that I could not run compiler.py when I wanted to add support for D and noticed that several parts of the code don't fit together as if they were taken from different sources/dates. 2011-04-07T11:40:05 Frontier: you are right that I did put that together from a few other files. 2011-04-07T11:40:15 The changes were to support compiling on the workers. 2011-04-07T11:40:35 Plus, to remove any language data from the database 2011-04-07T11:40:57 BTW, the leap seconds issue is a total joke :D 2011-04-07T11:41:38 *** Chris_0076 has joined #aichallenge 2011-04-07T11:42:04 McLeopold: I said it looks like an april fool's joke. The previous version of compiler.py looks more consistent. 2011-04-07T11:43:42 does the current version break for you? 2011-04-07T11:45:18 It breaks badly. Look at the main method for example. compile_anything requires an argument, but doesn't return a list of results 2011-04-07T11:45:37 (these are two separate issues) 2011-04-07T11:46:03 oh, I guess I should have removed the main, look at how the worker.py uses it 2011-04-07T11:46:05 It's really a crazy mix. 2011-04-07T11:47:08 hmm, it would be nice if I could have the main back to run tests with it 2011-04-07T11:47:46 let me fix it then 2011-04-07T11:47:54 also there is still a compile_anything.py that should be removed I think :) 2011-04-07T11:48:20 remove the mysql import while you are at it. 2011-04-07T11:50:16 *** amstan has joined #aichallenge 2011-04-07T11:50:16 *** ChanServ sets mode: +o amstan 2011-04-07T11:56:04 *** Kingpin13 has quit IRC (Quit: You're a kitty!) 2011-04-07T12:06:19 *** needsch has quit IRC (Quit: Leaving.) 2011-04-07T12:19:21 *** skynetz has joined #aichallenge 2011-04-07T12:28:09 *** delt0r has quit IRC (Ping timeout: 276 seconds) 2011-04-07T12:44:15 janzert/amstan: can you answer a compiling question? 2011-04-07T12:44:24 McLeopold: compiling what? 2011-04-07T12:44:25 maybe 2011-04-07T12:44:29 java 2011-04-07T12:44:42 ok, try me, but i doubt it 2011-04-07T12:45:03 in the last contest, did a java submission get compiled to a class file or jar file? 2011-04-07T12:46:35 i think it was class 2011-04-07T12:48:08 that's going to break my detect_language function 2011-04-07T12:54:17 amstan: language is detected by finding a MyBot.* file, when java is compiled I get a MyBot.class and MyBot.jar file, if I attempt to detect the language again, it fails because MyBot.class could be scala 2011-04-07T12:54:48 why are you detecting language after compilation? 2011-04-07T12:54:54 shouldn't this be done before? 2011-04-07T12:55:13 don't you have the same problem with compiled languages, like Haskell/C 2011-04-07T12:55:19 because we pulled the language info out of the database 2011-04-07T12:55:23 they both get compiled to a binary MyBot 2011-04-07T12:55:57 So, I need a worker database of submission_id's to languages, or we write a file to the dir that can help us detect 2011-04-07T12:56:11 if we write a file to the dir, a side affect is people can self declare the language 2011-04-07T12:56:23 maybe a good thing? 2011-04-07T12:57:27 Or, I could create a run.sh file after compilation, and not need to detect the language again? 2011-04-07T12:58:10 amstan: again, if we didn't overwrite, it could have interesting side affects 2011-04-07T12:58:47 i think that's fine... 2011-04-07T13:02:12 I thought Java was compiled to .jar, but look at compile_anything.py to know. Anyway why do we keep .o and .class files? If we cleaned them up we could detect the language again. A shell wrapper is ok I guess. I'd prefer to just load the executable for the compiled languages. It is the purist in me… 2011-04-07T13:04:05 *** mceier has joined #aichallenge 2011-04-07T13:04:29 Frontier: there could always be another conflict, it is better to just detect the language once and be done with it 2011-04-07T13:04:49 Also, I could just load the text from the run.sh and use that as my command directly 2011-04-07T13:06:20 :) 2011-04-07T13:17:41 *** smellyhippy has joined #aichallenge 2011-04-07T13:24:00 *** contestbot_ has joined #aichallenge 2011-04-07T13:24:01 *** contestbot has quit IRC (Disconnected by services) 2011-04-07T13:25:07 aichallenge: McLeopold epsilon * r814d2ba / (worker/compiler.py worker/worker.py): Compiler.py can now be run directly for testing. Added run.sh as output to compiling for language detection. - http://bit.ly/g5Pugp 2011-04-07T13:25:07 *** contestbot_ is now known as contestbot 2011-04-07T13:25:08 *** ChanServ sets mode: +o contestbot 2011-04-07T13:25:19 Frontier: ok, try the compiler.py now, let me know how it goes 2011-04-07T13:27:15 *** needsch has joined #aichallenge 2011-04-07T13:28:41 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 3.6.16/20110319135224]) 2011-04-07T13:31:11 *** amstan has quit IRC (Ping timeout: 248 seconds) 2011-04-07T13:51:55 you're not saving the compile output? 2011-04-07T13:53:03 well, saving as in holding onto in a variable until it can be emailed to the user 2011-04-07T14:00:13 *** andy____ has joined #aichallenge 2011-04-07T14:02:43 Zannick: do we really want to email compile output to the users? 2011-04-07T14:03:11 seemed like a good idea last contest 2011-04-07T14:03:19 My stuff always worked :) 2011-04-07T14:03:33 Not everyone's did 2011-04-07T14:03:34 I'm gathering error messages, which could be sent 2011-04-07T14:03:51 I figured the progress logging was worthless 2011-04-07T14:04:03 *** needsch has quit IRC (Quit: Leaving.) 2011-04-07T14:07:04 *** needsch has joined #aichallenge 2011-04-07T14:09:19 the main problems that output is useful for is when it doesn't compile as expected but succeeds. 2011-04-07T14:10:01 i can't think of an example 2011-04-07T14:10:03 :P 2011-04-07T14:10:27 *** [1]Kender <[1]Kender!~Kender@5ED3098D.cm-7-4a.dynamic.ziggo.nl> has joined #aichallenge 2011-04-07T14:10:35 i'm not going to worry about it 2011-04-07T14:10:44 <[1]Kender> good evening 2011-04-07T14:11:38 *** [1]Kender is now known as Kender2 2011-04-07T14:12:28 *** mceier has quit IRC (Quit: leaving) 2011-04-07T14:13:03 could the aichallenge worker, manager, etc code also be used for continous competition? 2011-04-07T14:13:19 rather than specific events 2011-04-07T14:19:48 *** choas has joined #aichallenge 2011-04-07T14:20:32 Are there even any other sites running this apart from ai-contest.com ? 2011-04-07T14:23:47 *** delt0r has joined #aichallenge 2011-04-07T14:30:21 for java coders: http://bitbath.org "A real time strategy game where your code does the fighting." 2011-04-07T14:31:32 I am interested in making it multi-language so I was wondering if the aichallenge software might be a fit? 2011-04-07T14:32:40 Kender2: it might be 2011-04-07T14:33:48 I'm digging through the repository now, but python is not my strongest language 2011-04-07T14:34:22 it looks like it depends on 2 java class methods to work 2011-04-07T14:34:39 It would be some effort to abstract that out into a text protocol for communication 2011-04-07T14:35:04 oh yes, I am thinking about doing a rewrite from scratch 2011-04-07T14:35:12 Plus, you would need the game engine to interface with the text protocol 2011-04-07T14:35:14 to support multiple programming languages 2011-04-07T14:35:24 ok 2011-04-07T14:35:33 and then I stumbled upon aichallenge / ai-contest.com 2011-04-07T14:35:44 Then, first, you want to define the bot communication protocol, then write a game class. 2011-04-07T14:36:25 look at ants/ants.py as an example of a game engine, also look at ants/game.py as a "interface" of sorts which will work with worker/engine.py 2011-04-07T14:37:31 As far as a continuous contest goes, we will eventually have an official tcp server, which may suit your needs better, but another ai-context instance would work too. 2011-04-07T14:37:32 I can't figure out if it supports adding new contestants while the manager is running though 2011-04-07T14:37:40 hmm... 2011-04-07T14:37:41 tcp server? 2011-04-07T14:37:58 You mean the game is always running and you just enter the arena? 2011-04-07T14:38:08 yes 2011-04-07T14:38:36 at bitbath.org a new entry is pitted against all existing bots to determine it's ranking 2011-04-07T14:38:43 no, this is written for specific game instances with a set number of players 2011-04-07T14:39:07 so more like a tournament 2011-04-07T14:39:34 ? 2011-04-07T14:40:34 bitbath is not continuous, it is specific games, so ai-contest code would work 2011-04-07T14:41:05 yes, we are more like a tournament 2011-04-07T14:41:34 the worker and manager together control the matchups and maintain the rankings 2011-04-07T14:42:41 yes, I see now how they could be expanded to support my arena-style contest 2011-04-07T14:43:19 *** amstan has joined #aichallenge 2011-04-07T14:43:19 *** ChanServ sets mode: +o amstan 2011-04-07T14:43:47 to be clear, there are individual game instances with only 2 contestants, which isn't really an arena style game 2011-04-07T14:43:53 true 2011-04-07T14:44:04 but anyone can enter a new bot at any time 2011-04-07T14:44:14 yeah, for ai-contest, we keep it running until the deadline 2011-04-07T14:44:25 we could keep it running forever if we wanted to 2011-04-07T14:45:02 ah, then I misunderstood a very basic concept 2011-04-07T14:45:27 I thought the competition (manager/worker) would not start running until everyone had submitted their programds 2011-04-07T14:47:05 Guess I'll have to brush up on my python to understand the code a bit better 2011-04-07T14:47:17 and to be able to contribute 2011-04-07T14:47:51 it runs continuously, workers are separate servers that request a game to run, then the manager gives them a set of players to put in a game 2011-04-07T14:48:23 if a new submission is entered, the manager will probably schedule them for a game sooner than later 2011-04-07T14:49:44 *** amstan has quit IRC (Ping timeout: 246 seconds) 2011-04-07T14:51:17 ok thanks. I'm going to have a think about the direction I'm going to take. And this has certainly given me something interesting to consider. 2011-04-07T14:52:30 McLeopold: I see your changes, now I git stashed my changes, pulled yours and have to figure out how to merge. 2011-04-07T14:52:54 *** Kender2 is now known as Kender_away 2011-04-07T14:54:21 *** boegel has joined #aichallenge 2011-04-07T14:59:09 Frontier: welcome to hell 2011-04-07T14:59:14 merge hell 2011-04-07T14:59:34 *** delt0r_ has joined #aichallenge 2011-04-07T14:59:51 Frontier: do you need your changes? I'd rather maintain control of the compiler.py 2011-04-07T15:00:30 *** delt0r has quit IRC (Read error: Operation timed out) 2011-04-07T15:01:02 *** helo_ has quit IRC (Ping timeout: 252 seconds) 2011-04-07T15:03:01 Oh it is just the addition for D. Nothing that conflicts with your work. But I'd like to test it thoroughly here instead of just telling you what to add. 2011-04-07T15:03:59 Are you currently working on that file? 2011-04-07T15:04:21 Okay, got worried for a second. Merge away, I'm done with it. 2011-04-07T15:05:02 Probably easier to just copy paste your dict entries into the current version 2011-04-07T15:06:27 Frontier: you probably want to unstash (or stash pop) or maybe rebase 2011-04-07T15:07:08 *** needsch has quit IRC (Ping timeout: 258 seconds) 2011-04-07T15:07:29 Zannick: did that, now I need a gui tool to have a good look at the conflicts. I could pick out my changes manually, but it is not as much fun in the long run. 2011-04-07T15:07:51 eurgh, yeah. 2011-04-07T15:13:34 *** needsch has joined #aichallenge 2011-04-07T15:24:25 I am using git-cola now. It is written in Python and you can select lines in the diff viewer, right click and choose to undo or keep them 2011-04-07T15:27:34 *** Accoun has quit IRC () 2011-04-07T15:30:25 aichallenge: Marco Leise epsilon * r1a1ca86 / worker/compiler.py : added compiler support for D - http://bit.ly/er8AEB 2011-04-07T15:30:44 it worked ok :) 2011-04-07T15:32:53 *** dlila has joined #aichallenge 2011-04-07T15:39:33 *** Accoun has joined #aichallenge 2011-04-07T15:47:55 *** choas has quit IRC (Quit: Lost terminal) 2011-04-07T16:21:07 *** boegel has quit IRC (Quit: Leaving) 2011-04-07T16:37:40 *** skynetz has quit IRC (Quit: .) 2011-04-07T16:37:47 *** amstan has joined #aichallenge 2011-04-07T16:37:47 *** ChanServ sets mode: +o amstan 2011-04-07T16:39:57 *** Naktibalda has joined #aichallenge 2011-04-07T16:53:12 *** amstan_ has joined #aichallenge 2011-04-07T16:53:12 *** ChanServ sets mode: +o amstan_ 2011-04-07T16:58:41 *** amstan_ has quit IRC (Ping timeout: 246 seconds) 2011-04-07T17:39:25 *** amstan_ has joined #aichallenge 2011-04-07T17:39:25 *** ChanServ sets mode: +o amstan_ 2011-04-07T17:44:45 *** amstan_ has quit IRC (Ping timeout: 276 seconds) 2011-04-07T17:55:56 *** sigh has joined #aichallenge 2011-04-07T17:59:53 *** stocha has joined #aichallenge 2011-04-07T18:08:14 *** stocha has quit IRC (Quit: Page closed) 2011-04-07T18:09:18 *** sigh has quit IRC (Remote host closed the connection) 2011-04-07T18:20:09 aichallenge: McLeopold epsilon * rd0d9f0f / (sql/2_generate_matchup.sql website/api_get_task.php): on demand trueskill matchmaking - http://bit.ly/faWdmD 2011-04-07T18:33:57 aichallenge: McLeopold epsilon * recd7cba / (85 files in 4 dirs): cleanup of maps - http://bit.ly/hvs8qu 2011-04-07T18:37:12 antimatroid: why'd you nuke my .gitignore files? 2011-04-07T18:54:17 McLeopold: i got rid of one ages ago so i could add .txt files to the maps directory 2011-04-07T18:54:22 i'm not aware of doing it since then 2011-04-07T18:54:52 i can't remember who, but at the time i was like wtf to someone and they realised it was that and said just get rid of it for the moment :) 2011-04-07T19:00:33 you don't need to nuke them, just change them 2011-04-07T19:01:23 antimatroid: I'm renamed your .txt to .map 2011-04-07T19:01:29 *I've* 2011-04-07T19:01:58 McLeopold: consider adding parameters to map files for the "map generator used" 2011-04-07T19:02:03 and possibly even for row and col translate 2011-04-07T19:02:16 YES! I have thought of that. 2011-04-07T19:02:23 i've gotta head to the uni now 2011-04-07T19:02:33 I want to provide the seed used so you can regenerate the exact same map if needed. 2011-04-07T19:03:14 hmmm, for asymmetry i would still prefer if food spawning was the same for each map each game 2011-04-07T19:03:33 are we talking about spawning? 2011-04-07T19:03:39 otherwise it's even harder to tell how fair a map is 2011-04-07T19:03:43 yeah 2011-04-07T19:03:55 you could embed the seed in the map 2011-04-07T19:04:05 but then the food spawn becomes predictable 2011-04-07T19:04:16 only predictable if you play the same map over and over 2011-04-07T19:04:29 in the final contest new maps with seeds can be used 2011-04-07T19:04:36 I think we are talking about 2 things. Antimatroid wants to give hints to the engine for best food method, I want reproducible map generation. 2011-04-07T19:04:47 oh, and not provide the seed to the players 2011-04-07T19:04:56 i dunno why i thought that 2011-04-07T19:07:55 *** FireFly has quit IRC (Quit: swatted to death) 2011-04-07T19:09:00 "git status = nothing to commit" feels as good as "inbox = 0" 2011-04-07T19:09:20 OBTW 2011-04-07T19:09:35 aichallenge: McLeopold epsilon * r45a16a9 / (11 files): start of js engine - http://bit.ly/hlROnA 2011-04-07T19:09:39 aichallenge: McLeopold epsilon * r09ec0a4 / (3 files): some changes to submission test - http://bit.ly/gwgGFa 2011-04-07T19:09:41 aichallenge: McLeopold epsilon * readade8 / sql/matchups.sql.txt : cleanup of sql - http://bit.ly/guYHGE 2011-04-07T19:09:42 TLDR 2011-04-07T19:10:24 antimatroid: btw .gitignore just tells git what to ignore by default, you can still explicitly git add ignored-file 2011-04-07T19:13:35 my inbox has over 600 unread emails 2011-04-07T19:13:36 it's a mess 2011-04-07T19:13:37 *** amstan has quit IRC (Ping timeout: 248 seconds) 2011-04-07T19:37:40 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 3.6.16/20110323143040]) 2011-04-07T19:40:39 *** Eruonen has quit IRC (Ping timeout: 264 seconds) 2011-04-07T19:41:15 *** Eruonen has joined #aichallenge 2011-04-07T19:44:28 *** amriedle has joined #aichallenge 2011-04-07T19:51:35 *** Frontier has quit IRC (Quit: Leaving.) 2011-04-07T19:55:14 *** Keth has quit IRC (Read error: Connection reset by peer) 2011-04-07T19:55:23 *** Keth has joined #aichallenge 2011-04-07T20:01:02 *** Frontier has joined #aichallenge 2011-04-07T20:17:05 *** McLeopold has left #aichallenge 2011-04-07T20:19:48 *** amstan has joined #aichallenge 2011-04-07T20:19:48 *** ChanServ sets mode: +o amstan 2011-04-07T20:23:36 *** sigh_ is now known as sigh 2011-04-07T20:26:17 *** Mathnerd314 has joined #aichallenge 2011-04-07T20:42:11 *** needsch has quit IRC (Quit: Leaving.) 2011-04-07T21:24:38 *** dlila has quit IRC (Quit: Leaving) 2011-04-07T21:31:54 *** Eruonen has quit IRC () 2011-04-07T21:33:27 *** Kingpin13 has joined #aichallenge 2011-04-07T21:38:08 *** McLeopold has joined #aichallenge 2011-04-07T22:10:26 *** antimatroid has quit IRC (Ping timeout: 246 seconds) 2011-04-07T22:21:29 *** amstan_ has joined #aichallenge 2011-04-07T22:21:29 *** ChanServ sets mode: +o amstan_ 2011-04-07T22:22:29