2011-05-08T00:05:16 amstan: same as a main server basically but run worker_setup.py instead of server_setup.py 2011-05-08T00:05:44 it should work if i execute it on the same server with the manager, right? 2011-05-08T00:05:59 I believe so but I haven't tested that in a while 2011-05-08T00:06:40 also the cron job and worker startup script mentioned in the setup don't actually work 2011-05-08T00:06:54 it is still pointing to the old planetwars scripts 2011-05-08T00:06:58 API Base url? would be http://hypertriangle.com:13080/ right? 2011-05-08T00:07:08 yep 2011-05-08T00:07:15 where do i get the api key from? 2011-05-08T00:07:17 where api_* pages are found 2011-05-08T00:07:36 you set it in the main server server_info.php 2011-05-08T00:07:55 it can be any string basically 2011-05-08T00:08:01 what's the key for that property? 2011-05-08T00:08:07 I'd recommend a random 128bit hex number :) 2011-05-08T00:08:15 there's no line with key in there 2011-05-08T00:08:36 I don't remember what the variable name is offhand... 2011-05-08T00:09:10 looks like it is 'api_key' 2011-05-08T00:09:28 actually I'm wrong 2011-05-08T00:10:29 the api_key in the server_info.php is used for api_server_setup.php and it generates a specific worker key 2011-05-08T00:10:39 it is the worker key that the worker_setup.py needs 2011-05-08T00:10:52 you'll need to manually insert that into the main server database 2011-05-08T00:11:24 so, what do i need to do then? run api_server_setup.php? 2011-05-08T00:11:52 basically pick a key then "insert into worker SET api_key = 'workerkey', ip_address = '';" 2011-05-08T00:12:11 the api_server_setup.php is for automatic worker setup with ec2 2011-05-08T00:12:18 ok.. first.. i have to get api_key into the server_info.php 2011-05-08T00:12:27 i wanted to fix that first, on the template ones too 2011-05-08T00:13:01 you should be able to actually skip the api_key in server_info.php unless you'll be doing automatic worker deployment 2011-05-08T00:13:24 yeah, but i'm both verifying that everything goes smooothly and setting this up 2011-05-08T00:13:27 it should only be used in api_server_setup 2011-05-08T00:13:35 so i want to have the repo either working nicely, or documented at the end 2011-05-08T00:14:12 and api_server_setup.php is broked too 2011-05-08T00:14:32 obviously since it still points at the old svn repo :P 2011-05-08T00:14:36 yeah 2011-05-08T00:14:38 i noticed 2011-05-08T00:16:17 aichallenge: Alexandru Stan epsilon * rc5bb76d / (4 files in 4 dirs): deleted duplicate sample files(all the templates are now in the setup folder), added the api_key option in the sample server_info.php - http://bit.ly/muCk0Z 2011-05-08T00:16:18 k.. so.. 2011-05-08T00:16:21 now there's that.. 2011-05-08T00:16:30 i need to make a random 128 bit string 2011-05-08T00:18:19 there we go: print "%x"%random.randint(0,2**128) 2011-05-08T00:21:29 *** bhasker has joined #aichallenge 2011-05-08T00:21:40 sigh.. 2011-05-08T00:21:42 [Sun May 08 00:21:22 2011] [error] [client 192.168.0.1] PHP Fatal error: Call to a member function get() on a non-object in /home/contest/aichallenge/website/index.php on line 15 2011-05-08T00:21:46 memcached is installed, running 2011-05-08T00:21:48 but i still get that 2011-05-08T00:24:11 *** bhasker has quit IRC (Client Quit) 2011-05-08T00:29:17 aichallenge: Alexandru Stan epsilon * r3c9b663 / website/index.php : added check so index.php doesn't crash when memcached is not properly working(need help on this too) - http://bit.ly/lnhI8x 2011-05-08T00:32:57 aichallenge: Alexandru Stan epsilon * r08f7467 / website/.gitignore : added forums, phpmyadmin to the gitignore in website - http://bit.ly/kyddgr 2011-05-08T00:35:25 janzert: what's with $server_info["api_create_key"] ? 2011-05-08T00:36:22 sorry, yes 'api_create_key' is the correct variable name in server_info.php 2011-05-08T00:36:27 ugh.. 2011-05-08T00:36:36 no idea about memcached 2011-05-08T00:37:00 janzert: the real problem with memcached is that it didn't have a check if it was loading like the other files, my commit fixed that 2011-05-08T00:38:59 but it still isn't working correctly? 2011-05-08T00:39:38 from what McLeopold said before it sounded like the package just needed to be installed on the server and no further configuration was necessary 2011-05-08T00:40:02 janzert: it worked on my main hypertriangle server, but on the test 13000 server it didn't 2011-05-08T00:40:10 seems like the module is not active or something 2011-05-08T00:40:17 even though i did a2enmod 2011-05-08T00:40:25 maybe the firewall is still active or something 2011-05-08T00:40:34 i don't think ufw is disabled 2011-05-08T00:40:57 it'd be rather weird for a firewall to be blocking localhost communication 2011-05-08T00:41:27 aichallenge: Alexandru Stan epsilon * rcf1e0ac / setup/server_info.php.template : real key for the setting of api_create_key - http://bit.ly/isFTZS 2011-05-08T00:44:13 janzert: why is server_info.php owned by root? is that my bad or is it supposed to be that way? 2011-05-08T00:47:21 janzert: ok.. so i have the server_key setup properly.. i have a key from server_setup.php 2011-05-08T00:47:30 though it's giving me all kinds of other random stuff 2011-05-08T00:47:42 from the svn, that's not valid anymore 2011-05-08T00:48:03 i guess we'll have to fix that sometime 2011-05-08T00:48:22 what does Install startup cronjob? [no] do? 2011-05-08T00:50:50 oh my.. all these packages 2011-05-08T00:51:24 * amstan wonders why /var/cache/apt/archives is not used, they're all already in there 2011-05-08T00:51:43 would cut down installation time a lot 2011-05-08T00:52:14 the startup cronjob would install a cronjob to start the worker script on a reboot 2011-05-08T00:52:29 how do i do this manually? 2011-05-08T00:52:34 all packages are being installed in a chroot so the base system isn't available 2011-05-08T00:52:44 what part? 2011-05-08T00:52:56 the worker.py script 2011-05-08T00:53:12 mcleopold gave me an example, but i just want to make sure we're all on the same wavelength 2011-05-08T00:53:21 I haven't actually run it 2011-05-08T00:53:22 something like ./worker.py 0 2011-05-08T00:53:24 oh 2011-05-08T00:53:30 but just look at the bottom of the file 2011-05-08T00:53:40 you can see the optparse stuff pretty easily 2011-05-08T00:53:55 it's something like worker.py -t something something :} 2011-05-08T00:54:20 what does worker_init.py do? 2011-05-08T00:54:35 it has something to do with that .tgz file, when do i run that? 2011-05-08T00:54:41 or did it already? 2011-05-08T00:54:43 you don't 2011-05-08T00:55:05 it's used for the automatic deploy to setup the environment to run worker_setup.py 2011-05-08T00:55:16 well shit: http://pastebin.com/z2JEyaQY 2011-05-08T00:55:18 lol 2011-05-08T00:55:26 oh.. right.. 2011-05-08T00:55:32 that might be one issue, i don't have amd64 2011-05-08T00:55:33 e.g. create the worker, download that .tgz file, unpack it in the right spot 2011-05-08T00:55:52 hmm 2011-05-08T00:56:23 you can try just changing it to i386 and see if it works I suppose 2011-05-08T00:56:52 although your system is going to be in a state that worker_setup probably can't continue from at this point 2011-05-08T00:56:54 full paste: http://pastebin.com/fMg60AbN 2011-05-08T00:57:04 anything that i can delete? 2011-05-08T00:57:28 /srv/chroot/* and let me check if there's anything else 2011-05-08T00:57:53 there's an aic-base 2011-05-08T00:57:56 i could delete that 2011-05-08T00:59:26 yep, I think just 'rm -rf /srv/chroot' should do it 2011-05-08T00:59:52 from where the script crashed there 2011-05-08T01:00:14 any idea how to make ubuntu regurgitate what arch and version it is? 2011-05-08T01:00:23 i could try looking in sources.list, but that's a hack 2011-05-08T01:00:30 i think there was a file in /etc/ or something 2011-05-08T01:00:37 looking for? 2011-05-08T01:00:54 oh, missed the first question 2011-05-08T01:00:58 no idea 2011-05-08T01:01:22 there's debian_version 2011-05-08T01:01:30 but I'm not sure how to check the ubuntu version 2011-05-08T01:01:41 /etc/lsb-release => DISTRIB_CODENAME=natty 2011-05-08T01:04:16 on memcache again, did php5-memcached and memcached packages both get installed? 2011-05-08T01:04:32 janzert: yes, i have them both installed, the extension is working 2011-05-08T01:04:40 ahh 2011-05-08T01:05:39 janzert: i think "if (extension_loaded ('memcache')) {" fails 2011-05-08T01:08:10 hmm, that's a php extension nothing to do with apache btw 2011-05-08T01:08:41 oh 2011-05-08T01:08:45 darn 2011-05-08T01:08:59 maybe printing the output of get_loaded_extensions() to see if it is in the list 2011-05-08T01:09:09 s/maybe/maybe try 2011-05-08T01:09:11 can i do it with php-cli? 2011-05-08T01:09:30 I'm not positive if that loads the same extensions or not 2011-05-08T01:09:33 array.. ugh.. 2011-05-08T01:09:54 janzert: how about a php_info 2011-05-08T01:10:35 that should list it too 2011-05-08T01:10:39 http://hypertriangle.com:13080/phpinfo.php 2011-05-08T01:10:54 it's in there.. 2011-05-08T01:11:40 looks like it's called memcached their and memcache in the code? 2011-05-08T01:12:13 McLeopold told me those were the packages he was using but now I wonder 2011-05-08T01:12:31 'cause I know he was going back and forth between a few different ones at one point 2011-05-08T01:12:33 well.. presumably the d at the end is fro daemon 2011-05-08T01:13:00 yes, but the extension still seems to be named memcached 2011-05-08T01:13:25 at least I don't see a plain memcache extension in that list 2011-05-08T01:13:40 http://home.hypertriangle.com/phpinfo.php 2011-05-08T01:13:42 you're right.. 2011-05-08T01:13:49 the other server i have is just memcache, without d 2011-05-08T01:14:03 both running natty? 2011-05-08T01:14:20 wonder if it's something that changed 2011-05-08T01:14:22 nope, my home.hypertriangle is karmic 2011-05-08T01:15:05 lucid* 2011-05-08T01:15:58 *** sigh has joined #aichallenge 2011-05-08T01:21:16 *** sigh has quit IRC (Ping timeout: 258 seconds) 2011-05-08T01:27:28 *** sigh has joined #aichallenge 2011-05-08T01:32:00 *** NESTEA has quit IRC (Ping timeout: 252 seconds) 2011-05-08T01:33:14 aichallenge: Alexandru Stan fix-32bit-worker * radbcb5d / setup/install_tools.py : added get_ubuntu_release_info - http://bit.ly/kLxX1V 2011-05-08T01:33:15 janzert: heheh ^ 2011-05-08T01:36:27 aichallenge: Alexandru Stan fix-32bit-worker * r6d66382 / setup/worker_setup.py : made worker_setup to install the same ubuntu in the chroot as the host is, using get_ubuntu_release_info - http://bit.ly/lmwp9o 2011-05-08T01:36:29 janzert: may i merge that? 2011-05-08T01:37:15 https://github.com/aichallenge/aichallenge/compare/epsilon...fix-32bit-worker 2011-05-08T01:42:14 hmm 2011-05-08T01:42:41 I'd rather just match the arch 2011-05-08T01:43:02 janzert: what? 2011-05-08T01:43:06 you need to change more than just that when switching ubuntu versions 2011-05-08T01:43:29 e.g. see my commit a while back to update to natty 2011-05-08T01:43:43 and if you go older than maverick the changes are much greater 2011-05-08T01:44:03 grep -R natty . only return worker_setup.py about a broken package and your sources.list 2011-05-08T01:44:10 returns* 2011-05-08T01:44:29 yeah, but I'd still rather just stick with a known configuration 2011-05-08T01:44:48 we could put a check in the installer, a warning if it's not natty 2011-05-08T01:45:03 but i would like to have less things hardcoded 2011-05-08T01:45:04 pre-maverick the chroot configuration moves location a couple of times 2011-05-08T01:45:22 I'd rather hardcode so we can say, this is what it is 2011-05-08T01:45:29 my commits there don't imply we'll be able to use anything but natty 2011-05-08T01:45:53 what about putting a check for that? 2011-05-08T01:46:04 less magic, more explicit in other words 2011-05-08T01:46:21 and throw an error if it's not natty? 2011-05-08T01:46:27 yeah 2011-05-08T01:46:30 that would be fine 2011-05-08T01:46:35 assert version==natty 2011-05-08T01:47:08 janzert: about that sources.list, how about copying it from host? 2011-05-08T01:47:15 also have you checked to see if changing the arch actually gives you a working system? 2011-05-08T01:47:29 I'd rather not 2011-05-08T01:47:32 janzert: i'm installing i386 right now 2011-05-08T01:47:51 since the host may have extra sources 2011-05-08T01:47:59 what's wrong with extra sources? 2011-05-08T01:48:04 and we may also want to add some ppa in the future 2011-05-08T01:48:16 we could end up with a different version of a package than we expect 2011-05-08T01:48:29 is that a problem for the chroot machine? 2011-05-08T01:48:52 it's what the bots are compiled and run under 2011-05-08T01:49:14 shouldn't that comp have access to the same software as the host? 2011-05-08T01:49:17 i don't understand 2011-05-08T01:49:19 nope 2011-05-08T01:49:31 the chroot only has access to what is installed in it 2011-05-08T01:49:38 also, aren't we installing the languages and stuff on the host? 2011-05-08T01:49:42 it can technically be a whole different distribution 2011-05-08T01:49:43 nope 2011-05-08T01:50:33 well.. shit: http://pastebin.com/ZYdt15Xg 2011-05-08T01:50:44 my lsb_release doesn't exist in --chroot-setup mode 2011-05-08T01:50:59 i should probably move that line, to somewhere where i'm sure it's only ubuntu 2011-05-08T01:51:25 ahh, I didn't notice that was setting global variables 2011-05-08T01:51:37 janzert: any suggestions? 2011-05-08T01:51:42 yeah, just move that line into setup_base_chroot 2011-05-08T01:51:59 should i just do it in setup_base_chroot, where it's needed? 2011-05-08T01:52:18 yp, setup_base_chroot won't be executed in the chroot itself 2011-05-08T01:52:21 yep 2011-05-08T01:52:55 you're only about halfway through the chroot setup btw ;) 2011-05-08T01:53:02 yes, i noticed 2011-05-08T01:56:38 *** Chris_0076 has joined #aichallenge 2011-05-08T01:57:37 janzert: i was thinking of: if version!="natty": raise Exception("This contest framework was designed to work on Ubuntu Natty(11.04) only.") 2011-05-08T01:58:08 sure, put that right at the start of main 2011-05-08T01:58:26 then also shove the arch into options 2011-05-08T01:58:37 janzert: which options? 2011-05-08T01:58:50 *** nullkuhl has quit IRC (Read error: No route to host) 2011-05-08T01:59:08 it's the options object that gets passed around for everything to pull the setup configuration out of 2011-05-08T01:59:09 *** nullkuhl has joined #aichallenge 2011-05-08T01:59:09 *** nullkuhl has joined #aichallenge 2011-05-08T01:59:25 *** Cyndre has quit IRC (Read error: Connection reset by peer) 2011-05-08T01:59:42 *** Cyndre has joined #aichallenge 2011-05-08T01:59:46 it's called opts in main 2011-05-08T02:01:20 so move your get_ubuntu_release_info call to the top of main, throw an exception if not natty and the set opts.arch = ubuntu_arch after the get_options call 2011-05-08T02:01:37 ok 2011-05-08T02:01:50 but I also thought of a problem since lsb_release isn't in the chroot you'll need to catch that exception 2011-05-08T02:02:27 probably just set the version to natty if lsb_release has an error 2011-05-08T02:03:53 it'll allow a false positive if run on a distribution without lsb_release but it's going to get pretty intertwined to get everything working correctly otherwise 2011-05-08T02:05:07 janzert: the chroot has /etc/lsb-release 2011-05-08T02:05:11 so i might change to that.. 2011-05-08T02:11:47 *** Mathnerd314 has quit IRC (Read error: Connection reset by peer) 2011-05-08T02:16:09 aichallenge: Alexandru Stan fix-32bit-worker * rc5748f0 / setup/install_tools.py : made get_release_info more stable, added check_ubuntu_version - http://bit.ly/klp8Og 2011-05-08T02:16:29 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-05-08T02:18:01 janzert: opts is a dict, right? 2011-05-08T02:28:42 aichallenge: Alexandru Stan epsilon * r5fdc4d6 / ants/dist/starter_bots/ruby/ants.rb : 2011-05-08T02:28:43 aichallenge: Merge pull request #85 from MatmaRex/epsilon 2011-05-08T02:28:43 aichallenge: Ruby starter - helper functions, no more r lines - http://bit.ly/kyBxcR 2011-05-08T02:29:28 janzert: so.. see that instal_tools.py commit, it shouldn't crash anymore 2011-05-08T02:29:47 janzert: do you still want me to do those opt things? 2011-05-08T02:30:40 opts is an optparse values object 2011-05-08T02:30:51 never worked with them before 2011-05-08T02:30:54 can i use it as a dict? 2011-05-08T02:31:27 basically for what you're doing just use it like a regular object and set attributes on it 2011-05-08T02:32:04 oh, so.. opt.ubuntu_arch, opt_ubuntu_version = check_ubuntu_version() 2011-05-08T02:32:14 basically 2011-05-08T02:32:16 opt.* 2011-05-08T02:32:42 but I think you only need to put the arch into it 2011-05-08T02:32:57 what about the bit from worker_setup.py? 2011-05-08T02:33:00 just do the version check directly in the main function and raise an exception if it's wrong 2011-05-08T02:33:18 i need the version in there.. 2011-05-08T02:33:35 this is all in worker_setup.py 2011-05-08T02:33:56 in the main function first line do version, arch = check_ubuntu_version() 2011-05-08T02:34:14 then compare the version to natty and raise exception if not 2011-05-08T02:34:27 then call get_options to get the opts object 2011-05-08T02:34:39 after that call set opts.ubuntu_arch = arch 2011-05-08T02:35:06 then use that in the chroot setup function 2011-05-08T02:35:16 yes, but what about %s http://us.archive.ubuntu.com/ubuntu/" % (ubuntu_arch, ubuntu_version, base_chroot_dir,)) 2011-05-08T02:35:20 it needs ubuntu_version 2011-05-08T02:35:24 and that's not in the same scope 2011-05-08T02:35:26 hardcode it to natty 2011-05-08T02:35:29 aww 2011-05-08T02:35:32 after all that work, lol 2011-05-08T02:35:44 since that is all everything will work with anyway 2011-05-08T02:36:10 if it's anything but natty the sources.list that gets installed will cause problems later on 2011-05-08T02:37:10 janzert: well.. i was just wondering if we could not hardcode it to natty there, we already have the exception for that 2011-05-08T02:37:21 it's pretty much the same functionality 2011-05-08T02:38:16 it makes the code simpler to hardcode it 2011-05-08T02:38:29 yeah.. ok 2011-05-08T02:38:31 and not hardcoding it just opens up the opportunity for more errors later on 2011-05-08T02:39:43 aichallenge: Alexandru Stan fix-32bit-worker * r2d0550c / (3 files): only included the check in the main functions, hardcoded natty back into the debootstrap - http://bit.ly/jFq6Qs 2011-05-08T02:40:03 like that? 2011-05-08T02:40:27 full diff: https://github.com/aichallenge/aichallenge/compare/epsilon...fix-32bit-worker 2011-05-08T02:42:06 janzert: ^ 2011-05-08T02:42:33 *** kaemo has joined #aichallenge 2011-05-08T02:44:01 yep looks fine 2011-05-08T02:45:28 well 2011-05-08T02:45:32 i wanted to install using that.. 2011-05-08T02:45:34 but.. 2011-05-08T02:45:45 it somehow tought it was done the base chroot 2011-05-08T02:45:49 and now i have a bunch of users 2011-05-08T02:45:57 how do i reverse it? 2011-05-08T02:47:15 do i just delete the users? 2011-05-08T02:47:27 and /srv/chroot again? 2011-05-08T02:49:23 yeah 2011-05-08T02:49:39 I'm pretty sure that should do it 2011-05-08T02:50:34 janzert: oh man.. fstab is full too 2011-05-08T02:51:06 oh yeah, if it added the users 2011-05-08T02:52:56 that's the really nice thing about ec2 and using a vm for testing, it's very easy to start over fresh again :} 2011-05-08T02:53:27 yeah, i would use them if i had access to them 2011-05-08T02:54:41 aichallenge: Alexandru Stan epsilon * radbcb5d / setup/install_tools.py : added get_ubuntu_release_info - http://bit.ly/kLxX1V 2011-05-08T02:54:42 aichallenge: Alexandru Stan epsilon * r2d0550c / (3 files): only included the check in the main functions, hardcoded natty back into the debootstrap - http://bit.ly/jFq6Qs 2011-05-08T02:54:44 aichallenge: Alexandru Stan epsilon * r6d66382 / setup/worker_setup.py : made worker_setup to install the same ubuntu in the chroot as the host is, using get_ubuntu_release_info - http://bit.ly/lmwp9o 2011-05-08T02:54:47 aichallenge: Alexandru Stan epsilon * rc5748f0 / setup/install_tools.py : made get_release_info more stable, added check_ubuntu_version - http://bit.ly/klp8Og 2011-05-08T02:55:56 k.. running setup 2011-05-08T02:55:58 and there's an error 2011-05-08T02:56:00 hmm 2011-05-08T02:58:18 aichallenge: Alexandru Stan epsilon * r1624e21 / setup/worker_setup.py : fixed typo - http://bit.ly/iAhAWD 2011-05-08T02:58:56 i cannot catch a break.. 2011-05-08T02:59:07 :{ 2011-05-08T02:59:18 setup_base_chroot does not have opts in it.. 2011-05-08T02:59:29 it's options? 2011-05-08T02:59:40 i guess 2011-05-08T02:59:43 is that the same thing? 2011-05-08T03:01:48 * amstan is getting irritated about those are you sure messages, lol 2011-05-08T03:04:22 ok.. maybe it's time for me to go to bed 2011-05-08T03:04:36 i don't care what happens to that install process now, i'll reset it tomorrow if i have to 2011-05-08T03:07:38 *** sigh has quit IRC (Remote host closed the connection) 2011-05-08T03:08:30 *** sigh has joined #aichallenge 2011-05-08T03:09:26 :/ 2011-05-08T03:09:51 my copying of .deb files seemed to have sped up the process 2011-05-08T03:09:53 i might wait 2011-05-08T03:10:00 it's getting java now 2011-05-08T03:10:05 yeah it's basically only called opts in main because I was too lazy to spell it out 2011-05-08T03:10:19 you'll see it gets passed down as a parameter to the various functions 2011-05-08T03:10:46 yeah, i s/*/options/g 'd it 2011-05-08T03:11:13 i should really stop doing --amends and push -f 2011-05-08T03:11:49 fonts... why would it need fonts? lol 2011-05-08T03:11:58 java demands it :P 2011-05-08T03:12:17 actually I don't know what the dependency is that pulls 'em in 2011-05-08T03:12:32 we should make a cd 2011-05-08T03:12:39 with the contest framework 2011-05-08T03:12:48 just install everything from there 2011-05-08T03:12:49 lol 2011-05-08T03:12:57 it'll have debs too 2011-05-08T03:13:23 i really like this sandbox system 2011-05-08T03:13:40 you can pretty much try different versions of ubuntu in it, assuming you don't care about drivers 2011-05-08T03:14:09 s/ubuntu/anydistro?/g 2011-05-08T03:19:28 and now i'm compiling go 2011-05-08T03:20:05 we should really get that repo for haskell and the others 2011-05-08T03:20:17 just make our own debs with the compiled stuff 2011-05-08T03:24:12 yeah, that'd be nice 2011-05-08T03:24:43 and yeah I think so long as the distro can use whatever kernel is currently in use you should be able to put it in a chroot 2011-05-08T03:33:47 janzert: yay.. it's done 2011-05-08T03:33:59 sweet 2011-05-08T03:34:20 try running play_one_game and see if it works 2011-05-08T03:35:49 janzert: where from? 2011-05-08T03:36:06 in the ants directory 2011-05-08T03:36:17 it just runs a game with some of the sample bots 2011-05-08T03:36:47 it launched w3m on me 2011-05-08T03:37:22 contest@aichallengebeta:~/aichallenge/ants$ cat game_logs/0.replay 2011-05-08T03:37:23 {"error": "'repo_path'"} 2011-05-08T03:37:45 ugh, that must be one of the changes McLeopold made with his recent checkins 2011-05-08T03:37:50 the w3m thing that is 2011-05-08T03:38:31 hmm, that's a rather unhelpful error 2011-05-08T03:38:49 maybe check the repo_path setting in the worker/server_info.py 2011-05-08T03:38:49 yeah.. 2011-05-08T03:39:17 it should be relative to the root_path 2011-05-08T03:39:24 "root_path": "/home/contest", 2011-05-08T03:39:25 not an absolute path by the way 2011-05-08T03:39:29 there's no repo_path 2011-05-08T03:39:48 really 2011-05-08T03:40:09 in the worker directory? 2011-05-08T03:40:15 yeah... that's weird.. 2011-05-08T03:40:24 should i add it manually? 2011-05-08T03:40:42 sure, but also is there one in the template in your setup directory? 2011-05-08T03:40:44 the server_info.py file in worker seems to be really skeletal 2011-05-08T03:40:52 setup/worker_server_info.py.template 2011-05-08T03:40:54 setup has it properly 2011-05-08T03:40:55 yeah 2011-05-08T03:41:11 should i attempt to fix this? 2011-05-08T03:41:35 * janzert checks to see if a server_info.py got checked in somehow 2011-05-08T03:41:50 nope 2011-05-08T03:41:56 janzert: i don't dismiss the possibility of me having had that file before 2011-05-08T03:41:59 yeah, you could just fix it manually 2011-05-08T03:42:15 probably just copy over the template and fill in the values 2011-05-08T03:42:25 yeah.. my server_info has stuff like db_username in it 2011-05-08T03:42:39 i think it's a manager server_info file 2011-05-08T03:42:41 it must be left over from a while ago 2011-05-08T03:42:47 ahh, that could be 2011-05-08T03:44:31 janzert: play one game is definatelly doing ants stuff now 2011-05-08T03:44:36 great 2011-05-08T03:44:41 ran about 8 jailusers 2011-05-08T03:44:46 and it's turning now.. 2011-05-08T03:44:48 woot 2011-05-08T03:44:59 the sandbox chroots should be working then 2011-05-08T03:45:56 k.. wtf? http://hypertriangle.com:13080/visualizer.php?game_id=0 2011-05-08T03:46:54 http://hypertriangle.com:13080/ants/games/0/0.replay 2011-05-08T03:47:16 heh 2011-05-08T03:47:48 ok.. editing it a little 2011-05-08T03:47:56 did McLeopold change the replay format in his last batch of checkins? 2011-05-08T03:48:21 yay 2011-05-08T03:48:22 http://hypertriangle.com:13080/visualizer.php?game_id=0 2011-05-08T03:48:24 there we go 2011-05-08T03:48:44 and that game was just ran in the sandbox.. 2011-05-08T03:48:45 yay 2011-05-08T03:48:47 cool 2011-05-08T03:49:03 so.. bugs discovered today while setting up the sandbox 2011-05-08T03:49:05 very few 2011-05-08T03:49:08 janzert: good job 2011-05-08T03:49:41 well what you did is what I've been doing locally for testing :) 2011-05-08T03:50:17 contestbot: later tell mcleopold I have the sandbox ready on the hypertriangle beta server, when you have time we can work on that 2011-05-08T03:50:17 amstan: Ready to serve, my lord. 2011-05-08T03:50:26 the only thing i'm affraid now is the sql state.. 2011-05-08T03:50:32 ? 2011-05-08T03:50:41 oh just the overall state of the sql 2011-05-08T03:50:43 i have a feeling my schema is out of date 2011-05-08T03:50:47 on the server 2011-05-08T03:50:49 ahh 2011-05-08T03:51:06 and i have no idea what clearing it and reinstalling it would do 2011-05-08T03:51:37 the thing I'm dreading right now is that to make it work at all efficiently I've gotta change the way the sandbox class works a bit 2011-05-08T03:51:53 for compilation that is 2011-05-08T03:52:20 would that be a big problem to change? 2011-05-08T03:52:49 not really it's just that I really should change the API of the class and that means I'll need to change the engine to match 2011-05-08T03:52:56 I'd rather not touch too much 2011-05-08T03:53:05 *** sigh_ has joined #aichallenge 2011-05-08T03:56:47 *** sigh has quit IRC (Ping timeout: 240 seconds) 2011-05-08T03:58:18 *** Zepp has quit IRC (Ping timeout: 240 seconds) 2011-05-08T03:58:23 *** Zepp has joined #aichallenge 2011-05-08T04:02:08 *** sigh_ has quit IRC (Read error: Connection reset by peer) 2011-05-08T04:04:24 *** Kingpin13 has joined #aichallenge 2011-05-08T04:04:29 *** Kingpin13 has joined #aichallenge 2011-05-08T04:10:10 *** berak has joined #aichallenge 2011-05-08T04:12:15 *** sigh has joined #aichallenge 2011-05-08T04:26:23 *** amstan has quit IRC (Ping timeout: 240 seconds) 2011-05-08T05:08:13 *** mcstar has joined #aichallenge 2011-05-08T05:20:57 *** qacek has quit IRC (Ping timeout: 252 seconds) 2011-05-08T05:21:21 *** qacek has joined #aichallenge 2011-05-08T05:29:18 *** sigh_ has joined #aichallenge 2011-05-08T05:31:55 *** delt0r___