2011-10-06T00:00:02 *** dvladim has joined #aichallenge 2011-10-06T00:00:29 Is it ok and my time wasn't come yet? It could be because my old version got a game right before new one compiled. 2011-10-06T00:08:48 need to go 2011-10-06T00:08:50 *** UncleVasya has quit IRC (Remote host closed the connection) 2011-10-06T00:13:33 *** amstan has quit IRC (Ping timeout: 276 seconds) 2011-10-06T00:19:45 *** dvladim has quit IRC (Ping timeout: 255 seconds) 2011-10-06T00:21:22 *** eashoka has joined #aichallenge 2011-10-06T00:22:00 *** delt0r_ has quit IRC (Read error: Operation timed out) 2011-10-06T00:36:33 *** delt0r_ has joined #aichallenge 2011-10-06T00:38:14 *** mleyen has quit IRC (Quit: ~ Trillian Astra - www.trillian.im ~) 2011-10-06T00:50:47 mleise: shh, the languages on the server aren't meant for distribution and aren't linked to anywhere. They are meant solely for workers so as not to put unnecessary load on the originating 2011-10-06T00:51:37 I know it's probably a rather stretch, but I figured we can just pull 'em if any group complained 2011-10-06T00:51:58 janzert: It is all ok. We have official permission. 2011-10-06T00:55:09 regarding fpc package did you checkin a changed worker_setup.py that includes it? 2011-10-06T00:57:00 ahh, think I see fp-compiler 2011-10-06T00:57:20 the workers will pick it up next time ec2 decides to kill them and start new ones 2011-10-06T01:13:40 janzert: have you looked at the ants tutorial yet? 2011-10-06T01:14:07 *** McLeopold1 is now known as McLeopold 2011-10-06T01:14:12 no 2011-10-06T01:15:03 something specific you want looked at? 2011-10-06T01:15:43 no, just a general review 2011-10-06T01:15:53 just got done with a round trip 7 hours of driving so probably not going to dive into much tonight 2011-10-06T01:15:59 I have my Pascal starter bot now, but it acts weird. It makes the engine not ... oh wait. Why do I keep telling people to *flush* the IO buffers all the time... 2011-10-06T01:16:04 ok, I'll try and read it tomorrow 2011-10-06T01:16:15 mleise: :P 2011-10-06T01:16:27 janzert: from my perspective, it is tomorrow for you :P 2011-10-06T01:16:50 for me tomorrow is whatever time I go to sleep and get back up from more than a nap :) 2011-10-06T01:17:17 I don't have a bad sleep schedule, I have no sleep schedule 2011-10-06T01:17:22 then again there is seemingly no flush routine in freepascal for stdout 2011-10-06T01:17:28 ouch 2011-10-06T01:17:54 so either it always flushes or the language wants to make fun of me 2011-10-06T01:18:39 wtf? http://www.freepascal.org/docs-html/rtl/system/flush.html 2011-10-06T01:19:34 there example is really scary 2011-10-06T01:20:00 apparently it also does internal buffering of file objects? 2011-10-06T01:24:48 too bad the engine can't invoke the flushing of buffers 2011-10-06T01:26:03 It is alright if they buffer, but I should not be forced to assign a dummy file name ('') to a text file pointer to get stdout to flush 2011-10-06T01:26:42 usually you have functions for using files that take a file pointer and for stdout that dont take a file pointer, but there is no overloaded flush() with no arguments 2011-10-06T01:26:48 anyway it works now :) 2011-10-06T01:27:02 just have to figure out why the ants dont move 2011-10-06T01:35:04 lack of pascal-fu. I have to pass dynamic arrays by ref to change them inside a function 2011-10-06T01:36:51 aichallenge: McLeopold epsilon * r6365b71 / ants/dist/starter_bots/vb/Ants.vb : fix for vb starter package - http://git.io/P7j5WQ 2011-10-06T01:36:52 aichallenge: McLeopold epsilon * rbbc3d41 / setup/worker_setup.py : Merge branch 'epsilon' of github.com:aichallenge/aichallenge into epsilon - http://git.io/Sg7J4A 2011-10-06T02:16:46 *** janzert1 has joined #aichallenge 2011-10-06T02:19:14 *** janzert1 has quit IRC (Read error: No route to host) 2011-10-06T02:19:28 *** janzert has quit IRC (Ping timeout: 248 seconds) 2011-10-06T02:21:11 *** UncleVasya has joined #aichallenge 2011-10-06T02:21:37 *** janzert has joined #aichallenge 2011-10-06T02:21:42 Still no games for my bot. 2011-10-06T02:21:44 *** UncleVasya has quit IRC (Remote host closed the connection) 2011-10-06T02:21:54 *** okayzed is now known as okay 2011-10-06T02:23:23 McLeopold: What does compiler.py do when it encounters output on stderr? 2011-10-06T02:24:43 assumes that it is compiler error output, fails the compile and displays the first and last bit to the user 2011-10-06T02:25:32 *** ltriant has quit IRC (Quit: Computer has gone to sleep) 2011-10-06T02:33:30 aichallenge: McLeopold epsilon * rcca172f / (5 files): update to vb starter package - http://git.io/v-RJDQ 2011-10-06T02:38:03 janzert: Is it possible to check for the generation of an executable instead? 2011-10-06T02:39:12 hmm, I don't think it is currently 2011-10-06T02:39:57 I'm not opposed to it in any way, just whoever wrote it way back when didn't write it that way :/ 2011-10-06T02:42:43 although it is rather nice that currently warnings aren't ignored 2011-10-06T02:43:12 probably helps stop some occasional crashers from slipping through 2011-10-06T02:44:08 UncleVasya: If you see this, next time you mention a possible problem with your account please mention your account name too 2011-10-06T02:44:22 since it seems not to contain either Uncle or Vasya 2011-10-06T02:47:50 *** okay is now known as okayzed 2011-10-06T02:49:09 aichallenge: McLeopold epsilon * r1453186 / ants/dist/starter_bots/vb/Ants.vb : update to vb error messages - http://git.io/_p81tg 2011-10-06T02:50:33 janzert: I believe UncleVasya is Delin_psm_Vizov 2011-10-06T02:52:01 looks like that would fit his problem description too 2011-10-06T02:53:03 janzert: there are 3 matchups without players... 2011-10-06T02:53:26 looks like it's about time to clean out the matcup table again :/ 2011-10-06T02:53:54 btw, any idea why we aren't getting a traceback for that Nonetype and str +=? 2011-10-06T02:54:07 haven't really looked at it yet 2011-10-06T02:54:53 I guess check the code for a caught exception without re-raising? 2011-10-06T02:58:29 looks like Vasya's submission is seeded in the currently running game 2011-10-06T02:59:49 janzert: Well, so far I've seen D output 'scheduled for deprecation' warnings and freepascal makes the linker print a warning that is expected and does no harm 2011-10-06T02:59:52 yes, I set the matchups to a negative worker 2011-10-06T03:00:06 I think the sql failed to find opponents, and left it in a limbo state 2011-10-06T03:00:27 ahh and hmm 2011-10-06T03:05:34 aichallenge: McLeopold epsilon * rc51c654 / worker/engine.py : turn on full trackback for workers - http://git.io/ywBkiw 2011-10-06T03:05:56 janzert: there's the culprit, I was thinking at one time that a traceback might be a security issue 2011-10-06T03:06:13 janzert: if you're on the workers, can you update? 2011-10-06T03:07:32 heh and I saw that a little while back and thought I already ripped it out :/ 2011-10-06T03:08:23 update worker-src.tgz already? 2011-10-06T03:08:48 no 2011-10-06T03:09:01 ok, I got it 2011-10-06T03:10:47 aichallenge: Marco Leise epsilon * r8f34143 / (6 files): added Pascal starter bot - http://git.io/zDgRLg 2011-10-06T03:10:48 aichallenge: Marco Leise epsilon * r2c70abd / (6 files in 2 dirs): Merge branch 'epsilon' of github.com:aichallenge/aichallenge into epsilon - http://git.io/3jPBMg 2011-10-06T03:11:08 hmm, something seems to have gone wrong with worker_setup.py 2011-10-06T03:11:29 at least the file permissions are messed up on the current worker 2011-10-06T03:12:11 what the... how is the current worker even running 2011-10-06T03:13:03 oh, whoops 2011-10-06T03:13:20 heh, I'm apparently logged into a worker that ec2 just brought up 2011-10-06T03:14:40 hmm, seems worker_setup.py has crashed though 2011-10-06T03:15:42 the pascal ":=" gives me warm fuzzies... 2011-10-06T03:17:08 aichallenge: McLeopold epsilon * r421c8da / sql/2_generate_matchup.sql : add back abort logic to generate matchup - http://git.io/ZEIvSw 2011-10-06T03:17:08 aichallenge: McLeopold epsilon * r7df2ee9 / (9 files in 3 dirs): Merge branch 'epsilon' of github.com:aichallenge/aichallenge into epsilon - http://git.io/1kQwcw 2011-10-06T03:17:40 *** boegel has joined #aichallenge 2011-10-06T03:17:53 does it log an error somewhere if it ends up aborting? 2011-10-06T03:17:59 no 2011-10-06T03:18:03 should I? 2011-10-06T03:18:22 or maybe leave a matchup entry with negative worker 2011-10-06T03:18:39 I could do that, then update it with an appropriate error message 2011-10-06T03:18:40 so we'll see if it is happening unexpectedly and/or frequently 2011-10-06T03:19:04 I just don't like letting errors pass silently when we can help it 2011-10-06T03:21:55 aichallenge: McLeopold epsilon * rd500aeb / sql/2_generate_matchup.sql : add abort reason - http://git.io/VF_e5w 2011-10-06T03:22:38 *** Blkt has joined #aichallenge 2011-10-06T03:22:56 janzert: okay, it's in 2011-10-06T03:23:09 janzert: is the worker up? 2011-10-06T03:23:33 not yet, one is initializing the other will stop after this game to update 2011-10-06T03:23:38 good morning everyone 2011-10-06T03:24:27 * janzert waves 2011-10-06T03:27:00 ok, first one is updated and running 2011-10-06T03:29:54 *** aerique has joined #aichallenge 2011-10-06T03:30:30 both up and running 2011-10-06T03:35:49 ok 2011-10-06T03:38:03 *** mceier has joined #aichallenge 2011-10-06T03:38:09 aichallenge: Marco Leise epsilon * r3ef37ae / (sql/1_data.sql worker/compiler.py): support compilation of Pascal submissions - http://git.io/GIZu3A 2011-10-06T03:40:02 janzert: my guess is it's my probably rank code 2011-10-06T03:41:22 when the cutoff is None 2011-10-06T03:42:03 oh was I a second too late? 2011-10-06T03:44:19 mleise: yep 2011-10-06T03:44:31 aichallenge: McLeopold epsilon * r2531d6b / ants/ants.py : fix for cutoff reason issue - http://git.io/exddzw 2011-10-06T03:44:31 aichallenge: McLeopold epsilon * r6ac5157 / (sql/1_data.sql worker/compiler.py): Merge branch 'epsilon' of github.com:aichallenge/aichallenge into epsilon - http://git.io/c5C4WA 2011-10-06T03:45:20 you both finished with worker changes for at least a half hour ;) 2011-10-06T03:45:35 maybe...let me test vb... 2011-10-06T03:46:06 If compile.py really doesn't check for the existence of the output file or the return value of the compiler then Pascal wont work. 2011-10-06T03:46:40 it can check for an output file or files 2011-10-06T03:46:54 instead of stderr? 2011-10-06T03:46:54 just can't ignore stderr basically 2011-10-06T03:47:28 in fact I think it requires checking for the final bot executable 2011-10-06T03:47:37 ld will *always* print: "warning: release/link.res contains output sections; did you forget -T?" 2011-10-06T03:50:06 *** TTE has joined #aichallenge 2011-10-06T03:50:13 I know this is a pita. Maybe stderr should be displayed, but not stop the bot from participating? 2011-10-06T03:50:40 script languages are already different enough? 2011-10-06T03:50:42 *** TTE has quit IRC (Client Quit) 2011-10-06T03:50:50 they don't produce output files usually 2011-10-06T03:51:00 except for coffeescript 2011-10-06T03:51:07 you don't run a compiler on them either 2011-10-06T03:51:33 mleise: I would suggest adding an option to ExternalCompiler ignore stderr 2011-10-06T03:52:12 mleise: what happens if there is a compiler error? 2011-10-06T03:52:26 the compiler returns != 0 2011-10-06T03:52:36 (i would assume that at least) 2011-10-06T03:54:12 I think the authors of compilers wouldn't mind stdout, stdinfo, stdwarn and stderr at their disposal :) 2011-10-06T03:54:58 I'm about to become the #1 ranked vb bot in the contest :) 2011-10-06T03:57:07 oh well... let me upload a pascal package 2011-10-06T04:06:01 workers updated again 2011-10-06T04:06:18 probably going to sleep soon so... ;) 2011-10-06T04:09:04 hmm it doesn't recognize Pascal sources yet 2011-10-06T04:10:08 if you uploaded before about 30 seconds before the last updated message then maybe just try uploading again 2011-10-06T04:11:01 i just tried again 2011-10-06T04:11:10 *** ibdknox has quit IRC (Remote host closed the connection) 2011-10-06T04:11:40 *** ibdknox has joined #aichallenge 2011-10-06T04:11:41 i can run compile.py locally and it works 2011-10-06T04:12:51 It looks like the old compile.py is in use on the worker... Pascal is just not listed among the supported languages in the error log 2011-10-06T04:13:34 ahh, I see what I did wrong 2011-10-06T04:14:50 i think i could add a bash script that wraps the FreePascal compiler and swallows the ld error line 2011-10-06T04:15:37 ok, try it now 2011-10-06T04:16:00 *** ibdknox has quit IRC (Ping timeout: 255 seconds) 2011-10-06T04:16:41 ok... it is still unpacking and compiling 2011-10-06T04:17:08 janzert: "Output file MyBot was not created." Is that expected? 2011-10-06T04:18:02 only if the executable doesn't exist after compilation 2011-10-06T04:18:25 sh.. how can I debug that? 2011-10-06T04:19:23 does the pascal compiler put error messages on stdout by chance? 2011-10-06T04:19:55 it prints the version, yes 2011-10-06T04:20:29 where does it print the message when it can't compile the source? 2011-10-06T04:20:37 stdout or stderr? 2011-10-06T04:21:11 also only the MyBot.pas file needs to be passed to the compiler? 2011-10-06T04:21:14 stdout, lol 2011-10-06T04:21:55 Yes the compiler gets angry if you pass more than one source file 2011-10-06T04:22:35 and all the code for the program must be in that one file or the compiler picks up the rest automatically? 2011-10-06T04:22:52 it picks the rest automatically 2011-10-06T04:23:31 ok 2011-10-06T04:24:01 aichallenge: janzert epsilon * r846683a / worker/compiler.py : Next iteration of pascal support - http://git.io/CZha3A 2011-10-06T04:24:25 And the process seems to work if I just invoke the compiler script in the bot directory on my computer 2011-10-06T04:25:05 didn't we already have the out_file in the languages list? 2011-10-06T04:26:01 that will make it send stdout errors back to the user 2011-10-06T04:26:06 try it again 2011-10-06T04:26:14 cool 2011-10-06T04:27:12 will it print the program version and errors, but ignore the warning from ld now? 2011-10-06T04:27:22 nope 2011-10-06T04:27:29 doesn't effect stderr handling 2011-10-06T04:27:47 hmmm.... "Output file MyBot was not created." + a *blank* line 2011-10-06T04:28:00 heh 2011-10-06T04:28:05 did I do something really bad? 2011-10-06T04:28:33 I replaced every "MyBot" with the BOT variable 2011-10-06T04:29:28 but that should not have any effect 2011-10-06T04:29:32 shouldn't matter 2011-10-06T04:30:04 I think the blank line means it never got any stdout from the compiler 2011-10-06T04:30:41 can you run "fpc -Mdelphi MyBot.pas" on the worker? 2011-10-06T04:30:43 heh, I just realized at least one of these workers is probably old enough that it doesn't have the pascal package installed 2011-10-06T04:30:57 yeah, about to 2011-10-06T04:31:03 yes, I also think it is missing the compiler 2011-10-06T04:31:03 *** Blkt` has joined #aichallenge 2011-10-06T04:31:47 what's the package name 2011-10-06T04:31:59 *** sigh has joined #aichallenge 2011-10-06T04:32:13 fp-compiler 2011-10-06T04:32:15 *** Blkt` has quit IRC (Client Quit) 2011-10-06T04:32:47 *** Blkt has quit IRC (Quit: ERC Version 5.3 (IRC client for Emacs)) 2011-10-06T04:33:08 yeah, neither one had it 2011-10-06T04:33:19 lol :-) 2011-10-06T04:33:29 Free Pascal Compiler version 2.4.0-2ubuntu3.1 [2011/06/17] for x86_64 2011-10-06T04:33:42 *** Blkt has joined #aichallenge 2011-10-06T04:33:54 Scala is version 2.9.0.1? I had someone complain about the ancient 2.7.7 listed on the starter packages page 2011-10-06T04:34:36 yes 2011-10-06T04:35:10 starter packages page even says python 2.5 2011-10-06T04:35:15 :/ 2011-10-06T04:35:37 starter packages page is months old 2011-10-06T04:36:01 I always update my entries. Look DMD is up to date :) 2011-10-06T04:36:07 yeah, obviously before we moved to natty even 2011-10-06T04:36:30 "/usr/bin/ld: warning: link.res contains output sections; did you forget -T?" 2011-10-06T04:36:41 yay, it works :P 2011-10-06T04:36:49 ok, so we are finally where we expected to end up. thank you for staying up so long :) 2011-10-06T04:37:02 np 2011-10-06T04:37:10 but yeah it really is time for me to go 2011-10-06T04:37:14 past 2011-10-06T04:49:33 *** kaemo has quit IRC (Ping timeout: 260 seconds) 2011-10-06T04:58:14 *** mcstar has joined #aichallenge 2011-10-06T05:00:43 i can't log into the beta server anymore, otherwise I would have checked the versions for the starter packages page 2011-10-06T05:01:10 aerique: did you forget your pass? 2011-10-06T05:01:41 No, it's in a pw manager. I think we agreed to just disable my access a while ago but I'm not sure. 2011-10-06T05:02:26 I could just go by Ubuntu Natty's packages list on their website, that would get most of the versions right I think. 2011-10-06T05:02:35 yeah, your not in passwd 2011-10-06T05:02:53 what's the check command? 2011-10-06T05:03:07 check command? 2011-10-06T05:03:15 command to check the version 2011-10-06T05:03:34 depends on the programming language, for SBCL it's "sbcl --version" 2011-10-06T05:03:56 think it's 2011-10-06T05:04:08 "--version" for most of them. "mono --version" 2011-10-06T05:04:52 SBCL 1.0.45.0.debian 2011-10-06T05:05:23 ought to be scripted actually so starter_packages.php will be generated automatically, or at least the table 2011-10-06T05:05:42 I mentioned that, janzert thinks it's not worth the effort 2011-10-06T05:07:15 so you want to go through all of them with me? :) 2011-10-06T05:18:50 *** sir_macelon has joined #aichallenge 2011-10-06T05:18:59 hi 2011-10-06T05:19:06 hi 2011-10-06T05:22:56 aichallenge: aerique epsilon * r907264d / website/starter_packages.php : Update SBCL version number. - http://git.io/VS4x-Q 2011-10-06T05:24:12 *** eashoka has quit IRC (Quit: Page closed) 2011-10-06T05:38:57 Scala is @ 2.9.0.1, DMD is up-to-date, FreePascal not ready for public use at the moment. 2011-10-06T05:42:18 somebody have to port slime to python 2011-10-06T05:43:41 *** antimatroid has quit IRC (Ping timeout: 255 seconds) 2011-10-06T05:46:03 *** antimatroid has joined #aichallenge 2011-10-06T05:51:10 *** antimatroid1 has joined #aichallenge 2011-10-06T05:51:18 *** kaemo has joined #aichallenge 2011-10-06T05:52:34 *** antimatroid has quit IRC (Ping timeout: 258 seconds) 2011-10-06T05:52:44 *** onensora has joined #aichallenge 2011-10-06T05:55:21 *** trewhella has joined #aichallenge 2011-10-06T05:58:24 *** trewhella has quit IRC (Client Quit) 2011-10-06T06:06:41 *** berak has joined #aichallenge 2011-10-06T06:16:27 *** Parsley_ has joined #aichallenge 2011-10-06T06:20:56 *** olexs has joined #aichallenge 2011-10-06T06:20:58 *** Parsley_ has quit IRC (Ping timeout: 252 seconds) 2011-10-06T06:21:26 *** parsley_ has joined #aichallenge 2011-10-06T06:36:49 *** onensora has quit IRC (Ping timeout: 260 seconds) 2011-10-06T06:38:05 *** onensora has joined #aichallenge 2011-10-06T06:39:34 why does a 3xnested iteration ~40x10^6 core-calls with a bit of numpy core take ages? 2011-10-06T06:41:16 i guess it has to do with index regeneration 2011-10-06T06:42:11 should i just generate once the index array, and then slice it for the loops? 2011-10-06T06:42:40 McLeopold: maybe you could answer since you swear by python 2011-10-06T06:42:50 if youre not too busy 2011-10-06T06:46:53 mcstar: what is slime? 2011-10-06T06:47:11 superior lisp interaction mode for emacs 2011-10-06T06:47:43 basically, this is what makes lisp development great 2011-10-06T06:47:50 mcstar: also python is quite slow... tight loops with little on the inside is where this becomes very obvious 2011-10-06T06:48:07 i think ive fallen victim of this 2011-10-06T06:48:38 i think it is running for more than 30 minutes 2011-10-06T06:49:24 I find SBCL a pita to install however 2011-10-06T06:49:35 really? 2011-10-06T06:49:46 Think i will try clousure --the one that runs on a JVM 2011-10-06T06:49:52 i compiled/installed it wo problems 2011-10-06T06:49:57 clojure 2011-10-06T06:50:01 and THATS the pita 2011-10-06T06:50:04 :D 2011-10-06T06:50:08 I have never had a clean compile 2011-10-06T06:50:29 right... well i would prefer a pure Common lisp implementation 2011-10-06T06:50:40 clojure is not common lisp in any way 2011-10-06T06:50:51 if you want cl on jvm, use armed bear common lisp 2011-10-06T06:50:58 there was going to be on for the jvm --but it never got finished 2011-10-06T06:51:18 yea that was it... it is finished? or to spec? now? 2011-10-06T06:51:59 ive not used it, but i think its not beta or anything 2011-10-06T06:52:02 its stable 2011-10-06T06:52:10 but dont take my word for it 2011-10-06T06:52:22 delt0r_: sbcl is really easy to install/use 2011-10-06T06:52:29 when did you have trouble with it? 2011-10-06T06:52:36 not in my experance... 2011-10-06T06:52:44 but when? 2011-10-06T06:52:45 Always get compile errors 2011-10-06T06:52:54 was it 10 years ago? 2011-10-06T06:52:56 about 6 months ago was last time i tried 2011-10-06T06:53:01 :) 2011-10-06T06:53:16 whats wrong with using the package manager? 2011-10-06T06:53:20 I can't spend a lot of time on it... i still need to mostly use java for work 2011-10-06T06:53:22 are you using gentoo? 2011-10-06T06:53:27 slackware 2011-10-06T06:53:37 so you have to build it from source? 2011-10-06T06:53:42 package managers suck money ---leaf sacs 2011-10-06T06:53:57 slack does a package thing -- 2011-10-06T06:54:18 but yea you often build from source... most of the time it works --and you get 64bit 2011-10-06T06:54:31 hey, its finished 2011-10-06T06:54:33 can't belive most things still are 32 bit 2011-10-06T06:54:56 like 2001 called and wants its 32bit cpus back 2011-10-06T06:55:07 i dont know what youre talking about 2011-10-06T06:55:11 what is 32 bits? 2011-10-06T06:55:51 all my software is 64bit 2011-10-06T06:55:58 yea me too 2011-10-06T06:56:12 but a lot of stuff out there is not build for 64 yet 2011-10-06T06:56:15 i think even flashplugin has a 64bit version 2011-10-06T06:56:25 (yeah, i think i have the 32bit instaleld) 2011-10-06T06:56:28 just installing humble indie bundle 2011-10-06T06:57:28 well, arch has a multilib repository, i can install 32-bit versions of libraries if its needed 2011-10-06T06:57:56 slack has a multilib you can install too... which i usally do 2011-10-06T06:58:03 damit 2011-10-06T06:58:17 what? 2011-10-06T06:58:32 delt0r_: can't you just grab a prebuilt from: http://www.sbcl.org/platform-table.html? 2011-10-06T06:58:34 Error Could not find a compatible OpenGL display resolution 2011-10-06T06:58:49 :S 2011-10-06T06:58:58 aerique: didn't work last time...don't remember why now thou 2011-10-06T06:59:07 like i said i didn't have a lot of time 2011-10-06T06:59:40 delt0r_: give me shell access to your machine and i'll try to get it running 2011-10-06T06:59:51 yea right 2011-10-06T07:00:03 you said you couldnt compile it, but a binary package need not compiling 2011-10-06T07:00:04 root password right... :/ 2011-10-06T07:00:10 me too 2011-10-06T07:00:18 or a permanent account 2011-10-06T07:00:20 whatever 2011-10-06T07:01:16 at anyrate --i would not take up the offer... i need to know what its like to install dev envourment and what deployment issue may be like 2011-10-06T07:01:19 delt0r_: if you already java, clojure might fit you well, but be prepared, its not that easy to get it running in emacs 2011-10-06T07:01:22 delt0r_: could be a shared screen session. i don't see why SBCL should be so hard to get running for you 2011-10-06T07:01:36 *** sir_macelon has quit IRC (Quit: Page closed) 2011-10-06T07:01:38 mcstar: I don't use emacs anymore 2011-10-06T07:01:47 sorry 2011-10-06T07:01:57 delt0r_: try the binary package again sometime then 2011-10-06T07:02:18 aerique: you mean "screen" session? 2011-10-06T07:02:19 yea i will 2011-10-06T07:02:32 right now i don't have the time... too busy playing games and IRC 2011-10-06T07:02:40 I mean working 2011-10-06T07:02:45 delt0r_: basically unzip and you can do a test run even without further installation 2011-10-06T07:02:51 mcstar: yeah 2011-10-06T07:03:13 *** kaemo has quit IRC (Ping timeout: 244 seconds) 2011-10-06T07:03:16 aerique: have you tried tmux? thats what im using recently, can do horizontal splits too 2011-10-06T07:03:28 but better is a clisp for the jvm then i know deployment will be easy and i have the standard java libs 2011-10-06T07:03:43 mcstar: no but i hear good things about it. any recent screen can do horizontal splits as well though 2011-10-06T07:03:53 delt0r_: that would be ABCL 2011-10-06T07:04:05 yea... will check it out 2011-10-06T07:04:14 really? mine cant 2011-10-06T07:04:18 and its recent 2011-10-06T07:04:19 delt0r_: http://common-lisp.net/project/armedbear/ 2011-10-06T07:04:24 i read it needs a patch 2011-10-06T07:04:28 I need to be able to run on windows, mac, sun and linux 2011-10-06T07:04:46 right now for anything non trivial java is still the easiest way to do that 2011-10-06T07:04:52 *** eashoka has joined #aichallenge 2011-10-06T07:05:00 which is not the same as easy... but its not bad 2011-10-06T07:05:36 seriously 2011 and still most langs don't have a TCP/IP socket lib as a standard library? WTF 2011-10-06T07:05:39 Good morning/evening to all! 2011-10-06T07:06:21 i thin kyou can compile clojure, you dont have to bundle the clojure.jar, but im not sure, but i dont think its particularly hard to distribute a clojure program, if you dont want to provide sources 2011-10-06T07:11:24 *** eashoka_ has joined #aichallenge 2011-10-06T07:12:18 *** eashoka has quit IRC (Ping timeout: 252 seconds) 2011-10-06T07:16:45 my god, i got a 1.4GB ascii dump 2011-10-06T07:17:40 every line must be solved as a 3x3 linear system 2011-10-06T07:18:15 aerique: you know linal libraries for cl? 2011-10-06T07:18:19 linalg. 2011-10-06T07:19:38 mcstar: no i don't 2011-10-06T07:22:34 *** eashoka_ has quit IRC (Ping timeout: 252 seconds) 2011-10-06T07:49:00 *** amstan has joined #aichallenge 2011-10-06T07:49:00 *** ChanServ sets mode: +o amstan 2011-10-06T07:52:09 *** kincody has joined #aichallenge 2011-10-06T08:07:16 mcstar: tried using Lua? 2011-10-06T08:07:48 we discussed lua already, couple months back, remember? 2011-10-06T08:07:58 i said its too hot where it comes from 2011-10-06T08:08:05 :D 2011-10-06T08:08:22 lua is Awesome's config language, thats as much as i use it 2011-10-06T08:08:25 why do you recommend? 2011-10-06T08:10:42 mleise: 2011-10-06T08:12:32 *** amstan has quit IRC (Ping timeout: 248 seconds) 2011-10-06T08:13:01 *** amstan has joined #aichallenge 2011-10-06T08:13:01 *** ChanServ sets mode: +o amstan 2011-10-06T08:13:32 *** parsley_ has quit IRC (Ping timeout: 252 seconds) 2011-10-06T08:13:37 I don't know much about Lua, but as a math language that is well known, I thought it might be useful here 2011-10-06T08:15:43 *** parsley_ has joined #aichallenge 2011-10-06T08:15:44 aichallenge: Marco Leise epsilon * rc909726 / worker/compiler.py : added new Compiler class for Pascal code that filters stdout/stderr - http://git.io/8fAlwQ 2011-10-06T08:15:56 oh 2011-10-06T08:16:01 math language? 2011-10-06T08:16:11 i thought its a simple scripting lang. 2011-10-06T08:16:19 like ruby for example 2011-10-06T08:16:23 maybe faster/lighter 2011-10-06T08:16:50 does it have fast numerical routines? or arrays like numpy? 2011-10-06T08:17:00 cause if it doesnt, its out of the question 2011-10-06T08:17:54 (on a personal note, im somewhat saturated with all the different f'ing syntaxes of all these languages, my heart belongs to the lispy ones) 2011-10-06T08:18:42 I only used it once, because I needed dynamic equations inside a C app 2011-10-06T08:19:33 like plotting for example? the user could assemble the function to plot? 2011-10-06T08:19:50 or? 2011-10-06T08:20:30 yes, like that 2011-10-06T08:24:17 pascal is still in business? i remember they tought that to us in high school 2011-10-06T08:24:19 maybe you are right, it is just a simple scripting language 2011-10-06T08:24:24 *** delt0r_ has quit IRC (Ping timeout: 255 seconds) 2011-10-06T08:24:56 mcstar: OpenSource lives on and Delphi is still on the market, yes 2011-10-06T08:25:39 mcstar: doesn't macsyma (or one of those systems for CL) include linalg? 2011-10-06T08:25:45 its just that, as far as i can remember that early pascal was like qbasic, my memory is hazy on this one though 2011-10-06T08:26:09 Pascal was always closer to the silicon 2011-10-06T08:26:15 aerique: sure, it does 2011-10-06T08:26:30 but i only skimmed maxima 2011-10-06T08:26:42 Maxima would be my next best bet. 2011-10-06T08:26:45 it's syntax is not lisp-liek 2011-10-06T08:27:01 but, yeah, i probably could use it 2011-10-06T08:27:04 At least I *know* that it has an integrated solver 2011-10-06T08:27:19 yes, it has many things, im not sure if its fast though 2011-10-06T08:27:28 *** Palmik has joined #aichallenge 2011-10-06T08:27:31 how much time do you have? 2011-10-06T08:27:37 the thing is, that in the meantime my python code finished 2011-10-06T08:27:54 its a write once, run twice code 2011-10-06T08:28:15 mleise: for what? 2011-10-06T08:28:20 in general? 2011-10-06T08:28:25 janzert, McLeopold: I kinda solved the Pascal problem 2011-10-06T08:28:33 im going to die in 2075 march ~27 or some like that 2011-10-06T08:28:48 mcstar: there's enough maths hits on Cliki but I assume you already checked there? (http://www.cliki.net/Mathematics, http://www.cliki.net/linear%20algebra, https://github.com/blindglobe/lisp-matrix, https://github.com/OdonataResearchLLC/linear-algebra) 2011-10-06T08:28:51 mcstar: in that case you can let maxima run for a few years 2011-10-06T08:29:01 :) 2011-10-06T08:29:43 but i have the code in python now, and i started it in cl, but realized i should use python, cause has built-in vector instructions, like dot product 2011-10-06T08:29:48 *** olexs1 has joined #aichallenge 2011-10-06T08:30:08 and i can just copy from Mathematica the 3x3 matrix inversion code 2011-10-06T08:30:21 i dont need the general case 2011-10-06T08:30:31 aerique: yes, after i asked you 2011-10-06T08:30:49 *** olexs has quit IRC (Ping timeout: 260 seconds) 2011-10-06T08:31:15 *** kaemo has joined #aichallenge 2011-10-06T08:31:52 *** kincody has quit IRC (Ping timeout: 252 seconds) 2011-10-06T08:37:27 *** delt0r_ has joined #aichallenge 2011-10-06T08:47:21 aichallenge: Marco Leise epsilon * r0e46b9f / website/starter_packages.php : added Pascal starter bot to the list - http://git.io/sbNcRA 2011-10-06T08:47:23 *** onensora has quit IRC () 2011-10-06T08:53:35 aichallenge: Marco Leise epsilon * r2fd7945 / website/starter_packages.php : corrected alphabetical order - http://git.io/3bqTqw 2011-10-06T09:00:16 *** cichociemno has joined #aichallenge 2011-10-06T09:13:00 *** tobym has quit IRC (Remote host closed the connection) 2011-10-06T09:42:29 *** djstrong has joined #aichallenge 2011-10-06T09:48:44 could you tell me when is the start of the contest scheduled? 2011-10-06T09:55:16 djstrong: I'm also interested, but as for now they removed link to betaserver from IRC topic. 2011-10-06T09:56:20 djstrong: Last week there was a plan to launch on 3rd October, but I thing it was changed ;-) 2011-10-06T09:56:36 steve jobs died 2011-10-06T10:03:11 McLeopold: i really don't like that header change 2011-10-06T10:04:29 steve jobs did what?! 2011-10-06T10:04:43 yes 2011-10-06T10:04:50 yesterday 2011-10-06T10:04:50 no you are kidding me 2011-10-06T10:04:58 cichociemno: thx 2011-10-06T10:05:01 www.apple.com 2011-10-06T10:05:10 yesterday a german tv show made fun of his religion 2011-10-06T10:05:14 and then he dies? 2011-10-06T10:06:03 i dont think the 2 things are related 2011-10-06T10:06:04 mleise: not exactly ;) He died after 7year of cancer. 2011-10-06T10:06:37 he is even mention at http://www.google.com/ 2011-10-06T10:06:49 i will watch pirates of silicon valley in his memory 2011-10-06T10:06:50 oh i didn't know that... 2011-10-06T10:07:39 huh, i use in-addressbar search 2011-10-06T10:07:52 i didnt see the google thing, i found it on wikipedia's main page 2011-10-06T10:08:27 facebook, g+ and many other services is full of this information, 2011-10-06T10:08:49 i dont use any of them :) 2011-10-06T10:08:57 but the most ironic is over 200k of "likes" under the information about his death 2011-10-06T10:09:15 is that good? 2011-10-06T10:09:32 mcstar: I have no idea ;-\) 2011-10-06T10:09:35 why does somebody write like to his death? 2011-10-06T10:09:50 its just a button 2011-10-06T10:09:56 they dont write that literally 2011-10-06T10:10:06 there are 200k buttons under his name? 2011-10-06T10:10:25 nope, there is one "like" button, but it was pressed 200k times;) 2011-10-06T10:10:29 jeez, i don't talk to you again, ever :) 2011-10-06T10:10:41 mleise: sorry 2011-10-06T10:11:07 at my workplace there is lot of people who doesn't use any "social" thing;) 2011-10-06T10:11:35 i am antisocial, and its enough to keep up with my handful of friends 2011-10-06T10:11:46 sometimes that is too much too 2011-10-06T10:12:47 anyway, when i read that news, i had a cold feeling 2011-10-06T10:13:02 one that i will feel again, soon 2011-10-06T10:13:28 all the great people will die, all my favourite actors for example 2011-10-06T10:15:06 its like, when stargate sg-1 finished, now, that was bad, real bad, i felt blue for couple of weeks 2011-10-06T10:20:37 *** tobym has joined #aichallenge 2011-10-06T10:21:50 have you watched Firefly ?:> 2011-10-06T10:22:36 yes 2011-10-06T10:22:46 have you watched castle? 2011-10-06T10:24:04 firefly is exciting, but it was too short for any kiind of emotional attachment, at least for me 2011-10-06T10:27:13 yeap, because Rupert Murdoch cancel it.. No i'm tryint to not watch serials;) Firefly cancelation joke from big bang theory http://www.youtube.com/watch?v=vB9iAuw8ZCE 2011-10-06T10:28:15 *** onensora has joined #aichallenge 2011-10-06T10:28:28 i mentioned castle, because thats where you can find nathan fillion nowadays 2011-10-06T10:30:07 yes, I saw some trailer ;-) 2011-10-06T10:32:19 mcstar: Do you know if there are any plans of launch? 2011-10-06T10:32:41 they will launch as soon as they can 2011-10-06T10:32:51 there are some sponsoring issues to overcome 2011-10-06T10:33:17 at LEAST one week away 2011-10-06T10:35:33 as for me, they don't have to hurry up :-) 2011-10-06T10:36:28 I am experimenting with some tools like unit tests, mocking and test coverage, so;) 2011-10-06T10:39:31 *** sigh has quit IRC (Remote host closed the connection) 2011-10-06T11:00:31 *** aerique has quit IRC (Quit: ...) 2011-10-06T11:03:22 *** mceier_ has joined #aichallenge 2011-10-06T11:03:35 *** mceier has quit IRC (Disconnected by services) 2011-10-06T11:03:45 *** mceier_ is now known as mceier 2011-10-06T11:05:08 *** george has joined #aichallenge 2011-10-06T11:06:45 hi, quick question - in the starter package from hypertriangle ants page ... 2011-10-06T11:07:09 is the test case there ( test_bot.sh ) 2011-10-06T11:07:15 the only test i need to pass? 2011-10-06T11:07:32 (i submitted, but apparently failed test cases on server) 2011-10-06T11:07:50 was wondering if the test_bot.sh was the same test? 2011-10-06T11:08:59 ought to be the same 2011-10-06T11:09:49 thanks for your reply - when i visualise the test_bot.sh - game seems to run fine 2011-10-06T11:10:05 (ok, one player is eliminated, but this is just how the game goes) 2011-10-06T11:12:00 im not sure how exactly the that archive you downloaded is generated, i always recommend you test against the latest github version of the toolkit 2011-10-06T11:12:17 ah, github is still the best? 2011-10-06T11:12:22 yes 2011-10-06T11:12:28 ok, i try that then. thanks. 2011-10-06T11:13:02 (i'd got the impression that contest start was near, so assumed they'd be same now). 2011-10-06T11:13:17 it's near but things are updating quite quickly atm 2011-10-06T11:13:56 ok, thanks. 2011-10-06T11:13:56 the c++ bot hasn't changed recently though i don't think 2011-10-06T11:13:56 or are you using c now? :P 2011-10-06T11:15:37 still c++ 2011-10-06T11:16:26 (computer fried some months back - and i'd been somewhat remiss with my backups :/ ) so starting from scratch ... 2011-10-06T11:16:45 (not that i'd got that far before...) 2011-10-06T11:17:04 another reminder to back up my home 2011-10-06T11:20:58 i'm terrible about backing anything up 2011-10-06T11:21:06 but i do have my thesis stuff drop boxed, not risking that one 2011-10-06T11:21:17 *** UncleVasya has joined #aichallenge 2011-10-06T11:22:40 mleise: workers should be up to date with your latest compiler changes 2011-10-06T11:23:28 lol, i did that too 2011-10-06T11:23:47 mcstar: Hi. I've made an update for bot, now it's on place 12. 2011-10-06T11:24:11 11 places to go 2011-10-06T11:24:37 :D 2011-10-06T11:25:30 at least you beat solifugid 2011-10-06T11:26:35 luda and roadkill has updated their's too, now my bot is not the best lisp bot :( 2011-10-06T11:28:16 *** boegel has quit IRC (Ping timeout: 244 seconds) 2011-10-06T11:28:48 wow, you have a bot there! I always thought you are waiting for the Start. 2011-10-06T11:29:13 *** McLeopold has left #aichallenge 2011-10-06T11:29:41 i had one since may(?) 2011-10-06T11:30:49 Does it do something but going North? :) 2011-10-06T11:32:33 it did well, before pre-launch-beta(without hills and maze maps), collects food optimally, and has a good inflation ratio 2011-10-06T11:32:50 reached 36th place than, as i recall 2011-10-06T11:49:25 *** onensora has quit IRC (Ping timeout: 240 seconds) 2011-10-06T11:56:00 *** onensora has joined #aichallenge 2011-10-06T12:01:17 *** McLeopold has joined #aichallenge 2011-10-06T12:01:42 mcstar: Is it your bot? http://aichallengebeta.hypertriangle.com/profile.php?user=39 2011-10-06T12:06:28 39, seems to be 2011-10-06T12:07:30 *** cichociemno has quit IRC (Quit: Page closed) 2011-10-06T12:11:51 *** berak has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2011-10-06T12:28:41 http://imgur.com/ZSvgH 2011-10-06T12:28:52 heres is the result of an etire day's work 2011-10-06T12:28:56 *** Blkt has quit IRC (Quit: going home...) 2011-10-06T12:28:57 its ugly 2011-10-06T12:29:42 what is it, a fermi surface? 2011-10-06T12:31:08 is something called a Wulff construction, if you have the surface free energy anisotropy, you can calculate the minimum energy surface, the equilibrium shape of a crystal 2011-10-06T12:31:52 a fermi surface qould be analytical 2011-10-06T12:31:56 would* 2011-10-06T12:34:07 btw, it is wrong, i have to do it again 2011-10-06T12:35:18 um, i'd noticed, but didnt like to mention it .) 2011-10-06T12:36:53 you spot the asymmetry? 2011-10-06T12:39:14 george: were you having issues with the tools or compiling today? 2011-10-06T12:39:26 hi McLeopold 2011-10-06T12:39:31 i am ok now 2011-10-06T12:40:06 ok 2011-10-06T12:40:34 i have some c++ problem, that i dont understand, related to the #ifndef DEBUG in the Bug.h file 2011-10-06T12:40:51 so i just modified everywhere that i was using the Bug class 2011-10-06T12:41:01 so, it works locally, but not on the server? 2011-10-06T12:41:16 not something for organisers to worry about 2011-10-06T12:41:24 unless many c++ players start asking 2011-10-06T12:41:52 I will worry about you getting good feedback from the compilation process, but that's as far as I go :) 2011-10-06T12:42:03 no, it is same on both - so not organsier's problem 2011-10-06T12:43:19 but naively using the supplied Bug.h got me in trouble (seems my c++/motivation isnt good enough to understand why) 2011-10-06T12:43:21 *** okayzed is now known as okay 2011-10-06T12:44:13 McLeopold: Hi. Do I have a possibility to add new organization? 2011-10-06T12:44:27 UncleVasya: yes, it's automatic now 2011-10-06T12:44:38 Plus, you can use UTF8 2011-10-06T12:48:37 *** foRei has joined #aichallenge 2011-10-06T12:48:51 I can only choose from existing variants, cannot find where to write institute by my own. 2011-10-06T13:06:20 *** ibdknox has joined #aichallenge 2011-10-06T13:06:56 UncleVasya: looks like it only works on sign up 2011-10-06T13:07:37 ok, thanks 2011-10-06T13:14:22 *** berak has joined #aichallenge 2011-10-06T13:18:04 *** UncleVasya has quit IRC (Read error: Connection reset by peer) 2011-10-06T13:25:49 *** UncleVasya has joined #aichallenge 2011-10-06T13:30:40 *** ibdknox has quit IRC (Remote host closed the connection) 2011-10-06T13:33:49 *** mleyen has joined #aichallenge 2011-10-06T13:44:20 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-10-06T14:00:02 *** k333es has quit IRC (Ping timeout: 252 seconds) 2011-10-06T14:05:59 *** okay is now known as okayzed 2011-10-06T14:11:55 *** amstan_ has joined #aichallenge 2011-10-06T14:11:55 *** ChanServ sets mode: +o amstan_ 2011-10-06T14:15:18 *** amstan has joined #aichallenge 2011-10-06T14:15:18 *** ChanServ sets mode: +o amstan 2011-10-06T14:15:58 *** amstan has quit IRC (Remote host closed the connection) 2011-10-06T14:16:06 *** amstan has joined #aichallenge 2011-10-06T14:16:07 *** ChanServ sets mode: +o amstan 2011-10-06T14:23:30 *** UncleVasya has quit IRC (Read error: Connection reset by peer) 2011-10-06T14:24:26 *** Accoun has quit IRC () 2011-10-06T14:24:28 *** okayzed is now known as okay 2011-10-06T14:26:01 *** Apophis has joined #aichallenge 2011-10-06T14:28:22 *** Apophis_ has quit IRC (Ping timeout: 256 seconds) 2011-10-06T14:37:44 *** Accoun has joined #aichallenge 2011-10-06T14:41:43 *** b0lly has joined #aichallenge 2011-10-06T14:48:43 *** okay is now known as okayzed 2011-10-06T14:52:45 *** Apophis has quit IRC (Ping timeout: 255 seconds) 2011-10-06T14:52:48 *** tobym has quit IRC (Read error: No route to host) 2011-10-06T14:54:17 *** Apophis has joined #aichallenge 2011-10-06T14:54:18 I am trying to run the python playgame script on a centOS 5 (python 2.7) but I get a syntax error in a print statement. Am I missing something obvious? 2011-10-06T14:58:15 you sure its 2.7? 2011-10-06T14:58:35 (check the #! line) 2011-10-06T15:05:04 Ok you are right: the shebang is '#!/usr/bin/env python' which mistakenly I thought was calling the 2.7 since I have installed and made an alias for python to call 2.7 .In the prompt if a just call 'python -V' it reports the 2.7 version 2011-10-06T15:05:38 so what kind of magic does the shebang do that calls the old python? 2011-10-06T15:06:22 *** mleyen has quit IRC (Quit: ~ Trillian Astra - www.trillian.im ~) 2011-10-06T15:07:17 where is that alias? 2011-10-06T15:07:23 in the system folders? 2011-10-06T15:07:32 in roots .bash_profile 2011-10-06T15:07:39 i sourced it and even re loged in 2011-10-06T15:09:03 /root/.bash_profile shouldnt run for any other user 2011-10-06T15:09:04 do you mean /etc/bash_profile ? 2011-10-06T15:09:04 playgame runs as another user? 2011-10-06T15:09:04 yes in /root/.bash_profile, not etc 2011-10-06T15:09:04 of course not, as the one that executes it 2011-10-06T15:09:15 i mean what you dont in /root folder has no effect on any other user 2011-10-06T15:09:19 dont->do 2011-10-06T15:09:23 i know 2011-10-06T15:09:47 so 2011-10-06T15:09:47 then i dont understand why did you make that alias there 2011-10-06T15:10:14 because I run play_one_game.sh as root 2011-10-06T15:10:18 lol 2011-10-06T15:10:30 why on earth would you do that 2011-10-06T15:10:35 and I expect the effective environment to take root's .bash_profile into account? 2011-10-06T15:11:01 anyway, env doesnt expand aliases 2011-10-06T15:11:03 because I just setup a VM and went straight to test the ai environmet :) 2011-10-06T15:11:15 aha moment 2011-10-06T15:11:16 thanks and sorry 2011-10-06T15:11:20 and alias in bash for example only expands in literal bash commands 2011-10-06T15:12:14 so I should just useradd and take care of pyhton2.7 to be first in $PATH? 2011-10-06T15:12:23 in users env? 2011-10-06T15:12:31 if you dont mind, you should rewrite the shebang line in the startup script, i.e. play_one_game.sh to use python2.7 2011-10-06T15:12:50 #!/usr/bin/env python2.7 2011-10-06T15:12:52 thats all 2011-10-06T15:13:05 I tries it 2011-10-06T15:13:08 tried* 2011-10-06T15:13:30 well, you can leave out env altogether 2011-10-06T15:13:36 which python2.7 2011-10-06T15:13:42 but It doesnt work (although I have put puthon2.7 in the PATH) 2011-10-06T15:13:45 and put that path after #! 2011-10-06T15:13:47 that should work 2011-10-06T15:13:51 Ok let me not bother you any more. Thanks a lot 2011-10-06T15:14:01 np 2011-10-06T15:14:06 I m sure I will figure it out 2011-10-06T15:14:11 thanks a lot 2011-10-06T15:14:45 didnt know aliases wherent axpanded by env 2011-10-06T15:15:04 there's another thing with sebangs 2011-10-06T15:15:11 that i just recently discovered 2011-10-06T15:15:27 you cant parameterize your called program, i.e. python in this case 2011-10-06T15:15:42 #!/usr/bin/env python arg1 arg2 2011-10-06T15:16:02 if you want python to recognize arg1 arg2 as two separate parameters, thats not going to work 2011-10-06T15:16:27 python will receive them as 1 parameter, "arg1 arg2" 2011-10-06T15:17:05 thanks a lot :) 2011-10-06T15:17:08 sure 2011-10-06T15:39:15 *** amstan has quit IRC (Ping timeout: 252 seconds) 2011-10-06T15:40:18 *** okayzed is now known as okay 2011-10-06T15:44:29 b0lly: are you still having issues? 2011-10-06T16:04:21 amstan_: the tools Make file should probably compile the java bots into jar files 2011-10-06T16:11:23 *** olexs1 has quit IRC (Quit: Leaving.) 2011-10-06T16:19:28 *** nicky has joined #aichallenge 2011-10-06T16:24:04 *** nicky has quit IRC (Quit: Verlassend) 2011-10-06T16:24:33 *** delt0r_ has quit IRC (Ping timeout: 255 seconds) 2011-10-06T16:35:31 *** berak has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2011-10-06T16:37:35 *** delt0r_ has joined #aichallenge 2011-10-06T16:45:16 *** okay is now known as okayzed 2011-10-06T16:47:58 *** parsley_ has quit IRC (Ping timeout: 252 seconds) 2011-10-06T16:48:13 *** Palmik has quit IRC (Remote host closed the connection) 2011-10-06T16:56:53 *** mceier has quit IRC (Quit: leaving) 2011-10-06T17:03:41 *** foRei has quit IRC (Quit: Bye) 2011-10-06T17:04:39 *** ltriant has joined #aichallenge 2011-10-06T17:49:10 now this is interesting 2011-10-06T17:49:44 im in the process of porting that python code, which i was talking about earlier, to common lisp 2011-10-06T17:50:25 a part of it, that took at least 30 minutes for python, run in 15secs under sbcl, 5 secs of which was spent on garbage collection 2011-10-06T17:56:53 mcstar: what code? 2011-10-06T17:57:12 you werent here than 2011-10-06T17:58:41 its a 3x nested loop, ~40M evaluations of the core function, which consists of vector-products and trigonometric functions 2011-10-06T17:58:57 you might want to use numpy for that 2011-10-06T17:59:00 i am 2011-10-06T17:59:06 still slow? 2011-10-06T17:59:12 sure :) 2011-10-06T17:59:14 120x 2011-10-06T17:59:36 it takes me 3-4x more time to write it in common lisp 2011-10-06T17:59:42 but its very fast 2011-10-06T18:00:12 i have to write elementary operations, and have to declare types 2011-10-06T18:00:40 the other parts, that are yet to come will be more interesting 2011-10-06T18:00:51 mcstar: well.. i guess it depends on how much of the code is in python and how much numpy takes care of it 2011-10-06T18:01:08 the problem is with the loop i think 2011-10-06T18:01:15 can i see the code? 2011-10-06T18:01:40 in the first version i used range() for indexing, but i suspect thats bad, cause range() recreates the indices 2011-10-06T18:01:56 and its better to generate them once, and then slice it up for the nested loops 2011-10-06T18:01:58 xrange is much faster 2011-10-06T18:02:06 this will be faster than xrange 2011-10-06T18:02:14 xrange is still an iterator 2011-10-06T18:02:40 but, there might be no difference, i dont know python internals 2011-10-06T18:03:00 i will show you the code later 2011-10-06T18:03:08 i want to finish up my lisp code 2011-10-06T18:03:21 and i need the results, its for a project 2011-10-06T18:03:48 if you are here tomorrow and still interested ill show you 2011-10-06T18:03:58 sure 2011-10-06T18:04:13 python iterators are "supposed" to be optimized 2011-10-06T18:04:51 how does lisp garbage collect? 2011-10-06T18:07:00 what do you mean? 2011-10-06T18:07:17 well, i guess :) 2011-10-06T18:07:53 garbage is usually created when we cons much 2011-10-06T18:08:19 i.e. creating lists on the fly for temporary purposes 2011-10-06T18:08:36 for example in a loop, im collecting the reults 2011-10-06T18:08:52 yeah, i did some scheme in first year cs 2011-10-06T18:08:58 than i coerce it to a vector, now the list becomes obsolete 2011-10-06T18:09:21 i didnt care much about the GC, i had no problem with it 2011-10-06T18:09:27 i know that its configurable 2011-10-06T18:09:32 but its implementation dependent 2011-10-06T18:09:58 you can delay, or schedule its collection, whatever 2011-10-06T18:10:11 so what method does it use for garbage collecting? i know python does referrence counting for each object 2011-10-06T18:10:18 and when it reaches 0, it gets deleted 2011-10-06T18:10:51 well, in lisp, for lists, theres a great deal of data sharing 2011-10-06T18:11:06 you cant delete something just because theres no reference to it 2011-10-06T18:11:16 parts of other lists can still point to it 2011-10-06T18:11:36 that would still count as a referrence in python 2011-10-06T18:11:41 so that kind of pythonic reference counting will not work 2011-10-06T18:11:57 well, if you mean reference as ANY reference as in a pointer 2011-10-06T18:11:59 *** ibdknox_ has joined #aichallenge 2011-10-06T18:12:03 yes 2011-10-06T18:12:21 but usually by reference we mean symbols, that reference an object in memory 2011-10-06T18:13:04 so when one part of an object references to another, but that part wasnt named, its an indirect reference 2011-10-06T18:13:06 i think 2011-10-06T18:13:12 anyway, i dont know the internals 2011-10-06T18:16:36 once thing that i know, is that the GC runs periodically 2011-10-06T18:17:00 there can be some "hickupps", when the program pauses to g.c. 2011-10-06T18:17:45 and there was some talk about writing a good GC for sbcl, in a "donationware" fashion or something, Nikodemus came up with it 2011-10-06T18:18:07 he waits for donations or something, than writes the software 2011-10-06T18:18:22 they have another name for it but i dont remember what this development stly was called 2011-10-06T18:18:27 style* 2011-10-06T18:19:27 ah, its called crowdfunding 2011-10-06T18:20:12 they improved on SBCL's threading capabilities 2011-10-06T18:23:48 *** olexs has joined #aichallenge 2011-10-06T18:25:08 amstan_: solved 3379120 linear 3x3 systems in 40 secs 2011-10-06T18:25:21 for python it took ~40 minutes 2011-10-06T18:25:33 python code? 2011-10-06T18:25:45 numpy.linalg.solve 2011-10-06T18:26:48 hmm, yeah.. this thing can't really be used with a lot of data at once 2011-10-06T18:26:53 http://pastebin.com/GB3rq27V 2011-10-06T18:27:02 function call overhead is a lot 2011-10-06T18:27:34 mcstar: line 9, ret.append(sol) 2011-10-06T18:27:50 what's norm? 2011-10-06T18:28:04 sqrt(v.v) 2011-10-06T18:28:27 yeah... don't do this in python 2011-10-06T18:28:31 :) 2011-10-06T18:28:42 but in python i can write the code faster 2011-10-06T18:29:00 i will show you the sbcl code, wont be pretty 2011-10-06T18:29:03 yes, but then you should go get a coffee while it calcs 2011-10-06T18:29:21 i had a couple of tbbt's, and was talking here 2011-10-06T18:30:02 what i found surprising, is that numpy doesnt provide looping constructs 2011-10-06T18:30:43 mcstar: yeah, i'm sure there's some ways to group stuff together 2011-10-06T18:31:02 amstan_: why would .append() be faster? 2011-10-06T18:31:14 since you're not creating another list 2011-10-06T18:31:37 why doesnt the compiler recognize this? 2011-10-06T18:31:56 += basically does: for i in toadd: self.append(i) 2011-10-06T18:32:13 because it doesn't have a jit, it's not a compiler 2011-10-06T18:32:15 i think 2011-10-06T18:32:25 this is a trivial situation 2011-10-06T18:32:42 i think the code=sytanx should directly expand to a .append() than 2011-10-06T18:32:43 well.. let me check in #python 2011-10-06T18:32:51 code/syntax 2011-10-06T18:34:49 mcstar: what about array+=[newobj1, newobj2] 2011-10-06T18:34:59 you have to do 2 appends for that 2011-10-06T18:35:12 why? 2011-10-06T18:35:20 why not just append the new list? 2011-10-06T18:35:30 *** okayzed is now known as okay 2011-10-06T18:35:31 i dont know how python stores this internally 2011-10-06T18:35:35 is it a vector? 2011-10-06T18:36:20 mcstar: because this: http://pastebin.com/r9kSkk3z 2011-10-06T18:36:25 mcstar: yes, it pretty much is 2011-10-06T18:37:05 huh? 2011-10-06T18:37:24 what's that supposed to explain? 2011-10-06T18:37:53 mcstar: note the last elements in b and a 2011-10-06T18:38:00 in b, you have c inside a sublist 2011-10-06T18:38:05 i noticed 2011-10-06T18:38:09 in a, c's elements are part of a 2011-10-06T18:38:17 in b, c is itself part of a 2011-10-06T18:38:29 thats why i use a+=[4] 2011-10-06T18:38:31 for example 2011-10-06T18:39:06 thing is that you can't convert a+=c to a.append(c) because they do 2 separate things 2011-10-06T18:39:14 so when you append, put only a number in it 2011-10-06T18:39:16 it'll have to be a+=[c] 2011-10-06T18:39:25 sure 2011-10-06T18:39:25 but then you're making a new list 2011-10-06T18:39:37 that could easily be avoided 2011-10-06T18:39:46 yeah, pypy would probably fix it 2011-10-06T18:40:12 nice suggestions from the python guys 2011-10-06T18:41:08 i left 2011-10-06T18:41:24 lol, yeah, it's a weird channel 2011-10-06T18:41:42 when i ask similar question about lisp, on #lisp, i usually get the answer to use the profiler 2011-10-06T18:41:46 you cannot say lol inthere for some reason, they'll kick you out 2011-10-06T18:41:54 oh 2011-10-06T18:42:05 easier than typing "/close 2011-10-06T18:43:01 a data-structure that can grow, has to be some kind of linked list 2011-10-06T18:43:24 it's probably a tree or something 2011-10-06T18:43:29 nothing guarantees that you can linearly address that structure for all time 2011-10-06T18:43:31 but it acts like a vector of objects 2011-10-06T18:44:05 rich hickey did some good implementations of vectors and hash-maps using trees 2011-10-06T18:44:16 i dont know how much you know about clojure 2011-10-06T18:44:25 nothing 2011-10-06T18:44:34 the data structures in clojure are immutable 2011-10-06T18:44:38 i just know it's lisp under jvm 2011-10-06T18:45:04 thres only one type that is mutable, thats the reference, that you can change using concurrency semantics 2011-10-06T18:45:28 so all this data structures, lists, vectors, maps, they share data 2011-10-06T18:46:00 *** tobym has joined #aichallenge 2011-10-06T18:46:10 and the usual time-complexities still hold 2011-10-06T18:46:32 so he made them immutable, but still fast 2011-10-06T18:46:43 question.. how do you create/add to vectors if they're immutable? 2011-10-06T18:46:54 basically a new vector is created 2011-10-06T18:47:03 and you can change your reference to point to it 2011-10-06T18:47:19 but the thing is, that it will share data with the previous one, its not copied 2011-10-06T18:47:20 isn't that O(n)? 2011-10-06T18:47:26 oh, ok 2011-10-06T18:47:56 so your reference consist of a pointer and a length? 2011-10-06T18:48:05 i have no idea 2011-10-06T18:48:12 bt no 2011-10-06T18:48:14 probably 2011-10-06T18:48:24 that's so weird 2011-10-06T18:48:50 the basic memory model that it uses is a software multi-version concurrency model 2011-10-06T18:49:27 you can reference any snapshot of your data structure 2011-10-06T18:50:01 anyway, i only skimmed clojure, looks very promising though 2011-10-06T18:53:51 yeah, i have a bunch of languages i'm looking at 2011-10-06T18:54:24 mostly haskell 2011-10-06T18:54:32 i should get on that sometime 2011-10-06T18:55:54 ugh.. no #python, you don't have to know why.. 2011-10-06T18:56:37 so.. i was trying to build this ac circuit calculator in python 2011-10-06T18:57:10 since i was mainly using it in the interpreter, it tought it would be a good idea to change the default printing styles for the builtin float and complex 2011-10-06T18:57:15 yeah, turned out it was a bad idea 2011-10-06T18:57:28 because it's like impossible to do 2011-10-06T18:57:56 then the why came 2011-10-06T18:59:49 sry im tired, i dont follow 2011-10-06T18:59:59 amstan_: what do you mean no #python? 2011-10-06T19:00:24 i was just annoyed at their attitude to more eclectic questions 2011-10-06T19:00:33 right now? 2011-10-06T19:00:38 in general 2011-10-06T19:00:41 ok 2011-10-06T19:01:08 well, cl lets you customize nearly everything 2011-10-06T19:01:20 i dont know if you know, but you can modify readtables 2011-10-06T19:01:29 you can actually add syntax to the language 2011-10-06T19:01:48 like in other languages, a combination of characters means an operator 2011-10-06T19:02:02 and you can change how things are printed as well 2011-10-06T19:02:27 >>> float.__str__=lambda self: "zero all the way" 2011-10-06T19:02:29 TypeError: can't set attributes of built-in/extension type 'float' 2011-10-06T19:02:42 if it's coded in C, you can't.. 2011-10-06T19:02:55 cant you supreclass it? 2011-10-06T19:03:00 you can 2011-10-06T19:03:17 let's say you make 'a' an instance of that superclass 2011-10-06T19:03:31 the moment you do a+1.0, it turns back into the original float class 2011-10-06T19:04:04 in c++ you have to overload the + 2011-10-06T19:04:12 but it also controls what type it returns 2011-10-06T19:04:20 how is this different from python's 2011-10-06T19:04:21 yeah, so then i have to overload a ton of things 2011-10-06T19:04:25 ok 2011-10-06T19:04:33 its expected 2011-10-06T19:04:42 including +, -, *, /, the immediate versions too 2011-10-06T19:05:46 it's just a pain though 2011-10-06T19:06:20 i tried making a decorator that overrode a bunch of attributes 2011-10-06T19:06:48 i recently heard about @decorators, but im not sure what they're for 2011-10-06T19:07:16 they're alternative syntax for function=decorator(function) 2011-10-06T19:07:52 for example, you can make a decorator that adds all the decorated functions to an array 2011-10-06T19:08:27 it's useful for stuff like making commands in an interpreter 2011-10-06T19:10:05 argh 2011-10-06T19:10:16 are those the same functions? 2011-10-06T19:10:29 or fun1=decorator(fun2) 2011-10-06T19:10:59 it redefines fun to be the decorated fun 2011-10-06T19:11:12 see this: http://www.python.org/dev/peps/pep-0318/ 2011-10-06T19:11:56 *** onensora has quit IRC () 2011-10-06T19:12:10 ill read it butnot now 2011-10-06T19:12:35 amstan_: my important question is how do you edit python code? 2011-10-06T19:12:39 *** amstan has joined #aichallenge 2011-10-06T19:12:39 *** ChanServ sets mode: +o amstan 2011-10-06T19:12:51 amstan: my important question is how do you edit python code? 2011-10-06T19:13:41 mcstar: what do you mean? 2011-10-06T19:14:07 what is your development style/what editor you use? 2011-10-06T19:14:39 i use kate for editing, and i exec the script until i get the desired results 2011-10-06T19:15:02 sometimes i throw it at the interpreter to look around at vars 2011-10-06T19:15:08 nothing too advanced 2011-10-06T19:15:56 thats horrible 2011-10-06T19:16:00 sorry 2011-10-06T19:16:03 :) 2011-10-06T19:16:41 developing lisp is really a good feeling, you are in control, and i miss that feeling when i edit python code 2011-10-06T19:16:52 it seems that emacs can do a lot in this regard 2011-10-06T19:17:04 it is? 2011-10-06T19:17:14 in control how? 2011-10-06T19:17:19 i want to write function, and send it to the runnig interpreter 2011-10-06T19:17:24 try it out instantly 2011-10-06T19:17:34 redefining anything on the fly 2011-10-06T19:17:58 reloading the whole file every time is very rigid 2011-10-06T19:18:19 the problem is that somehow my python-emacs bridge is a bit broken 2011-10-06T19:18:31 i have to setup my python/ipython mode a bit better 2011-10-06T19:18:47 yeah, you can't really do that in python, since a bunch of stuff gets in the way 2011-10-06T19:18:53 but i couldnt go back to that development style with a dynamic language, that youre talking about 2011-10-06T19:19:10 i used kate too in my early days 2011-10-06T19:19:28 i liked that it displayed the open files on the side, and had sessions 2011-10-06T19:19:43 +1 for embeded terminal! 2011-10-06T19:19:45 and can do code folding 2011-10-06T19:20:03 i never used the e.terminal 2011-10-06T19:20:28 hm, i have 22 windows open 2011-10-06T19:20:32 not much 2011-10-06T19:20:36 i use it a lot 2011-10-06T19:20:45 maan, i can only recommend tiling wms 2011-10-06T19:20:58 use whichever you want, but it will help your productivity 2011-10-06T19:21:06 i like to keep some kind of small tests at the end of each file that executes when i run the file directly 2011-10-06T19:21:34 i have dual screen, so it partially helps 2011-10-06T19:21:46 thats nice too 2011-10-06T19:21:54 how big are they? 2011-10-06T19:21:57 i hear kde has some tiling stuff, but i never got around to using it 2011-10-06T19:22:13 1680xsomething and 1280x1024 2011-10-06T19:22:14 (i tried but i manly disappointed, but that was in march) 2011-10-06T19:22:38 i have the latter 2011-10-06T19:22:56 im very much considering a 24" screen 2011-10-06T19:23:20 but i think it will remain a dream until i get a job 2011-10-06T19:25:01 OMG 2011-10-06T19:25:19 wth happened 2011-10-06T19:25:24 hmm? 2011-10-06T19:25:36 somehow the python one went faster? 2011-10-06T19:25:59 couple of days ago, when i wante dto run my cuda app, it crashed with "cant allocate emmory" 2011-10-06T19:26:12 turned out, my system stopped seeing the cuda card 2011-10-06T19:26:13 *** k333es has joined #aichallenge 2011-10-06T19:26:23 i even pulled the card, and put it back 2011-10-06T19:26:26 no change 2011-10-06T19:26:41 now for the fun, i opend nvidia-settings, and the card show up 2011-10-06T19:27:47 maybe it got lonely 2011-10-06T19:27:59 wanted to talk to the driver 2011-10-06T19:28:05 sure 2011-10-06T19:28:09 :) 2011-10-06T19:28:22 cheap pc hardware 2011-10-06T19:33:19 oh man... reddit.. so hilarious 2011-10-06T19:34:33 *** olexs has quit IRC (Quit: Leaving.) 2011-10-06T19:38:57 i hope youre not talking about that horse 2011-10-06T19:39:22 horse? 2011-10-06T19:39:32 *** okay is now known as okayzed 2011-10-06T19:39:41 ok 2011-10-06T19:39:47 so whats funny than? 2011-10-06T19:40:31 the one with rihanna 2011-10-06T19:40:41 i saw the original post too 2011-10-06T19:41:44 sry, im not interested in the pop world 2011-10-06T19:42:14 but there a horse that passes out during sex 2011-10-06T19:43:37 i'm not either, but it was funny 2011-10-06T19:43:46 ok, i must see this now.. 2011-10-06T19:49:36 *** okayzed is now known as okay 2011-10-06T19:53:20 amstan: was the pw website using zend? 2011-10-06T19:53:45 i don't think so 2011-10-06T19:58:26 so, the zend code has been there and commented out for all time? 2011-10-06T20:03:24 aichallenge: McLeopold epsilon * rd1517cd / (7 files): refactor game_list to use better caching - http://git.io/trOCGQ 2011-10-06T20:03:24 aichallenge: McLeopold epsilon * r47f30fd / (website/starter_packages.php worker/compiler.py): Merge branch 'epsilon' of github.com:aichallenge/aichallenge into epsilon - http://git.io/ZNQigA 2011-10-06T20:04:58 *** mrpancakes has joined #aichallenge 2011-10-06T20:06:04 *** McLeopold has quit IRC (Read error: Connection reset by peer) 2011-10-06T20:19:05 amstan: http://paste.lisp.org/+2OK6 2011-10-06T20:19:08 this is the cl code 2011-10-06T20:19:24 its more or less finished, ill will add an exporter tomorrow 2011-10-06T20:19:32 -will 2011-10-06T20:20:05 mcstar: that is long 2011-10-06T20:20:32 welll 2011-10-06T20:20:39 it does a lot of things 2011-10-06T20:20:54 and its fast :D 2011-10-06T20:20:55 mcstar: maybe try c next time? 2011-10-06T20:20:59 nooo 2011-10-06T20:21:07 wouldnt be much shorter in c 2011-10-06T20:21:25 triple? 2011-10-06T20:21:32 is that another data structure? 2011-10-06T20:21:55 its just a vector 2011-10-06T20:22:09 oh, sorry 2011-10-06T20:22:12 i want to optimize, so i typed my vectors 2011-10-06T20:22:13 i tought it was builtin 2011-10-06T20:22:53 its a vector, of 3, 16bit signed bytes 2011-10-06T20:23:36 make-triple is for making these little vectors without consing 2011-10-06T20:24:13 than i manually compute permutations of not necessarily distinct integers 2011-10-06T20:24:40 but you wrote all this code from scratch? 2011-10-06T20:24:43 yes 2011-10-06T20:24:48 not using any math libraries? 2011-10-06T20:24:53 no 2011-10-06T20:24:59 look at the solve method 2011-10-06T20:25:12 you could try doing that in python next time, then doing pypy on it 2011-10-06T20:25:23 problem with pypy is that it doesn't do numpy 2011-10-06T20:25:41 pypy is the jit? 2011-10-06T20:25:57 ah, thats the python implementation written in python 2011-10-06T20:25:59 right? 2011-10-06T20:27:18 yeah, which is surprisingly faster than python itself 2011-10-06T20:27:36 btw there are packages that provide matrix/vector operations for cl 2011-10-06T20:28:01 next time, or if i do something serious, you can be rest assured, i wont be coding it liek this 2011-10-06T20:29:18 i wonder how long that code is ni J 2011-10-06T20:29:45 ive seen couple of rosetta-code examples in J, really succint 2011-10-06T20:35:11 *sigh* http://marco-leise.homedns.org/antgui/v3/ I started that in december 2010. We've come a long way from there :) 2011-10-06T20:35:25 mleise: you should put it in the repo 2011-10-06T20:36:00 it's stuck at drawing static scenery for me 2011-10-06T20:36:38 yeah, unlike chrome, where just the tab is stuck, your page takes down the whole firefox 2011-10-06T20:36:39 amstan_: could be.. i just had to hack a bit, because of serious bit-rot 2011-10-06T20:36:56 those cosmic rays are a bitch 2011-10-06T20:37:02 cool I guess I have a virus of some sort 2011-10-06T20:38:25 haha 2011-10-06T20:38:42 i could stop the script through a poop up window 2011-10-06T20:38:53 yeah, later 2011-10-06T20:39:05 lol, i wrote poop up 2011-10-06T20:39:22 mleise: why are you still up? 2011-10-06T20:39:26 2:38 here 2011-10-06T20:40:13 oh right, because I am an internet junkie 2011-10-06T20:41:51 right.. i should do my laundry 2011-10-06T20:41:58 do garbage, and wash the dishes 2011-10-06T20:42:32 mleise, mleise: oh, what do you guys think of the header minimization? 2011-10-06T20:43:18 me and aarossig hate it 2011-10-06T20:43:22 amstan_: it could use some polish. now the word ants is harder to read on the black ants logo 2011-10-06T20:44:07 2x him or 1x me? 2011-10-06T20:44:24 mcstar: huh? 2011-10-06T20:44:35 are you asking mleise only? 2011-10-06T20:44:41 mcstar: both 2011-10-06T20:45:00 mcstar: oh man.. i did ask him twice, lol 2011-10-06T20:45:15 i was confused for a second there on why you were so humble 2011-10-06T20:45:26 :) 2011-10-06T20:45:49 i think the sign-in sign-up buttons should be on the top of the sidebar 2011-10-06T20:46:28 its good that the headline is smaller, gives room for the content 2011-10-06T20:47:24 i think it's too small right now though 2011-10-06T20:47:35 janzert: toughts? 2011-10-06T20:48:34 @later tell janzert what do you think of the smaller header on the website? 2011-10-06T20:48:34 amstan_: I come to serve. 2011-10-06T20:48:47 i guess that should be @later ask, lol 2011-10-06T20:48:48 k 2011-10-06T20:48:50 leaving 2011-10-06T20:49:06 me too 2011-10-06T20:49:14 nite/afternoon whatever 2011-10-06T20:49:21 *** mcstar has left #aichallenge ("WeeChat 0.3.5") 2011-10-06T20:52:45 amstan_: sorry for locking up your browser. i defeated the bitrot now (I was switching to Apache which produces a different directory listing) 2011-10-06T20:53:04 *** amstan has quit IRC (Ping timeout: 248 seconds) 2011-10-06T20:54:28 yes, but now there's no replay 2011-10-06T20:54:42 nvm 2011-10-06T20:54:50 yeah 2011-10-06T20:54:52 i remember this 2011-10-06T20:55:48 mleise: omg do they leave trails? 2011-10-06T20:56:24 I just pulled it out again, because someone in a forum asked about it. Yes they left trails :) 2011-10-06T20:56:44 is the forum clear of spammers 2011-10-06T20:58:13 *** mrpancakes has quit IRC (Quit: Leaving) 2011-10-06T21:00:11 *** ibdknox_ has quit IRC (Remote host closed the connection) 2011-10-06T21:00:35 *** ibdknox_ has joined #aichallenge 2011-10-06T21:05:14 *** ibdknox_ has quit IRC (Ping timeout: 260 seconds) 2011-10-06T21:07:12 amstan_: It could be so much worse: http://www.blackdeathgame.com/forum/viewforum.php?f=3&sid=1b8461907844e59b9b95c5e12804dd64 2011-10-06T21:45:21 *** okay is now known as okayzed 2011-10-06T22:05:03 cool blog if you're into math/puzzles: http://nbickford.wordpress.com/ 2011-10-06T22:15:40 *** McLeopold has joined #aichallenge 2011-10-06T23:30:42 *** amstan has joined #aichallenge 2011-10-06T23:30:42 *** ChanServ sets mode: +o amstan 2011-10-06T23:33:30 *** amstan_ has quit IRC (Ping timeout: 260 seconds) 2011-10-06T23:48:19 *** okayzed is now known as okay 2011-10-06T23:49:27 aichallenge: McLeopold epsilon * r260b18c / website/sql.php : fix next game in logic - http://git.io/qDmKFw 2011-10-06T23:55:33 *** McLeopold has quit IRC (Read error: Connection reset by peer)