2011-04-24T00:00:19 *** Mathnerd314 has quit IRC (Ping timeout: 250 seconds) 2011-04-24T00:00:26 McLeopold: that's the right idea. On the landing page it suffices to say something like "you have ants that fight other ants". 2011-04-24T00:00:30 amstan: what do you want me to do about LaTeX on the specs page? 2011-04-24T00:00:33 should i do my best without it? 2011-04-24T00:00:39 antimatroid: idk 2011-04-24T00:00:43 antimatroid: what about it? 2011-04-24T00:00:50 it isn't rendering it when you convert 2011-04-24T00:00:56 look at the distance section on the specs page 2011-04-24T00:01:10 antimatroid: PDF? 2011-04-24T00:01:33 j3camero: you can embed some LaTeX (they don't have all the packages i would like) into github wikis 2011-04-24T00:02:03 *** Chris_0076 has quit IRC (Ping timeout: 250 seconds) 2011-04-24T00:02:17 Welcome to Google AI Challenge where you have a chance to write artificial intelligence programs with the utmost basic programming skills. You will be writing programs to compete in video games against other programs. 2011-04-24T00:02:20 *video* 2011-04-24T00:02:27 aichallenge: Alexandru Stan wwwdesign * re8bc5b1 / (9 files in 4 dirs): added favicon, added menu - http://bit.ly/h1Amna 2011-04-24T00:02:55 maybe i've gone too simple.. 2011-04-24T00:05:18 also, i created that ant on the new website design in inkscape. the svg file should be in the repo (afaik, amstan added it) 2011-04-24T00:05:30 we can do whatever with it, large versions could be fun 2011-04-24T00:05:41 computerwiz_222: to generic, more like mine, it needs to be "salesy" 2011-04-24T00:06:22 "Ants is a multi-player strategy game where colonies of ants fight for domination and you get to control one of the colonies. You participate by writing a computer program in a language of your choice that inputs data about the current game state and outputs which moves it would like to make. The objective is to eliminate the enemy colonies while taking over all of the resources." 2011-04-24T00:06:26 with a sample game below it 2011-04-24T00:06:37 aichallenge: Alexandru Stan epsilon * r4be86f3 / (20 files in 7 dirs): added a css-only version, moved the old version to site-draft - http://bit.ly/hfj4Ll 2011-04-24T00:06:38 aichallenge: Alexandru Stan epsilon * re8bc5b1 / (9 files in 4 dirs): added favicon, added menu - http://bit.ly/h1Amna 2011-04-24T00:06:40 aichallenge: Alexandru Stan epsilon * r698a3a3 / (12 files in 4 dirs): Merge branch 'wwwdesign' into epsilon - http://bit.ly/fC6yv9 2011-04-24T00:06:43 somebody please ruin my poor english skills 2011-04-24T00:06:51 even though it's my only language 2011-04-24T00:07:25 oh man.. 2011-04-24T00:07:27 git merge 2011-04-24T00:07:28 http://hypertriangle.com:13080/ 2011-04-24T00:07:33 it did all the magic 2011-04-24T00:07:34 "Ants is a multi-player strategy game where colonies of ants fight for domination. You participate by writing a computer program in a language of your choice that inputs data about the current game state and outputs which moves it would like to make. The objective is to eliminate the enemy colonies while taking over all of the resources." 2011-04-24T00:07:37 "New to programming? No problem. We can help you get started." 2011-04-24T00:07:39 nicely done antimatroid 2011-04-24T00:07:50 computerwiz_222: http://hypertriangle.com:13080/quickstart.php 2011-04-24T00:07:54 i just removed something that i felt was out of place.. idk why 2011-04-24T00:07:54 The Google AI Challenge is all about creating artificial intelligence, whether you are a complete beginner or an expert. Using the easy-to-use starter kit, you will create a computer program that controls a simulated colony of ants. Submit your program online to watch your ant colony fight for domination against colonies created by other people from around the world. It only takes 5 minutes to get started. 2011-04-24T00:08:21 amstan: would you look at this... this new design is working well 2011-04-24T00:08:23 computerwiz_222: can you try playing around with a "dirt" coloured background where white is? 2011-04-24T00:08:38 i like the design, but do feel a few of mcleopolds sentiments from yesterday 2011-04-24T00:08:45 the blue still doesn't feel quite right 2011-04-24T00:08:56 antimatroid: he tried white, it didn't look as cool 2011-04-24T00:08:57 antimatroid: i tried a few things... from an accessabiltiy point, i don't want to make it difficult to read the content text 2011-04-24T00:09:06 yep 2011-04-24T00:09:13 antimatroid: to be honest, it depends vastly on the monitor 2011-04-24T00:09:17 at work, it looks like ass 2011-04-24T00:09:23 (well.. the old version) 2011-04-24T00:09:30 i adjusted the blue 2011-04-24T00:09:39 j3camero: i'm going to morph our two explanations 2011-04-24T00:09:41 computerwiz_222: give me the commit with the white, i'll send a link 2011-04-24T00:09:47 um.. 2011-04-24T00:09:51 antimatroid: yeah for sure. I was just drafting. 2011-04-24T00:09:59 amstan: little help 2011-04-24T00:10:02 computerwiz_222: git log 2011-04-24T00:10:17 git checkout commitnumber to see it 2011-04-24T00:10:21 git checkout master when done 2011-04-24T00:11:52 computerwiz_222: can the phrase "Planet Wars" easily be put in place of "Ants" and still look good? I'm think of if the next contest has a longer name. 2011-04-24T00:12:03 http://home.hypertriangle.com/~andrew/aicontest/ 2011-04-24T00:12:06 antimatroid: in general I think we have to be careful to distinguish between "sales" materials and "reference" materials on the website. Most of the page hits will be on the sales materials. With the sales sections, the objective is purely to drive the traffic and convert them to the next stage. It's even okay to lie a little, like reassuring people that it will be really easy. 2011-04-24T00:12:31 McLeopold: I'm sure that something can be arranged. Possibly reduce the size of the font, have 2 lines and maybe offset them. 2011-04-24T00:12:42 McLeopold: It's not a deal breaker, that's for sure. 2011-04-24T00:12:51 Okay, here comes some nitpicking... 2011-04-24T00:13:23 Can you make the "Sign Up" larger than the "Sign In"? Regestered users will already know where to look. 2011-04-24T00:13:42 we could.. it might look funny though 2011-04-24T00:13:48 it will yes 2011-04-24T00:13:49 I think brown would be a better background 2011-04-24T00:13:53 Medium Text: Small Text | Large Text 2011-04-24T00:14:12 j3camero: i like the part where you explicitly add "will" to "you will" 2011-04-24T00:14:15 :p 2011-04-24T00:15:54 antimatroid: totally. That first paragraph has to reassure people that they really can do it. People have to be able to see themselves creating AI. Basically, you've got ten seconds to convince a person that they really can make AI, even though they are sure they cannot. 2011-04-24T00:16:14 antimatroid: j3camero: when you guys are done tell me when, i can update the test server with the text 2011-04-24T00:16:48 amstan: would you consider an upvote/downvote mod for the forums? 2011-04-24T00:16:55 "The Google AI Challenge is all about creating artificial intelligence, whether you are a complete beginner or an expert. Using one of the easy-to-use starter kits, you will create a computer program in a language of your choice that controls a simulated colony of ants which fights among other colonies for domination." 2011-04-24T00:17:14 McLeopold: if you can find me a mod that does it, and that doesn't take ages to install(manually editing files), then sure 2011-04-24T00:17:20 i really want to point out the fact that they're just having to read data from input, go to town deciding which moves to make, then output those moves 2011-04-24T00:17:23 McLeopold: but yes.. i agree, the reddit way is the best 2011-04-24T00:17:42 can we use reddit for our forums? 2011-04-24T00:17:48 it's all open source yeah? 2011-04-24T00:17:51 yes 2011-04-24T00:17:53 is it easily used for other forums 2011-04-24T00:17:55 that would be cool 2011-04-24T00:18:08 anyone know french? is this spam? http://ai-contest.com/forum/viewtopic.php?f=5&t=1371 2011-04-24T00:18:18 or we could just make an aichallenge subreddit? :P 2011-04-24T00:18:35 antimatroid: and post what in it? 2011-04-24T00:18:44 lol wtf yes its spam 2011-04-24T00:18:45 amstan: he lives in vancouver with a few kids 2011-04-24T00:18:50 our site is cool apparently 2011-04-24T00:18:51 i guess forums really want sections 2011-04-24T00:18:56 amstan: that is spam 2011-04-24T00:18:56 which is what subreddits are 2011-04-24T00:18:58 computerwiz_222: orly? 2011-04-24T00:19:03 computerwiz_222: it's spam lol 2011-04-24T00:19:03 i wish they would implement a tree structure for subreddits 2011-04-24T00:19:07 amstan: * 2011-04-24T00:19:31 ok, now it's gone 2011-04-24T00:19:48 j3camero: do you think we can attempt i18n? 2011-04-24T00:20:05 McLeopold: when we get django yes 2011-04-24T00:20:13 McLeopold: till then, god no, not with php 2011-04-24T00:20:13 McLeopold: I think we could yes, but only by using Django. 2011-04-24T00:20:36 we would need a team of translators 2011-04-24T00:21:02 We could use Django this time around, if we have somebody who already is pretty decent at it. Barring that, I think we should kick the i18n and l10n stuff one more contest down the road. 2011-04-24T00:21:03 I doesn't hurt to start now, and maybe make one page language specific, just as a test to see if people are willing 2011-04-24T00:21:42 maybe the 5-minute-starter could be special? 2011-04-24T00:21:50 and help with conversion rates 2011-04-24T00:22:26 We could offer translations of the most popular pages on a case by case basis. That would be really easy. 2011-04-24T00:22:35 As for full i18n, I think probably not. 2011-04-24T00:22:52 Let's try with the 5-min-starter only, just to see if we can get commited translators 2011-04-24T00:23:21 * amstan doesn't want to see the urls for those pages 2011-04-24T00:23:36 use HTTP_ACCEPT_LANGUAGE 2011-04-24T00:23:55 h'okay people, criticise this 2011-04-24T00:23:56 "The Google AI Challenge is all about creating artificial intelligence, whether you are a complete beginner or an expert. Using one of the easy-to-use starter kits, you will create a computer program in a language of your choice that controls a simulated colony of ants which fights among other colonies for domination. 2011-04-24T00:23:57 It only takes 5 minutes to submit one of the starter kits to the website and watch your ant colony fight for domination against colonies created by other people from around the world. From there check out the tutorials on how to locally run your bot and begin programming!" 2011-04-24T00:24:06 How about this: created a subdirectory fr/ in the website. In there we will do our best to expand a french translation of the most high traffic pages. Anything that isn't translated yet just gets linked out to the english version. 2011-04-24T00:24:23 http://ietherpad.com/dmVS6JTTSp 2011-04-24T00:25:10 there we go 2011-04-24T00:25:19 start hacking at it for me :P 2011-04-24T00:25:29 I say french because I believe we have several french speakers inn here, including myself. And I am guessing computerwiz_222 as well, just based on his name (Rossignol) 2011-04-24T00:25:59 stocha is from france 2011-04-24T00:26:18 But we can also start expanding other translations as well. Particularly, we want to do Polish, Russian, Chinese, and Spanish. Google Anlytics will tell us which languages are most important. 2011-04-24T00:26:42 j3camero: maintaining translation strings is best practice, not separate sets of web pages 2011-04-24T00:26:56 lol, i was going to say google should let websites provide their own versions translted 2011-04-24T00:27:01 then realised how bad that would be 2011-04-24T00:27:11 j3camero: i'm not a french speaker, but i can read/understand enough to get by 2011-04-24T00:27:32 McLeopold: yes, for sure. I am just trying to bootstrap what we have now, avoiding building any machinery to deal with translation strings and _() and that sort of thing. 2011-04-24T00:28:28 A partial french translation is something we can build out in a few minutes or hours with no programming required. 2011-04-24T00:28:44 i like those cute little flags 2011-04-24T00:28:51 we can use those on our site to select language 2011-04-24T00:28:51 j3camero: I'm willing to put minimal hooks in 1 php page. 2011-04-24T00:28:54 i want to know where asian programmers hang out 2011-04-24T00:29:00 asia? 2011-04-24T00:29:04 lol 2011-04-24T00:29:05 online 2011-04-24T00:29:33 j3camero: join the fun http://ietherpad.com/dmVS6JTTSp 2011-04-24T00:29:37 there are meant to be lots of really smart asians (meh screw pc, it's the stereotype) yet not so many for these contests, i want to attract them 2011-04-24T00:29:46 Yeah I really really want to tap China and India. I know there are huge communities out there, but we haven't broken into them so far. Russia, yes. Poland, yes. Germany, yes. But zilch from China and India. 2011-04-24T00:29:53 and korea 2011-04-24T00:30:07 and japan (although they might be a bit busy atm) 2011-04-24T00:30:17 i will give the subreddits a go when we launch 2011-04-24T00:30:19 we had a south korean fight me for the top spot last year 2011-04-24T00:30:23 but other than that i don't really know where to look 2011-04-24T00:30:35 yeah, but that was a single person 2011-04-24T00:30:37 i remember them 2011-04-24T00:30:43 kcm1700 2011-04-24T00:30:44 we had oldman from china too 2011-04-24T00:30:49 we should rent a van and go cross country : 2011-04-24T00:30:55 didn't know he was from china 2011-04-24T00:30:58 :P 2011-04-24T00:31:01 good luck picking me up 2011-04-24T00:31:06 It seems we have only broken into countries whhere there are a decent number of people who also speak english. That's not surprising, since we have been english-only up until now. 2011-04-24T00:31:11 aichallenge: Alexandru Stan epsilon * raef56d7 / website/css/typo.css : fixed header position - http://bit.ly/h40U7s 2011-04-24T00:31:19 j3camero: we've tapped reddits market :P 2011-04-24T00:31:26 Pretty much yeah. 2011-04-24T00:31:57 lol we are driving across australia in a van, showing the contest to ALL THE PEOPLE. *cricket* *cricket* 2011-04-24T00:32:00 what about contacting univesities directly and sending them promotional stuff like posters 2011-04-24T00:32:04 Are there crickets in the desert? 2011-04-24T00:32:07 like.. csc's at other schools 2011-04-24T00:32:21 j3camero; i'm on the island off the bottom, you'll need to get another boat ride from the mainland :P 2011-04-24T00:32:23 computerwiz_222: problem is getting people to do it 2011-04-24T00:32:30 computerwiz_222: we did this in the past 2011-04-24T00:32:59 i know.. i mentioned it at McMaster last year and a couple friends of mine made accounts 2011-04-24T00:33:06 they didn't do anything.. but they still found out about it 2011-04-24T00:33:12 computerwiz_222: I did a lot of that during Tron. Google as a whole was pumping it hard at many campuses. We got some limited success, but nothing ultra-successful. We got like a dozen here and five there. Nothing like the way it catches fire in internet communities (forums, reddit, etc). 2011-04-24T00:33:36 j3camero: i half-expected that, worth a shot i guess 2011-04-24T00:33:50 j3camero: i was in #proggit a few times, the mere mention of the contest spurred a few people's interest 2011-04-24T00:34:07 I really really did it too. I had a team of people liasing with reps at some 200 universities around the world. 2011-04-24T00:34:13 j3camero: if our sponsors will pay for some nice magnets i'll put them on my car 2011-04-24T00:34:23 2010 Kia Forte, not making the contest look too bad :P 2011-04-24T00:34:38 Awww yeah 2011-04-24T00:35:09 antimatroid: do we have an "using the tools" wiki page? 2011-04-24T00:35:50 amstan: no 2011-04-24T00:35:54 Okay, it is time for me to sleep. 2011-04-24T00:36:00 i have been waiting for people to decide on how we are to use them :P 2011-04-24T00:36:04 antimatroid: it still says planet wars in a few places on it 2011-04-24T00:36:11 j3camero: Same for me. 2011-04-24T00:36:12 oh, ok, lol 2011-04-24T00:36:14 I am pumped about this website design. We are going to look less sketchy than we ever have before. 2011-04-24T00:36:19 Night everyone 2011-04-24T00:36:27 computerwiz_222: cya 2011-04-24T00:36:29 the last design wasn't sketchy 2011-04-24T00:36:30 j3camero: awesome 2011-04-24T00:36:40 and it's 2 36pm you fools 2011-04-24T00:36:48 antimatroid: first thing i saw.. those vertical white lines on the header 2011-04-24T00:37:00 antimatroid: except we're in not in australia 2011-04-24T00:37:06 antimatroid: amstan showed me that removal of those lines actually makes it worse 2011-04-24T00:37:36 computerwiz_222 haha yes. I remember fiddling with those damn lines before the last launch. Grr. 2011-04-24T00:38:01 j3camero: Bye for now, I'll be online tomorrow more than likely 2011-04-24T00:38:08 *** computerwiz_222 has quit IRC (Remote host closed the connection) 2011-04-24T00:38:09 j3camero: what did amreidle do? 2011-04-24T00:38:32 amstan: another problem :P 2011-04-24T00:38:46 the example map etc. fonts aren't whatever gives constant width for characters 2011-04-24T00:38:49 which is a problem 2011-04-24T00:39:06 antimatroid: yes, i should fix that 2011-04-24T00:39:10 antimatroid: make an issue 2011-04-24T00:39:19 i wouldn't have a clue how :P 2011-04-24T00:39:39 antimatroid: https://github.com/aichallenge/aichallenge/issues 2011-04-24T00:39:44 antimatroid: press New Issue 2011-04-24T00:40:00 antimatroid: put a nice description, then assign it to me 2011-04-24T00:41:12 done 2011-04-24T00:41:59 antimatroid: only this page? http://hypertriangle.com:13080/specification.php 2011-04-24T00:42:36 amstan: the old one was done in a wiki yeah? 2011-04-24T00:42:37 http://ai-contest.com/specification.php 2011-04-24T00:42:41 they managed to get that out 2011-04-24T00:42:57 including by the looks of it the latex stuff 2011-04-24T00:43:05 idk, didn't you make that? 2011-04-24T00:43:23 not for planet wars 2011-04-24T00:43:34 janzert: was that you? 2011-04-24T00:43:45 hmm? 2011-04-24T00:44:02 oh, the old planetwars spec? 2011-04-24T00:44:12 I think that was written by jmcarthur 2011-04-24T00:44:50 jmcarthur: ping me sometime :P 2011-04-24T00:45:34 antimatroid: you might want to use later tell 2011-04-24T00:45:54 meh :P 2011-04-24T00:45:55 antimatroid: also.. it was on the forums, that's where he first posted it 2011-04-24T00:46:43 amstan: didn't he actually write in the google code wiki though? 2011-04-24T00:46:47 I'm not really sure 2011-04-24T00:46:54 maybe 2011-04-24T00:47:18 the wiki from google code was transferred: https://github.com/aichallenge/aichallenge/wiki/Planet-Wars-Game-Specification 2011-04-24T00:48:13 antimatroid: check the svn logs 2011-04-24T00:48:31 antimatroid: it seems like the one who transferred it over added the image 2011-04-24T00:48:46 antimatroid: git log*, it should be in the history 2011-04-24T00:49:15 from the looks of that page, maybe i should just go through and try a similar format 2011-04-24T00:49:43 how quickly can you update your local copy to what is saved on the wiki? 2011-04-24T00:49:53 antimatroid: https://github.com/aichallenge/aichallenge/commits/pw/planet_wars/www/specification.php 2011-04-24T00:50:01 antimatroid: i just git pull, then ./setup.py 2011-04-24T00:51:31 antimatroid: https://github.com/aichallenge/aichallenge/commit/d22dc9b6f7eddd7f6e7d12daa01e63e68151ab69 2011-04-24T00:51:32 actually, the pages aren't exactly the same 2011-04-24T00:52:38 antimatroid: pong 2011-04-24T00:52:51 antimatroid: i updated the test server with the new wiki contents 2011-04-24T00:52:51 i don't think we need you now sorry :P 2011-04-24T00:52:58 you didn't do the specs page for planet wars did you? 2011-04-24T00:53:04 yes i did 2011-04-24T00:53:14 i wrote it using markdown and pandoc 2011-04-24T00:53:25 pandoc? 2011-04-24T00:53:25 oh, so it wasn't just converted from a wiki? 2011-04-24T00:54:37 amstan: I have a challenge for you :P embed frontiers visualiser onto that main page 2011-04-24T00:54:46 antimatroid: yes, i have to work on that too 2011-04-24T00:55:41 https://github.com/aichallenge/aichallenge/issues/78 2011-04-24T01:03:56 antimatroid: no it wasn't just a conversion 2011-04-24T01:04:14 bugger :P 2011-04-24T01:04:19 antimatroid: i actually wrote the majority of both versions. i don't remember which was first, even 2011-04-24T01:04:29 *** needsch has quit IRC (Quit: Leaving.) 2011-04-24T01:04:37 antimatroid: if the wiki has a fairly common format then pandoc can probably read it 2011-04-24T01:04:48 markdown, textile, etc. 2011-04-24T01:05:03 it's using whatever github wikis use :P 2011-04-24T01:05:08 jmcarthur: what is this pandoc then? 2011-04-24T01:05:24 pandoc is markup sex 2011-04-24T01:05:38 all the markup languages getting together and having babies 2011-04-24T01:05:42 jmcarthur: so can it do githubish markdown + latex from it into html? 2011-04-24T01:05:49 probably 2011-04-24T01:06:08 how does github markdown differ from vanilla markdown? 2011-04-24T01:06:19 jmcarthur: i shall change this then to use pandoc: https://github.com/aichallenge/aichallenge/blob/epsilon/website/setup.py 2011-04-24T01:06:28 jmcarthur: a few things, i'm not sure offhand 2011-04-24T01:06:30 latex i guess 2011-04-24T01:06:41 pandoc supports latex in markdown 2011-04-24T01:07:13 it can even import and export latex :D 2011-04-24T01:08:47 *** needsch has joined #aichallenge 2011-04-24T01:13:41 *** sids_aquarius has quit IRC (Ping timeout: 276 seconds) 2011-04-24T01:25:49 *** sids_aquarius has joined #aichallenge 2011-04-24T01:35:38 amstan/janzert: are you familiar with the php "guard" code? $JPC_CONTEST_SESSION_PHP__ 2011-04-24T01:35:49 no 2011-04-24T01:35:59 nope 2011-04-24T01:36:26 I think it's a poor man's require_once 2011-04-24T01:36:29 at least doesn't ring a bell at all 2011-04-24T01:38:04 hmm, yeah looks like someone familiar with c headers wrote it 2011-04-24T01:38:56 *** amstan_ has joined #aichallenge 2011-04-24T01:38:56 *** ChanServ sets mode: +o amstan_ 2011-04-24T01:38:56 *** amstan has quit IRC (Read error: Connection reset by peer) 2011-04-24T01:39:11 *** amstan_ has quit IRC (Remote host closed the connection) 2011-04-24T01:41:47 I also see session.php is currently included with "include 'session.php';" everywhere, which also seems like an error since a missing include will only cause a warning not an error 2011-04-24T01:42:16 and require, better require_once, would be what we really want 2011-04-24T01:47:56 janzert: all the includes on the php site are ridiculous. it all gets linked in through spagetti code 2011-04-24T01:48:37 header -> session -> mysql_login -> server_info 2011-04-24T01:50:00 I think at some point you'd want a web request to only connect to the sql server if it needs info from it, but that may be every request? 2011-04-24T01:50:13 I think so 2011-04-24T01:50:53 doesn't the session information get pulled from the db? 2011-04-24T01:51:52 hmm, actually maybe it doesn't 2011-04-24T01:52:28 it looks like the only part of session.php that touches the database is to activate a user 2011-04-24T01:52:35 I'm looking at that now. I should at least unhook the header.php from connecting to the database unless needed 2011-04-24T01:53:08 the header does need the session though 2011-04-24T01:53:20 session doesn't need mysql 2011-04-24T01:53:24 right 2011-04-24T01:54:11 at least if you refactor activate_user out of it 2011-04-24T01:54:30 which it doesn't seem to really belong there anyway 2011-04-24T01:54:34 I don't see an "active_user"? 2011-04-24T01:54:43 activate 2011-04-24T01:54:52 last function in the file 2011-04-24T01:55:03 grant_permission_to_user 2011-04-24T01:55:32 oh, I already factored that out, but didn't commit yet :) 2011-04-24T01:55:35 hmm, at line 79 for me 2011-04-24T01:55:37 ahh 2011-04-24T01:56:30 I hope you're getting rid of all the /deprecated/ functions the just return false too? 2011-04-24T01:56:46 it might be dangerous to put an include before the yes, I am 2011-04-24T01:57:38 ug, and header needs footer or you have open div tags :( 2011-04-24T01:58:05 yes, fairly unavoidable with the way it is written 2011-04-24T01:58:30 okay, stop me from doing too much, I've only started this to fix uploads :/ 2011-04-24T01:58:51 *** ahmedhelal has quit IRC (Read error: Connection reset by peer) 2011-04-24T01:59:34 *** ahmedhelal has joined #aichallenge 2011-04-24T01:59:35 yeah, I wouldn't go too much further 2011-04-24T02:00:14 or you're looking at a whole rewrite of the way the html is produced, (e.g. using a really template system such as smarty) 2011-04-24T02:00:26 s/really/real 2011-04-24T02:17:28 janzert: do you know php? 2011-04-24T02:17:45 not really 2011-04-24T02:18:00 hmm, do you think you could debug a scoping issue? 2011-04-24T02:18:19 I can try and help at least ;) 2011-04-24T02:18:36 okay, I've got check_submit.php, which includes submission.php 2011-04-24T02:19:04 I'm using the function create_new_submission_for_current_user, which calls an sql insert statement. 2011-04-24T02:19:14 I've put the sql statement into sql.php 2011-04-24T02:19:35 mysql_query(sprintf($sql["insert_new_submission"],current_user_id())) 2011-04-24T02:19:51 $sql is an associative array 2011-04-24T02:20:20 I get a "Query was empty" error when I use the $sql variable, but if I just put the text in there, it works? 2011-04-24T02:21:14 If, instead of calling cureate_new_submission_for_current_user, I just call the mysql_query directly from check_submit, the $sql variable works? 2011-04-24T02:21:35 my first debug step with php would be to stick an 'echo sprintf($sql["insert_new_submission"]...' line just before the call 2011-04-24T02:22:13 McLeopold: it is slightly harder to give people a vector of food information now :P 2011-04-24T02:22:29 * antimatroid is updating his starter bot 2011-04-24T02:22:31 submission.php is including sql.php in some way? (require, etc.) 2011-04-24T02:23:12 check_submit -> mysql_login -> sql 2011-04-24T02:23:26 I've added directly to submission just for kicks 2011-04-24T02:29:48 slightly different topic, but I wonder if it wouldn't be a good idea to add contest_query function in mysql_login that takes a query name and the associated arguments and does the mysql_query(sprintf($sql[$query_name], arguments... call 2011-04-24T02:34:09 if we are doing that, then we could switch to prepared statements easily 2011-04-24T02:34:22 also, my problem was I can't use a global in a function 2011-04-24T02:34:48 I'm so use to scope chains 2011-04-24T02:34:52 ahh 2011-04-24T02:35:11 actually, that one gets me in python all the time 2011-04-24T02:35:15 you need "global $sql" 2011-04-24T02:35:30 yeah, so I might as well create the contest_query function 2011-04-24T02:37:34 something like http://pastebin.com/QzgeeGuR 2011-04-24T02:37:40 should do it 2011-04-24T02:38:01 I can just include the sql.php right in that function 2011-04-24T02:38:05 although completely untested and in particular unsure if it works for a query without additional parameters 2011-04-24T02:38:30 yeah, if it's not needed anywhere else 2011-04-24T02:43:59 janzert: http://pastebin.com/CisU4kvW 2011-04-24T02:44:04 aichallenge: Nick Ham epsilon * r34a936c / (3 files): updated the c++ starter bot - http://bit.ly/eIZhtf 2011-04-24T02:44:21 i think i might know what my problem with git ssh keys is 2011-04-24T02:44:31 did you have a problem? 2011-04-24T02:44:38 would having a dynamic ip and resetting it cause me issues? 2011-04-24T02:44:44 it shouldn't 2011-04-24T02:45:11 if you are on ubuntu, you should have a ".ssh" dir, and that should be good enough to authenticate 2011-04-24T02:45:15 i used to find i always needed to add my email address to the end of my key on the website whenever i wanted to do something 2011-04-24T02:45:29 but recently i got a new router and haven't been needing to reset it as much, anyway, something has changed 2011-04-24T02:45:36 yeah, my public part ends in my email address 2011-04-24T02:46:04 yeah, on the website it strips it, but i used to not be able to connect after x amount of time unless i manually readded that to the end 2011-04-24T02:46:09 but it just worked without me having to do that 2011-04-24T02:46:30 anyway, the c++ bot keeps track of food information now, and only gives people information about whether a square is visible or not 2011-04-24T02:46:59 why did you say food is harder now? 2011-04-24T02:48:41 it's hard to have say a vector food; in state 2011-04-24T02:48:53 well, unless you do a full scan of the map to find current foo 2011-04-24T02:49:11 if you just add locations with an 'f' line, you will add some food more than once 2011-04-24T02:49:31 the best bet is to use a set 2011-04-24T02:49:59 oh, yeah, I looked at that in my code and decided to solve it later :) 2011-04-24T02:50:10 i just got rid of the food and enemies vector :P 2011-04-24T02:50:34 let them deal with it 2011-04-24T02:50:52 I was thinking of creating a second food list in python, then for every 'f' line, scan the old list to determine known food and potential food 2011-04-24T02:51:20 let them do that 2011-04-24T02:51:35 we can cover these kinds of options in the beginners tutorial 2011-04-24T02:51:45 so they have quite a few simple things to play around with to get familira 2011-04-24T02:51:47 Well, I still think it's okay for the starter bots to handle the state updates 2011-04-24T02:52:35 The tutorial should cover things like, "attempt to return to seen food", not "rewrite the update function to remember seen food" 2011-04-24T02:53:09 The tutorial should cover possible strategies and tactics for the game, not programming techniques. 2011-04-24T02:53:33 but using the information differently is different strategy/tactics 2011-04-24T02:54:16 I'm making the information available. The tutorial can discuss using it. Others can re-write the implementation for speed or other reasons. 2011-04-24T02:54:21 if you start people with the stupider idea while making it just as easy to switch, then you can make it easy for people to see their bot beat the starter bot, and from there they're hooked 2011-04-24T02:54:33 but yeah, i really don't mind 2011-04-24T02:54:43 just voicing my opinion on starter bots 2011-04-24T02:55:10 well, you're in charge of them, but I really think we should include this 2011-04-24T02:56:21 include what? :P 2011-04-24T02:56:25 just to be clear 2011-04-24T02:57:05 Include updating food information to know the difference between visable food and non-visable potential food. 2011-04-24T02:57:20 *visible* that is 2011-04-24T02:58:08 So, included remembered food in the food list, but have a flag or maintain 2 lists 2011-04-24T03:07:24 *** needsch has quit IRC (Quit: Leaving.) 2011-04-24T03:09:59 but do they need a list? 2011-04-24T03:10:20 well, the starter bot writer can pick the data structure 2011-04-24T03:10:22 can't it be enough that they can check a location in the grid as to whether it's food and/or visible 2011-04-24T03:10:39 sure 2011-04-24T03:19:37 *** stocha has joined #aichallenge 2011-04-24T03:19:58 antimatroid: i think we said we always assume that unseen food is not there anymore. 2011-04-24T03:20:16 stocha: when? 2011-04-24T03:20:20 antimatroid: and that the R stuff wasn't desirable. Although it was not very annoying either 2011-04-24T03:20:34 it was desirable for some and doesn't affect others 2011-04-24T03:20:34 antimatroid: by we, i mean the guy you asked and me :p 2011-04-24T03:20:39 giving net deirability :P 2011-04-24T03:20:43 it's desireconomics 2011-04-24T03:20:54 antimatroid: now you see that it raises questions. Wich is "bad" :p 2011-04-24T03:21:09 Also i think there is way for implementing the game of life without indexes 2011-04-24T03:21:31 although i wouldn't have been able too when i was 14 :p 2011-04-24T03:21:31 lots of variables :P 2011-04-24T03:21:59 and without per say arrays 2011-04-24T03:22:54 Because you can use bit-operiations for the game of life. Where indices won't matter anymore. 2011-04-24T03:23:23 especially if it's 15x15 in size :p 2011-04-24T03:23:46 i don't think that was what they meant when they didn't use arrays 2011-04-24T03:24:22 i don't know, i did'nt follow. 2011-04-24T03:24:32 i just wanted to point out it wasn't impossible. 2011-04-24T03:25:20 you can do 15x15 with 15 (16bit) registers. 2011-04-24T03:25:45 multiplayer game of life would have been cool 2011-04-24T03:26:01 you choose your starting position ? 2011-04-24T03:26:06 yep 2011-04-24T03:26:12 like everyone in turn put up a piece 2011-04-24T03:26:13 and you can move your squares each turn 2011-04-24T03:26:16 and there's no draws 2011-04-24T03:26:31 *** McLeopold has left #aichallenge 2011-04-24T03:26:39 you can move like in ant ? 2011-04-24T03:26:44 yeah 2011-04-24T03:26:49 well that's one option 2011-04-24T03:26:55 so the life rule apply with everyone being same color ? 2011-04-24T03:27:05 sort of 2011-04-24T03:27:08 you can still use it for combat it ant. 2011-04-24T03:27:16 and breeding. 2011-04-24T03:27:21 for a new square to spawn, it needs 3 enemies exactly, so you can't draw for majority to claim spawned square 2011-04-24T03:27:33 well, 3 adjacent ants 2011-04-24T03:27:56 if you can move your ant, you will reproduce too fast. 2011-04-24T03:28:02 unless you also need at least a food. 2011-04-24T03:28:08 we fleshed out a version for fun a few months ago, but decided it wouldn't make for a great ai game 2011-04-24T03:28:12 and the food go when the spawning is done. 2011-04-24T03:28:26 i think ant is better. 2011-04-24T03:28:32 the game has a lot of options. 2011-04-24T03:28:38 i still like my ants game with the home bases more 2011-04-24T03:28:45 although some would need bit investment into visualizer. 2011-04-24T03:29:09 and also the game would be better with more time. And maybe the fog part off. 2011-04-24T03:29:14 for an ai challenge. 2011-04-24T03:29:16 fog is good 2011-04-24T03:29:23 like 10 minutes per move, with no fog at all. 2011-04-24T03:29:34 10 minutes per move? 2011-04-24T03:29:34 and maybe 1vs1 also. 2011-04-24T03:29:41 the contest can run for a couple of years too? 2011-04-24T03:29:45 with a probability of food appearing at each square 2011-04-24T03:29:52 so we have all informations. 2011-04-24T03:30:10 hey you have google for a sponsor. 2011-04-24T03:30:24 it could run in 10 minutes :p 2011-04-24T03:31:09 another option is to use the computing power of the participating and willing to lend computing power. 2011-04-24T03:31:28 although you have to have games with seed. So you can submit two calculation for each match. and verify they match. 2011-04-24T03:31:31 we've considered and rejected that option 2011-04-24T03:31:40 we want all bots in the final tournament to have the same resources 2011-04-24T03:32:07 Yeah you have to get rid of the "time" measurement. and find more clever metric. 2011-04-24T03:32:30 like everything is done in a virtual machine, and you virtual clock every instructions. 2011-04-24T03:32:51 i don't say it's simple. i just state it feasible from a theoritic point of view. 2011-04-24T03:33:11 it's not feasible practically with the amount of people offering time 2011-04-24T03:33:35 computing time. Or computer engineering time ? 2011-04-24T03:33:53 how could that effect computing time? 2011-04-24T03:34:17 you would have to start a gnu project of ai competition distributed framework. With luck you could have it ready in 3 years. 2011-04-24T03:34:43 i don't get your question. 2011-04-24T03:35:13 i have to go anyway. :p 2011-04-24T03:35:53 *** stocha has quit IRC (Quit: Page closed) 2011-04-24T03:39:21 *** berak has joined #aichallenge 2011-04-24T03:45:40 *** needsch has joined #aichallenge 2011-04-24T04:09:13 aichallenge: McLeopold epsilon * rbe196c1 / ants/dist/starter_bots/python/ants.py : added parsing of 'r' lines (in progress) - http://bit.ly/hduUXc 2011-04-24T04:09:14 aichallenge: McLeopold epsilon * r068dd8f / (31 files): update php includes/requires, added contest_query, fixed submission upload - http://bit.ly/g9R04i 2011-04-24T04:12:48 *** _0x47 <_0x47!~slash@p3EE0BEF1.dip.t-dialin.net> has joined #aichallenge 2011-04-24T04:14:29 *** Sunhay has quit IRC (Ping timeout: 240 seconds) 2011-04-24T04:15:16 *** Sunhay has joined #aichallenge 2011-04-24T04:15:59 *** sids_aquarius has quit IRC (Ping timeout: 250 seconds) 2011-04-24T04:28:53 *** sids_aquarius has joined #aichallenge 2011-04-24T04:31:19 *** boegel has joined #aichallenge 2011-04-24T04:36:53 *** needsch has quit IRC (Quit: Leaving.) 2011-04-24T04:40:03 how convenient 2011-04-24T04:40:15 my other bot is already using a set for food information 2011-04-24T04:50:51 *** mceier has joined #aichallenge 2011-04-24T04:54:23 A GNU project of ai competition distributed framework would surely increase the reputation of the Waterloo university ;) 2011-04-24T05:11:47 *** boegel has quit IRC (Quit: This computer has gone to sleep) 2011-04-24T05:13:27 *** smellyhippy has quit IRC (Quit: Take your stinking clause off me you damn dirty predicate!) 2011-04-24T05:16:35 Frontier: we're no longer associated with waterloo? 2011-04-24T05:16:57 well, not formally 2011-04-24T05:18:15 *** smellyhippy has joined #aichallenge 2011-04-24T05:19:03 *** Naktibalda has joined #aichallenge 2011-04-24T05:29:22 *** needsch has joined #aichallenge 2011-04-24T05:30:25 So this is really a project managed by Jeff, Alex and some other students? 2011-04-24T05:31:13 there may be more non-students than students at this point 2011-04-24T05:31:15 Now that I think about it, it has been like that for a long time now. I just always felt like there is some big organization behind it 2011-04-24T05:31:41 Janzert, sorry that I ignored you :p 2011-04-24T05:32:07 heh, well I was more thinking of sigh, Mcleopold and yourself 2011-04-24T05:32:45 awww... janzert is thinking of me 2011-04-24T05:32:54 It is good that we live in one of those western industry countries where ppl have too much free time, hehe 2011-04-24T05:40:42 <_0x47> hi, is there an already implemented way of running a game without sandboxing? because my bot crashes for no reason and i can't debug with the sandbox. (currently using playgame.py) 2011-04-24T05:41:14 <_0x47> at least it says "bot 0 died" at an totally impossible point 2011-04-24T05:41:39 the current sandbox is a fake sandbox, i.e. it isn't doing any sandboxing 2011-04-24T05:42:03 <_0x47> weird, but I still can't attach to it's process. :/ then I have to check again 2011-04-24T05:45:02 You can write out to a separate file for example or use stderr to get debug info on screen during a run if sigh implemented that already 2011-04-24T06:09:11 *** Redgis has joined #aichallenge 2011-04-24T06:19:38 *** needsch has quit IRC (Quit: Leaving.) 2011-04-24T06:36:08 *** kaemo has joined #aichallenge 2011-04-24T06:36:11 *** kaemo0 has joined #aichallenge 2011-04-24T06:44:35 <_0x47> Frontier: that doesn't work either. I modified the python files, so that all bot output is directly forwarded to stdout (for testing). i can't get any reason why it crashes. the starter bot from antimatroid uses a loop like 2011-04-24T06:44:40 <_0x47> while(std::cin >> state) 2011-04-24T06:44:41 <_0x47> { 2011-04-24T06:44:41 <_0x47> state.update_vision_information(); 2011-04-24T06:44:41 <_0x47> make_moves(); 2011-04-24T06:44:41 <_0x47> end_turn(); 2011-04-24T06:44:41 <_0x47> } 2011-04-24T06:45:16 <_0x47> now, 5 to 10 round everything works fine and then suddenly - nothing 2011-04-24T06:45:50 <_0x47> end_turn() is executed just fine (next round is initiated), but it doesn't even update the "state" any more 2011-04-24T06:45:57 <_0x47> just sais bot 0 died 2011-04-24T06:46:53 <_0x47> and even if I put a __builtin_trap(); in my code, i can't get the debugger to hit. somehow the python wrapping prevents it (i guess) 2011-04-24T06:51:25 <_0x47> Frontier: Maybe, if this was already implemented, I could at least get a backtrace from the python script to the bot's log file? 2011-04-24T06:53:39 *** delt0r___ has joined #aichallenge 2011-04-24T06:55:11 *** delt0r_ has quit IRC (Ping timeout: 240 seconds) 2011-04-24T06:57:34 *** FireFly has joined #aichallenge 2011-04-24T06:59:12 _0x47: what is it doing? 2011-04-24T07:03:21 <_0x47> antimatroid: it's not your fault. :) i just used your starter bot and after my changes it fails for no reason. 2011-04-24T07:03:34 i changed the starter bot today btw :P 2011-04-24T07:03:41 we added a new type of input 2011-04-24T07:03:53 <_0x47> type of input? 2011-04-24T07:04:21 you now get a line "r row col" if a food square no longer exists at a location that did contain a food square the last time you saw it 2011-04-24T07:04:38 so in clear in square, you can get rid of isFood = 0 and process the r line 2011-04-24T07:04:59 and you'll want to use a set for where food is 2011-04-24T07:05:29 *** skynetz has joined #aichallenge 2011-04-24T07:05:32 <_0x47> i see, but i didn't update my git yet, so I guess I'll try to fix my issues first 2011-04-24T07:05:33 *** Eruonen has joined #aichallenge 2011-04-24T07:05:42 <_0x47> but thanks for pointing that out 2011-04-24T07:05:44 <_0x47> :) 2011-04-24T07:16:14 the specs page has been updated 2011-04-24T07:16:17 i'll make a forum post 2011-04-24T07:18:47 *** sids_aquarius has quit IRC (Ping timeout: 240 seconds) 2011-04-24T07:35:58 *** Naktibalda has quit IRC (Remote host closed the connection) 2011-04-24T07:36:17 *** Naktibalda has joined #aichallenge 2011-04-24T07:37:42 <_0x47> lolz i found my error 2011-04-24T07:38:00 <_0x47> iterating through a vector and doing an erase() without break; afterwards 2011-04-24T07:38:09 * _0x47 is a giant douche 2011-04-24T07:38:49 <_0x47> and i figured how to attach to my process. first: use plain gdb, second: be root 2011-04-24T07:39:36 i've found it's mostly better to just use sets with this 2011-04-24T07:39:48 although i have vectors for new and dead ants on a square 2011-04-24T07:41:09 <_0x47> well, it's the food vecotr, as I need to somehow track my already addressed food, this seems pretty comfortable, as it's renewed every round. 2011-04-24T07:47:05 if you make use of the new information and stop wiping it 2011-04-24T07:47:13 using a vector like the starter package used to wont work 2011-04-24T07:47:15 use a set 2011-04-24T07:48:55 sigh: do you know what is different about vector? 2011-04-24T07:49:57 antimatroid: it packs them into bits 2011-04-24T07:50:10 why? 2011-04-24T07:50:30 friend and i were talking about the lameness of not being able to call matrix[location] 2011-04-24T07:51:12 you can't? 2011-04-24T07:51:52 no, because it needs to return a reference to a vector of bools 2011-04-24T07:52:00 "inline T& operator[](const Location &loc) const" 2011-04-24T07:52:55 and what happens when you try to do that? 2011-04-24T07:53:30 <_0x47> antimatroid: thanks, i'll check that out then 2011-04-24T07:53:30 give me a sec, i'll get the error 2011-04-24T07:54:59 sigh: http://pastebin.com/a5j1syc4 2011-04-24T07:55:13 works fine for a matrix of any type but bool 2011-04-24T07:56:02 make a template specialisation for book 2011-04-24T07:56:04 *bool 2011-04-24T08:01:56 how would you do that? :P 2011-04-24T08:02:03 "inline bool& operator[](const Location &Loc) const" "can't be overloaded 2011-04-24T08:03:03 why not Matrix::operator[] ? 2011-04-24T08:03:18 Matrix::operator[] 2011-04-24T08:03:19 hmm, i'll try 2011-04-24T08:04:14 that will probably still error, but show me the message 2011-04-24T08:06:00 i need to relocate them somehow 2011-04-24T08:06:09 i had them inside the struct in the .h file 2011-04-24T08:06:20 trying to work out how to do the operator outside the struct definition 2011-04-24T08:06:45 i'm not sure i can 2011-04-24T08:08:43 yeah no idea :\ 2011-04-24T08:09:31 sigh: http://pastebin.com/usdrfFHV :P 2011-04-24T08:10:57 try: inline vector::reference& operator[](const Location &loc) const { return elements[loc.row][loc.col]; } 2011-04-24T08:12:33 *** sids_aquarius has joined #aichallenge 2011-04-24T08:13:42 doesn't like the overload with a bit of changing 2011-04-24T08:14:45 actually... yeah, what I wrote won't make sense in this context... oh wel 2011-04-24T08:15:24 i might just have to live with matrix[loc.row][loc.col] for matrices of bools 2011-04-24T08:25:03 *** Stocha has joined #aichallenge 2011-04-24T08:29:09 *** MatmaRex has joined #aichallenge 2011-04-24T08:29:30 *** MatmaRex has joined #aichallenge 2011-04-24T08:30:30 deutch will be important as well 2011-04-24T08:30:39 frontier speaks that. 2011-04-24T08:33:11 can't you use facebook or i don't know world of warcraft to find one indian/asian geek. once you have one, he will most likely tell you where to go to advertise to more. 2011-04-24T08:33:42 there is probably some in america as well, with good connection to their home country. 2011-04-24T08:38:29 who are you talking to? I cannot find the context. 2011-04-24T08:39:44 hey guys, would somebody care to look at Ruby starter for Ants, or tell me what do I do with it? 2011-04-24T08:39:53 I posted it to forums a few days ago and received no reply 2011-04-24T08:39:54 http://ai-contest.com/forum/viewtopic.php?p=8978#p8978 2011-04-24T08:39:58 *** kaemo has quit IRC (Ping timeout: 240 seconds) 2011-04-24T08:40:10 *** kaemo0 has quit IRC (Ping timeout: 260 seconds) 2011-04-24T08:41:19 you do realize you posted it as an answer to a game idea thread? 2011-04-24T08:41:27 j3camero: was talking about getting some indian and asian into the contest. And also about translating some page of the website. 2011-04-24T08:41:35 i do realize, since that's where i was asked for it 2011-04-24T08:41:44 that's also why i ask once again where to post it 2011-04-24T08:42:11 i guess you should ask antimatroid or sigh matmarex 2011-04-24T08:43:21 normaly you would have to commit it using git. i don't know how the whole thing works though. 2011-04-24T08:45:28 Why do you want to used Rad2=4 rather than the old Rad2=5 (old civilization city radius shape) for the fighting radius2. I think 5 is more interesting. 2011-04-24T08:45:36 MatmaRex: Others created a git account, uploaded their starter bot there and made a "push request", so that the code gets merged into the main repository by someone with write access 2011-04-24T08:47:08 the warlight thread was probably not the best place to post it. 2011-04-24T08:48:11 *** Stocha has quit IRC (Quit: Page closed) 2011-04-24T08:50:09 Frontier: i've only ever used git to upload things to heroku (a Ruby website hosting) 2011-04-24T08:50:17 although i think i even have a github acc 2011-04-24T08:52:10 it is no problem, antimatroid is taking care of the starter bots and he can commit your code to the repository 2011-04-24T08:55:33 alright. so, what do i do? 2011-04-24T08:56:02 *** ahmedhelal has quit IRC (Read error: Connection reset by peer) 2011-04-24T08:56:21 can i just send somebody the files, or do you want me to fork, pull it to my PC, edit, commit, push it back to github, and then send a pull request? (that's how you do it, right?) 2011-04-24T08:56:29 *** ahmedhelal has joined #aichallenge 2011-04-24T09:09:48 *** kaemo has joined #aichallenge 2011-04-24T09:09:52 *** kaemo0 has joined #aichallenge 2011-04-24T09:22:22 MatmaRex: the proper way is to pull, make your changes... then ask other devs to pull your changes or submit a patch from that 2011-04-24T09:22:49 however I don't know what they have decided to do for this comp 2011-04-24T09:23:15 so basically, what i said above? 2011-04-24T09:23:19 you shouldn't generally push without "push permission" 2011-04-24T09:23:28 i mean 2011-04-24T09:23:32 push it to my fork 2011-04-24T09:23:36 oh right 2011-04-24T09:23:37 yes 2011-04-24T09:23:39 perfect 2011-04-24T09:23:43 its the git way 2011-04-24T09:23:54 I miss understood 2011-04-24T09:23:54 by the way, i just spent 20 minutes trying to pull that stuff to my PC 2011-04-24T09:24:05 that stuff is crazy. 2011-04-24T09:24:13 is it large? 2011-04-24T09:24:20 or your internets slow? 2011-04-24T09:24:26 or are you on windows? 2011-04-24T09:24:34 no, yes, yes 2011-04-24T09:24:38 but it first took me 10 minutes to set up the remote 2011-04-24T09:24:45 git kinda has not been very windows friendly.. 2011-04-24T09:24:56 then it took 5 minutes after i told it to "git pull Github" 2011-04-24T09:25:01 its blindingly fast on linux 2011-04-24T09:25:02 and broke saying i didnt specify branch 2011-04-24T09:25:12 it took me 5 more minutes trying to find out how to specify the damn branch 2011-04-24T09:25:25 sounds right... but then you should have some kind a fecth_head 2011-04-24T09:25:32 turns out i just had to do "git pull Github epsilon" 2011-04-24T09:25:38 FETCH_HEAD 2011-04-24T09:25:38 and it took like 5 seconds 2011-04-24T09:25:40 eh. 2011-04-24T09:25:52 but it seems to work now 2011-04-24T09:26:02 yea.. cus the first 5 min stuff whould have been cached 2011-04-24T09:26:03 now, 20 more minutes, and i should have it up on github... 2011-04-24T09:27:00 oh and sometimes github is slow for "ssh reason" IIRC 2011-04-24T09:27:06 reasons 2011-04-24T09:27:37 I run all my own servers for my git stuff.. and it really fast for the most part 2011-04-24T09:41:07 woohoo, works. i had to kill the whole folder and clone it again first, lol 2011-04-24T09:41:10 https://github.com/MatmaRex/aichallenge/commit/48f11e9dc2a9a106ae1bd338c987b19802daab10 2011-04-24T09:41:30 so, now i click "pull request" and I'm done? 2011-04-24T09:42:44 delt0r___: help? :) 2011-04-24T09:43:44 err 2011-04-24T09:43:57 don't know... i only ever use command line... pull where? 2011-04-24T09:44:18 dam i going off python in a big way latetly 2011-04-24T09:44:27 i can see a shiny button saying "Pull Request" here: https://github.com/MatmaRex/aichallenge 2011-04-24T09:44:35 will anything blow when I click it? 2011-04-24T09:44:56 MatmaRex: so when you pull ---you pull from somewhere to where you are... so in this context --i am unsure where "here" is 2011-04-24T09:45:13 *** sids_aquarius has quit IRC (Ping timeout: 260 seconds) 2011-04-24T09:45:20 *** skynetz has quit IRC (Quit: .) 2011-04-24T09:45:27 but its git... it can't blow... you can always just go back one commit 2011-04-24T09:45:33 *** siddharth has joined #aichallenge 2011-04-24T09:45:42 alright then, lets see what happens 2011-04-24T09:46:01 it will typically fail if something is going to get lost, and forces you to use "force" options 2011-04-24T09:48:05 looks like it worked. https://github.com/aichallenge/aichallenge/pulls 2011-04-24T09:48:13 worst case, somebody is going to be angry with me 2011-04-24T09:48:39 yea--that about right... and since you have permission... well 2011-04-24T09:49:11 MatmaRex: still i don't like githubs UI... its not clear whats getting pulled onto what 2011-04-24T09:50:02 i think i don't like git at all. ;) 2011-04-24T09:50:04 MatmaRex: looks fine 2011-04-24T09:50:40 git is fantastic... but the idea of distrubuted version control takes some getting use to 2011-04-24T09:50:55 *** siddharth has quit IRC (Ping timeout: 252 seconds) 2011-04-24T09:50:56 distributed version control i can live with 2011-04-24T09:51:06 it's the branches and merging and everything that kills me 2011-04-24T09:51:07 i am not sure github is the best place to have your first experience with it 2011-04-24T09:51:16 well thats the point 2011-04-24T09:51:29 heh 2011-04-24T09:51:29 without branching and merging its not distributed 2011-04-24T09:51:30 ;) 2011-04-24T09:51:42 and git does do it very well 2011-04-24T09:52:34 i'm going to just believe you. 2011-04-24T09:52:58 lol 2011-04-24T09:53:22 MatmaRex: what other version control system have you used? 2011-04-24T09:54:09 none. i used git before, but only the simple parts 2011-04-24T09:54:27 right.. well then not much to compare it too 2011-04-24T09:54:36 i mean, "git add", "git commit", "git push" was all i needed 2011-04-24T09:54:45 yeah 2011-04-24T09:55:11 well push is just a little evil if you don't know what it does 2011-04-24T09:55:27 you should only ever ever push to a bare repo... 2011-04-24T09:55:46 github has done this in the background, but its far from clear 2011-04-24T09:55:57 i was using it with heroku, so actually, that would be "git push heroku master" - i havent even wondered what it does, it just worked 2011-04-24T09:56:22 push can bork a repo thats not bare 2011-04-24T09:56:47 bare means there is nothing checked out... aka no local version 2011-04-24T09:56:59 i'm going to just believe you once again ;) 2011-04-24T09:58:38 *** kaemo0 has quit IRC (Ping timeout: 240 seconds) 2011-04-24T09:58:54 *** kaemo has quit IRC (Ping timeout: 260 seconds) 2011-04-24T10:03:40 *** amstan has joined #aichallenge 2011-04-24T10:03:40 *** ChanServ sets mode: +o amstan 2011-04-24T10:07:56 *** siddharth has joined #aichallenge 2011-04-24T10:10:17 aichallenge: MatmaRex epsilon * r48f11e9 / (2 files): Ruby starter bot. - http://bit.ly/dWXOv6 2011-04-24T10:13:08 whoa. 2011-04-24T10:13:10 :D 2011-04-24T10:13:10 *** amriedle has joined #aichallenge 2011-04-24T10:13:22 MatmaRex: thanks for that 2011-04-24T10:14:51 :) 2011-04-24T10:15:07 i just hope it actually works 2011-04-24T10:15:21 but it didnt seem to crash, so it should be okay 2011-04-24T10:16:20 MatmaRex: you might want to stick around to ensure that, we don't want to find out when we launch that it crashes 2011-04-24T10:17:04 it worked for me 2011-04-24T10:17:19 although i didnt test it on ruby 1.8 2011-04-24T10:18:44 MatmaRex: yeah.. we're going to be using that 2011-04-24T10:19:04 ruby 1.8.7 (2010-08-16 patchlevel 302) [i686-linux] 2011-04-24T10:19:15 hm. why not 1.9? 2011-04-24T10:19:25 that's what's in ubuntu 2011-04-24T10:19:40 isnt 1.9 available? 2011-04-24T10:19:56 idk about the main repo, might be in a ppa 2011-04-24T10:20:19 see https://github.com/aichallenge/aichallenge/issues/3 2011-04-24T10:22:08 *** 16WAAB8IE <16WAAB8IE!~mad5ci@aeog138.neoplus.adsl.tpnet.pl> has joined #aichallenge 2011-04-24T10:22:08 *** kaemo has joined #aichallenge 2011-04-24T10:22:59 i rarely use linux, but 1.8 and 1.9 should both be available as different packages 2011-04-24T10:23:12 since there are some incompatibilities between them 2011-04-24T10:23:20 oh yes.. you're right 2011-04-24T10:23:22 like between python 2.x and 3.0 2011-04-24T10:23:25 there's a way to install 1.9.1 2011-04-24T10:23:40 anyway.. 2011-04-24T10:23:41 g2g 2011-04-24T10:28:30 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-04-24T11:03:27 *** sigh has quit IRC (Ping timeout: 248 seconds) 2011-04-24T11:07:08 *** needsch has joined #aichallenge 2011-04-24T11:09:43 *** berak has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2011-04-24T11:17:57 *** computerwiz_222 has joined #aichallenge 2011-04-24T11:23:37 *** boegel has joined #aichallenge 2011-04-24T11:51:06 *** choas has joined #aichallenge 2011-04-24T12:03:35 *** boegel has quit IRC (Quit: Leaving) 2011-04-24T12:34:26 *** siddharth has quit IRC (Ping timeout: 276 seconds) 2011-04-24T12:50:09 *** siddharth has joined #aichallenge 2011-04-24T12:51:01 *** MatmaRex has left #aichallenge 2011-04-24T13:10:31 *** 16WAAB8IE <16WAAB8IE!~mad5ci@aeog138.neoplus.adsl.tpnet.pl> has quit IRC (Ping timeout: 240 seconds) 2011-04-24T13:10:55 *** kaemo has quit IRC (Ping timeout: 248 seconds) 2011-04-24T13:27:55 *** kaemo has joined #aichallenge 2011-04-24T13:27:57 *** kaemo0 has joined #aichallenge 2011-04-24T13:37:32 *** orivej has joined #aichallenge 2011-04-24T13:52:40 *** Frontier has quit IRC (Quit: Leaving.) 2011-04-24T13:58:37 *** computerwiz_222 has quit IRC (Remote host closed the connection) 2011-04-24T14:05:38 *** amriedle has quit IRC (Quit: Lost terminal) 2011-04-24T14:07:58 *** berak has joined #aichallenge 2011-04-24T14:13:14 *** Mathnerd314 has joined #aichallenge 2011-04-24T14:34:07 *** _0x47 <_0x47!~slash@p3EE0BEF1.dip.t-dialin.net> has quit IRC (Ping timeout: 258 seconds) 2011-04-24T14:54:13 *** delt0r_ has joined #aichallenge 2011-04-24T14:55:56 *** delt0r___ has quit IRC (Ping timeout: 276 seconds) 2011-04-24T15:25:57 the workers already install ruby 1.9.2 2011-04-24T15:35:42 *** Chris_0076 has joined #aichallenge 2011-04-24T15:49:02 *** Accoun has quit IRC () 2011-04-24T16:00:40 *** Accoun has joined #aichallenge 2011-04-24T16:02:52 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 3.6.16/20110323143040]) 2011-04-24T17:03:18 *** _0x47 <_0x47!~slash@p3EE0A3EC.dip.t-dialin.net> has joined #aichallenge 2011-04-24T17:19:03 *** berak has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2011-04-24T17:50:13 *** choas has quit IRC (Quit: leaving) 2011-04-24T17:58:12 *** Kingpin13 has quit IRC (Quit: money can't buy you love, but it can if you have enough) 2011-04-24T18:02:41 *** siddharth has quit IRC (Ping timeout: 240 seconds) 2011-04-24T19:10:23 *** mceier has quit IRC (Quit: leaving) 2011-04-24T20:00:44 *** _0x47 <_0x47!~slash@p3EE0A3EC.dip.t-dialin.net> has quit IRC (Quit: Leaving.) 2011-04-24T20:11:11 *** FireFly has quit IRC (Quit: swatted to death) 2011-04-24T20:41:20 *** needsch has quit IRC (Quit: Leaving.) 2011-04-24T20:41:39 *** bhasker has joined #aichallenge 2011-04-24T20:45:39 amstan 2011-04-24T20:45:43 merged with epsilon 2011-04-24T20:45:48 can you try the pull again? 2011-04-24T21:01:52 btw in c++ can we use boost will it work on the contest server? 2011-04-24T21:02:04 bhasker: congrats? :P 2011-04-24T21:02:13 :) thanks:D 2011-04-24T21:02:20 just got home today finally after 3 days 2011-04-24T21:02:29 ah nice 2011-04-24T21:02:31 she is a joy to be around 2011-04-24T21:02:34 boy/girl? name? 2011-04-24T21:02:37 first one's answered 2011-04-24T21:02:38 girl 2011-04-24T21:02:41 akshara :) 2011-04-24T21:02:47 :) 2011-04-24T21:03:21 (11:02:04) bhasker: she is a joy to be around 2011-04-24T21:03:21 you say that for now :P 2011-04-24T21:03:48 i don't think boost is available (could be wrong), it might be nice if it was 2011-04-24T21:03:54 i am sure there will be times when i will say the contrary:) but she is a joy neverthless 2011-04-24T21:03:58 ah k 2011-04-24T21:05:10 i was hoping to use some filter classes to filter standard containers 2011-04-24T21:05:12 and get a modified 2011-04-24T21:05:14 sequence 2011-04-24T21:08:05 i was trying to work out before how i could best remove stuff from sets 2011-04-24T21:08:50 if i have say sets of vectors of ints, maybe I want to remove everything like (0, -1, _, _) from the set 2011-04-24T21:10:46 hmm 2011-04-24T21:11:08 so each set element is a vector itself? 2011-04-24T21:11:34 i guess you could write a function that returns true or false for each set element 2011-04-24T21:11:53 let me pull up somestuff holdonf 2011-04-24T21:12:18 (i'm off to a surprise 21st bbq soon, but i'll read stuff later otherwise (:) 2011-04-24T21:12:21 what you need is the remove_if fn from algorithms 2011-04-24T21:12:31 ah cheers 2011-04-24T21:12:42 it takes a start/stop iterator and a predicate 2011-04-24T21:12:47 i can use that for finding non-collision moves for sets of ants 2011-04-24T21:12:51 http://www.cplusplus.com/reference/algorithm/remove_if/ 2011-04-24T21:14:38 it basically walks through the list and overwrites any that match with the next element 2011-04-24T21:15:26 it seems like it might not be straight forward to pass the predicate something to match with? 2011-04-24T21:15:39 u can use a binder 2011-04-24T21:15:43 i don't have time to play now though, so i'll look later 2011-04-24T21:15:44 read up on STL binder 2011-04-24T21:15:50 will do :) 2011-04-24T21:15:58 you can basically pass a function object 2011-04-24T21:16:01 which binds a value 2011-04-24T21:16:09 and pass that as a predicate 2011-04-24T21:16:17 instead of a simple function 2011-04-24T21:16:46 *** sigh has joined #aichallenge 2011-04-24T21:16:48 the idea is for each ant, find the set of moves not into water 2011-04-24T21:16:59 then take the cartesian product of those sets to get moves in a game tree 2011-04-24T21:17:07 but then remove all moves that result in cycles/collisions 2011-04-24T21:17:19 that just went over my head 2011-04-24T21:17:38 oh wait i think i understand 2011-04-24T21:17:43 :) 2011-04-24T21:17:48 cartesian set will yield every possible combination 2011-04-24T21:17:54 a cartesian product is just the set of all ordered tuples from your sets 2011-04-24T21:17:58 yep 2011-04-24T21:18:00 wont that be a really big ass tree 2011-04-24T21:18:05 when there are 500 ants 2011-04-24T21:18:10 depends 2011-04-24T21:18:20 i plan to partition them into sets of battle ants that can affect each other 2011-04-24T21:18:21 you better do massive pruning i guess 2011-04-24T21:18:25 ah 2011-04-24T21:18:25 then iteratively do those smaller trees 2011-04-24T21:18:41 but how will you coordinate strategy 2011-04-24T21:18:46 i will see where i get with that 2011-04-24T21:18:56 i am only using a tree for "battles" 2011-04-24T21:19:01 ah k 2011-04-24T21:19:02 everything else i'm using heuristics for 2011-04-24T21:19:14 sounds like a good idea 2011-04-24T21:19:35 anyway, ttyl :) 2011-04-24T21:19:46 l8r 2011-04-24T21:19:51 *** bhasker has quit IRC (Quit: bhasker) 2011-04-24T21:23:52 *** amstan has joined #aichallenge 2011-04-24T21:23:52 *** ChanServ sets mode: +o amstan 2011-04-24T21:32:31 'ello 2011-04-24T21:52:56 *** siddharth has joined #aichallenge 2011-04-24T21:53:24 *** jaycobcoleman has joined #aichallenge 2011-04-24T21:54:03 *** Eruonen has quit IRC () 2011-04-24T22:31:27 *** Ttech has quit IRC (Quit: Your nick was overruled.) 2011-04-24T22:34:20 *** siddharth has quit IRC (Ping timeout: 264 seconds) 2011-04-24T22:35:07 animatroid: What's new? 2011-04-24T22:38:28 *** Ttech has joined #aichallenge 2011-04-24T22:43:47 contestbot: later tell mcleopold Can you get the test server with a few sample games? i would like to test the visualizer and the leaderboards's look 2011-04-24T22:43:47 amstan: OK 2011-04-24T22:47:18 *** siddharth has joined #aichallenge 2011-04-24T22:49:08 *** chris__0076 has joined #aichallenge 2011-04-24T22:51:49 *** Chris_0076 has quit IRC (Ping timeout: 250 seconds) 2011-04-24T22:52:08 *** Redgis has quit IRC (Ping timeout: 260 seconds) 2011-04-24T22:54:09 *** delt0r___ has joined #aichallenge 2011-04-24T22:55:43 *** delt0r_ has quit IRC (Ping timeout: 250 seconds) 2011-04-24T23:06:50 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-04-24T23:10:13 *** jaycobcoleman has quit IRC (Ping timeout: 252 seconds) 2011-04-24T23:18:55 *** chris__0076 has quit IRC (Quit: Leaving) 2011-04-24T23:36:23 *** jaycobcoleman has joined #aichallenge