2011-12-12T00:04:03  <BenJackson> huh, drand48 is pretty damn expensive
2011-12-12T00:10:57  <thestinger> BenJackson: http://en.cppreference.com/w/cpp/numeric/random not sure if gcc/libstdc++ have this stuff yet
2011-12-12T00:11:36  <BenJackson> I'm just using drand48() as parto f my simulated annealing
2011-12-12T00:11:48  <BenJackson> it's surprisingly slow
2011-12-12T00:12:07  <thestinger> I think the C++11 rand stuff lets you pick a generator
2011-12-12T00:12:32  <BenJackson> I'm just going to switch from exp() < drand48() to exp()*RAND_MAX < random()
2011-12-12T00:14:58  <BenJackson> random() is about 10x faster than drand48()
2011-12-12T00:16:24  *** u_ <u_!~u_@adsl-75-57-180-214.dsl.emhril.sbcglobal.net> has quit IRC (Quit: u_)
2011-12-12T00:17:04  <thestinger> the drand48 man page says not to use it :P
2011-12-12T00:19:49  *** Areks <Areks!~Areks@rs.gridnine.com> has joined #aichallenge
2011-12-12T00:19:58  <BenJackson> not mine on FreeBSD :)
2011-12-12T00:20:16  <dragonfyre13> ok, been at this since friday, time to watch a tv show and go to bed. Tired....
2011-12-12T00:20:19  <dragonfyre13> night all
2011-12-12T00:20:38  <thestinger> BenJackson: the linux rand() is the same generator as random() according to the man page
2011-12-12T00:21:02  <thestinger> but it says not to use it if you want it to be portable, cause the old rand() is crap
2011-12-12T00:21:13  <thestinger> anyway doesn't matter since the contest server is modern linux
2011-12-12T00:21:35  <BenJackson> oh I know way too much about random
2011-12-12T00:21:45  <BenJackson> I even filed an obscure bug against the glibc one
2011-12-12T00:21:59  <BenJackson> a friend of mine got interested in attacks on random() on the nethack server
2011-12-12T00:22:14  <BenJackson> someone noticed they were using srandom(time())
2011-12-12T00:22:27  <BenJackson> so to make a point he started a game at just the right second of the day so there would be a wand of wishing on the first level
2011-12-12T00:22:30  <BenJackson> and then kicked it
2011-12-12T00:22:32  <BenJackson> (that kills you)
2011-12-12T00:22:42  <BenJackson> so it would show he died from kicking a wand of wishing on lvl 1
2011-12-12T00:23:05  <BenJackson> the server operators changed it to use srandom( 4 bytes from /dev/urandom, which is truly random in some sense)
2011-12-12T00:23:14  <BenJackson> so still pseudorandom, but you can't game the seed
2011-12-12T00:23:25  <BenJackson> my friend had worked out ways in which the game leaked out bits of random
2011-12-12T00:23:39  <thestinger> oh, so he figured out the state?
2011-12-12T00:23:41  <BenJackson> and had written some code to search every seed (2^32) to figure out which one he was in
2011-12-12T00:24:00  <BenJackson> he came to me when he thought he wanted to put it in an FPGA
2011-12-12T00:24:19  <BenJackson> ultimately I wrote a version that searched the seed space 50x faster
2011-12-12T00:24:23  <BenJackson> something like 25M seeds/sec
2011-12-12T00:24:37  <BenJackson> basically in a prototype of how I'd do it in an FPGA, but in C
2011-12-12T00:24:46  *** TheLinker <TheLinker!~TheLinker@200.117.99.103> has quit IRC (Quit: Bye)
2011-12-12T00:24:52  <BenJackson> and after that I computed a sort of rainbow table
2011-12-12T00:24:55  <BenJackson> it made a 16G file
2011-12-12T00:25:01  <BenJackson> and could tell you the seed in a fraction of a second
2011-12-12T00:25:10  <BenJackson> ANYWAY
2011-12-12T00:25:21  <BenJackson> glibc interpreted negative seeds wrong
2011-12-12T00:25:37  <BenJackson> which I discovered because I reimplemented random/srandom in a much faster way for that project
2011-12-12T00:25:43  <thestinger> glibc is crap though
2011-12-12T00:26:07  <BenJackson> btw that is the only project where I've ever seen a HUGE difference between 32 and 64-bit performance
2011-12-12T00:26:21  <BenJackson> that 25M seed/sec program built for 32 bit on the same CPU is only about 50-60% as fast
2011-12-12T00:27:03  <thestinger> with -march=native?
2011-12-12T00:27:24  <BenJackson> http://sourceware.org/ml/glibc-bugs/2009-03/msg00010.html
2011-12-12T00:27:29  <BenJackson> there's my obscure bug
2011-12-12T00:27:46  <BenJackson> just -m32
2011-12-12T00:28:46  <BenJackson> that program also had very careful balancing of "likely" and "unlikely"
2011-12-12T00:28:52  <BenJackson> some (true) likelys actually slowed it down
2011-12-12T00:28:54  *** TheLinker <TheLinker!~TheLinker@200.117.99.103> has joined #aichallenge
2011-12-12T00:29:01  *** Extrarius <Extrarius!~Extrarius@cpe-70-125-154-130.satx.res.rr.com> has quit IRC (Ping timeout: 276 seconds)
2011-12-12T00:30:14  <thestinger> clang can almost compile my bot
2011-12-12T00:30:23  <thestinger> it just doesn't support C++11 lambdas
2011-12-12T00:30:30  <thestinger> but apparently it _does_ support lambdas
2011-12-12T00:31:02  <thestinger> well, it has objective-C blocks as a C language extension or something like that
2011-12-12T00:31:48  <thestinger> not sure how to build it with support for them though, the Arch package for clang is already barely hacked together
2011-12-12T00:32:31  <bqf> BenJackson: "(...) how I'd do it in an FPGA, but in C" <--- with the limited amount of FPGA programming i have done, the first thing that would come to mind would be parallelizing it. since i can't imagine that that approach would be 50x faster in C (unless you have 50 cores), what was your idea?
2011-12-12T00:32:35  <thestinger> I tried some bits of my code in clang though, and the performance differences are very neat
2011-12-12T00:33:39  <BenJackson> bqf: if you look at how random() is implemented and convolve it to avoid repeated operations you can go much much faster
2011-12-12T00:34:23  <BenJackson> actually I don't think an FPGA implementation would be better
2011-12-12T00:34:45  <bqf> BenJackson: oh, of course... if you're trying to search the whole space there's no need to start from scratch every time.
2011-12-12T00:35:24  <BenJackson> in particular if you are willing to search the seeds in the order of srandom()'s own internal LFSR you can eliminate a ton of math
2011-12-12T00:35:55  <BenJackson> the next step of srandom is actually invoking random() 300ish times to stir the pot of its internal state
2011-12-12T00:36:05  <BenJackson> to avoid some initial degeneracy I think
2011-12-12T00:36:19  <BenJackson> so it stirs up its 31 (I think it is) terms of random state
2011-12-12T00:36:36  <BenJackson> whereas I precomputed the coefficients of the resulting 31 terms that apply to the initial seed
2011-12-12T00:37:40  <BenJackson> and finally since you are testing several sequential outputs of random when searching the seeds it is able to pipeline those
2011-12-12T00:38:16  <BenJackson> since a huge amount is shared between the first output of a given seed and the second output of the *previous* seed (given you're searching them in the order I described)
2011-12-12T00:40:07  <bqf> i really shouldn't be talking to you. every time i do (or listen in on conversations you're having here), you make me want to research some obscure, but incredibly interesting topic.
2011-12-12T00:40:29  <BenJackson> that was a really interesting project
2011-12-12T00:40:41  <BenJackson> the adjunct project (figure out how to exploit the known seed) never got done, though :)
2011-12-12T00:40:51  <bmh> BenJackson: I tend to hate everyone's RNG.
2011-12-12T00:41:20  <BenJackson> a friend of mine was trying to work out (given the seed) how many times to do some action that burned 1 random() before you did something like make a wish or pray
2011-12-12T00:41:21  <bmh> In particular, I have it in for Mersenne's Twister
2011-12-12T00:44:28  <bqf> BenJackson: if the adjunct project caused you to dig that deep into a topic and stay interested, then it has served its purpose. i guess i need new friends, when mine come to me with a technology related problem, it is mostly "how do i cheat in game X?" or "can you get me this movie?" (no, i can't. go pay for it.)
2011-12-12T00:45:39  <BenJackson> technically this WAS "how to cheat at nethack"
2011-12-12T00:46:44  <bmh> BenJackson: I do a tiny bit of development on the game Dungeon Crawl (along with Zannick) -- I ended up rewriting the RNG to prevent exactly that sort of attack
2011-12-12T00:47:26  <BenJackson> btw if you like nethack you should look at the nethack AIs that are out there
2011-12-12T00:47:42  <BenJackson> saiph in particular
2011-12-12T00:47:52  <BenJackson> http://www.youtube.com/watch?v=vDAmUNofE00
2011-12-12T00:47:55  <BenJackson> as an example
2011-12-12T00:48:20  <BenJackson> the neat thing about it is that it's very modular, so if you solve some part of nethack (eg sokoban or the castle or whatever) you canplug it in
2011-12-12T00:48:47  <BenJackson> plus it can use those strategies you're too lazy to use, like engraving Elbereth all the damn time
2011-12-12T00:49:10  <bmh> That's cute. The problem is that nethack is tedious and there's an 'optimal' strategy
2011-12-12T00:52:08  <bqf> i never really got into nethack - mainly because i could instantly recognize its potential to kill weeks of my time. botting it does sound interesting though, so once again, i _really_ shouldn't be talking to you ;)
2011-12-12T00:52:32  <bmh> I quit crawl when I was 18 and thankfully didn't discover crawl until 5 years later.
2011-12-12T00:52:34  <BenJackson> nethack is amazing for the way everything works together
2011-12-12T00:52:42  <bmh> er %s/crawl/nethack/
2011-12-12T00:52:53  <BenJackson> everything interacts in appropriate ways
2011-12-12T00:53:01  <bmh> BenJackson: try crawl :)
2011-12-12T00:53:09  <BenJackson> bmh: I'm playing Skyrim :)
2011-12-12T00:53:49  <bmh> mid-game I'm finding that my goals become instable and ants near the hill never really run away from the hill. Suggestions?
2011-12-12T00:54:14  <BenJackson> same thing happened to me
2011-12-12T00:54:38  <BenJackson> couple ideas:  change your idea of explored ever to explored "recently"
2011-12-12T00:54:45  <BenJackson> so it keeps refreshing
2011-12-12T00:54:55  <BenJackson> or add the refreshing one as a secondary goal
2011-12-12T00:55:03  <BenJackson> V7 did something like that, I think
2011-12-12T00:56:07  <bmh> v2 beats v3. I think I might have made a mistake
2011-12-12T00:56:23  <BenJackson> that's why I keep all the old ones around
2011-12-12T00:56:33  <bmh> That's why I use git and tags
2011-12-12T00:56:41  <BenJackson> yep
2011-12-12T00:58:08  <bmh> v2 uses integer values for goals and each ant moves toward its nearest goal. v3 transforms those scores 1/sqrt(n) and then sums over the directions
2011-12-12T00:58:13  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has quit IRC (Ping timeout: 248 seconds)
2011-12-12T00:58:21  *** skalogryz <skalogryz!ae63307a@gateway/web/freenode/ip.174.99.48.122> has joined #aichallenge
2011-12-12T00:58:28  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has joined #aichallenge
2011-12-12T00:59:16  <BenJackson> I used integer values for goals
2011-12-12T00:59:24  <BenJackson> but then I scaled them
2011-12-12T00:59:34  <BenJackson> something like if > threshold use *mul/div
2011-12-12T00:59:49  <BenJackson> so local goals would use 1/n to make it uninteresting far away
2011-12-12T01:00:01  <BenJackson> errr n/1
2011-12-12T01:00:10  <BenJackson> and goals like hills would use 1/n so it seemed "closer"
2011-12-12T01:00:50  <bmh> oh, I see the problem. Some of my BFSs are blocked by ants. Ants go through choke points near the hill, which causes other ants to reverse course
2011-12-12T01:00:55  <BenJackson> I'm pretty much convinced now that despite my attempts to avoid it the only way to win is going to be to assign to each individual bot
2011-12-12T01:01:38  <BenJackson> yeah I wrote path costing but I've yet to use it
2011-12-12T01:01:47  <BenJackson> I'm afraid of how... emergent... its behavior coudl be
2011-12-12T01:02:42  <bmh> my costing is implicit. The BFS gets staunched at an ant, so the scores traveling through an ant are something like: n, n+1, n+3
2011-12-12T01:02:53  *** skalogryz <skalogryz!ae63307a@gateway/web/freenode/ip.174.99.48.122> has quit IRC (Ping timeout: 258 seconds)
2011-12-12T01:05:01  <bmh> BenJackson: are you still running v4 on my server?
2011-12-12T01:05:15  <BenJackson> I think it's running now
2011-12-12T01:05:24  <BenJackson> I stopped V7 when someone mentioned it was beating them a lot
2011-12-12T01:06:11  <AntDroid_> do you guys use an ant to block BFS? i currently don't, makes my exploration weaker
2011-12-12T01:06:30  <BenJackson> not sure it does, really
2011-12-12T01:06:33  <BenJackson> one ant doesn't block much
2011-12-12T01:06:51  <BenJackson> even if the cost of the ant is 1000 the cost of walking around the ant is only like 5
2011-12-12T01:07:31  <BenJackson> plus let's say it DOES work (there's a chokepoint).  do you really mean that THAT ant is responsible for everything on the other side?
2011-12-12T01:07:38  <BenJackson> since it is "on the way" over there
2011-12-12T01:08:26  <AntDroid_> yes. my problem is that at chokepoints my ants just pile up there
2011-12-12T01:08:42  <bmh> AntDroid_: I block BFS depending on the goal type -- 'enemy ant,' yes, to get my ants to flow around and try to envelop the target
2011-12-12T01:08:53  <AntDroid_> so i wanted to limit the number of ants there at 1 layer
2011-12-12T01:09:23  <AntDroid_> the problem is that at mazes my exploration doesn't work well
2011-12-12T01:11:12  <bmh> I still need combat. oops.
2011-12-12T01:12:21  <BenJackson> bmh: did you try the "simple" one on the forum?
2011-12-12T01:12:38  <bmh> BenJackson: nope. Haven't looked at the forum lately
2011-12-12T01:12:49  <AntDroid_> bmh: i used to block BFS when enemy ant is encountered, i cant remember anymore why my current implementation just subtracts a value whne an ant is encountered
2011-12-12T01:13:20  <AntDroid_> BenJackson: is that the one by Memetix?
2011-12-12T01:15:13  <BenJackson> "yeah
2011-12-12T01:15:31  <bmh> BenJackson: for doing so many dumb things, v3 is running over your v4 http://bhickey.net:2080/replay.7228
2011-12-12T01:17:05  <AntDroid_> i've managed to implement the calculations on the technique but what i can't get is how to command my ants to do combat with that, the only thing i could use it for is to determine if a square you're stepping to is safe, draw, or not
2011-12-12T01:18:18  <BenJackson> yes, well, v4 is terrible :)
2011-12-12T01:19:29  <BenJackson> bmh: if you're runnign out of disk space you coudl go do a query to delete all the v4 vs v5 games
2011-12-12T01:19:32  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has quit IRC (Ping timeout: 252 seconds)
2011-12-12T01:19:37  <BenJackson> which must be 1000s
2011-12-12T01:19:41  <bmh> BenJackson: I took down the machine and gave it an extra 4gb
2011-12-12T01:19:49  *** shleb <shleb!48c598ce@gateway/web/freenode/ip.72.197.152.206> has joined #aichallenge
2011-12-12T01:19:55  <bmh> It ran out of space because I installed every Haskell library
2011-12-12T01:20:45  <bmh> BenJackson: http://bhickey.net:2080/replay.7234 I think I might be on to something.
2011-12-12T01:21:03  *** gazpachoking <gazpachoking!~gazpachok@c-67-163-203-66.hsd1.pa.comcast.net> has left #aichallenge ("Leaving")
2011-12-12T01:21:07  *** ltriant <ltriant!~ltriant@lithium.mailguard.com.au> has quit IRC (Quit: Computer has gone to sleep)
2011-12-12T01:24:13  *** Jak_o_Shadows <Jak_o_Shadows!~hayden@CPE-144-136-214-52.sa.bigpond.net.au> has joined #aichallenge
2011-12-12T01:25:02  <shleb> Im following the tutorial, anyone answer a python question me?
2011-12-12T01:25:12  <shleb> *mind answering
2011-12-12T01:25:59  <BenJackson> bmh: wow that's hard to watch
2011-12-12T01:26:02  <BenJackson> those bots are terrible :)
2011-12-12T01:26:30  <bmh> BenJackson: three weeks ago, I'd be in the top 20.
2011-12-12T01:29:41  *** shleb <shleb!48c598ce@gateway/web/freenode/ip.72.197.152.206> has quit IRC (Quit: Page closed)
2011-12-12T01:34:18  <bmh> BenJackson: does my exploration (apart from the food bug at 365) look ok?
2011-12-12T01:35:39  <McLeopold> can some one give me a sanity check?
2011-12-12T01:35:51  <McLeopold> I've got a combat situation I want a review on
2011-12-12T01:36:02  <McLeopold> a a _ _ b
2011-12-12T01:36:39  <McLeopold> my thought is the "correct" move is the close a goes up or down, and the furthur slides in it's place
2011-12-12T01:39:40  <matthewd> McLeopold: Agree
2011-12-12T01:40:00  <McLeopold> so, shouldn't a1k0n's method be able to solve that?
2011-12-12T01:41:21  <bmh> yes.
2011-12-12T01:41:40  <McLeopold> bmh: I wonder what I'm doing wrong then...
2011-12-12T01:42:52  <McLeopold> bmh: what is your food bug?
2011-12-12T01:43:17  <bmh> McLeopold: I only consider food that can be seen at the moment.
2011-12-12T01:44:17  <roflmao> what the
2011-12-12T01:44:17  <roflmao> hell
2011-12-12T01:44:20  <roflmao> java is being RETARDED
2011-12-12T01:44:36  <roflmao> zoneMap = combatMap;
2011-12-12T01:44:43  <roflmao> where zonemap is int[][] zoneMap
2011-12-12T01:44:51  <roflmao> = complete bot freeze
2011-12-12T01:44:51  <roflmao> :|
2011-12-12T01:49:12  *** Harpyon <Harpyon!~Harpyon@ti0031a380-0316.bb.online.no> has quit IRC (Quit: Textual IRC Client: http://www.textualapp.com/)
2011-12-12T01:50:38  <thestinger> Fluxid: btw, your tcp server has slowed to a crawl for some reason
2011-12-12T01:50:53  <thestinger> http://ants.fluxid.pl/player/strcat_push
2011-12-12T01:51:20  <thestinger> well
2011-12-12T01:51:24  <thestinger> I guess it's not that bad :P
2011-12-12T01:51:35  <thestinger> there are just a lot of good bots now so games take a long time
2011-12-12T01:52:13  <thestinger> there seem to be a lot of timeouts now though
2011-12-12T01:56:24  *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal)
2011-12-12T02:01:06  *** epicmonkey <epicmonkey!~epicmonke@178.130.5.103> has joined #aichallenge
2011-12-12T02:07:59  <bmh> bug fixed. My bot has object permanence.
2011-12-12T02:09:45  *** raemde_ is now known as raemde
2011-12-12T02:10:04  *** epicmonkey <epicmonkey!~epicmonke@178.130.5.103> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T02:15:04  <McLeopold> arg, my battle simulator assumes ants can't move on to each other :(
2011-12-12T02:15:18  <McLeopold> considering I wrote the dang engine, that is pretty stupid :(
2011-12-12T02:15:51  *** TheLinker <TheLinker!~TheLinker@200.117.99.103> has quit IRC (Quit: Bye)
2011-12-12T02:16:02  <bmh> McLeopold: why do we pick up food by walking next to it?
2011-12-12T02:16:20  <McLeopold> instead of on it?
2011-12-12T02:16:24  <bmh> yes
2011-12-12T02:16:45  <McLeopold> we originally had a radius, so you only had to be within 3 squares
2011-12-12T02:16:54  <McLeopold> then we started playing with the numbers
2011-12-12T02:17:21  <McLeopold> at one point it was highly likely that 2 ants could both be within range, it which case the food would vanish
2011-12-12T02:18:08  <McLeopold> we lowered it to 1, which at the time was the closest you could get, because we wanted the map in a matrix and you couldn't have 2 things at the same spot
2011-12-12T02:18:24  <AntDroid_> what would happen if             a*b ?
2011-12-12T02:18:29  *** roflmao <roflmao!~jstrong@189.229.217.18> has left #aichallenge
2011-12-12T02:18:50  <McLeopold> AntDroid_: food goes away, no one gets it
2011-12-12T02:18:57  <McLeopold> but it can't happen
2011-12-12T02:18:58  <AntDroid_> who gets the food? is it vanished or both ants will kill each other?
2011-12-12T02:19:10  <McLeopold> the attack will happen first, and the ants will die
2011-12-12T02:19:30  <janzert> McLeopold: can still happen only if the food spawns next to a hill at the same time an ant spawns on the hill
2011-12-12T02:19:38  <McLeopold> also, we had battle types that allowed ants to live if they were near each other, so we needed the rule for that as well
2011-12-12T02:19:46  <janzert> ahh, right
2011-12-12T02:19:50  <McLeopold> janzert: nope
2011-12-12T02:20:07  <McLeopold> well, yes, you can see it, but the gather happens after the deaths
2011-12-12T02:20:14  <McLeopold> so the next turn will be _*_
2011-12-12T02:20:52  <AntDroid_> ah. okay.
2011-12-12T02:24:15  *** Garf <Garf!~Garf@78-22-80-252.access.telenet.be> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T02:25:13  *** bmh <bmh!ada4e181@fsf/member/bmh> has quit IRC (Quit: Page closed)
2011-12-12T02:34:20  <antimatroid1> McLeopold: can two ants not be in range of food after battle?
2011-12-12T02:34:29  <antimatroid1> ooh, janzert answered
2011-12-12T02:34:47  <antimatroid1> bmh: ants also used to spawn at food before ant hills, so it wouldn't have worked very well to spawn ants tehre
2011-12-12T02:38:01  *** GeorgeSebastian <GeorgeSebastian!~georgeseb@unaffiliated/georgesebastian> has joined #aichallenge
2011-12-12T02:39:38  *** Jak_o_Shadows <Jak_o_Shadows!~hayden@CPE-144-136-214-52.sa.bigpond.net.au> has quit IRC (Read error: Connection reset by peer)
2011-12-12T02:41:37  *** Jak_o_Shadows <Jak_o_Shadows!~hayden@CPE-144-136-214-52.sa.bigpond.net.au> has joined #aichallenge
2011-12-12T03:00:25  *** mviel <mviel!~mviel@236.227.73.86.rev.sfr.net> has joined #aichallenge
2011-12-12T03:02:31  *** Garf <Garf!~Garf@78-22-80-252.access.telenet.be> has joined #aichallenge
2011-12-12T03:07:30  *** Hexren <Hexren!~hexren@46.59.205.23> has joined #aichallenge
2011-12-12T03:07:32  *** Hexren <Hexren!~hexren@46.59.205.23> has left #aichallenge
2011-12-12T03:09:49  *** mj41 <mj41!~mj41_nby@195.39.106.218> has joined #aichallenge
2011-12-12T03:14:54  *** ikaros <ikaros!~ikaros@dslb-084-059-073-057.pools.arcor-ip.net> has joined #aichallenge
2011-12-12T03:17:27  <janzert> hk
2011-12-12T03:37:09  *** Blkt <Blkt!~user@89-96-199-46.ip13.fastwebnet.it> has joined #aichallenge
2011-12-12T03:47:48  <BenJackson> @rankings
2011-12-12T03:47:49  <contestbot> BenJackson: Top 10 players: pguillory(93.5), xathis(92.7), GreenTea(91.4), ChrisH(90.3), lazarant(89.9), delineate(89.7), Memetix(88.2), FlagCapper(88.1), a1k0n(87.7), BenJackson(87.2)
2011-12-12T03:47:56  *** Blkt <Blkt!~user@89-96-199-46.ip13.fastwebnet.it> has quit IRC (Remote host closed the connection)
2011-12-12T03:48:15  <Garf> my bot dropped like 50 places this weekend
2011-12-12T03:48:22  <Garf> people must be submitting new ones like crazy
2011-12-12T03:48:38  <BenJackson> I'm trying to figure out if I moved back up or if some people resub'd
2011-12-12T03:48:40  <Wraithan> They want to get testing in before the final
2011-12-12T03:48:45  <bhasker> my bot is slowly climbing the ranks
2011-12-12T03:48:47  <bhasker> at 807 now!
2011-12-12T03:48:58  <bhasker> if only i could make my battle code faster
2011-12-12T03:49:02  <bhasker> stupid thing timesout all the time
2011-12-12T03:49:08  <bhasker> my move generator for minmax sucks
2011-12-12T03:49:16  <bhasker> maybe i should just switch to the battle described in the forums
2011-12-12T03:49:32  <BenJackson> it's obviously "enough"
2011-12-12T03:49:42  <BenJackson> since that guy is Memetix
2011-12-12T03:49:43  <bhasker> did it work for you ben?
2011-12-12T03:49:55  <Wraithan> My bot blows and I am at 1400s
2011-12-12T03:49:56  <BenJackson> no, I wrote mine before that and it's much different
2011-12-12T03:50:03  <bhasker> ah k
2011-12-12T03:50:08  <Wraithan> Hope to have a new bot ready tonight to resub and see where it can go
2011-12-12T03:50:12  <BenJackson> I don't think combat is differentiating that much in the top 10
2011-12-12T03:50:13  <bhasker> a1k0n is also doing quite well
2011-12-12T03:50:19  <BenJackson> I think in the top 10 it's back to strategy
2011-12-12T03:50:20  <bhasker> so mostly exploration?
2011-12-12T03:50:24  <bhasker> and when to fight
2011-12-12T03:50:33  <BenJackson> look at xathis on fluxid
2011-12-12T03:50:41  <BenJackson> his positioning of ants is crazy good
2011-12-12T03:50:48  * Wraithan started late so his goal is to break into the top 500
2011-12-12T03:50:58  <BenJackson> wraithan: beware creeping goals
2011-12-12T03:51:06  <BenJackson> I was happy with top 25 or so until I broke into the top 10 :)
2011-12-12T03:51:19  <Wraithan> BenJackson: heh
2011-12-12T03:51:26  <Wraithan> I'm already spending all my spare time on my bot
2011-12-12T03:51:33  <Wraithan> but I just got started last weekend
2011-12-12T03:51:54  <BenJackson> so at this instant I'm #10.  my vague goal would be to be solidly in the top 10 so as to be sure to stay there in the finals
2011-12-12T03:52:00  <bhasker> nice ben
2011-12-12T03:52:07  *** olexs <olexs!~Adium@s0491.dyn.hrz.tu-darmstadt.de> has joined #aichallenge
2011-12-12T03:53:09  <Garf> BenJackson: that can be hard, you have no idea what will be submitted the 17th :P
2011-12-12T03:53:42  <BenJackson> that's why being #10 (or say #13 like I was earlier this afternoon) is no good!
2011-12-12T03:54:04  <Wraithan> If you don't have a top ranking bot is it worth playing on fluxid?
2011-12-12T03:54:21  <BenJackson> sure, there's a mix of bots
2011-12-12T03:54:30  <janzert> http://pastebin.com/QUxumwqt current top 20 mu
2011-12-12T03:54:33  <BenJackson> if you're at 1500 you might try bhasker.net instead
2011-12-12T03:54:39  <Garf> fluxid has insaly good bots
2011-12-12T03:54:44  <Wraithan> BenJackson: thanks
2011-12-12T03:54:52  <bhasker> bhasker.net?? lol
2011-12-12T03:54:53  *** Blkt <Blkt!~user@89-96-199-46.ip13.fastwebnet.it> has joined #aichallenge
2011-12-12T03:54:56  <Garf> server spot = fluxid spot / 2   :P
2011-12-12T03:55:03  <bhasker> did i just host a server there?
2011-12-12T03:55:04  <BenJackson> err bhickey.net
2011-12-12T03:55:13  <BenJackson> bh*something ;)
2011-12-12T03:56:13  <BenJackson> janzerT: you should make that a page
2011-12-12T03:56:33  <BenJackson> anyway that list by mu is pretty close to what's in my mind
2011-12-12T03:56:46  <BenJackson> I don't expect to be xathis, pguillory or ChrisH
2011-12-12T03:56:50  <janzert> people would pretty certainly generally take the wrong idea from such a page
2011-12-12T03:56:54  <BenJackson> haven't been watching GreenTea or Flag lately
2011-12-12T03:57:04  *** Jak_o_Shadows <Jak_o_Shadows!~hayden@CPE-144-136-214-52.sa.bigpond.net.au> has quit IRC (Remote host closed the connection)
2011-12-12T03:57:12  <BenJackson> wow my sigma fell below xathis?
2011-12-12T03:57:26  <Garf> isn't that expected?
2011-12-12T03:57:35  <Garf> it's harder for xanthis to find close opponents
2011-12-12T03:57:41  <BenJackson> he's got like 2x my games though
2011-12-12T04:33:42  *** Fandekasp <Fandekasp!~Fandekasp@p26153-ipngn100102yosemiya.okinawa.ocn.ne.jp> has quit IRC (Ping timeout: 248 seconds)
2011-12-12T04:35:18  *** Hexren <Hexren!~hexren@46.59.205.23> has joined #aichallenge
2011-12-12T04:35:19  *** Hexren <Hexren!~hexren@46.59.205.23> has left #aichallenge
2011-12-12T04:47:42  <AntDroid_> thestinger: you are saved by an unfortunate timeout of my bot here: http://ants.fluxid.pl/replay.27605
2011-12-12T04:48:21  <AntDroid_> my combat algo sucks but it is being compensated by sheer number
2011-12-12T04:50:29  *** ikaros <ikaros!~ikaros@dslb-084-059-073-057.pools.arcor-ip.net> has quit IRC (Quit: Ex-Chat)
2011-12-12T04:54:27  *** liberforce <liberforce!~lmenina@LLagny-156-36-6-198.w80-14.abo.wanadoo.fr> has joined #aichallenge
2011-12-12T05:03:38  *** replore <replore!~replore@203.152.213.161.static.zoot.jp> has quit IRC (Read error: Connection reset by peer)
2011-12-12T05:06:08  *** pairofdice <pairofdice!~pod@dsl-hkibrasgw2-ff67c300-50.dhcp.inet.fi> has joined #aichallenge
2011-12-12T05:09:38  *** amstan <amstan!~amstan@69-196-130-230.dsl.teksavvy.com> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T05:12:34  *** raemde_ <raemde_!~raemde@dslb-088-073-235-117.pools.arcor-ip.net> has joined #aichallenge
2011-12-12T05:15:16  *** raemde <raemde!~raemde@dslb-088-073-235-117.pools.arcor-ip.net> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T05:18:05  *** mviel_ <mviel_!~mviel@213.163.164.204> has joined #aichallenge
2011-12-12T05:18:54  *** mviel <mviel!~mviel@236.227.73.86.rev.sfr.net> has quit IRC (Read error: Connection reset by peer)
2011-12-12T05:20:34  <Fluxid> go to the first 100 over the night.. three non-1st-place games out of 15
2011-12-12T05:21:45  *** olexs <olexs!~Adium@s0491.dyn.hrz.tu-darmstadt.de> has quit IRC (Quit: Leaving.)
2011-12-12T05:23:56  *** AntDroid_ <AntDroid_!781c39b6@gateway/web/freenode/ip.120.28.57.182> has quit IRC (Ping timeout: 258 seconds)
2011-12-12T05:26:21  <Garf> is it possible that visionradius2 changes in the final contest??
2011-12-12T05:28:09  <Fluxid> it may change
2011-12-12T05:29:14  <janzert> pretty certainly at this stage none of the radius variables will change
2011-12-12T05:33:52  *** vjacob <vjacob!~vjacob@94-195-174-165.zone9.bethere.co.uk> has quit IRC (Quit: Leaving)
2011-12-12T05:44:22  *** nha <nha!~prefect@imamac13.epfl.ch> has joined #aichallenge
2011-12-12T05:50:21  *** jstemmer <jstemmer!~cheetah@mrpwn.stemmertech.com> has joined #aichallenge
2011-12-12T05:58:00  *** raemde_ is now known as raemde
2011-12-12T06:03:31  *** ikaros <ikaros!~ikaros@studpool-wlan-74-86.fs.fbi.h-da.de> has joined #aichallenge
2011-12-12T06:04:15  *** sigh <sigh!~sigh@ppp114-153.static.internode.on.net> has joined #aichallenge
2011-12-12T06:11:22  *** Kurnevsky <Kurnevsky!~Errins@217.21.41.205> has joined #aichallenge
2011-12-12T06:11:34  *** Fandekasp <Fandekasp!~Fandekasp@nt10-ppp1812.west.sannet.ne.jp> has joined #aichallenge
2011-12-12T06:11:38  *** raemde <raemde!~raemde@dslb-088-073-235-117.pools.arcor-ip.net> has quit IRC (Read error: Connection reset by peer)
2011-12-12T06:19:38  *** JakeFisher <JakeFisher!5777b983@gateway/web/freenode/ip.87.119.185.131> has joined #aichallenge
2011-12-12T06:29:11  *** JakeFisher <JakeFisher!5777b983@gateway/web/freenode/ip.87.119.185.131> has quit IRC (Quit: Page closed)
2011-12-12T06:31:24  *** Kurnevsky1 <Kurnevsky1!~Errins@217.21.41.205> has joined #aichallenge
2011-12-12T06:31:24  *** Kurnevsky <Kurnevsky!~Errins@217.21.41.205> has quit IRC (Read error: Connection reset by peer)
2011-12-12T06:58:20  *** g0llum <g0llum!~g0llum@10.106.113.82.net.de.o2.com> has joined #aichallenge
2011-12-12T07:14:35  *** AntDroid <AntDroid!70c6400f@gateway/web/freenode/ip.112.198.64.15> has quit IRC (Quit: Page closed)
2011-12-12T07:18:33  *** replore <replore!~replore@ntkngw268248.kngw.nt.ftth.ppp.infoweb.ne.jp> has joined #aichallenge
2011-12-12T07:19:33  *** replore <replore!~replore@ntkngw268248.kngw.nt.ftth.ppp.infoweb.ne.jp> has quit IRC (Remote host closed the connection)
2011-12-12T07:26:37  *** Nitro <Nitro!~nitro@unaffiliated/nitro> has joined #aichallenge
2011-12-12T07:28:24  *** Nitro <Nitro!~nitro@unaffiliated/nitro> has quit IRC (Client Quit)
2011-12-12T07:29:16  *** TL_ <TL_!70c64009@gateway/web/freenode/ip.112.198.64.9> has joined #aichallenge
2011-12-12T07:29:46  *** Conorach <Conorach!~cono@90.170.87.124> has joined #aichallenge
2011-12-12T07:32:11  *** ikaros <ikaros!~ikaros@studpool-wlan-74-86.fs.fbi.h-da.de> has quit IRC (Ping timeout: 268 seconds)
2011-12-12T07:33:54  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has joined #aichallenge
2011-12-12T07:38:54  *** praveen <praveen!7aac106a@gateway/web/freenode/ip.122.172.16.106> has joined #aichallenge
2011-12-12T07:45:04  *** ikaros <ikaros!~ikaros@studpool-wlan-74-86.fs.fbi.h-da.de> has joined #aichallenge
2011-12-12T07:48:57  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has quit IRC (Ping timeout: 255 seconds)
2011-12-12T07:49:36  *** ikaros <ikaros!~ikaros@studpool-wlan-74-86.fs.fbi.h-da.de> has quit IRC (Remote host closed the connection)
2011-12-12T08:10:11  *** ikaros <ikaros!~ikaros@studpool-wlan-74-86.fs.fbi.h-da.de> has joined #aichallenge
2011-12-12T08:23:07  *** mviel__ <mviel__!~mviel@213.163.164.204> has joined #aichallenge
2011-12-12T08:26:43  *** mviel_ <mviel_!~mviel@213.163.164.204> has quit IRC (Ping timeout: 245 seconds)
2011-12-12T08:39:46  *** liuw <liuw!~liuw@firewall.ctxuk.citrix.com> has joined #aichallenge
2011-12-12T08:44:37  *** u_ <u_!~u_@adsl-75-57-180-214.dsl.emhril.sbcglobal.net> has joined #aichallenge
2011-12-12T08:45:22  *** u__ <u__!~u_@adsl-75-57-180-214.dsl.emhril.sbcglobal.net> has joined #aichallenge
2011-12-12T08:45:23  *** wuil <wuil!3ec81602@gateway/web/freenode/ip.62.200.22.2> has joined #aichallenge
2011-12-12T08:46:39  *** wuil <wuil!3ec81602@gateway/web/freenode/ip.62.200.22.2> has quit IRC (Client Quit)
2011-12-12T08:48:47  *** u_ <u_!~u_@adsl-75-57-180-214.dsl.emhril.sbcglobal.net> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T08:48:47  *** u__ is now known as u_
2011-12-12T08:50:40  *** raemde <raemde!~raemde@dslb-088-073-196-235.pools.arcor-ip.net> has joined #aichallenge
2011-12-12T08:53:19  *** keolpkojihu <keolpkojihu!~keolpkoji@ec2-176-34-249-220.eu-west-1.compute.amazonaws.com> has joined #aichallenge
2011-12-12T08:54:05  <antimatroid1> quiet channel is quiet
2011-12-12T08:54:15  *** UncleVasya <UncleVasya!kvirc@31.144.168.130> has joined #aichallenge
2011-12-12T08:59:04  *** flowenol <flowenol!LW@bl4-237-93.dsl.telepac.pt> has joined #aichallenge
2011-12-12T09:01:35  *** mviel_ <mviel_!~mviel@236.227.73.86.rev.sfr.net> has joined #aichallenge
2011-12-12T09:04:39  *** |UncleVasya| <|UncleVasya|!kvirc@31-12-143-177.dialup.umc.net.ua> has joined #aichallenge
2011-12-12T09:05:13  *** mviel__ <mviel__!~mviel@213.163.164.204> has quit IRC (Ping timeout: 252 seconds)
2011-12-12T09:06:18  *** UncleVasya <UncleVasya!kvirc@31.144.168.130> has quit IRC (Ping timeout: 252 seconds)
2011-12-12T09:13:13  *** TheLinker <TheLinker!~TheLinker@200.117.99.103> has joined #aichallenge
2011-12-12T09:15:02  *** avdg <avdg!~avdg@78-22-168-200.access.telenet.be> has joined #aichallenge
2011-12-12T09:18:40  *** Anilm3 <Anilm3!~anil@150.Red-83-44-168.dynamicIP.rima-tde.net> has joined #aichallenge
2011-12-12T09:19:05  <Anilm3> Hi guys
2011-12-12T09:19:53  <pairofdice> Hewwo
2011-12-12T09:21:57  *** sigh <sigh!~sigh@ppp114-153.static.internode.on.net> has quit IRC (Remote host closed the connection)
2011-12-12T09:28:05  *** QuirionPT <QuirionPT!~quirion@gyf5.kyla.fi> has joined #aichallenge
2011-12-12T09:29:30  *** keolpkojihu <keolpkojihu!~keolpkoji@ec2-176-34-249-220.eu-west-1.compute.amazonaws.com> has quit IRC (Quit: Leaving)
2011-12-12T09:33:13  <Fluxid> http://aichallenge.org/visualizer.php?game=205479&user=757 this game... those bots are few whole levels above
2011-12-12T09:35:25  *** tobym <tobym!~tobym@pool-173-56-80-136.nycmny.fios.verizon.net> has joined #aichallenge
2011-12-12T09:36:19  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has joined #aichallenge
2011-12-12T09:39:07  *** mviel__ <mviel__!~mviel@213.163.164.204> has joined #aichallenge
2011-12-12T09:40:27  *** Areks <Areks!~Areks@rs.gridnine.com> has quit IRC (Ping timeout: 268 seconds)
2011-12-12T09:42:36  <pairofdice> Big walls
2011-12-12T09:42:55  *** mviel_ <mviel_!~mviel@236.227.73.86.rev.sfr.net> has quit IRC (Ping timeout: 268 seconds)
2011-12-12T09:47:02  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has quit IRC (Ping timeout: 252 seconds)
2011-12-12T09:50:55  *** roflmao <roflmao!~jstrong@189.229.217.18> has joined #aichallenge
2011-12-12T09:54:19  *** xathis <xathis!~xathis@g231126239.adsl.alicedsl.de> has joined #aichallenge
2011-12-12T09:57:13  <nha> impressive fighting skills
2011-12-12T10:02:41  <Anilm3> Fluxid: pguillory is very aggressive
2011-12-12T10:05:18  <Anilm3> he's like a spartan
2011-12-12T10:05:22  <rwest> lol
2011-12-12T10:05:22  <Anilm3> or a klingon
2011-12-12T10:05:34  <rwest> he wins!
2011-12-12T10:05:45  <Anilm3> Qapla'!!!
2011-12-12T10:06:07  *** roflmao <roflmao!~jstrong@189.229.217.18> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T10:08:55  <Fluxid> i have no idea if and how he intentionaly isolates enemy
2011-12-12T10:09:13  <Fluxid> it looks like he actually work on isolating anemy and killing one by one
2011-12-12T10:13:14  <Anilm3> Certainly looks that way
2011-12-12T10:13:48  <Anilm3> another thing I noticed
2011-12-12T10:14:06  <Anilm3> is that he takes into account enemy ants when attacking to a third enemy
2011-12-12T10:14:37  <Anilm3> I don't know if it's intentional or just something emergent
2011-12-12T10:15:24  <avdg> maybe its trying to make hidden formations
2011-12-12T10:15:30  *** Palmik <Palmik!~palmik@man-208-200.pks.muni.cz> has joined #aichallenge
2011-12-12T10:15:37  <Anilm3> what do you mean?
2011-12-12T10:16:07  *** praveen <praveen!7aac106a@gateway/web/freenode/ip.122.172.16.106> has quit IRC (Quit: Page closed)
2011-12-12T10:18:17  <avdg> I have the impression he is trying to squeeze enemies, though I have to see more games
2011-12-12T10:20:41  <avdg> hmm, nah, I think he is just exploring every corner, go as far as the ants can go, then pick enemies that are squeezed
2011-12-12T10:23:00  *** BEASTMODE <BEASTMODE!u515@gateway/web/irccloud.com/x-zfhsthufluhjbduu> has quit IRC (Remote host closed the connection)
2011-12-12T10:23:00  *** nickjohnson <nickjohnson!u789@gateway/web/irccloud.com/x-urfxairbfnugldrh> has quit IRC (Remote host closed the connection)
2011-12-12T10:23:14  *** BEASTMODE <BEASTMODE!u515@gateway/web/irccloud.com/x-npqxsiphuuchyjmw> has joined #aichallenge
2011-12-12T10:26:31  *** nickjohnson <nickjohnson!u789@gateway/web/irccloud.com/x-prenyxqtofanzsef> has joined #aichallenge
2011-12-12T10:32:55  *** mviel__ is now known as mviel
2011-12-12T10:35:04  *** amstan <amstan!~amstan@69-196-130-230.dsl.teksavvy.com> has joined #aichallenge
2011-12-12T10:35:04  *** ChanServ sets mode: +o amstan
2011-12-12T10:58:27  *** nha <nha!~prefect@imamac13.epfl.ch> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T11:16:52  *** bergmark_ <bergmark_!bergmark@nat/ibm/x-hsmpsngayakvczht> has joined #aichallenge
2011-12-12T11:20:27  *** dvladim <dvladim!~dvladim@85.113.131.65> has joined #aichallenge
2011-12-12T11:24:46  *** jstrong <jstrong!~jstrong@189.229.217.18> has joined #aichallenge
2011-12-12T11:31:30  *** JorgeB <JorgeB!~textual@209.66.74.34> has joined #aichallenge
2011-12-12T11:32:29  <Anilm3> Hola JorgeB
2011-12-12T11:32:48  <JorgeB> morning
2011-12-12T11:33:08  *** flowenol <flowenol!LW@bl4-237-93.dsl.telepac.pt> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T11:41:52  <Fluxid> http://aichallenge.org/visualizer.php?game=206254&user=757 achievment unlocked: handled 432 ants without timeout and still working "combat"
2011-12-12T11:43:38  <avdg> I wonder about games like this: http://aichallenge.org/visualizer.php?game=199516
2011-12-12T11:43:39  <pairofdice> Beautifull guerrilla action
2011-12-12T11:44:38  <avdg> (my bot isn't as good as I wanted)
2011-12-12T11:45:11  <Fluxid> avdg: your bot is "gauss bot"
2011-12-12T11:45:22  <Fluxid> almost gaussian distrubution :)
2011-12-12T11:45:40  <a1k0n> central limit theorem and all that
2011-12-12T11:45:41  <Anilm3> I don't have time for developing my bot
2011-12-12T11:45:41  <avdg> meh, I wish I didn't had a lot of frustrations
2011-12-12T11:45:48  <Anilm3> it's so sad to see it rot
2011-12-12T11:46:16  <Fluxid> SIX MORE DAYS
2011-12-12T11:46:20  <avdg> :p
2011-12-12T11:46:21  <Minthos> at least it ranks high on ants in the hive
2011-12-12T11:46:26  <Anilm3> :(
2011-12-12T11:46:33  <avdg> js -> slow c++ -> no skillz
2011-12-12T11:47:27  <Minthos> 204 ants and 2115 in the hive, I don't think I've seen that much ever before
2011-12-12T11:48:01  <avdg> the hive could be a lot higher
2011-12-12T11:48:13  <pairofdice> It's important to have reserves :)
2011-12-12T11:48:23  <avdg> :p
2011-12-12T11:53:55  <pairofdice> Java bots seem to do well
2011-12-12T11:54:26  <pairofdice> c++ has a strong showing aswell
2011-12-12T11:54:49  <|UncleVasya|> And do not forget OCaml!
2011-12-12T11:55:33  <pairofdice> I don't even know what that is ;)
2011-12-12T11:55:40  *** bergmark_ <bergmark_!bergmark@nat/ibm/x-hsmpsngayakvczht> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T11:56:21  * a1k0n shakes his fist at Memetix
2011-12-12T12:05:14  *** Harpyon <Harpyon!~Harpyon@ti0031a380-0316.bb.online.no> has joined #aichallenge
2011-12-12T12:05:19  <amstan> Fluxid: 460ms, cutting it close eh?
2011-12-12T12:06:11  <amstan> i like the way easydoesit moves
2011-12-12T12:07:14  *** Dlayne <Dlayne!~qw@95.79.79.59> has left #aichallenge
2011-12-12T12:08:29  *** bergmark_ <bergmark_!bergmark@nat/ibm/x-momdrvrazyijhztf> has joined #aichallenge
2011-12-12T12:13:43  *** Ashoka <Ashoka!70875b1a@gateway/web/freenode/ip.112.135.91.26> has joined #aichallenge
2011-12-12T12:16:19  <a1k0n> holy crap xathis resubmitted?
2011-12-12T12:16:44  <rwest> a1k0n: run
2011-12-12T12:17:18  <Anilm3> @rankings
2011-12-12T12:17:20  <contestbot> Anilm3: Top 10 players: pguillory(93.6), GreenTea(91.3), FlagCapper(91.1), ChrisH(90.4), lazarant(90.0), delineate(89.5), Memetix(88.9), a1k0n(88.8), BenJackson(87.4), Hammerok(86.8)
2011-12-12T12:17:25  <Anilm3> :O
2011-12-12T12:17:59  <Anilm3> interesting development!!!
2011-12-12T12:18:57  <pairofdice> oh
2011-12-12T12:19:09  *** SMJ <SMJ!~janne@dyn2-212-50-133-246.psoas.suomi.net> has joined #aichallenge
2011-12-12T12:19:18  <pairofdice> xathis takes out the secret weapon
2011-12-12T12:19:39  <SMJ> I was wondering where xathis disappeared
2011-12-12T12:19:53  <Anilm3> nuclear powered ants!
2011-12-12T12:20:02  <SMJ> from the ranking
2011-12-12T12:20:09  <hkraal^afk> amstan: Went do the finals start?
2011-12-12T12:20:19  <a1k0n> hm i'm on the homepage again, always weird to see that
2011-12-12T12:20:20  *** hkraal^afk is now known as hkraal_
2011-12-12T12:21:03  <SMJ> actually I'm still wondering
2011-12-12T12:21:09  <avdg> tcp?
2011-12-12T12:21:47  *** GeorgeSebastian <GeorgeSebastian!~georgeseb@unaffiliated/georgesebastian> has quit IRC (Read error: Connection reset by peer)
2011-12-12T12:21:51  <Anilm3> xathis's profile must be the most visited page of today
2011-12-12T12:22:41  <SMJ> I see, a new version
2011-12-12T12:23:18  <avdg> probably of all time
2011-12-12T12:23:26  <avdg> (except homepage)
2011-12-12T12:23:27  <Anilm3> pguillory's F5 must be on fire
2011-12-12T12:23:40  *** besh <besh!835e7324@gateway/web/freenode/ip.131.94.115.36> has joined #aichallenge
2011-12-12T12:24:05  * avdg f5's his page (bot in game)
2011-12-12T12:26:13  *** Blkt <Blkt!~user@89-96-199-46.ip13.fastwebnet.it> has quit IRC (Quit: ERC Version 5.3 (IRC client for Emacs))
2011-12-12T12:26:14  <avdg> hmm, there are 5 players ahead now and I still didn't play a game :/
2011-12-12T12:26:26  <avdg> I wonder what happened
2011-12-12T12:26:29  <Anilm3> avdg: check if your submission is active
2011-12-12T12:26:39  <avdg> its just submitted
2011-12-12T12:27:07  <avdg> ah
2011-12-12T12:27:15  <avdg> game finished
2011-12-12T12:30:16  <BenJackson> xathis resubmitted?  wow
2011-12-12T12:30:30  <McLeopold> dang, he played starter bots :(
2011-12-12T12:31:06  <avdg> :p
2011-12-12T12:31:27  <avdg> I wonder whats next, probably timeout bots
2011-12-12T12:31:38  <a1k0n> well duh, everyone gets to beat up on starter bots
2011-12-12T12:32:00  <McLeopold> but they have  both been deactivated for a month
2011-12-12T12:32:08  <BenJackson> did that change the mu distribution a lot?  suddenly I'm alone at 87.5
2011-12-12T12:32:12  <BenJackson> I'm the only person in the 87s
2011-12-12T12:32:13  <avdg> doesn't mean they can't play games
2011-12-12T12:32:31  <BenJackson> yesterday I was so close to RVeerdonk I was bouncing back and forth every game
2011-12-12T12:33:29  * avdg wonders when xathis will play
2011-12-12T12:33:35  <Anilm3> haahaha
2011-12-12T12:33:37  <Anilm3> me too
2011-12-12T12:34:08  <a1k0n> i keep switching places with Memetix
2011-12-12T12:34:30  <a1k0n> so i guess it is indeed fair to say our approaches are broadly comparable
2011-12-12T12:34:45  <a1k0n> except his takes 1/1000th the CPU power
2011-12-12T12:34:49  <rwest> haha
2011-12-12T12:35:05  <McLeopold> a1k0n: I think he also has a lot of logic he didn't describe
2011-12-12T12:35:16  <a1k0n> perhaps.  i guess i do too
2011-12-12T12:35:55  <McLeopold> a1k0n: I implemented formations and shoehorned it into your method
2011-12-12T12:36:01  <McLeopold> it's helped out alot
2011-12-12T12:36:07  <a1k0n> formations eh
2011-12-12T12:36:10  <McLeopold> I don't have to do the dependant ant thing
2011-12-12T12:36:23  <a1k0n> meaning you consider moves of ants in formation together?
2011-12-12T12:36:32  <McLeopold> kind of
2011-12-12T12:36:43  <a1k0n> you're not going to submit your bot?
2011-12-12T12:36:50  <McLeopold> after moving 1 ant, I get a hash of only my ants and their formation
2011-12-12T12:36:54  <McLeopold> a1k0n: no
2011-12-12T12:36:56  <avdg> I wish I could do big array operations (preferably with a simple syntax)
2011-12-12T12:37:40  <McLeopold> for each formation formed by a single ant's best move, inc the counter by one
2011-12-12T12:37:42  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has joined #aichallenge
2011-12-12T12:37:46  <a1k0n> how do you represent formations?
2011-12-12T12:38:06  <McLeopold> then instead of returning individual moves, I return formations, with the instructions to reach them
2011-12-12T12:38:15  <rwest> I hate when I break my code completely
2011-12-12T12:38:19  *** pguillory <pguillory!62f8eec0@gateway/web/freenode/ip.98.248.238.192> has joined #aichallenge
2011-12-12T12:38:31  <McLeopold> this also help avoid ant swaping
2011-12-12T12:38:44  <McLeopold> it's all in python and very slow
2011-12-12T12:38:51  <McLeopold> a fomation is a tuple of tuples :)
2011-12-12T12:38:55  <BenJackson> I just accepted ant swapping
2011-12-12T12:38:58  <avdg> my implementation doesn't allow swapping
2011-12-12T12:39:06  <BenJackson> because it was going to be more expensive to eliminate it
2011-12-12T12:39:09  <a1k0n> could add a peephole optimizer to un-swap them :)
2011-12-12T12:39:17  <BenJackson> and sometimes a whole bunch of meaningless moves actually do something
2011-12-12T12:39:22  <avdg> except if I have a version to move 2 ants at the same time (which requires me to write an other function)
2011-12-12T12:39:44  <McLeopold> I do allow swaping, I just don't consider different ways to rotate into the same position as different
2011-12-12T12:40:07  <besh> My bot had more than 600 ants but fortunately it didn't time out. But all of its ants were piled up at choke points.
2011-12-12T12:40:12  <McLeopold> so, whichever random set of moves made the formation the first time, that is all I store
2011-12-12T12:40:14  <a1k0n> so i don't get it.  how do you decide which ants are part of which formation?
2011-12-12T12:40:19  <besh> How do you avoid that ?
2011-12-12T12:40:38  <McLeopold> a formation is a list of final destinations for my ants only
2011-12-12T12:40:41  *** GeorgeSebastian <GeorgeSebastian!~georgeseb@unaffiliated/georgesebastian> has joined #aichallenge
2011-12-12T12:40:44  <a1k0n> for all of your ants?
2011-12-12T12:40:55  <avdg> should work fine
2011-12-12T12:40:57  <McLeopold> oh, I separate ants into combat theaters
2011-12-12T12:41:00  <pairofdice> I was hoping you would say eigenvectors ;)
2011-12-12T12:41:01  <a1k0n> oh ok
2011-12-12T12:41:08  <a1k0n> oh shit that's a good idea, though
2011-12-12T12:41:08  *** ikaros <ikaros!~ikaros@studpool-wlan-74-86.fs.fbi.h-da.de> has quit IRC (Quit: Ex-Chat)
2011-12-12T12:41:10  <avdg> pairofdice: :p
2011-12-12T12:41:26  <McLeopold> a1k0n: it helped me solve this a a _ _ b
2011-12-12T12:42:02  <McLeopold> the left ant moving east is good most of the time, the right ant moving up or down to line up with it is good half the time
2011-12-12T12:42:19  <rwest> wow, I almost won a game with no food collecting code working
2011-12-12T12:42:20  <McLeopold> the same is with moving backwards, right goes west, left goes up or down
2011-12-12T12:42:22  <rwest> kinda funny
2011-12-12T12:42:45  <McLeopold> so for each individual ant, moving e or w was the best move, so they just swapped places
2011-12-12T12:42:55  <avdg> funny, my only cares about collecting food
2011-12-12T12:43:18  <rwest> avdg: I broke the base code that processed food tiles hah
2011-12-12T12:43:31  <pairofdice> I think it was spectral clustering....
2011-12-12T12:45:04  <a1k0n> McLeopold: ah yes, heh
2011-12-12T12:45:07  <BenJackson> is anyone using any more sophisticated distance heuristics than manhattan?
2011-12-12T12:45:14  <BenJackson> like waypoints or whatnot?
2011-12-12T12:46:01  <McLeopold> a1k0n: so, if a combat set is large enough, there is no way I can hit all the formations, and I might not hit a formation more than a couple of times
2011-12-12T12:46:11  <McLeopold> a1k0n: that's where your sampling method comes in
2011-12-12T12:46:18  *** Ashoka <Ashoka!70875b1a@gateway/web/freenode/ip.112.135.91.26> has quit IRC (Ping timeout: 258 seconds)
2011-12-12T12:46:40  <BenJackson> the pousse program I lost to was clever about searching the move tree
2011-12-12T12:46:53  <BenJackson> they hashed the whole board and if any tree hit that board again they knew the result already
2011-12-12T12:47:07  <a1k0n> zobrist hashing
2011-12-12T12:47:10  <BenJackson> there's a hashing function named aftloses
2011-12-12T12:48:00  <Anilm3> McLeopold: do you have a replay of your strategy?
2011-12-12T12:48:13  <McLeopold> Anilm3: yeah, hold on
2011-12-12T12:49:04  <jstrong> YESSS!!
2011-12-12T12:49:06  <jstrong> bug DOWN
2011-12-12T12:50:03  <jstrong> two big bugs down
2011-12-12T12:50:26  <Anilm3> xathis:  is taking a long time for his next game...
2011-12-12T12:50:31  <Anilm3> -:
2011-12-12T12:50:51  <BenJackson> jstrong: yesterday I found my entire exploration frontier was one square to the east of where it should have been
2011-12-12T12:50:54  <BenJackson> sadly not a game changer :)
2011-12-12T12:51:44  <avdg> I expect that xathis will play in around 20 min
2011-12-12T12:51:56  <jstrong> bj: hah nice
2011-12-12T12:52:08  <jstrong> bj: today I discovered that 1. i was double counting enemies for my new island system
2011-12-12T12:52:17  <jstrong> which is why the whole prediction went to the toilet and my bot skill went to the toilet as well
2011-12-12T12:52:17  <rwest> wow, I just sat here for 5 minutes debugging my combat
2011-12-12T12:52:27  <rwest> to realize I commented out the call to run it
2011-12-12T12:52:33  <jstrong> and 2. I was using information about ENEMY ANTS for combat prediction moves
2011-12-12T12:52:39  <jstrong> when i should have been suing the maps for MY ants
2011-12-12T12:52:53  <jstrong> creating stupid prediction moves for my enemies also screwing around with the payoff matrix
2011-12-12T12:53:09  <McLeopold> http://paste.aichallenge.org/Ek04K/
2011-12-12T12:53:44  <jstrong> which explains why i would get into so many 1v1s
2011-12-12T12:53:54  <jstrong> because the "offensive" move I'd predict for my enemy didn't actually pit him against me
2011-12-12T12:54:06  <jstrong> making the me defense he offense score be a nice 0
2011-12-12T12:54:09  <jstrong> meaning a neutral move
2011-12-12T12:54:16  <jstrong> meaning my bot would suicide into enemy ants sometimes
2011-12-12T12:56:40  <jstrong> now maybe this will make my bot worse
2011-12-12T12:56:44  <jstrong> since it will make it more cautious
2011-12-12T12:56:55  <jstrong> meaning i gotta tweak some of the other stuff to implement intentional aggressivity xD
2011-12-12T12:57:41  <Fluxid> <@amstan> Fluxid: 460ms, cutting it close eh? ← sorry... is that bad? i try to iterate until i find best solution, sometimes it hits time limit, when tried to limit it to loop count it failed hard...
2011-12-12T12:57:52  <amstan> Fluxid: no, i'm just commenting
2011-12-12T12:58:02  <Fluxid> k
2011-12-12T12:58:24  *** Israfel <Israfel!~jeff@99-0-14-82.lightspeed.frokca.sbcglobal.net> has quit IRC (Ping timeout: 268 seconds)
2011-12-12T12:58:45  <Fluxid> 18:16:18 < a1k0n> holy crap xathis resubmitted? ← HE GONNA BEAT OUR ASSES HARD
2011-12-12T12:58:50  <jstrong> wow its been 24 hours and my bot hasn't played a single game on official
2011-12-12T12:58:52  <Anilm3> hahaha
2011-12-12T12:59:02  <Anilm3> jstrong: you have to be the only one :P
2011-12-12T12:59:04  <pairofdice> jstrong: deactivated?
2011-12-12T12:59:14  <jstrong> pairofdice: that might've been it
2011-12-12T12:59:17  <jstrong> i thought bots auto reactivated
2011-12-12T12:59:20  <jstrong> when you uploaded new code
2011-12-12T12:59:24  <pairofdice> 10 games per minute atm
2011-12-12T12:59:31  <pairofdice> That's quite a bit
2011-12-12T12:59:40  <McLeopold> jstrong: 22 mintues ago?
2011-12-12T12:59:41  <pairofdice> They do
2011-12-12T13:00:02  <jstrong> acutally its -only- been 12 hrs
2011-12-12T13:00:10  <McLeopold> jstrong: what is your bot name?
2011-12-12T13:00:22  <jstrong> oh, I'm roflmao
2011-12-12T13:00:42  <jstrong> i just couldn't secure that nickname on nickserv and my irc client is too stupid to have alternate nicknames
2011-12-12T13:00:43  <McLeopold> 48 seconds ago :)
2011-12-12T13:00:46  <jstrong> soo i have to change it manually
2011-12-12T13:00:49  *** jstrong is now known as roflmao
2011-12-12T13:01:39  <pairofdice> What kind of a irc clinet is that. All I've ever tried have alt nicks
2011-12-12T13:01:47  <pairofdice> client*
2011-12-12T13:02:22  <pairofdice> Except maybe webclients
2011-12-12T13:02:37  <roflmao> well hang on
2011-12-12T13:02:44  <roflmao> by my "irc client is too stupid"
2011-12-12T13:02:57  <roflmao> i really mean "i'm too lazy to figure it out"
2011-12-12T13:03:23  <rwest> it's usually default to have a _ after your original nick
2011-12-12T13:03:30  *** McLeopold <McLeopold!~Scott@c-98-247-248-39.hsd1.wa.comcast.net> has quit IRC (Quit: Leaving.)
2011-12-12T13:05:42  <roflmao> it was under preferences (duh)
2011-12-12T13:06:47  *** bhasker <bhasker!~bhasker@99-162-151-97.lightspeed.sntcca.sbcglobal.net> has quit IRC (Ping timeout: 252 seconds)
2011-12-12T13:07:34  *** mj41 <mj41!~mj41_nby@195.39.106.218> has quit IRC (Ping timeout: 276 seconds)
2011-12-12T13:10:10  *** Israfel <Israfel!~jeff@99-0-14-82.lightspeed.frokca.sbcglobal.net> has joined #aichallenge
2011-12-12T13:11:13  <Anilm3> xathis got a new fight
2011-12-12T13:12:02  <avdg> 49 steps only :/
2011-12-12T13:12:04  <rwest> 49 turns lol
2011-12-12T13:13:32  * avdg looks back at his prediction :p
2011-12-12T13:13:43  <Fluxid> http://fluxid.pl/misc/2011-12-12-191327_250x183_scrot.png Why do i see this only with js disabled
2011-12-12T13:14:30  <avdg> my js is enabled, which browser are you using?
2011-12-12T13:14:43  <Anilm3> mine too
2011-12-12T13:14:46  <rwest> I see it with js enabled
2011-12-12T13:15:01  <avdg> and that logo is quite new
2011-12-12T13:15:05  <avdg> only a few days
2011-12-12T13:15:18  *** liberforce <liberforce!~lmenina@LLagny-156-36-6-198.w80-14.abo.wanadoo.fr> has left #aichallenge
2011-12-12T13:15:36  <Fluxid> it disappears immediately after page finishes loading, and stays when i disable js (Opera 11.60)
2011-12-12T13:15:53  <rwest> Fluxid: you have a plugin that does it maybe?
2011-12-12T13:16:06  <pairofdice> adblock or something?
2011-12-12T13:17:16  <rwest> I see it in opera with js enabled
2011-12-12T13:17:33  <Fluxid> hm... then "Thanks to additional server sponsor" shuldn't dissapear, but lemme check
2011-12-12T13:17:49  <Fluxid> right, adblock...
2011-12-12T13:18:04  <Fluxid> sorry!
2011-12-12T13:18:18  <Fluxid> anyway, i think xathis implemented enemy hill positon prediction
2011-12-12T13:18:31  <Anilm3> xathis: tell us your new secrets
2011-12-12T13:19:03  <rwest> Hey anybody tried seeing what happens when 3 ants move on the same tile?
2011-12-12T13:19:18  <pairofdice> nuclear fusion
2011-12-12T13:19:29  <avdg> whats next, four ants?
2011-12-12T13:19:30  <Fluxid> i'm pretty sure he did, his ants rush in group to hill that wasn't seen
2011-12-12T13:19:30  <avdg> :p
2011-12-12T13:19:52  <Fluxid> rwest: WHAT ABOUT SIX?
2011-12-12T13:20:45  <pairofdice> Will it blend?!
2011-12-12T13:20:54  <Fluxid> move six ants to the same tile and every enemy dies and you get +100 bonus points
2011-12-12T13:20:57  <rwest> pairofdice: I like the ipad one haha
2011-12-12T13:21:28  <rwest> Fluxid: I was thinking like what Darhuuk  said on forums, maybe you can take enemy hills like that if 1 ant is left
2011-12-12T13:21:55  <Fluxid> rwest: what do you mean?
2011-12-12T13:22:17  <Fluxid> you can raze hill even with ant on top of it
2011-12-12T13:22:18  <rwest> say the enemy hill has ants in it
2011-12-12T13:22:27  <rwest> maybe moving 2 ants onto it
2011-12-12T13:22:35  <rwest> will cause 2 to die from collision and leave you 1
2011-12-12T13:22:42  <rwest> to take it
2011-12-12T13:22:43  <Fluxid> you don't need to spawnkill all the ants in hill
2011-12-12T13:23:10  <rwest> I dunno if it works, that's why I asked if anybody has seen what happens when 3 ants stack
2011-12-12T13:23:51  <pairofdice> You do, unless the ant moves
2011-12-12T13:23:55  <pairofdice> Is how I understood it
2011-12-12T13:25:39  *** TheLinker <TheLinker!~TheLinker@200.117.99.103> has quit IRC (Quit: Bye)
2011-12-12T13:25:40  <|UncleVasya|> All ants are moved to their new positions.
2011-12-12T13:25:41  <|UncleVasya|> Any ants which occupy the same square are killed.
2011-12-12T13:25:55  <|UncleVasya|> rwest: this is a commentary from engine source code
2011-12-12T13:26:33  *** |UncleVasya| is now known as UncleVasya
2011-12-12T13:26:45  <Anilm3> Hi Uncle
2011-12-12T13:28:56  <rwest> UncleVasya: Yea I just read the source, it adds them to a list and kills them
2011-12-12T13:29:03  <rwest> after all moves
2011-12-12T13:29:17  <UncleVasya> Anilm3: Hi
2011-12-12T13:29:23  *** TheLinker <TheLinker!~TheLinker@200.117.99.103> has joined #aichallenge
2011-12-12T13:29:33  *** Jacob_Strauss <Jacob_Strauss!bc4eb12c@gateway/web/freenode/ip.188.78.177.44> has joined #aichallenge
2011-12-12T13:32:16  *** dvladim <dvladim!~dvladim@85.113.131.65> has quit IRC (Ping timeout: 276 seconds)
2011-12-12T13:32:18  <roflmao> hmm
2011-12-12T13:32:22  <roflmao> i have this weird problem with my islands
2011-12-12T13:32:26  <roflmao> for combat resolution
2011-12-12T13:32:31  <roflmao> its not....local enough
2011-12-12T13:32:37  <roflmao> i'm just using AR
2011-12-12T13:32:50  <roflmao> but it can create this kind of huge chain creating a huge global island
2011-12-12T13:32:56  <roflmao> that really contains like 3-4 mini conflicts inside
2011-12-12T13:32:58  <roflmao> :-\
2011-12-12T13:35:47  <ronchaine> Damn combat.  It's like pick two:  effective, original, not-slow. :|
2011-12-12T13:36:18  <UncleVasya> I've tuned my prunning hard so I do not afraid big fights anymore.
2011-12-12T13:36:27  <Fluxid> 59th place
2011-12-12T13:36:38  <UncleVasya> Even more, I like big fights more now.
2011-12-12T13:36:46  <avdg> picking c or c++ would get ride of at least 1 of them
2011-12-12T13:37:02  <avdg> at least, if programming isn't a problem
2011-12-12T13:37:07  <Fluxid> i think i'll take wednesday free at work and speedhack missing parts of my bot
2011-12-12T13:37:11  <UncleVasya> Or OCaml :D
2011-12-12T13:37:52  <Fluxid> guys, will you publish your code after submissions are closed?
2011-12-12T13:38:20  <UncleVasya> It is a honor to do so.
2011-12-12T13:38:27  <pairofdice> I hope some of the top bots do so I can have my revenge
2011-12-12T13:38:32  <avdg> I'll do it, though it will serve no purpose imo
2011-12-12T13:38:50  <Fluxid> i want to know the rocket science stuff behind top bots
2011-12-12T13:39:02  <Fluxid> (i'll publish mine too of course)
2011-12-12T13:39:11  <UncleVasya> But you also need to make some cleanup in code to make it look not as ugly as it is :)
2011-12-12T13:39:32  * avdg would probably use uglify.js :p
2011-12-12T13:39:33  <UncleVasya> And write several lines about your strategy...
2011-12-12T13:39:35  *** roflmao <roflmao!~jstrong@189.229.217.18> has quit IRC (Quit: Leaving.)
2011-12-12T13:39:45  <UncleVasya> Maybe I'll be too lazy for it :)
2011-12-12T13:39:54  <avdg> well, me too actually
2011-12-12T13:40:11  <avdg> but I already have written an api
2011-12-12T13:40:13  <avdg> wiki
2011-12-12T13:40:39  *** roflmao <roflmao!~jstrong@189.229.217.18> has joined #aichallenge
2011-12-12T13:40:40  <avdg> I can't believe how many small problems prevents me to build a good bot
2011-12-12T13:40:52  <BenJackson> roflmao: how are you doing "islands"?
2011-12-12T13:41:19  <roflmao> i compile a map of ints
2011-12-12T13:41:29  <roflmao> with all the tiles that are in the combat radius of any ant on the map
2011-12-12T13:41:30  <UncleVasya> btw, after last contest bots of different ranks were published. While being #310 I have found something interesting even in some bot with rank over 600.
2011-12-12T13:41:39  <roflmao> then I start a BFS from enemy tiles
2011-12-12T13:41:52  <roflmao> make sure they haven't been used for an already existing island
2011-12-12T13:42:06  <roflmao> and then ignore tiles that have a value of 0 on the map of ints
2011-12-12T13:42:11  <roflmao> (ergo they aren't part of a combat radius)
2011-12-12T13:42:17  <roflmao> so it grabs all clusters of combat radius
2011-12-12T13:42:23  <roflmao> .. but
2011-12-12T13:42:23  <roflmao> http://minus.com/m0NCqWkzA#1
2011-12-12T13:42:41  <roflmao> the pistachio ant way over on the left
2011-12-12T13:42:42  <avdg> UncleVasya: its also a different ranking system if I'm right, but still an interesting fact
2011-12-12T13:42:44  <roflmao> is part of that one huge combat
2011-12-12T13:42:54  <roflmao> since there is an actual link between all the radiuses of that entire area
2011-12-12T13:43:03  <roflmao> when in reality they are two completely different combat zones
2011-12-12T13:43:15  <roflmao> so my ants on the bottom wont attack because the ant dying on top is reducing the score of an offensive move
2011-12-12T13:43:27  <roflmao> (on the left better said)
2011-12-12T13:44:31  <roflmao> maybe I should use attackradius MINUS 1?
2011-12-12T13:44:34  <BenJackson> my method:
2011-12-12T13:44:36  <roflmao> for the int compilation
2011-12-12T13:44:40  <BenJackson> 1) find all enemy ants threatened by me
2011-12-12T13:44:48  <BenJackson> 2) find all my ants threatened by enemy
2011-12-12T13:45:13  <BenJackson> into a common map label each combat radius of both sets using union-find to merge overlaps
2011-12-12T13:45:56  <BenJackson> then for each label that was created, if it is the root of a union-find tree, pick all ants and enemies that are in that "island"
2011-12-12T13:46:37  <BenJackson> then for each label if there were any ants in it (some will not have ants because they're not the root of a tree) run combat just like I did before
2011-12-12T13:46:40  <BenJackson> but only considering those ants
2011-12-12T13:46:52  <roflmao> not sure what you mean by "label"
2011-12-12T13:47:29  *** tdubellz_ <tdubellz_!~tdubellz@freenode/staff/tdubellz> has joined #aichallenge
2011-12-12T13:47:45  <roflmao> ah i see
2011-12-12T13:47:49  <BenJackson> for every ant in combat (both sides): in a grid of ints assign 'n' to every square in its combat radius
2011-12-12T13:47:56  <BenJackson> if you find a pre-existing n, use union-find to merge it
2011-12-12T13:48:26  <BenJackson> you could do it other ways, union-find is just the "best" way
2011-12-12T13:48:39  <UncleVasya> roflmao: I thought about dividing big combats into subcombats.
2011-12-12T13:48:52  <BenJackson> but you could use maps or anything to say equiv[n] = some other n you found overlapping
2011-12-12T13:48:55  *** iglo <iglo!~xy@xdsl-92-252-113-46.dip.osnanet.de> has joined #aichallenge
2011-12-12T13:49:04  <BenJackson> so long as you have a way to find some COMMON n for ALL n that got unioned
2011-12-12T13:49:04  <avdg> I wonder if it wouldn't be interested to just create arrays of it with a number how high the threat would be
2011-12-12T13:49:26  <UncleVasya> It can increase your speed (at the cost of little decreasing of accuracy)/
2011-12-12T13:49:48  <avdg> it can be used for many things for simply 1 map
2011-12-12T13:49:54  <roflmao> right
2011-12-12T13:49:57  <roflmao> so the overlap of the combat zones
2011-12-12T13:50:04  <UncleVasya> But I found that optimizing move generator is much easier.
2011-12-12T13:50:25  <roflmao> using simple +1 on the int map this would mean a row/col value greater than 1
2011-12-12T13:50:35  <roflmao> (if you do it properly
2011-12-12T13:51:00  *** tdubellz_ is now known as tdubz
2011-12-12T13:51:12  <roflmao> and then you collect
2011-12-12T13:51:16  *** GeorgeSebastian <GeorgeSebastian!~georgeseb@unaffiliated/georgesebastian> has quit IRC (Read error: Connection reset by peer)
2011-12-12T13:51:17  <roflmao> ants that are on unioned ground?
2011-12-12T13:51:51  <roflmao> no that can't be right
2011-12-12T13:53:45  *** risico <risico!~risico@79.117.179.115> has joined #aichallenge
2011-12-12T13:55:01  *** Scryer <Scryer!4a645367@gateway/web/freenode/ip.74.100.83.103> has joined #aichallenge
2011-12-12T13:56:14  <Scryer> Woohoo - Xathis v2 is on its way!  Looking at his first two games, we may need to get a new category... instead of just number of games to get into the top few, the total number of *moves* to get there.
2011-12-12T13:58:01  *** tdubz <tdubz!~tdubellz@freenode/staff/tdubellz> has quit IRC (Read error: Connection reset by peer)
2011-12-12T13:58:24  <avdg> we are watching xathis very closely, probably even closer than you think Scryer ;-)
2011-12-12T13:58:55  <xathis> now i feel stalked
2011-12-12T13:59:38  <Scryer> Phear us. :)
2011-12-12T14:02:02  <pairofdice> haha
2011-12-12T14:04:02  *** flowenol <flowenol!LW@bl4-237-93.dsl.telepac.pt> has joined #aichallenge
2011-12-12T14:05:13  <BenJackson> Scryer: fast games aren't *that* impressive.  Against starter kits it's the turn you took your first hill
2011-12-12T14:06:47  <Anilm3> xathis: haha
2011-12-12T14:09:20  *** kire <kire!~kire@78-22-125-201.access.telenet.be> has joined #aichallenge
2011-12-12T14:09:32  *** TheLinker <TheLinker!~TheLinker@200.117.99.103> has quit IRC (Remote host closed the connection)
2011-12-12T14:10:33  <roflmao> bj one last question
2011-12-12T14:10:37  <roflmao> do assign a unique label to every ant?
2011-12-12T14:10:48  <roflmao> ergo three ants are threatened in the same battle
2011-12-12T14:10:58  <roflmao> so you assign labels '1' '2' and '3'
2011-12-12T14:11:04  <roflmao> then have 3 union finds that you go through
2011-12-12T14:11:18  <roflmao> since by definition for them to be in the same battle there is overlap between their combat zones
2011-12-12T14:11:27  <BenJackson> yes
2011-12-12T14:11:36  <BenJackson> during the labelling every ant is assigned a new id
2011-12-12T14:11:50  <BenJackson> as you iterate over its combat area you may find other ids
2011-12-12T14:11:53  <BenJackson> you union with those
2011-12-12T14:12:05  <roflmao> hm
2011-12-12T14:12:08  <BenJackson> union find has two operations, union and find (!)
2011-12-12T14:12:22  <roflmao> because i think it will still cause a union that I don't want
2011-12-12T14:12:25  <iglo> have they changed the timeout deadline?
2011-12-12T14:12:30  <BenJackson> so while labelling if grid(loc) is already labelled, union(my new label, grid(loc))
2011-12-12T14:12:46  <BenJackson> later for all ants, the group it's in is find(grid(loc))
2011-12-12T14:13:02  <BenJackson> iglo: no, but there are new timeout issues
2011-12-12T14:13:04  <BenJackson> see the forum
2011-12-12T14:13:07  *** rmmh <rmmh!~none@host-80-87.flgaroy.clients.pavlovmedia.com> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T14:14:33  <iglo> which topic?
2011-12-12T14:15:02  *** rmmh <rmmh!~none@host-80-87.flgaroy.clients.pavlovmedia.com> has joined #aichallenge
2011-12-12T14:19:10  <UncleVasya> maybe we should change topic to something like 'xathis v2 released! *link to profile*' :D
2011-12-12T14:19:34  <thestinger> lol :)
2011-12-12T14:20:04  *** aarossig <aarossig!~andrew@209-195-125-110.cpe.distributel.net> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T14:20:06  <avdg> or a bot mentioning when xathis played a game (or any top 5 players?)
2011-12-12T14:20:09  <thestinger> his bot that's been on fluxid's tcp server is probably pretty close to that
2011-12-12T14:20:45  <thestinger> seems to care about hills a lot more than his last one
2011-12-12T14:22:10  <g0llum> btw, how does the contestbot get its rangings ?
2011-12-12T14:22:30  <avdg> g0llum: check microsofts trueskill system
2011-12-12T14:22:49  <UncleVasya> In his next version he will teach his ants to spawn directly on the enemy hill.
2011-12-12T14:22:50  <g0llum> no, i mean, where does it get the data from ?
2011-12-12T14:23:04  <avdg> I think game position
2011-12-12T14:23:17  <BenJackson> UncleVasya: at one point I wondered what the game would be like with "powerups"
2011-12-12T14:24:49  * avdg guesses many people should look at xathis stats in 10 min
2011-12-12T14:25:29  <xathis> my profile page says 5 mins remaining
2011-12-12T14:25:30  <BenJackson> he'll win at least 4 in a row
2011-12-12T14:25:30  <g0llum> next contest should come with some rest-api for bots, twitter feeds, etc
2011-12-12T14:25:37  <BenJackson> I think all of my bots have won at least 4 or 5 in a row
2011-12-12T14:25:52  <BenJackson> xathis can probably win 6 and get in the top 100 that way
2011-12-12T14:25:57  <BenJackson> he'll be in the top 100 by 7 games for sure
2011-12-12T14:26:02  <BenJackson> not going to be too interesting until then
2011-12-12T14:27:08  <UncleVasya> it's like a totalizator :D
2011-12-12T14:27:20  <pairofdice> A what?
2011-12-12T14:27:59  <UncleVasya> wait a sec, I'll open my vocabulary
2011-12-12T14:28:11  <pairofdice> Terminator?
2011-12-12T14:28:27  *** Hexren <Hexren!~hexren@46.59.205.23> has joined #aichallenge
2011-12-12T14:28:30  *** Hexren <Hexren!~hexren@46.59.205.23> has left #aichallenge
2011-12-12T14:29:21  <UncleVasya> no
2011-12-12T14:29:21  <UncleVasya> A place where people make stackes
2011-12-12T14:29:22  <UncleVasya> like '100$ for case when horse a come first'
2011-12-12T14:29:22  <UncleVasya> * horse A
2011-12-12T14:29:24  <g0llum> more like T1001
2011-12-12T14:29:44  <BenJackson> gambling
2011-12-12T14:29:49  <BenJackson> bookmaking
2011-12-12T14:29:53  <BenJackson> wagering
2011-12-12T14:29:57  <BenJackson> placing bets
2011-12-12T14:30:11  <avdg> I wish I had money to do that ;-)
2011-12-12T14:30:12  <pairofdice> Ahh
2011-12-12T14:30:27  <pairofdice> Casino
2011-12-12T14:30:29  <UncleVasya> casino?
2011-12-12T14:30:43  <BenJackson> a casino is not really associated with wagering on events
2011-12-12T14:30:51  <BenJackson> at a casino you play games mostly against the house
2011-12-12T14:31:05  <BenJackson> bookies (bookmakers) take bets on arbitrary events
2011-12-12T14:31:11  <BenJackson> like who will win a game
2011-12-12T14:31:14  <pairofdice> Yeah, more some sports events
2011-12-12T14:31:25  <avdg> hmm 3 min earlier than normal
2011-12-12T14:33:14  <besh> are there strong bots left that have never been submitted to the official tournament ?
2011-12-12T14:33:27  <avdg> meh, xathis almost had world domination (1 bot alive at the end of the game)
2011-12-12T14:33:42  <besh> I see too many of them on fluxid with weird names ..
2011-12-12T14:34:15  *** JorgeB <JorgeB!~textual@209.66.74.34> has quit IRC (Ping timeout: 252 seconds)
2011-12-12T14:35:33  *** Paradoxial <Paradoxial!~Paradoxia@pool-108-28-22-94.washdc.fios.verizon.net> has joined #aichallenge
2011-12-12T14:36:44  <pguillory> i think xathis' submission will win straight games until it hits the top 10
2011-12-12T14:39:41  <Anilm3> haha
2011-12-12T14:40:35  <xathis> anyone know who plays as Q on the fluxid tcp?
2011-12-12T14:41:00  <Anilm3> Q is an omnipotent being
2011-12-12T14:41:23  <Anilm3> from the Q continuum
2011-12-12T14:41:28  <Anilm3> you can't win against him
2011-12-12T14:41:36  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has quit IRC (Ping timeout: 255 seconds)
2011-12-12T14:41:40  <Anilm3> I think he shouldn't be allowed to participate
2011-12-12T14:41:51  <Anilm3> but how could you refuse...
2011-12-12T14:42:33  <Fluxid> because Q is near A i guess it's FlagCapper ;P
2011-12-12T14:42:36  <Fluxid> but no idea
2011-12-12T14:42:49  <Fluxid> In Game: Playing in a game right now.
2011-12-12T14:42:53  <Fluxid> wish me luck
2011-12-12T14:42:56  <Anilm3> gl
2011-12-12T14:43:07  <Fluxid> f5f5f5f5
2011-12-12T14:43:13  <Anilm3> haha
2011-12-12T14:43:49  <Fluxid> i want to know if my newest submission is better than previous one
2011-12-12T14:43:50  <UncleVasya> popcorn, F5 �  InsaneMalkavian
2011-12-12T14:44:08  <UncleVasya> it never happens
2011-12-12T14:44:16  <Anilm3> haha
2011-12-12T14:46:31  <Fluxid> fffffffffffoiled again
2011-12-12T14:46:55  <Anilm3> what?
2011-12-12T14:47:05  *** aarossig <aarossig!~andrew@209-195-125-110.cpe.distributel.net> has joined #aichallenge
2011-12-12T14:47:06  <UncleVasya> this is simple formula
2011-12-12T14:47:51  <Fluxid> fourth place in 9p map, but another game immediately after soooo
2011-12-12T14:47:54  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has joined #aichallenge
2011-12-12T14:48:47  <avdg> xathis: do you have a dead-ant detector?
2011-12-12T14:49:06  <UncleVasya> if you have a good submission then you will beat it not earlier than cur_submission_num + N where n is 2 or more.
2011-12-12T14:49:19  <xathis> avdg: you mean bots that timeouted?
2011-12-12T14:49:23  <avdg> yeah
2011-12-12T14:50:08  <xathis> avdg: i look for ants that stay on the same position for multiple turns
2011-12-12T14:50:20  <UncleVasya> xathis: I'm sure somebody already asked but will you share your code after deadline?
2011-12-12T14:50:36  <xathis> UncleVasya: i will, but it's a mess
2011-12-12T14:50:57  <Fluxid> it doesn't matter!
2011-12-12T14:51:03  <avdg> ah, that would probably explain why you only attack certain groups of ants, but not all ants of 1 enemie
2011-12-12T14:51:23  <UncleVasya> on Planet Wars one _very_ interesting bot wasn't open sourced because guys who made it used some of their commerial secrets in it.
2011-12-12T14:52:14  <Zannick> i remember that
2011-12-12T14:52:23  <Zannick> that was the genetic programming bot
2011-12-12T14:52:32  <Zannick> *a* genetic programming bot, i suppose
2011-12-12T14:52:51  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has quit IRC (Ping timeout: 252 seconds)
2011-12-12T14:53:07  <Fluxid> http://aichallenge.org/visualizer.php?game=208209&user=757 proud to not die
2011-12-12T14:54:50  *** Accoun <Accoun!~1@static-213-88-39-97.chebnet.ru> has quit IRC ()
2011-12-12T14:55:09  <thestinger> I love the early my jump when I start a new bot on tcp
2011-12-12T14:55:15  <ronchaine> Someone tell me why this line doesn't work: if ( state->map[nloc.row][nloc.col].ant > 0 )
2011-12-12T14:55:38  <ronchaine> I get inside that if statement if the value is -1.
2011-12-12T14:55:46  <ronchaine> I fixed it by adding 2 to both sides.
2011-12-12T14:55:52  <ronchaine> But I'm still curious.
2011-12-12T14:56:27  <thestinger> http://ants.fluxid.pl/player/strcat_stupid hehe, early luck :)
2011-12-12T14:58:02  *** bhasker <bhasker!~bhasker@173-164-245-33-SFBA.hfc.comcastbusiness.net> has joined #aichallenge
2011-12-12T15:02:48  *** ikaros <ikaros!~ikaros@dslb-084-059-073-057.pools.arcor-ip.net> has joined #aichallenge
2011-12-12T15:05:05  *** sigh <sigh!~sigh@ppp114-153.static.internode.on.net> has joined #aichallenge
2011-12-12T15:05:10  *** bergmark_ <bergmark_!bergmark@nat/ibm/x-momdrvrazyijhztf> has quit IRC (Ping timeout: 248 seconds)
2011-12-12T15:08:48  *** treeform <treeform!~treeform@108.60.121.46> has joined #aichallenge
2011-12-12T15:09:05  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has joined #aichallenge
2011-12-12T15:12:03  *** Accoun <Accoun!~1@static-213-88-39-97.chebnet.ru> has joined #aichallenge
2011-12-12T15:12:44  *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge
2011-12-12T15:14:22  <UncleVasya> xathis: Have you found too that removing repeated positions of ants in battle is more time-consumig than process them again?
2011-12-12T15:15:04  <UncleVasya> like [30;22] [30;21] and [30;21] [3022]
2011-12-12T15:15:31  <_flag> Hello everyone
2011-12-12T15:15:52  <Anilm3> Hi _flag, are you Q on tcp?
2011-12-12T15:15:55  <UncleVasya> hi
2011-12-12T15:16:16  <UncleVasya> it will be good if you will answer my question too :)
2011-12-12T15:16:16  <xathis> UncleVasya: the way i implemented it removing those doesn't cost any time, it's a single if statement
2011-12-12T15:16:34  <_flag> Anilm3: I am
2011-12-12T15:16:46  <Anilm3> everybody's been asking
2011-12-12T15:16:59  <xathis> oh nice to know, _flag
2011-12-12T15:17:27  <_flag> xathis: I see you've been giving me a bit of a beating :(
2011-12-12T15:18:38  <xathis> _flag: i'm sorry ;)
2011-12-12T15:19:00  <xathis> i really hope all those timeouts are tcp-specific
2011-12-12T15:19:34  <_flag> _flag: I think they are, I timeout on tcp often but when I feed the input into my bot it does just fine
2011-12-12T15:19:44  *** TheLinker <TheLinker!~TheLinker@host80.200-117-105.telecom.net.ar> has joined #aichallenge
2011-12-12T15:19:49  <xathis> _flag: that's good
2011-12-12T15:22:36  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T15:26:55  <Fluxid> http://aichallenge.org/visualizer.php?game=208027&user=4513 almost extermination... but not so exciting
2011-12-12T15:29:00  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has joined #aichallenge
2011-12-12T15:29:02  <UncleVasya> when I'll share my code you will find probably the most commented bot source. Because it has several hundrets of *commented out* code ><
2011-12-12T15:29:24  *** bhasker <bhasker!~bhasker@173-164-245-33-SFBA.hfc.comcastbusiness.net> has quit IRC (Ping timeout: 244 seconds)
2011-12-12T15:29:55  *** sigh <sigh!~sigh@ppp114-153.static.internode.on.net> has quit IRC (Remote host closed the connection)
2011-12-12T15:31:21  <thestinger> UncleVasya: use version control :P
2011-12-12T15:33:13  *** Redgis <Redgis!~Nex@2a01:e34:ee3b:aa40:ccb:8a1d:94b0:4c6> has joined #aichallenge
2011-12-12T15:35:06  <pairofdice> How I Learned to Stop Worrying and Love the Git
2011-12-12T15:37:03  <amstan> lol
2011-12-12T15:39:07  <BenJackson> xathis: that's one of those tricks that as soon as I thought of it I spotted other top bots doing it
2011-12-12T15:39:22  <BenJackson> oops paged up more than I thought
2011-12-12T15:39:26  <BenJackson> that was re: attacking stationary bots
2011-12-12T15:40:43  <UncleVasya> So to win _xathis you must be able to pretend dead.
2011-12-12T15:40:54  <pairofdice> Trap sprung
2011-12-12T15:41:08  <xathis> that might work, go for it! :P
2011-12-12T15:41:31  <BenJackson> I actually have a small random component in that code
2011-12-12T15:41:36  <BenJackson> to avoid someone trying to game it
2011-12-12T15:42:19  *** Kurnevsky1 <Kurnevsky1!~Errins@217.21.41.205> has left #aichallenge
2011-12-12T15:43:31  <UncleVasya> xathis: Isn't it a secret how many kb is your source code?
2011-12-12T15:45:04  <xathis> UncleVasya: 7 classes, 100kb total
2011-12-12T15:45:29  <Garf> thats not much (certainy given that it must be java)
2011-12-12T15:46:05  <Fluxid> I'm sad, my new "combat" works rather poorly, and i don't know it needs tweaking or scrap whole design and restart
2011-12-12T15:46:42  <Anilm3> poor Fluxid
2011-12-12T15:46:55  <besh> Use your old combat code then ;)
2011-12-12T15:47:15  <Fluxid> my old combat isn't good enough
2011-12-12T15:47:33  <Fluxid> (old == current submission)
2011-12-12T15:48:30  <Garf> I spent quite some time implementing xathis-like map control. Now my mu dropped because my ants are getting isolated and killed off.
2011-12-12T15:48:40  <Garf> "almost but not quite" :P
2011-12-12T15:48:46  <besh> fluxid: what are you trying to implement in the new one ? my combat don't have probablities that a lot other bots seem to have
2011-12-12T15:49:25  <besh> my combat works best on timed out bots :) In other words it assumes they are static.
2011-12-12T15:50:05  <thestinger> Garf: yeah, my combat isn't good enough to pull off map control
2011-12-12T15:50:24  <Antimony> me too
2011-12-12T15:50:34  <Antimony> it turns into a genius agaisnt tmied out bots
2011-12-12T15:51:03  <Fluxid> besh: i try to predict how likely enemy will move in given area in given direction
2011-12-12T15:51:06  <Fluxid> and try to predict how much my ant actually can do in killing enemy
2011-12-12T15:51:26  <Fluxid> or something like that
2011-12-12T15:51:32  <BenJackson> my code is 84k, which is more than I expected
2011-12-12T15:52:02  *** epicmonkey <epicmonkey!~epicmonke@178.130.5.103> has joined #aichallenge
2011-12-12T15:52:11  <thestinger> gzipped source size is more meaningful
2011-12-12T15:52:13  <Fluxid> BenJackson: loc or chars?
2011-12-12T15:52:32  <BenJackson> 3k LOC 84kB
2011-12-12T15:52:40  <Fluxid> my sloccount: python:         933 (55.83%) ansic:          738 (44.17%)
2011-12-12T15:52:53  <thestinger> BenJackson: mine is pretty much 50% of that :P
2011-12-12T15:53:09  <thestinger> 1600 lines, 60kB
2011-12-12T15:53:10  <UncleVasya> 69.2kb current submission,  79.6kb coming version.
2011-12-12T15:53:11  <thestinger> so not really
2011-12-12T15:53:14  <BenJackson> the C++ starter is like 1000 by itself
2011-12-12T15:53:22  <thestinger> yeah but I deleted almost all of that
2011-12-12T15:53:57  <UncleVasya> 16.7 kb is my starterpack
2011-12-12T15:55:39  <besh> I would like to see some discussion on exploration instead of combat in the forum.
2011-12-12T15:55:54  <Antimony> My bot is 114kb in 25 files, but there's a lot of duplicated or unused code
2011-12-12T15:56:18  <BenJackson> I have mostly deleted my unused code
2011-12-12T15:56:23  <BenJackson> but I do have ifdef'd out visualizations
2011-12-12T15:56:36  <Antimony> For example, my bot has at least 2 BFSs and 2A*s in it
2011-12-12T15:56:54  <pairofdice> Holy ****
2011-12-12T15:56:55  *** ltriant <ltriant!~ltriant@lithium.mailguard.com.au> has joined #aichallenge
2011-12-12T15:56:57  <UncleVasya> http://paste.aichallenge.org/DzIEp/
2011-12-12T15:57:00  <Antimony> what?
2011-12-12T15:57:03  <BenJackson> I have 2 BFS but no A*
2011-12-12T15:57:33  <Fluxid> I have dragons
2011-12-12T15:57:33  <Antimony> the problem is that with slight differences between each search, it's easier to just duplicate the code, then to try to generalize it
2011-12-12T15:57:45  <pairofdice> Yeah, true
2011-12-12T15:57:55  <Antimony> my food search and hill search in particular are near copies
2011-12-12T15:58:01  <BenJackson> when my primary bfs got too many options I made it a template that took functors
2011-12-12T15:58:09  <BenJackson> and my code actually got smaller *and* faster
2011-12-12T15:58:23  <Antimony> I guess I should try that
2011-12-12T15:58:26  <BenJackson> I only have a second BFS because the second one is a C++ iterator
2011-12-12T15:58:47  <BenJackson> so Bfs<PassabilityFunctor> begin(loc), end
2011-12-12T15:58:53  <thestinger> Antimony: that's what higher order functions are for :P
2011-12-12T15:59:02  <Antimony> yeah but I'm lazy
2011-12-12T15:59:08  <thestinger> same with me
2011-12-12T15:59:11  <BenJackson> yeah if I was using std::functions then I could probably fold all mine into 1
2011-12-12T15:59:20  <BenJackson> because instead of an iterator I'd just pass the code as an argument
2011-12-12T15:59:41  *** besh <besh!835e7324@gateway/web/freenode/ip.131.94.115.36> has quit IRC (Quit: Page closed)
2011-12-12T15:59:43  <BenJackson> although the iterator is fast for what it does
2011-12-12T16:00:09  <BenJackson> it can tell you the distance and direction to the source but it doesn't have to record most of that like the map-making one does
2011-12-12T16:00:25  <Garf> xathis on TCP is timing out
2011-12-12T16:00:53  <ikaros> tcp has many timouts
2011-12-12T16:01:05  *** UncleVasya <UncleVasya!kvirc@31-12-143-177.dialup.umc.net.ua> has quit IRC (Read error: Connection reset by peer)
2011-12-12T16:01:09  <Antimony> I also have duplicated combat graph code because one version uses positions and one uses integer ids
2011-12-12T16:01:40  <Antimony> I should probably just stick with ints and make code to reassign ids to a graph
2011-12-12T16:02:54  <Antimony> is anyone using boost?
2011-12-12T16:03:43  <avdg> I would, but its making me more confused than its being helpful atm
2011-12-12T16:04:40  <avdg> and I think someone else on aichallenge's forum wanted to use boost as well (look it up)
2011-12-12T16:04:56  <ikaros> c++11 is kind of a boost :)
2011-12-12T16:05:05  <avdg> http://forums.aichallenge.org/viewtopic.php?f=26&t=1718&p=11954&hilit=boost#p11954
2011-12-12T16:05:20  <avdg> meh, I need a new laptop
2011-12-12T16:05:32  <ikaros> but yesterday i wanted to submit my first c++11 bot and had to realize the server uses gcc4.5 which doesnt implement lots of the nice stuff
2011-12-12T16:05:34  <avdg> osx is hurting me more than helping as well
2011-12-12T16:07:04  <iglo> hmm i like to implement a combat algo... but it would take too long time :/
2011-12-12T16:07:14  <Anilm3> avdg: linux power
2011-12-12T16:07:44  <avdg> Anilm3: yeah, I was thinking that too
2011-12-12T16:07:56  <Anilm3> you don't need to buy a new laptop for that!
2011-12-12T16:08:17  <avdg> no, but its too messy to do atm
2011-12-12T16:08:31  <Anilm3> Tell me about it, I have a mac
2011-12-12T16:08:56  <Anilm3> but I used osx for a while and hated it
2011-12-12T16:09:05  <avdg> my disk is almost full, and I have a server I wanted to install which has enough hd to store a backup
2011-12-12T16:09:31  <avdg> well, I loved osx simply because it was better than windows and easier than linux
2011-12-12T16:09:43  <avdg> but now I learned more, linux became more easy to use
2011-12-12T16:10:33  <Anilm3> avdg: in my experience
2011-12-12T16:10:40  <Anilm3> since 2005 or so
2011-12-12T16:10:55  <Anilm3> linux has not been difficult
2011-12-12T16:11:00  <Anilm3> before that, maybe
2011-12-12T16:11:11  <Anilm3> before that I had only tried sucky distros
2011-12-12T16:11:18  <avdg> :-)
2011-12-12T16:11:50  <Anilm3> also be careful with kde
2011-12-12T16:11:57  <Anilm3> and unity
2011-12-12T16:12:12  <Anilm3> they will scare the hell out of you!!
2011-12-12T16:12:27  <avdg> meh, I just need a browser and a console ;-)
2011-12-12T16:12:29  <ikaros> i dont like what they did to gnome too :)
2011-12-12T16:12:49  <Anilm3> ikaros: you mean the gnome shell stuff??
2011-12-12T16:12:53  <Anilm3> I don't like it either
2011-12-12T16:12:54  <ikaros> gnome3
2011-12-12T16:13:01  <ikaros> im glad my debian still has gnome2
2011-12-12T16:13:09  <avdg> yeah, I have plans for a new gui, but I'm always bad in implementing stuff (check my ai as proof)
2011-12-12T16:13:16  <ikaros> but on my laptop i have arch and there is gnome 3
2011-12-12T16:13:27  <avdg> and I have to learn everything on my own
2011-12-12T16:13:30  <ikaros> guess ill have to switch to xfce or such
2011-12-12T16:13:41  <Anilm3> go back to gnome2
2011-12-12T16:13:41  <Anilm3> !
2011-12-12T16:13:41  <ikaros> not sure yet
2011-12-12T16:13:44  <ikaros> well
2011-12-12T16:14:00  <avdg> maybe archlinux :p
2011-12-12T16:14:00  <ikaros> that's not so easy if your distro has switched to gnome 3
2011-12-12T16:14:33  <ikaros> im waiting for some guys that will fork gnome 2 and continue it :)
2011-12-12T16:14:37  <ikaros> cant take that long lol
2011-12-12T16:14:38  <avdg> (though I really would keep debian packages)
2011-12-12T16:14:51  <ikaros> debian and arch are both great
2011-12-12T16:14:59  <avdg> meh, I just hope I can create my own gui
2011-12-12T16:15:19  <avdg> I'm too good in theory, and bad in doing it
2011-12-12T16:15:24  <Anilm3> what do you mean create your own gui?
2011-12-12T16:15:28  <ikaros> i like how bleeding edge the arch packages are.. but it breaks more often :)
2011-12-12T16:15:56  <Anilm3> ikaros: me too, debian and arch are two sides of the same coin
2011-12-12T16:15:59  <avdg> I simply need a search bar to launch my app (and nothing more), a way to switch apps and my console :p
2011-12-12T16:16:01  <ikaros> if i wouldnt be so lazy i should probably use some *box and compiz on top
2011-12-12T16:16:04  <avdg> thats all I need
2011-12-12T16:16:09  <thestinger> arch only breaks if you break it :P
2011-12-12T16:16:20  <ikaros> hehe not really :)
2011-12-12T16:16:31  <ikaros> upgrades are enough sometimes :)
2011-12-12T16:16:41  <thestinger> I've run [testing] for 2 years and never had breakage
2011-12-12T16:16:48  <thestinger> you have to read the mailing list tho
2011-12-12T16:16:49  <ikaros> fascinating :)
2011-12-12T16:17:12  <ikaros> well.. i've had lots of -Syu which caused alot of trouble
2011-12-12T16:17:21  <ikaros> can be fixed always so no problem
2011-12-12T16:17:29  *** bergmark <bergmark!bergmark@nat/ibm/x-qaeyzjfoiememrdj> has joined #aichallenge
2011-12-12T16:17:55  <g0llum> @seen mcstar
2011-12-12T16:17:55  <contestbot> g0llum: mcstar was last seen in #aichallenge 2 weeks, 0 days, 22 hours, 11 minutes, and 58 seconds ago: <mcstar> i dont agree
2011-12-12T16:18:01  <avdg> :p
2011-12-12T16:18:18  <ikaros> but that are experiences that are a few years old i think
2011-12-12T16:18:25  <ikaros> didnt use arch for at least 3 years
2011-12-12T16:18:32  <thestinger> ah
2011-12-12T16:18:38  <thestinger> it's quite a bit different than it used to be
2011-12-12T16:18:43  <ikaros> until some weeks ago :) i decided to give it another shot on my laptop
2011-12-12T16:19:00  <thestinger> every package in [core] has to get sign-offs before they move it from [testing]
2011-12-12T16:19:02  <ikaros> funny thing was when i installed it switched to testing immediately
2011-12-12T16:19:09  <thestinger> so there's really no breakage in the core packages
2011-12-12T16:19:13  <ikaros> wanted to update with pacman 4
2011-12-12T16:19:24  <thestinger> pacman 4 is nice :)
2011-12-12T16:19:26  <ikaros> and the packages were signed and i hadnt the keys
2011-12-12T16:19:39  <ikaros> there was a script on the wiki to import them..
2011-12-12T16:19:43  <thestinger> yeah, there's no keyring package so it has to fetch them
2011-12-12T16:19:47  <thestinger> you don't need to use the script
2011-12-12T16:19:47  <ikaros> but it used elinks that wasnt installed lol
2011-12-12T16:19:53  <thestinger> just need to change the setting in pacman.conf
2011-12-12T16:19:55  <ikaros> was kind of funny to import this by hand..
2011-12-12T16:19:58  <ikaros> yea
2011-12-12T16:20:02  <ikaros> i realised that later
2011-12-12T16:20:03  <ikaros> lol
2011-12-12T16:20:40  <thestinger> imo, Ubuntu and Arch are probably equally unstable at any point in time
2011-12-12T16:20:44  <ikaros> thestinger, what wm do you use?
2011-12-12T16:20:55  <thestinger> but with Arch you deal with the upgrades slowly over time, not a massive change every 6 months
2011-12-12T16:20:58  <QuirionPT> is everyone allowed to participate in the final tournament or is it limited to TOP X players?
2011-12-12T16:20:58  <thestinger> ikaros: i3
2011-12-12T16:21:01  <ikaros> ubuntu has become a mess (imho)
2011-12-12T16:21:03  <ikaros> oh ok :)
2011-12-12T16:21:05  <thestinger> used awesomewm until recently
2011-12-12T16:21:11  <thestinger> i3 is amazing though :D
2011-12-12T16:21:19  <ikaros> when they changed to unity it became dirty
2011-12-12T16:21:27  <ikaros> unstable^3 unity is...
2011-12-12T16:21:37  <ikaros> QuirionPT, everyone
2011-12-12T16:21:42  <QuirionPT> ikaros, thanks
2011-12-12T16:21:43  <thestinger> I haven't really used Ubuntu since like 9.10
2011-12-12T16:21:59  <thestinger> I just installed it in a VM to see if a java zero-day worked in linux
2011-12-12T16:22:06  <thestinger> that was a few days ago
2011-12-12T16:22:08  <ikaros> its  installed on my gfs laptop :)
2011-12-12T16:22:34  <thestinger> in the 10 mins I used it, I ran into like 3 unity bugs
2011-12-12T16:22:52  <thestinger> and I checked out the software center thing, but it crashed when I tried to remove libreoffice
2011-12-12T16:23:02  <thestinger> so yeah... seems like they have taken a step backwards
2011-12-12T16:23:10  <QuirionPT> I have been using solely Ubuntu for the last 2 years. Has been quite okay and was a great help to avoid playing games and focusing on working/studying :)
2011-12-12T16:23:50  <thestinger> QuirionPT: so you haven't found dwarf fortress yet? :P
2011-12-12T16:24:14  <ikaros> hrhr
2011-12-12T16:24:41  <thestinger> I still have windows 7 on a drive but I haven't booted into it for a year
2011-12-12T16:24:41  <ikaros> doomrl for the coffeebreaks :)
2011-12-12T16:24:51  <ikaros> i had to some weeks ago
2011-12-12T16:24:58  <ikaros> and it needed a half day to install updates..
2011-12-12T16:26:53  *** Palmik <Palmik!~palmik@man-208-200.pks.muni.cz> has quit IRC (Remote host closed the connection)
2011-12-12T16:27:11  <avdg> meh, I need to do a major cleanup on my computer, I almost have no free space
2011-12-12T16:28:42  <QuirionPT> thestinger, doesn't seem to be my thing, fortunately! :)
2011-12-12T16:30:13  <pairofdice> There's no guarantee that finals include everyone
2011-12-12T16:30:42  *** mj41 <mj41!~mj41_nby@82.100.0.58> has joined #aichallenge
2011-12-12T16:31:04  <g0llum> pairofdice, explain
2011-12-12T16:31:18  <BenJackson> the final finals
2011-12-12T16:31:31  <BenJackson> they'll implement a cutoff after the finals run for a while
2011-12-12T16:31:51  <BenJackson> no need to churn the starter kits against each other for 100s of games
2011-12-12T16:32:02  <g0llum> ok.
2011-12-12T16:32:45  <ikaros> i would guess there will be a resubmission ? eliminating the starter kits
2011-12-12T16:35:00  * avdg wonders how long it takes to do 150 turns with relative fast starter bots
2011-12-12T16:37:58  <g0llum> how to eliminate the starter kits, crc32, md5hash, unsolved problem, afiik
2011-12-12T16:38:17  *** jab_bott <jab_bott!5163fc1b@gateway/web/freenode/ip.81.99.252.27> has joined #aichallenge
2011-12-12T16:38:38  <jab_bott> eh up chucks - I still haven't implemented combat yet
2011-12-12T16:39:14  <Anilm3> haha
2011-12-12T16:39:22  <pairofdice> g0llum, by cutting low mu, low sigma bots
2011-12-12T16:39:54  * avdg wonders how many games it would take to get a reliable mu and sigma score
2011-12-12T16:40:02  <janzert> for the finals the lower ranks are slowly eliminated
2011-12-12T16:40:05  <jab_bott> um.. so what's the deal? if I get something working by the 18th can I still submit it? or will it need to have time to rise in the ranks in order to take part or anything like that?
2011-12-12T16:40:27  <janzert> all rankings are completely reset at the beginning of finals
2011-12-12T16:40:28  <pairofdice> I don't think it will have to rise
2011-12-12T16:40:40  <avdg> jab_bott: there is an hour given on the homepage
2011-12-12T16:40:45  <janzert> so it doesn't matter where in the rankings you are at the time submissions close
2011-12-12T16:41:24  <janzert> just whatever your last submission is, is what will be used in the finals
2011-12-12T16:41:32  <ikaros> janzert, do we have to resubmit after deadline?
2011-12-12T16:41:33  <Anilm3> janzert: is the deadline fixed or can it be changed?
2011-12-12T16:41:36  <ikaros> ah
2011-12-12T16:41:39  <ikaros> forget that :)
2011-12-12T16:41:47  <janzert> the deadline is fixed
2011-12-12T16:41:48  <jab_bott> ok, phew, so I can frantically submit this week and try to improve it until Sunday!
2011-12-12T16:41:53  <Anilm3> :(
2011-12-12T16:42:02  <antimatroid1> janzert: are we resetting all the rankings? people were talking about just messing with the mu's the other day?
2011-12-12T16:42:16  <janzert> yes, their will be a complete reset
2011-12-12T16:42:25  <antimatroid1> good :)
2011-12-12T16:42:38  <jab_bott> I suppose I'd better be sure to not break something at the last minute
2011-12-12T16:42:39  <avdg> tip: just send your bot as early as possible, so most bugs/mistakes are out once the deadline passes ;-)
2011-12-12T16:43:12  <janzert> yeah, the very last submission made for planetwars ended up not compiling
2011-12-12T16:43:19  <janzert> poor guy
2011-12-12T16:43:20  <antimatroid1> counter tip: don't come up with a neat idea for your game tree like 7 hours before submissions close and then submit a buggy version of it
2011-12-12T16:43:40  <ikaros> :)
2011-12-12T16:43:55  <avdg> :p
2011-12-12T16:43:58  <pairofdice> counter-counter tip: please do exactly that
2011-12-12T16:44:01  <antimatroid1> i don't think my bot was actually any worse though, if anything it might have been better
2011-12-12T16:44:05  <jab_bott> janzert: wouldn't it revert to the last-known-good (i.e. compiled and passed tests) version in that case?
2011-12-12T16:44:29  <janzert> in planetwars it didn't, I think it might now but I'm not positive
2011-12-12T16:44:40  <jab_bott> I really wish I'd worked on combat much sooner now so I could have more time for feedback and refinement
2011-12-12T16:44:51  <janzert> don't remember for sure if that changed or not
2011-12-12T16:45:14  * avdg wonders if the site could crash just before the deadline
2011-12-12T16:45:22  *** bergmark <bergmark!bergmark@nat/ibm/x-qaeyzjfoiememrdj> has quit IRC (Ping timeout: 252 seconds)
2011-12-12T16:45:27  <jab_bott> I'm gonna submit the C-version of my bot now to see if it even compiles on the servers!
2011-12-12T16:45:32  <Garf> avdg: or overload, like tax-on-web :P
2011-12-12T16:45:33  <avdg> there are too many server crashes at ai-class already
2011-12-12T16:45:47  <avdg> garf: are you belgian?
2011-12-12T16:45:55  <avdg> ah you are
2011-12-12T16:46:11  <Garf> yes, thats why I made that reference
2011-12-12T16:46:21  <avdg> :-)
2011-12-12T16:48:21  <jab_bott> ahh.. another ai-class student :)
2011-12-12T16:48:31  <avdg> just a watcher
2011-12-12T16:48:39  <avdg> I'm too frustrated to study well
2011-12-12T16:48:41  <avdg> :-)
2011-12-12T16:49:03  <jab_bott> that's how I heard about this aichallenge, through searching since I've been doing the ai-class
2011-12-12T16:49:28  <avdg> I heared on twitter
2011-12-12T16:49:34  <jab_bott> my bot doesn't exactly employ much of the stuff from ai-class though
2011-12-12T16:50:12  <avdg> too many people promising they will complete the course, but I didn't hear any1 once week 2 started
2011-12-12T16:50:14  <pairofdice> Oh right, deadline in 2 hours
2011-12-12T16:50:17  <jab_bott> the c-version is now super-fast! I'm struggling with making my ants spread out though, they keep getting stuck to water
2011-12-12T16:50:36  <jab_bott> avdg: I'm still on for 100% on the course
2011-12-12T16:50:44  <pairofdice> Yeah 100% here too
2011-12-12T16:50:50  <pairofdice> It's a good course
2011-12-12T16:51:04  <avdg> I'm still ready to listen, I'm lacking math skills to complete the course
2011-12-12T16:51:09  <jab_bott> I got 98% and 95% on two of the homeworks, but they discard the two worst scores, so technically I could still get 100
2011-12-12T16:51:26  <pairofdice> There's hardly any math required, some basic division in Bayes
2011-12-12T16:51:37  <avdg> probability kills me
2011-12-12T16:52:00  <avdg> and rarely time
2011-12-12T16:52:07  <avdg> to study it peaceful
2011-12-12T16:52:20  <avdg> (though I have no job atm)
2011-12-12T16:53:05  <jab_bott> I keep getting annoyed by poorly specified questions, like that's the minimum value for something, and the *real* correct answer is that the value must be >14 and the answer used for marking is 15 - but nobody said it must be an integer - things like taht
2011-12-12T16:53:34  <avdg> yeah, that was especially at the start
2011-12-12T16:54:15  *** risico <risico!~risico@79.117.179.115> has quit IRC (Quit: Leaving)
2011-12-12T16:55:12  <jab_bott> I guess the videos take long to make and are hard to fix because they sometimes leave wrong information up and just put some text down the bottom to ay actually the video is wrong in some way
2011-12-12T16:55:21  <antimatroid1> "what's the minimum value of X" I would expect to be giving a unique element of X as an answer to that
2011-12-12T16:56:09  <jab_bott> antimatroid1: if X is a set, yes, but if it's a value...
2011-12-12T16:56:56  *** Andrew_Makar <Andrew_Makar!bca36c05@gateway/web/freenode/ip.188.163.108.5> has joined #aichallenge
2011-12-12T16:57:20  <antimatroid1> by X i mean the set of values they're referring to :P
2011-12-12T16:57:31  <antimatroid1> that meet the given condition
2011-12-12T16:57:32  *** Andrew_Makar <Andrew_Makar!bca36c05@gateway/web/freenode/ip.188.163.108.5> has quit IRC (Client Quit)
2011-12-12T16:57:37  *** g0llum <g0llum!~g0llum@10.106.113.82.net.de.o2.com> has quit IRC (Read error: Connection reset by peer)
2011-12-12T16:57:45  <antimatroid1> anyway, i'm off to submit a phd application
2011-12-12T16:58:05  *** g0llum <g0llum!~g0llum@10.106.113.82.net.de.o2.com> has joined #aichallenge
2011-12-12T16:58:30  <jab_bott> antimatroid1: show-off :P
2011-12-12T16:58:31  * avdg wishes he could study on a more peaceful location
2011-12-12T16:58:47  <avdg> and a little bit help :p
2011-12-12T16:59:07  <avdg> I hate being the smartest person in house
2011-12-12T16:59:47  <avdg> nobody to share ideas
2011-12-12T17:00:09  <Minthos> welcome to my life
2011-12-12T17:00:32  <Minthos> it's not that I'm smarter than everyone else, it's just that usually even the smart people have no clue what I'm talking about
2011-12-12T17:01:08  <avdg> oh well, I'm not smart on everything as well
2011-12-12T17:01:41  *** bergmark <bergmark!bergmark@nat/ibm/x-tecjkmzbspuwdzjy> has joined #aichallenge
2011-12-12T17:01:46  <avdg> just 108 iq, but yeah, iq doesn't tell a lot
2011-12-12T17:02:52  *** mj41 <mj41!~mj41_nby@82.100.0.58> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T17:03:40  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T17:04:55  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has joined #aichallenge
2011-12-12T17:13:42  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has quit IRC (Ping timeout: 255 seconds)
2011-12-12T17:29:36  *** Anilm3 <Anilm3!~anil@150.Red-83-44-168.dynamicIP.rima-tde.net> has quit IRC (Quit: Lost terminal)
2011-12-12T17:33:14  *** g0llum <g0llum!~g0llum@10.106.113.82.net.de.o2.com> has quit IRC (Read error: Connection reset by peer)
2011-12-12T17:37:09  <jab_bott> by the way - I've been testing my bot on this map: maps/cell_maze/cell_maze_p04_14.map - and I've noticed a significant advantage to playing in player-3 position - when I play against my older bots I can only just beat them when they are in spot 3 and when my latest bot plays in spot 3 it kicks-ass
2011-12-12T17:37:43  <jab_bott> aren't they supposed to be equal?
2011-12-12T17:38:31  <amstan> jab_bott: you have something wrong with your code then
2011-12-12T17:38:33  <pairofdice> The maps are equal, but your bots might not be
2011-12-12T17:38:48  <pairofdice> ie, the bots have a bias
2011-12-12T17:39:13  <amstan> jab_bott: your bot should play exactly the same from all positions given that you have the same opponents in the other positions(and given that they play the same too)
2011-12-12T17:39:21  *** bergmark <bergmark!bergmark@nat/ibm/x-tecjkmzbspuwdzjy> has quit IRC (Ping timeout: 255 seconds)
2011-12-12T17:39:44  <thestinger> jab_bott: you probably have a bias in how you pick directions
2011-12-12T17:40:22  <janzert> or a problem crossing the map wrap in the negative direction
2011-12-12T17:40:37  *** gazpachoking <gazpachoking!~gazpachok@c-67-163-203-66.hsd1.pa.comcast.net> has joined #aichallenge
2011-12-12T17:41:06  <janzert> well, I guess it's furthest from the vertical negative wrap but not the horizontal one
2011-12-12T17:41:14  <amstan> jab_bott: in your language, what does -3%10 equal to?
2011-12-12T17:41:16  <jab_bott> I pick directions by picking a random direction, then cycling from there until a max-desirable direction is found, so if none is found the left-over direction is random
2011-12-12T17:41:30  <BenJackson> I may have finally built a bot capable of consistently beating my previous version
2011-12-12T17:41:50  <jab_bott> amstan: for wrapping I do (x + width) % width and (y + height) % height
2011-12-12T17:41:56  *** ikaros <ikaros!~ikaros@dslb-084-059-073-057.pools.arcor-ip.net> has quit IRC (Quit: Ex-Chat)
2011-12-12T17:41:58  <amstan> that should be fine
2011-12-12T17:42:28  <BenJackson> I've run into maps like that
2011-12-12T17:42:33  <BenJackson> where the symmetry is all "north", too
2011-12-12T17:42:39  <BenJackson> so there's basically no difference
2011-12-12T17:42:43  <BenJackson> never did figure out why
2011-12-12T17:43:12  <jab_bott> do all bots start with the same player_seed?
2011-12-12T17:43:21  <BenJackson> they do if you pass it to the engine
2011-12-12T17:43:23  <BenJackson> not sure otherwise
2011-12-12T17:44:15  <BenJackson> wow 6 in a row and xathis is only 902nd
2011-12-12T17:44:28  <BenJackson> just didn't luck out and get a strong opponent to boost him up
2011-12-12T17:45:47  <jab_bott> ugh - my new C bot crashed in the test cases!
2011-12-12T17:46:19  <jab_bott> I thought I'd fixed all the bugs, must be a few lurking still - can I run the test cases locally so I can run it under valgrind?
2011-12-12T17:47:30  *** iglo <iglo!~xy@xdsl-92-252-113-46.dip.osnanet.de> has quit IRC (Remote host closed the connection)
2011-12-12T17:48:32  *** mj41 <mj41!~mj41_nby@82.100.0.58> has joined #aichallenge
2011-12-12T17:52:05  <thestinger> jab_bott: libmudflap is nicer than valgrind
2011-12-12T17:52:39  <pairofdice> There's the test_bot.sh/.cmd
2011-12-12T17:52:45  <jab_bott> thestinger: oh - never heard of it - will have to take a look
2011-12-12T17:52:56  <jab_bott> pairofdice: cheers - just found it :P
2011-12-12T17:53:12  *** ikaros <ikaros!~ikaros@dslb-084-059-073-057.pools.arcor-ip.net> has joined #aichallenge
2011-12-12T17:54:00  <thestinger> jab_bott: http://sprunge.us/gdEg
2011-12-12T17:54:04  <thestinger> that's how you use it
2011-12-12T17:54:29  <thestinger> valgrind doesn't catch that one
2011-12-12T17:54:45  <BenJackson> oops, tcpclient ate all my fd's again
2011-12-12T17:54:49  <thestinger> it basically catches _all_ out of bounds array access
2011-12-12T17:54:54  <thestinger> valgrind just catches some
2011-12-12T17:56:05  <thestinger> I guess it doesn't notice stuff on the stack like that
2011-12-12T17:56:34  <BenJackson> @seen McLeopold
2011-12-12T17:56:34  <contestbot> BenJackson: McLeopold was last seen in #aichallenge 4 hours, 55 minutes, and 50 seconds ago: <McLeopold> 48 seconds ago :)
2011-12-12T17:57:10  *** mj41 <mj41!~mj41_nby@82.100.0.58> has quit IRC (Ping timeout: 244 seconds)
2011-12-12T17:57:43  <thestinger> jab_bott: http://gcc.gnu.org/wiki/Mudflap_Pointer_Debugging there are a dozen switches to mess around with but it has pretty sane default, and it comes with gcc
2011-12-12T17:58:01  <thestinger> sadly it doesn't really work with C++
2011-12-12T17:58:14  *** Jacob_Strauss <Jacob_Strauss!bc4eb12c@gateway/web/freenode/ip.188.78.177.44> has quit IRC (Quit: Page closed)
2011-12-12T18:01:24  <jab_bott> thestinger: cool - I may try to get clang working on it too
2011-12-12T18:01:32  *** bluegaspode <bluegaspode!55b22105@gateway/web/freenode/ip.85.178.33.5> has joined #aichallenge
2011-12-12T18:01:33  *** pairofdice <pairofdice!~pod@dsl-hkibrasgw2-ff67c300-50.dhcp.inet.fi> has quit IRC (Quit: leaving)
2011-12-12T18:06:39  <Fluxid> http://aichallenge.org/visualizer.php?game=209877&user=757 this game stabilized pretty quickly
2011-12-12T18:08:17  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has joined #aichallenge
2011-12-12T18:09:03  <avdg> just afraid to attack :/
2011-12-12T18:11:41  <ikaros> yellow has a bit of traffic jam there lol
2011-12-12T18:12:47  *** hkraal_ <hkraal_!~hkraal@85.158.251.215> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T18:12:55  *** hkraal_ <hkraal_!~hkraal@85.158.251.215> has joined #aichallenge
2011-12-12T18:14:24  *** kire <kire!~kire@78-22-125-201.access.telenet.be> has quit IRC (Quit: Leaving)
2011-12-12T18:15:05  <amstan> lol! ai-class has a question: "Did you get this?"
2011-12-12T18:15:11  <amstan> and if you answer no, it says it's wrong, lol
2011-12-12T18:16:41  <avdg> yeah :-)
2011-12-12T18:17:02  <avdg> luckely I tried wrong at take 2
2011-12-12T18:17:11  <avdg> (well I hope it won't count in the results)
2011-12-12T18:17:53  <amstan> quizes don't count for your mark
2011-12-12T18:18:21  <avdg> well, I heard they record the first guess
2011-12-12T18:18:36  <ronchaine> Unless you're in basic class, it doesn't count towards your final score.
2011-12-12T18:18:37  <amstan> they do, but only for that progress thing
2011-12-12T18:19:29  <amstan> most of those quizes i fail at first try, because maybe i didn't get the question, have no idea how to do the question or i'm just too lazy to do it
2011-12-12T18:19:53  <amstan> they should really use the last attempt instead
2011-12-12T18:20:35  <ronchaine> Then you could just answer anything->see explanation->answer correct.
2011-12-12T18:21:07  <avdg> you don't learn much that way :p
2011-12-12T18:21:19  <amstan> yes, but i don't see how the quiz mark is relevant if some of the questions are designed to be impossible to do with no previous knowlege
2011-12-12T18:21:55  <avdg> they sometimes "measure" your preknowledge or thinking
2011-12-12T18:21:58  <amstan> if they want to have it as a student tool, students should be able to use it as they want, as a way to track progress of stuff they understand at present
2011-12-12T18:22:13  <amstan> not your knowlege before you were even taught the material
2011-12-12T18:22:48  *** Anilm3 <Anilm3!~anil@150.Red-83-44-168.dynamicIP.rima-tde.net> has joined #aichallenge
2011-12-12T18:30:30  *** Redgis <Redgis!~Nex@2a01:e34:ee3b:aa40:ccb:8a1d:94b0:4c6> has quit IRC (Read error: Connection reset by peer)
2011-12-12T18:32:17  *** Kaji <Kaji!koen@oege.ie.hva.nl> has quit IRC (Ping timeout: 252 seconds)
2011-12-12T18:32:47  *** epicmonkey <epicmonkey!~epicmonke@178.130.5.103> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T18:34:40  *** epicmonkey <epicmonkey!~epicmonke@178.130.5.103> has joined #aichallenge
2011-12-12T18:37:08  *** Redgis <Redgis!~Nex@2a01:e34:ee3b:aa40:ccb:8a1d:94b0:4c6> has joined #aichallenge
2011-12-12T18:41:58  *** Redgis <Redgis!~Nex@2a01:e34:ee3b:aa40:ccb:8a1d:94b0:4c6> has quit IRC (Ping timeout: 252 seconds)
2011-12-12T18:43:22  *** Kaji <Kaji!koen@oege.ie.hva.nl> has joined #aichallenge
2011-12-12T18:44:15  *** Anilm3 <Anilm3!~anil@150.Red-83-44-168.dynamicIP.rima-tde.net> has quit IRC (Quit: Lost terminal)
2011-12-12T18:47:27  *** epicmonkey <epicmonkey!~epicmonke@178.130.5.103> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T18:54:21  <BenJackson> amstan: when I was still following ai-class I got to the point where I just put in "0" for every quiz
2011-12-12T18:54:43  <amstan> BenJackson: done that a few times, to find out how to actually solve something
2011-12-12T18:54:47  <BenJackson> at least the ones that seemed to involve too much calculation for mid-lecture
2011-12-12T18:55:14  <BenJackson> I was down on the quality of the ai-class lectures and then a friend at Stanford showed me the REAL video lectures
2011-12-12T18:55:20  <BenJackson> then I was REALLY down on the ai-class ones
2011-12-12T18:55:39  <avdg> hmm, are these videos public?
2011-12-12T18:55:43  <BenJackson> no
2011-12-12T18:55:51  <BenJackson> the urls are not locked down, but they're not advertised
2011-12-12T18:55:52  <amstan> that would have been nice
2011-12-12T18:56:05  <BenJackson> same topics, same guys, but with prepared slides
2011-12-12T18:56:07  <BenJackson> working examples
2011-12-12T18:56:10  <amstan> still, the idea of most classes is that you use outside resources too, like books and stuff
2011-12-12T18:56:10  <avdg> well, I know they released only some of the videos
2011-12-12T18:56:15  <BenJackson> and a class of students pointing out confusing/wrong things
2011-12-12T18:56:18  <amstan> i have a bad habit of not using books
2011-12-12T18:56:19  <BenJackson> also high def
2011-12-12T18:56:22  <BenJackson> and not interrupted by quizzes
2011-12-12T18:56:29  <avdg> :p
2011-12-12T18:56:42  <BenJackson> I need to keep an eye on the used book price
2011-12-12T18:56:45  <avdg> or with other words, its a better way to fall asleep
2011-12-12T18:56:50  <BenJackson> I'm hoping it will crash after ai-class so I can pick it up
2011-12-12T18:57:57  <amstan> i'm still undecided what course to do next term
2011-12-12T18:58:29  <BenJackson> people who took ml-class really liked it
2011-12-12T18:58:34  <BenJackson> but there's overlap with ai-class
2011-12-12T18:58:54  <rwest_> is there any way to put the topics on the forum in some kind of logical order?
2011-12-12T18:59:01  <rwest_> like by date of last reply?
2011-12-12T18:59:05  <amstan> i eventually want to learn that stuff properly, but not for any short term stuff
2011-12-12T18:59:08  <Garf> BenJackson: yeah, I tried both, and quickly picked ml in favor of ai
2011-12-12T18:59:11  <BenJackson> http://forums.aichallenge.org/search.php?search_id=newposts
2011-12-12T18:59:16  <amstan> i want something that can benefit me in short term
2011-12-12T18:59:29  <Garf> ml-class pacing, questions, exercises are excellent imho
2011-12-12T18:59:37  <rwest_> BenJackson: I guess that works, wish it was just normal lol
2011-12-12T18:59:44  <Garf> wish it went deeper and broader (even if not in the class, but with pointers) sometimes
2011-12-12T19:00:22  <BenJackson> rwest_: I happened to just click on that when you asked
2011-12-12T19:01:34  <amstan> there was this webpage with all the free stanford courses for next term, anyone have a link?
2011-12-12T19:01:43  <amstan> there was also a subreddit on it, which might have that link
2011-12-12T19:02:08  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T19:03:16  <avdg> the teacher of ml-class is also caring a lot for quality and ways to educate stuff the "new" way ;-)
2011-12-12T19:03:40  <avdg> I'm sure he spend easily hours with it
2011-12-12T19:03:41  <BenJackson> so if I had stuck on my previous submission it would have slipped from top 20 to mid 40s by now
2011-12-12T19:04:06  <amstan> i just got a 50 in my latest homework, that was unexpected
2011-12-12T19:04:14  <avdg> :p
2011-12-12T19:04:47  <avdg> hmm, ai-class doesn't load
2011-12-12T19:05:45  <Garf> I guess Stanford can look at retention rates for both classes and draw a conclusion
2011-12-12T19:06:34  <avdg> :-) looks like I have 4/6 right (I didn't answer 2/6)
2011-12-12T19:06:36  <amstan> avdg: the homework just closed
2011-12-12T19:06:45  <amstan> so yeah, it's pretty slow
2011-12-12T19:07:43  * avdg wonders why we should learn the position of standford's team :p
2011-12-12T19:08:40  <avdg> oh, I did make mistakes, I looked at it wrong
2011-12-12T19:09:11  <avdg> dynamic state -> 1 under :/
2011-12-12T19:09:35  <amstan> come on... both particle questions i got wrong
2011-12-12T19:09:44  <amstan> and even #7
2011-12-12T19:09:50  <amstan> where i made a python program run it
2011-12-12T19:10:33  <amstan> avdg: did you get #7?
2011-12-12T19:10:38  <avdg> no
2011-12-12T19:11:06  <avdg> I didn't do anything that required a lot of math (on first view)
2011-12-12T19:11:28  <avdg> only 1, 2, 3 and 6
2011-12-12T19:11:43  <rwest_> http://www.bhickey.net:2080/replay.8921
2011-12-12T19:11:44  <amstan> oh man! i did get it wrong, i only did 15 steps, not 16
2011-12-12T19:11:51  <amstan> http://pastebin.com/BRQ4BY7W
2011-12-12T19:11:55  <rwest_> lol @ my ants near my hill at the end
2011-12-12T19:12:57  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has joined #aichallenge
2011-12-12T19:13:27  <avdg> rwest: too scared to kill?
2011-12-12T19:13:55  <rwest_> my combat is completely reactive
2011-12-12T19:13:56  <rwest_> haha
2011-12-12T19:14:07  <rwest_> I still need to fix that
2011-12-12T19:15:38  <avdg> well, if you count the coverage of enemies on 1 map and have a system to cover at least 2 ants on each "safe" direction, you would have a pretty good one to start with ;-)
2011-12-12T19:16:21  <rwest_> the main issue is I don't calculate safe moves only unsafe ones really
2011-12-12T19:17:55  <avdg> meh I wouldn't implement it in my current bot, because js is too slow to do it
2011-12-12T19:20:01  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has quit IRC (Ping timeout: 276 seconds)
2011-12-12T19:22:29  <BenJackson> js?
2011-12-12T19:22:52  <BenJackson> I guess we're getting to the point where we could look at the highest scoring bots for the obscure languages
2011-12-12T19:23:15  <rwest_> javascript I would assume
2011-12-12T19:23:18  <avdg> yeah, I really hoped it would be fast enough
2011-12-12T19:23:22  <rwest_> guy I work with is doing it in tcl
2011-12-12T19:23:32  <BenJackson> not that much variety in the top 100
2011-12-12T19:23:33  <avdg> and I'm not good enough to do it in c++
2011-12-12T19:23:41  <BenJackson> tcl is crazy talk
2011-12-12T19:23:49  <rwest_> That's what I said
2011-12-12T19:23:58  <rwest_> Our whole automation system is in tcl but still
2011-12-12T19:24:12  <BenJackson> modelsim?
2011-12-12T19:24:29  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has joined #aichallenge
2011-12-12T19:24:44  <rwest_> no
2011-12-12T19:24:47  <rwest_> home-grown
2011-12-12T19:25:00  <BenJackson> that's the only commercial system I know of where tcl took off the way it was intended
2011-12-12T19:25:32  <rwest_> http://www.bhickey.net:2080/replay.8925
2011-12-12T19:25:34  <BenJackson> a lonnng time ago I wrote a 3d wireframe tool in tcl/wish
2011-12-12T19:25:36  <rwest_> new explore code works
2011-12-12T19:25:37  <rwest_> lol
2011-12-12T19:25:39  <BenJackson> but I did implement matrix extensions in C
2011-12-12T19:26:35  <Antimony> oh hi ben
2011-12-12T19:26:45  <Antimony> Is your syemmtry system similar to the one on the forums?
2011-12-12T19:27:35  <BenJackson> nope!
2011-12-12T19:27:45  <BenJackson> I'd call that the brute force method
2011-12-12T19:28:00  <BenJackson> very nice visualization of how it works, though
2011-12-12T19:29:11  <Antimony> oh, so there's a better way? I was doing brute force too
2011-12-12T19:29:26  <avdg> rwest_: nice distribution :p
2011-12-12T19:29:39  <Antimony> hmm, looks like there's a bug in my stored food calculation
2011-12-12T19:30:02  <avdg> what kind of bug?
2011-12-12T19:30:49  <BenJackson> I implemented the calculation but still haven't used it
2011-12-12T19:31:37  <avdg> I have implemented it in my c++ bot, but not tested it yet, its currently underestimating
2011-12-12T19:32:11  <Antimony> In this replay, my bot thinks white has an extra food starting around turn 123http://aichallenge.org/visualizer.php?game=210389&user=3875
2011-12-12T19:32:16  <BenJackson> mine resets if it doesn't see an ant appear at a hill for a few turns
2011-12-12T19:32:31  <BenJackson> if it thinks it has food and no ant appears it thinks it knows my hills are dead
2011-12-12T19:32:34  <Antimony> mine concludes that a hill was razed because it expected spawning
2011-12-12T19:32:34  <BenJackson> but it also doesn't use that
2011-12-12T19:32:55  <Antimony> I guess I should make it fail safe, rather than fail deadly
2011-12-12T19:32:56  <avdg> hmm overestimating?
2011-12-12T19:33:06  <avdg> thats dangerous
2011-12-12T19:33:08  <Antimony> that's what I'm trying to figure out
2011-12-12T19:33:17  <Antimony> there must be a bug somewhere
2011-12-12T19:34:33  <avdg> do you take the number of spawned ants into account?
2011-12-12T19:34:42  <cyphase> http://codesprint.interviewstreet.com/recruit/challenges/
2011-12-12T19:34:44  *** replore_ <replore_!~replore@203.152.213.161.static.zoot.jp> has joined #aichallenge
2011-12-12T19:34:58  <avdg> you probably have to debug it with tcp and debug information
2011-12-12T19:35:27  <cyphase> "CodeSprint is a coding contest designed to streamline the job hunt for programmers. Instead of applying to 65 companies individually, work your noggin on a single set of interesting and tough coding challenges in a 48 hour period to get interview calls from these awesome companies!"
2011-12-12T19:36:06  * avdg doesn't believe in job hunts :-)
2011-12-12T19:36:17  <BenJackson> "login to see the fall 2011 problems"  mmmm... no.
2011-12-12T19:37:33  <avdg> most recruiters don't hire me, because they don't understand my brain
2011-12-12T19:38:07  <amstan> avdg: get in the top 100 for aichallenge and you'll notice quite a few people wanting to talk to you
2011-12-12T19:38:27  <avdg> oh, I already got my plans :p
2011-12-12T19:38:31  <Garf> top 100 is hard :P
2011-12-12T19:38:43  <BenJackson> get in the top 10 and they'll be afraid of you
2011-12-12T19:38:43  <avdg> and I have too many issues with my system to do good coding
2011-12-12T19:39:04  <amstan> well, that thing could probably work for less too
2011-12-12T19:39:14  <amstan> also.. contributing to the contest code is not a bad thing either
2011-12-12T19:39:30  <amstan> if the employers heard about it
2011-12-12T19:39:31  <avdg> :p I only added "extermination"
2011-12-12T19:39:45  <avdg> 2 loc
2011-12-12T19:40:42  <avdg> I'm probably better in finding bugs
2011-12-12T19:41:44  *** ltriant <ltriant!~ltriant@lithium.mailguard.com.au> has quit IRC (Read error: No route to host)
2011-12-12T19:42:07  <Garf> <amstan> avdg: get in the top 100 for aichallenge and you'll notice quite a few people wanting to talk to you   <- there's no way to find out who's who, right?
2011-12-12T19:42:13  <QuirionPT> is there a way to upload replays?
2011-12-12T19:42:24  <avdg> garf: just add a link?
2011-12-12T19:42:39  <amstan> Garf: ideally you want to use a name on the servers that you are known for
2011-12-12T19:42:42  <Garf> I mean for people who didn't specifically put something down
2011-12-12T19:42:52  <amstan> but url is good too
2011-12-12T19:43:16  <amstan> in the past years jeff sent the top epsilon to google
2011-12-12T19:43:19  <amstan> with their permission
2011-12-12T19:46:05  *** ltriant <ltriant!~ltriant@lithium.mailguard.com.au> has joined #aichallenge
2011-12-12T19:46:07  <avdg> google would probably a dream
2011-12-12T19:46:29  <avdg> but I know someone who rejected a job offer from google (but I also know why)
2011-12-12T19:46:40  <amstan> why?
2011-12-12T19:46:42  <Garf> 'cos he didn't want to move?
2011-12-12T19:46:53  <avdg> because he want his own business :p
2011-12-12T19:47:38  <avdg> now he is co-owner of a new business specialized in mobile applications
2011-12-12T19:49:49  <amstan> cool
2011-12-12T19:52:44  <bugnuts> is there a way to get the version of Go installed on the workers?
2011-12-12T19:58:11  <Wraithan> bugnuts: http://aichallenge.org/starter_packages.php
2011-12-12T19:58:12  <amstan> bugnuts: check setup/worker_setup.py
2011-12-12T19:58:22  <amstan> actually.. that link is easier
2011-12-12T19:58:32  <Wraithan> bugnuts: the site says what version you'll be running against
2011-12-12T19:58:46  <BenJackson> hey xathis played some "real" games
2011-12-12T19:58:51  <BenJackson> beat chrish
2011-12-12T19:59:19  <bugnuts> the version that reports isn't really a standard release afaict
2011-12-12T19:59:54  <bugnuts> I will have a look at worker_setup
2011-12-12T20:01:33  <QuirionPT> wow, the updateVision() function from the c++ starter package is really terrible :o
2011-12-12T20:01:51  <BenJackson> yep
2011-12-12T20:01:52  <QuirionPT> I was using it without much changes and now I realized it is taking more than half of the whole cpu usage
2011-12-12T20:01:53  <Antimony> yeah
2011-12-12T20:01:57  <BenJackson> yep
2011-12-12T20:02:01  <amstan> QuirionPT: antimatroid1 made it probably
2011-12-12T20:02:14  <Antimony> The C++ starter in general isn't very good
2011-12-12T20:02:20  <BenJackson> yep
2011-12-12T20:02:22  <BenJackson> :_
2011-12-12T20:02:22  <Antimony> by the way, did you get my updated timer?
2011-12-12T20:02:23  <BenJackson> )
2011-12-12T20:02:43  <QuirionPT> I am not sure
2011-12-12T20:02:43  <pguillory> BenJackson: ChrisH is having timing issues or something
2011-12-12T20:03:08  * avdg checks c++ timer
2011-12-12T20:03:29  <bugnuts> amstan: is there a way I can pull down the golang.deb you are installing...
2011-12-12T20:03:29  <pguillory> BenJackson: you can see a lot of it happening in that game, ants not moving
2011-12-12T20:03:48  <amstan> bugnuts: https://github.com/downloads/aichallenge/aichallenge/golang_60.1-9753~natty1_amd64.deb
2011-12-12T20:03:59  <bugnuts> excellent tyvm
2011-12-12T20:04:06  <avdg> Antimony: are you storyyeller?
2011-12-12T20:05:01  <Antimony> yes, why?
2011-12-12T20:05:04  <Antimony> oh on github
2011-12-12T20:05:06  <Antimony> yeah
2011-12-12T20:05:13  <BenJackson> pguillory: ah, didn't watch the game yet
2011-12-12T20:05:39  <Antimony> Ok I figured out where the discrepancy in stored food count came from
2011-12-12T20:05:45  <avdg> Antimony: you probably have to add a commit where you remove a file
2011-12-12T20:06:01  <Antimony> My bot moved an ant onto the hill, but the actual bot moved it away, so when a new ant spawned, my bot didn't recognize it as new
2011-12-12T20:06:24  <Antimony> oh ok
2011-12-12T20:06:44  <Antimony> I'm new to git. Why doesn't it automatically remove the old version if you rename?
2011-12-12T20:06:57  *** roflmao <roflmao!~jstrong@189.229.217.18> has quit IRC (Quit: Leaving.)
2011-12-12T20:07:07  <avdg> git only knows about renames by guessing and comparing files, renaming in git equals 2 "internal" operations (well, to keep things abstracted and simple)
2011-12-12T20:07:10  <BenJackson> rename in git is add+delete
2011-12-12T20:07:15  <Antimony> oh
2011-12-12T20:07:20  <BenJackson> renames are detected dynamically every time (every time you merge, log, etc)
2011-12-12T20:07:35  <BenJackson> there's a shortcut git command that does both steps for you at once
2011-12-12T20:07:41  <Wraithan> git mv
2011-12-12T20:08:35  <Antimony> I guess stored food detection is difficult if the moves being made are not what you expect
2011-12-12T20:08:49  <Antimony> at least it's not a bug in my boy
2011-12-12T20:08:52  <Antimony> my bot
2011-12-12T20:09:01  <Wraithan> My least favorite thing to explain to people who don't know git is that git has no idea what a directory is, it has paths to files and directories are an implementation detail. Which is why you can't add an empty dir lol.
2011-12-12T20:09:14  <avdg> there are some constrains, but once you know them, you shouldn't hit strange moves
2011-12-12T20:09:30  <Antimony> no, I'm running it on a replay from a different bot
2011-12-12T20:09:46  <avdg> ah, in that case, you are always underestimating
2011-12-12T20:09:58  <Antimony> or overestimating in this case
2011-12-12T20:10:16  <avdg> in the case you have a visual on all hills and see nothing spawns, you can reset the hive count
2011-12-12T20:10:20  <BenJackson> actually I don't rely on hive spawning at all for counting
2011-12-12T20:10:40  <BenJackson> I only observe the lack of ants on hills over time as indicating I should reset (in case of other bugs leading to drift)
2011-12-12T20:10:45  <avdg> overestimating would do more bad in most cases here
2011-12-12T20:11:01  <Antimony> but in this case, the bot thought there was no spawn when there was
2011-12-12T20:11:06  <Antimony> thus overestimation
2011-12-12T20:11:17  <Antimony> I'm not really sure how you can avoid that
2011-12-12T20:11:21  <BenJackson> why do you need spawn?
2011-12-12T20:12:03  <BenJackson> hive = food eaten - (dead ants + (live ants - original ants))
2011-12-12T20:12:46  <Antimony> you can't always see your ants deaths
2011-12-12T20:12:53  <BenJackson> yes you can
2011-12-12T20:13:02  <BenJackson> you see your ant deaths even if you don't see the square otherwise
2011-12-12T20:13:04  <avdg> and its also kinda tricky to determine if an ant was spawned from the hill
2011-12-12T20:13:13  <Antimony> really?
2011-12-12T20:13:16  <avdg> because you don't know the moves of the ant
2011-12-12T20:13:26  *** ikaros <ikaros!~ikaros@dslb-084-059-073-057.pools.arcor-ip.net> has quit IRC (Quit: Ex-Chat)
2011-12-12T20:13:44  <BenJackson>             # only include updates to squares which are visible
2011-12-12T20:13:44  <BenJackson>             # and the current players dead ants
2011-12-12T20:13:44  <BenJackson>             if v[row][col] or (ilk == 'd' and update[-1] == player)
2011-12-12T20:13:53  <Antimony> wow
2011-12-12T20:14:03  <BenJackson> that's why I just read the damn game engine
2011-12-12T20:14:06  <Antimony> I don't think that was in the specification
2011-12-12T20:14:09  <BenJackson> the specs are not precise enough
2011-12-12T20:14:55  <avdg> :p my assumption was right here
2011-12-12T20:15:04  <avdg> I hope I won't get a wrong assumption
2011-12-12T20:15:20  <avdg> I also wonder what happens with dead ants in view radius of the dead ant
2011-12-12T20:15:58  <avdg> for example you collide against an enemy, would you also see that you killed the enemy
2011-12-12T20:16:09  <Antimony> Oh wait, it is in the specs. I guess I fail at reading
2011-12-12T20:16:15  <Antimony> well that will make things ismpler
2011-12-12T20:16:19  <QuirionPT> when an ant moves and dies does it get to see the map anyway?
2011-12-12T20:16:28  <Antimony> no
2011-12-12T20:16:46  <BenJackson> no, just the death of that ant
2011-12-12T20:16:47  <Garf> just its own square, if I read that correctly
2011-12-12T20:16:52  <BenJackson> not even its square
2011-12-12T20:16:54  <BenJackson> just its death
2011-12-12T20:16:58  <Antimony> just the carcass
2011-12-12T20:17:01  <BenJackson> multiple deaths can occur on a square
2011-12-12T20:17:08  <BenJackson> and in the same turn something else (an ant, a hill) can be there
2011-12-12T20:17:10  *** NightExcessive <NightExcessive!~NightExce@unaffiliated/nightexcessive> has joined #aichallenge
2011-12-12T20:18:08  <Antimony> I guess I should redo my stored food count to use dead ants
2011-12-12T20:18:35  <BenJackson> I had ripped the dead ant list out of Square (in the C++ starteR) so I had to re-add counting them
2011-12-12T20:18:39  <BenJackson> but I only count, I don't save them
2011-12-12T20:19:05  <Wraithan> I removed dead ant stuff from the python lib
2011-12-12T20:19:22  <Antimony> For a while, I didn't even process dead ants
2011-12-12T20:19:40  <Antimony> they're only useful for keeping track of food
2011-12-12T20:21:12  <QuirionPT> it might be good to know where your ants died
2011-12-12T20:21:22  <QuirionPT> e.g. if your trying to assess the strength of an opponent ;)
2011-12-12T20:21:31  <QuirionPT> you're
2011-12-12T20:23:08  <Antimony> do you try to detect when your hills have been razed by a lack of spawning?
2011-12-12T20:23:38  *** amstan <amstan!~amstan@69-196-130-230.dsl.teksavvy.com> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T20:24:48  <bugnuts> so I spun up an m2.xlarge instance and benchmarked it and it turns out its almost exactly as fast as my desktop machine...
2011-12-12T20:27:36  <Antimony> as long as I'm fixing my bot to be able to run on replays of others, I guess I should change the dead ant set to a multiset
2011-12-12T20:27:46  <Antimony> in case the actula bot is dumb enough to allow collisions
2011-12-12T20:28:36  *** Garf <Garf!~Garf@78-22-80-252.access.telenet.be> has quit IRC (Quit: Make a new plan, Stan!)
2011-12-12T20:31:31  *** Conorach <Conorach!~cono@90.170.87.124> has quit IRC (Ping timeout: 276 seconds)
2011-12-12T20:33:08  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has quit IRC (Ping timeout: 240 seconds)
2011-12-12T20:37:50  <BenJackson> Antimony: I print it out in my debug but I don't do anything with it
2011-12-12T20:38:53  <rwest_> BenJackson: can you put your bots back on bhickey?
2011-12-12T20:40:01  *** TL__ <TL__!781c39b6@gateway/web/freenode/ip.120.28.57.182> has joined #aichallenge
2011-12-12T20:42:08  *** QuirionPT <QuirionPT!~quirion@gyf5.kyla.fi> has quit IRC (Quit: Leaving)
2011-12-12T20:49:04  <thestinger> nice, lots of servers now
2011-12-12T20:49:19  <thestinger> I just reuploaded earlier today and my bot has already played 5 games
2011-12-12T20:49:42  <thestinger> so around a game per hour
2011-12-12T20:50:31  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has joined #aichallenge
2011-12-12T20:54:59  <jab_bott> Antimony: why don't you use the fact that the hill no longer exists or the ant no longer exists to track ant/hill removal?
2011-12-12T20:55:29  <jab_bott> rwest_: how do you get your ants to spread out so nicely?
2011-12-12T20:56:31  <jab_bott> mine subtract values from the heat-map I use for exploration so that bunched up ants tend to move away from each other, but when it gets very crowded they get confused and start sticking to walls, causing further problems
2011-12-12T21:05:32  <thestinger> jab_bott: you're talking about filling space to search for food?
2011-12-12T21:05:45  <jab_bott> thestinger: yeah
2011-12-12T21:06:11  <jab_bott> walls==water :D
2011-12-12T21:06:21  <thestinger> I've experimented with it a bit and I can get a nice behavior to emerge now
2011-12-12T21:06:39  <thestinger> I started with just a multi-source BFS from all non-visible locations
2011-12-12T21:07:00  <jab_bott> I think perhaps they need to look further afield, rather than their immediate surroundings, because often they move one way then move straight back
2011-12-12T21:07:09  <thestinger> and I stop multiple ants from going to a location (I propagate locations _and_ distance through that BFS) and then tag the nearby locations of a target as also taken
2011-12-12T21:07:35  <thestinger> and with a bunch of random hacks and tweaks it turned into a nice spread for foraging
2011-12-12T21:08:19  <thestinger> basically I maximize the area that is visible
2011-12-12T21:08:33  <thestinger> but I get it to happen indirectly
2011-12-12T21:08:41  <jab_bott> thestinger: I have something similar for initial exploration, it works great, but after the outer ants have explored nicely, when the food collection is too much, basically when it gets too crowded around my hill the ones in the crowd get stuck
2011-12-12T21:08:59  <thestinger> oh, because they don't have a path to a target?
2011-12-12T21:10:06  <jab_bott> no, they're not searching for a path, it's just that with so many crowded up they move one way to get away from the crowd only to find there's another crowd in that direction too and so they move back again as the distances change
2011-12-12T21:10:20  <thestinger> ah, I don't actually make them spread out
2011-12-12T21:10:38  <thestinger> the behavior emerges because they won't get the same targets
2011-12-12T21:10:43  <thestinger> for food searching
2011-12-12T21:10:45  <jab_bott> and the ones on the edge stop spreading out once the exploration frontier is empty (i.e. the whole map has been seen at some point regardless of its current visibility)
2011-12-12T21:11:11  *** amstan <amstan!~amstan@69-196-130-230.dsl.teksavvy.com> has joined #aichallenge
2011-12-12T21:11:11  *** ChanServ sets mode: +o amstan
2011-12-12T21:11:30  <jab_bott> so the ones on the edge think everything is fine, they see no need to spread out further, they don't realize that crowds are building up near the hill
2011-12-12T21:12:16  *** SMJ <SMJ!~janne@dyn2-212-50-133-246.psoas.suomi.net> has quit IRC (Quit: Konversation terminated!)
2011-12-12T21:12:40  <thestinger> I have two types of exploration
2011-12-12T21:13:03  <thestinger> there's the food gathering part (which tries to maximize the visible area) and it doesn't really need very many ants
2011-12-12T21:13:15  <thestinger> so I have a bunch leftover at the end to attack enemy hills, or search for them
2011-12-12T21:13:52  <thestinger> I used to limit the number of ants used for foraging, but now that I mark off locations around a target as taken in addition to the target, it limits itself
2011-12-12T21:18:57  <jab_bott> ok.. I see. I think I'm going to try looking a bit further afield for a suitable direction, that might help the ants get past local valleys in the heat-map
2011-12-12T21:20:15  <jab_bott> or maybe just adding a bias away from the hill and requiring a greater step difference to override it
2011-12-12T21:21:01  <Antimony> maybe you aren't propogating your heat map enough
2011-12-12T21:25:26  <antimatroid1> how did languages other than c++ do vision
2011-12-12T21:25:48  <thestinger> antimatroid1: I copied the java method
2011-12-12T21:26:00  <thestinger> it basically comes up with the offsets to paint on info around the radius
2011-12-12T21:26:04  *** dmj111 <dmj111!~user@c-98-235-94-211.hsd1.pa.comcast.net> has joined #aichallenge
2011-12-12T21:26:05  <antimatroid1> yeah, i do that now
2011-12-12T21:26:07  <thestinger> like, on turn 0
2011-12-12T21:26:12  <thestinger> and then paints it on every turn
2011-12-12T21:26:17  <antimatroid1> oh well, starter bots were just a starting point :P
2011-12-12T21:26:21  <thestinger> so I made a higher-order function to do it for any radius info
2011-12-12T21:26:40  <antimatroid1> yeah, i was just reading comments earlier about the updateVision function i wrote ages ago :P
2011-12-12T21:27:05  <thestinger> well, you did save me the trouble of writing the 10 lines for a BFS once I gave up on A* :P
2011-12-12T21:28:57  <thestinger> oh wow, I have a major bug
2011-12-12T21:29:02  <BenJackson> rwest_: done
2011-12-12T21:29:14  <thestinger> I used std::function<bool(Location)>
2011-12-12T21:29:38  <thestinger> and I wanted it to return an int
2011-12-12T21:30:04  <thestinger>                 if (get_nattackers(new_loc) <= get_nattackers(ant)) {
2011-12-12T21:30:15  <thestinger> that just returns a bool atm...
2011-12-12T21:31:05  <antimatroid1> finding bugs rules when it makes your bot better
2011-12-12T21:31:11  <thestinger> I'm scared to fix this
2011-12-12T21:31:13  <thestinger> lol
2011-12-12T21:31:14  <antimatroid1> it's just plain annoying when it seems to make your bot worse
2011-12-12T21:31:30  <thestinger> I have to figure out what the hell my positioning code is actually doing now...
2011-12-12T21:32:06  <thestinger> I think this explains why my bot sometimes doesn't attack a lone ant
2011-12-12T21:32:54  <thestinger> wow, this is actually a huge bug
2011-12-12T21:33:15  <thestinger> I use this for coming up with offensive moves, positioning, etc.
2011-12-12T21:33:18  <BenJackson> I just found that bug with my bot with a testcase
2011-12-12T21:34:17  <thestinger> oh well, I'll test this fix with strcat_next :P
2011-12-12T21:34:46  <thestinger> I have 6 bots on tcp now though...
2011-12-12T21:35:44  <thestinger> http://ants.fluxid.pl/replay.29264 strcat_push fixes another silly bug - my ants would leave a gap whenever they attacked due to collision avoidance
2011-12-12T21:37:57  *** roflmao <roflmao!~jstrong@189.229.217.18> has joined #aichallenge
2011-12-12T21:38:02  <jab_bott> woah - do new submissions start playing all the lowest players first or something? my bot just played against a load of bots that basically did nothing
2011-12-12T21:38:17  <thestinger> jab_bott: yeah :)
2011-12-12T21:38:22  <thestinger> you start at skill 0
2011-12-12T21:38:33  <jab_bott> so they only play similar skills?
2011-12-12T21:39:11  <thestinger> not only, but you usually play people close to you in skill
2011-12-12T21:39:22  <thestinger> sometimes the matchmaking does inexplicable things :P
2011-12-12T21:39:37  <jab_bott> ahh well at least it will go up eventually and I can get an idea if it's better than my previous one that got to like place 550 or something around there
2011-12-12T21:41:15  <BenJackson> xathis just beat pguillory on his trek to the top
2011-12-12T21:41:52  <thestinger> jab_bott: _lots_ of games get played now, which is nice :)
2011-12-12T21:42:05  <thestinger> running both on tcp at the same time should give you a good idea of which is better
2011-12-12T21:42:10  <thestinger> if there's a significant improvement
2011-12-12T21:42:31  <BenJackson> oh, in a 3 player game he beat ThisIsNotABug and tied pguillory more like
2011-12-12T21:42:33  <thestinger> fluxid's tcp server is nice and crowded now
2011-12-12T21:42:58  <jab_bott> I'll probably be late for work quite a few days this week!!
2011-12-12T21:45:01  *** flowenol <flowenol!LW@bl4-237-93.dsl.telepac.pt> has quit IRC (Ping timeout: 244 seconds)
2011-12-12T21:45:12  <thestinger> I wonder if Rust will be somewhat stable in time for the next contest
2011-12-12T21:45:19  <thestinger> it seriously looks like such an awesome language
2011-12-12T21:45:36  <thestinger> http://www.rust-lang.org/
2011-12-12T21:46:01  <thestinger> I love all the features/decisions on the front page, and looking through the tutorial really makes me wish I was using it now :P
2011-12-12T21:46:59  <roflmao> Rust is a curly-brace, 
2011-12-12T21:47:05  <roflmao> i stopped reading there
2011-12-12T21:47:15  <thestinger> they haven't actually decided on syntax yet
2011-12-12T21:47:18  <thestinger> that doesn't really matter
2011-12-12T21:47:31  <thestinger> it will be compiled, so you can just have a preprocessor for python-like syntax or whatever
2011-12-12T21:47:42  *** timtamboy63 <timtamboy63!~spamalot6@C-61-68-245-105.per.connect.net.au> has joined #aichallenge
2011-12-12T21:48:01  <thestinger> utf8-strings (so sane, like perl/go)
2011-12-12T21:48:27  <timtamboy63> Hey guys, I'm trying to work out how to set up the Ants challenge on my local machine to work with PHP, but I'm not entirely sure how, anyone have any tips/advice/help?
2011-12-12T21:48:32  <thestinger> algebraic data types, type inference, C++-style generics but without the crap
2011-12-12T21:48:42  <BenJackson> I sure hope this isn't some kind of "local maximum" because my working bot just curb stomped two of my previous version
2011-12-12T21:49:03  <thestinger> and concurrency like erlang
2011-12-12T21:51:55  <thestinger> http://www.rust-lang.org/doc/tutorial/data.html so awesome :D
2011-12-12T21:52:39  <roflmao> bj:
2011-12-12T21:52:41  <roflmao> isn't that good?
2011-12-12T21:53:04  <BenJackson> roflmao: it is, I mean I hope I haven't just made a old-bot-killer
2011-12-12T21:53:12  <roflmao> yeah
2011-12-12T21:53:14  <roflmao> that happens
2011-12-12T21:53:22  <roflmao> so fricken hard to adequately test your bot
2011-12-12T21:53:24  <roflmao> in a global way
2011-12-12T21:53:38  <BenJackson> but with a near-top-10 bot as your baseline it's hard to make something that consistently wins
2011-12-12T21:53:39  <Antimony> BenJackson, how does your symmetry detection do on this map? http://aichallenge.org/visualizer.php?game=210389&user=3875
2011-12-12T21:53:52  <thestinger> that's why you run 20 versions on tcp to slowly optimize equations/constants :)
2011-12-12T21:54:02  <roflmao> offtopic: http://www.ted.com/talks/blaise_aguera_y_arcas_demos_photosynth.html
2011-12-12T21:54:23  <roflmao> I wonder when someone is going to grab that design a GUI for an OS around it
2011-12-12T21:54:28  <roflmao> no more workspaces
2011-12-12T21:54:48  <BenJackson> Antimony: no idea.  I didn't get that far in testing before deciding I didn't want to use it in my strategy
2011-12-12T21:54:54  <Antimony> oh
2011-12-12T21:55:10  <BenJackson> I only bothered with certain rotations
2011-12-12T21:55:11  <Antimony> now I'll never know if your method is really as good as you say it is
2011-12-12T21:55:13  <thestinger> roflmao: workspaces are silly, I never use them
2011-12-12T21:55:19  <BenJackson> to the poitn where I could see it work and then see my bot be worse
2011-12-12T21:55:25  <roflmao> thestinger: but ... what if you have a gazillion windows open?
2011-12-12T21:55:42  <roflmao> i usually dual monitor 6 workspaces
2011-12-12T21:55:42  <thestinger> yes, my window manager lets me make any windows tabbed or stacked together
2011-12-12T21:55:44  <roflmao> 12 total screens :P
2011-12-12T21:55:52  <roflmao> what wmanager do you use? xmonad?
2011-12-12T21:55:55  <BenJackson> I just use a 30" monitor at home
2011-12-12T21:55:56  <thestinger> i3
2011-12-12T21:55:59  <BenJackson> it seemed huge when I got it
2011-12-12T21:56:05  <BenJackson> just like my 24" seemed huge when I got *it*
2011-12-12T21:56:07  <thestinger> it was inspired by wmii, but it's a _lot_ different
2011-12-12T21:56:10  <BenJackson> now the 30" seems normal :)
2011-12-12T21:56:15  <roflmao> 30"
2011-12-12T21:56:16  <roflmao> wow
2011-12-12T21:56:17  <thestinger> each workspace is a tree of windows/containers
2011-12-12T21:56:17  <roflmao> I'm envious
2011-12-12T21:56:40  <roflmao> I'm running on 21"
2011-12-12T21:56:42  <thestinger> so any container (a workspace is basically a container) can be split horizontally/vertically and they can be stacked, tabbed, tiled or floating
2011-12-12T21:57:00  <thestinger> no layouts like dwm/awesome, you make your own
2011-12-12T21:57:01  *** Antimony <Antimony!~chatzilla@lawn-143-215-58-110.lawn.gatech.edu> has quit IRC (Quit: ChatZilla 0.9.87 [Firefox 8.0/20111104165243])
2011-12-12T21:57:09  <roflmao> sounds a lot like xmonad
2011-12-12T21:57:12  <roflmao> or the other tiling window managers
2011-12-12T21:57:30  <thestinger> well, it's a manual tiler basically
2011-12-12T21:57:35  <roflmao> ah cool
2011-12-12T21:57:40  <roflmao> does it integrate easily with gnome?
2011-12-12T21:57:44  <roflmao> i don't want to lose my taskbar
2011-12-12T21:57:48  <roflmao> with all my app shortcuts
2011-12-12T21:57:51  <thestinger> dunno
2011-12-12T21:57:56  <thestinger> try dmenu, it's better :P
2011-12-12T21:58:26  <roflmao> ugluy
2011-12-12T21:58:55  <roflmao> :P
2011-12-12T21:59:08  <thestinger> not ugly, you only see it when you launch something anyway
2011-12-12T21:59:39  <roflmao> thing I don't really liker about tiling managers
2011-12-12T21:59:43  <roflmao> is I have a 21" screen
2011-12-12T21:59:50  <roflmao> so if I tile things they'll all be super small
2011-12-12T22:00:00  <roflmao> so I end up just using the tiling manager workspaces
2011-12-12T22:00:04  <roflmao> which is identical to using the gnome ones
2011-12-12T22:00:20  <thestinger> you don't have to tile everything
2011-12-12T22:00:51  <roflmao> yeah but i still easily fill up all 1-0 keys of workspaces
2011-12-12T22:00:56  <thestinger> my monitor is wide enough that I pretty much always want it split down the middle
2011-12-12T22:00:58  <roflmao> then I'm like "uhh err which one was I on?"
2011-12-12T22:01:02  <thestinger> so I don't really "tile"
2011-12-12T22:01:09  <thestinger> I have a vertical split
2011-12-12T22:01:14  <thestinger> and then everything in stacks/tabs
2011-12-12T22:01:34  <roflmao> also
2011-12-12T22:01:37  <roflmao> what about app shortcuts?
2011-12-12T22:01:43  <thestinger> like what?
2011-12-12T22:01:46  <roflmao> using alt+everything = huge conflict possibilities
2011-12-12T22:01:48  <roflmao> like photoshop ;P
2011-12-12T22:01:53  <thestinger> super+d == dmenu
2011-12-12T22:02:10  <thestinger> or whatever
2011-12-12T22:02:18  <thestinger> so if I want to start gimp
2011-12-12T22:02:34  <thestinger> <SUPER>dgimp<ENTER>
2011-12-12T22:02:50  <thestinger> it has menu based (dmenu :P) autocompletion
2011-12-12T22:03:26  *** jab_bott <jab_bott!5163fc1b@gateway/web/freenode/ip.81.99.252.27> has quit IRC (Quit: Page closed)
2011-12-12T22:04:02  *** Paradoxial <Paradoxial!~Paradoxia@pool-108-28-22-94.washdc.fios.verizon.net> has quit IRC (Quit: :)
2011-12-12T22:04:03  <roflmao> nice
2011-12-12T22:04:06  <roflmao> that does sound nice :P
2011-12-12T22:04:42  <thestinger> it's a lightweight version of quicksilver basically
2011-12-12T22:04:46  <roflmao> ooooooooooooooooooh
2011-12-12T22:04:47  <roflmao> stacking
2011-12-12T22:04:48  <roflmao> i see
2011-12-12T22:04:50  <roflmao> thats fkin awesome
2011-12-12T22:05:38  <thestinger> http://i3wm.org/docs/userguide.html#_changing_the_container_layout
2011-12-12T22:05:46  <thestinger> their example is pretty silly
2011-12-12T22:06:11  <thestinger> you can also resize containers/splits manually ofc
2011-12-12T22:07:07  <thestinger> window hints are respected by default, so stuff like a firefox dialog (page info, preferences) is floating by default
2011-12-12T22:07:07  *** NightExcessive <NightExcessive!~NightExce@unaffiliated/nightexcessive> has quit IRC (Quit: Leaving)
2011-12-12T22:07:22  <roflmao> cool
2011-12-12T22:07:57  *** timtamboy63 <timtamboy63!~spamalot6@C-61-68-245-105.per.connect.net.au> has left #aichallenge
2011-12-12T22:08:56  <roflmao> anyways
2011-12-12T22:09:02  <roflmao> i can totally envision a touch tablet
2011-12-12T22:09:09  <roflmao> using that photosynth movement mechanism
2011-12-12T22:09:17  <roflmao> infinite real estate
2011-12-12T22:12:04  *** Jak_o_Shadows <Jak_o_Shadows!~hayden@CPE-144-136-214-52.sa.bigpond.net.au> has joined #aichallenge
2011-12-12T22:51:20  *** roflmao <roflmao!~jstrong@189.229.217.18> has quit IRC (Quit: Leaving.)
2011-12-12T22:53:28  <antimatroid1> on the discussion earlier about the same bots playing the same, early in beta i found it was very hard to do that
2011-12-12T22:53:50  <antimatroid1> the wrapping makes it hard for all bots to choose the same targets with equal paths
2011-12-12T22:53:54  *** ltriant <ltriant!~ltriant@lithium.mailguard.com.au> has quit IRC (Ping timeout: 255 seconds)
2011-12-12T22:54:33  <antimatroid1> although thinking about it, i don't see why for translationally symmetric maps
2011-12-12T22:58:25  *** TheLinker <TheLinker!~TheLinker@host80.200-117-105.telecom.net.ar> has quit IRC (Quit: Bye)
2011-12-12T23:02:20  *** dvladim <dvladim!~dvladim@85.113.131.65> has joined #aichallenge
2011-12-12T23:02:34  <avdg> hmm, could i3wm replace gnome or is it a layer on top of it?
2011-12-12T23:04:03  <avdg> meh, I have to go
2011-12-12T23:06:50  *** ltriant <ltriant!~ltriant@lithium.mailguard.com.au> has joined #aichallenge
2011-12-12T23:07:23  *** avdg <avdg!~avdg@78-22-168-200.access.telenet.be> has quit IRC (Quit: Leaving.)
2011-12-12T23:07:40  *** Fandekasp <Fandekasp!~Fandekasp@nt10-ppp1812.west.sannet.ne.jp> has quit IRC (Ping timeout: 268 seconds)
2011-12-12T23:10:28  *** keith__ <keith__!43a1fc8b@gateway/web/freenode/ip.67.161.252.139> has joined #aichallenge
2011-12-12T23:16:54  *** xathis <xathis!~xathis@g231126239.adsl.alicedsl.de> has quit IRC ()
2011-12-12T23:22:41  *** Scryer <Scryer!4a645367@gateway/web/freenode/ip.74.100.83.103> has quit IRC (Quit: Page closed)
2011-12-12T23:29:26  *** dvladim <dvladim!~dvladim@85.113.131.65> has quit IRC (Ping timeout: 252 seconds)
2011-12-12T23:42:16  *** treeform <treeform!~treeform@108.60.121.46> has quit IRC (Ping timeout: 252 seconds)
2011-12-12T23:46:41  *** keith__ <keith__!43a1fc8b@gateway/web/freenode/ip.67.161.252.139> has quit IRC (Quit: Page closed)
2011-12-12T23:47:00  *** keith__ <keith__!43a1fc8b@gateway/web/freenode/ip.67.161.252.139> has joined #aichallenge
2011-12-12T23:49:23  *** u_ <u_!~u_@adsl-75-57-180-214.dsl.emhril.sbcglobal.net> has quit IRC (Quit: u_)
2011-12-12T23:58:57  <BenJackson> oh  man, I have a condition
2011-12-12T23:58:59  <BenJackson> !foo < 8
2011-12-12T23:59:05  <BenJackson> dammit