2011-06-15T00:18:07 *** computerwiz_222 has quit IRC (Ping timeout: 240 seconds) 2011-06-15T00:37:10 *** UncleVasya has joined #aichallenge 2011-06-15T00:37:57 Hi. There are spamers on forums again: http://ai-contest.com/forums/viewtopic.php?f=25&t=1313&start=90 2011-06-15T00:38:31 UncleVasya: wanna be a mod? 2011-06-15T00:41:59 ok.. i'm not going to make you a mod without answering that 2011-06-15T00:42:02 and i want to go to bed 2011-06-15T00:42:03 cya 2011-06-15T00:42:06 amstan: 1. I will have no computer from July 6 to the September. 2. 1. I must to learn English better. (this is what I hope to do while will not have a computer) If in the September your offer will still remain valid I thin I'll accept it. Thank you. 2011-06-15T00:42:23 what's with the computerless state? 2011-06-15T00:43:22 I'm sduying and I'm live in very different places. Computer is only there where I am studying. 2011-06-15T00:44:15 I need to go too, cya. 2011-06-15T00:44:26 *** UncleVasya has quit IRC () 2011-06-15T00:50:06 Ah compile-time function execution is wonderful! 2011-06-15T00:51:18 *** delt0r_ has joined #aichallenge 2011-06-15T00:52:14 The compiler told me that my parser isn't fully implemented yet. At some point it could run into a line of text that has to be parsed into a struct that contains an enum and I haven't implemented enum parsing yet. 2011-06-15T00:52:41 And that's just from compiling the code, not loading any test data into the parser. 2011-06-15T00:52:51 mleise: sounds like haskell 2011-06-15T00:52:52 *** delt0r- has quit IRC (Ping timeout: 244 seconds) 2011-06-15T00:53:40 I've heard about haskel here a few times. I should take a look at that language some day 2011-06-15T00:54:13 yes, it's definatelly on my language todo list 2011-06-15T00:54:29 but first.. i want to learn django and get comfortable with c++ 2011-06-15T00:54:48 haha ok, what part of haskel reminds you of what i said? 2011-06-15T00:55:09 the magic stuff that happens in haskell all the time 2011-06-15T00:55:14 i don't think compile-time function execution will yield any results 2011-06-15T00:55:39 where function declaractions just look like formulas 2011-06-15T00:55:53 i can't really explain it, i just know it's awesome 2011-06-15T00:56:27 perhaps jmcarthur can give you a demo 2011-06-15T00:56:35 http://www.haskell.org/ghc/docs/6.10.1/html/users_guide/template-haskell.html <- this seems to do it 2011-06-15T00:59:01 But to answer your implicit question: I'm using D for now. I used C++ last year and Java, Delphi, VisualBasic before that. 2011-06-15T01:02:33 *** delt0r_ has quit IRC (Ping timeout: 264 seconds) 2011-06-15T01:05:09 amstan: Haskel would be a lot different from what I'm used to. There is no assignment to variables and everything is just the definition of functions that process input. If someone proposed such a language to me I would say he's crazy1 2011-06-15T01:05:39 mleise: yes... functional 2011-06-15T01:05:52 anyway. 2011-06-15T01:05:56 i'm going to bed 2011-06-15T01:05:58 cya 2011-06-15T01:09:41 night 2011-06-15T01:19:04 mleise, you can assign to a variable you just can't change it afterwards :P 2011-06-15T01:25:01 sigh: that's a degenerate case of setting an initial value for a function then, or not? 2011-06-15T01:25:27 Anyway look at my super cool text file parser: 2011-06-15T01:25:46 when I want an optional name in my text file like 'name "foo bar"' that must not be empty and will be stored in the field '_name' I write: 2011-06-15T01:25:55 "parser.read!('?', "name", "!value.empty")(_name)" 2011-06-15T01:31:46 It also loads lines like 'xyz "hello" 42 3.1415' into a struct { string, int, float } the same way to have a more complex example. And it understands repeated sections, like multiple turns with content of their own that is parsed in a different method. 2011-06-15T01:39:18 mleise, I suppose it's closer to a degenerate case of creating a constant function :P 2011-06-15T01:43:02 sigh: How is the program flow in a functional language? I imagine every byte that goes in is processed the same way by all the functions you connected with each other and the result of the outer function is put back onto the console 2011-06-15T01:43:52 But there must be ifs and switches in such a language, too. 2011-06-15T01:44:57 not everything is processed the same, the excution can take different paths for different data 2011-06-15T01:45:30 there are conditionals and in fact they are usually more powerful than in procedural languages 2011-06-15T01:46:21 it's just that if you have the *same* input then you will always get the same output 2011-06-15T01:46:37 so the same input always takes the same path :) 2011-06-15T01:50:11 mleise ^^ 2011-06-15T01:50:35 *** nann has quit IRC (Read error: Connection reset by peer) 2011-06-15T01:50:38 D also implements some ideas from functional languages, like map, filter, reduce and the like, but it is difficult to understand these inside a imperative language :p 2011-06-15T01:51:01 *** nann has joined #aichallenge 2011-06-15T01:51:35 mleise, you can write higher-order functions like map, filter and reduce in any language, although some languages make it easier :) 2011-06-15T01:51:44 for example, python comes with those functions builtin 2011-06-15T01:53:11 yes, but when the return type is of "Range(map(filter(uniq(int[])))" I don't exactly know what to do with it other than putting that into the array() function to get back a simple old-school array. 2011-06-15T01:53:56 These Ranges have attributes like isForwardRange, isRandomAccessRange, isInfinite etc... I have no idea what this stuff is good for. 2011-06-15T01:55:10 ah, I don't know D, but based on the names i assume isRandomAccessRange is to determine if certian operations are possible on a range (think array vs linked list) 2011-06-15T01:55:18 Seems like someone coming from a functional language will understand the infinite range concept better as functions can be iterated to infinity 2011-06-15T01:55:29 and isInfinite... well you know not to iterate over all the elements in an infinite list :P 2011-06-15T01:56:27 mleise, you don't need to come from a functional language... for example python has generators which can do the same thing 2011-06-15T01:56:40 but when would I want to use a function that can be iterated to infinity and is evaluated on demand? I never needed that, lol. 2011-06-15T01:56:55 ever read from a stream? 2011-06-15T01:57:20 when you run 'tail -f' that's a close description of what it does :) 2011-06-15T01:57:26 while ( ( line = stream.readLine() ) != null ) ... 2011-06-15T01:58:13 I mean... why would I bother with the philosophical concept of infinity when I can use a while loop? 2011-06-15T01:59:09 it is essentially a different way to think about the while loop... but it is more powerful 2011-06-15T01:59:27 for example, the execution can be paused and restarted and handed off to other parts of the programs 2011-06-15T01:59:57 Oo and why would I want to do THAT??? 2011-06-15T02:00:35 I can hardly tell my SocketStream to go back in time. 2011-06-15T02:00:55 And I can hand off data to other parts of my program using an if statement, no? 2011-06-15T02:00:56 eh? it's the same reason you would pass the stream object to another function 2011-06-15T02:01:57 what do you mean by pause? I think of the debugger there that stops the process, nothing special 2011-06-15T02:04:08 What do you think is the real benefit? Is functional code faster to write? Is it often more readable? Or can you mostly just type as you think? 2011-06-15T02:06:31 mleise, you are overthinking this... an infinite iterator is bascially a class that has this interface: http://pastebin.com/Gp3eeE9r 2011-06-15T02:08:21 ok, I think I get the idea why the language needs to have concepts like this. you cannot use loop over the length of such a 'range', but you could use a foreach that runs forever. 2011-06-15T02:08:29 functional programs make different guarentees... for example you know that your functions have no side effects if you put the same thing in you get the same thing out... it also means that the compiler can optimise heavily 2011-06-15T02:09:01 ok, so you pass EVERYTHING as parameter to the function 2011-06-15T02:09:02 mleise, you can loop over a range 2011-06-15T02:09:27 mleise, kind of... but there are ways of hiding global state if you need to 2011-06-15T02:09:31 *** Palmik has joined #aichallenge 2011-06-15T02:11:07 mleise, you can loop over a range, but sometimes the range is artificial... for example if you want to generate a list of n primes it is simpler to define an infinite list of primes (because hey, primes are infinite) and then only take the elements that you need 2011-06-15T02:11:13 It sounds like a matter of taste to me. I could say I use objects so I don't have to pass all parameters to my methods, because the object has an internal state that defines it and all methods may operate on it. 2011-06-15T02:11:53 of course it's a matter of taste, it's not like functional languages gives your computer magic powers 2011-06-15T02:12:19 Well, exception handling is one thing that I consider a magic moment 2011-06-15T02:12:37 heh 2011-06-15T02:14:03 sigh: But how can the compiler know how I use the prime number generator? Maybe it needs to cache values, maybe that would comsume too much memory 2011-06-15T02:15:01 For me its only kosher if I know how the implementation works. I don't like hitting hidden performance botllenecks. 2011-06-15T02:18:13 With the plain old procedural languages I always know where my data is, because I explicitly allocated it one way or another. 2011-06-15T02:18:45 fair enough, but that's just because you've worked with procedural languages and learnt their implementation details :P 2011-06-15T02:19:10 don't tell me you knew all the magic that happens when you do a malloc without being told 2011-06-15T02:19:55 Actually I read about memory allocation on the internet, yes :p 2011-06-15T02:20:31 And working with Delphi there was an OpenSource memory manager replacement that was supposed to be faster and also showed memory leaks. 2011-06-15T02:20:33 my point was that you had to invest time to learn it 2011-06-15T02:21:16 same with other things that are taken for granted like the stack and what actually happens when you call a function 2011-06-15T02:21:35 what is there to learn if you just use it? you get a portion of RAM assigned to your app and the pointer is the address. 2011-06-15T02:23:08 Ok, the kernel has to find a continuous block of the requested size, but that's about it on the user end I think :p All you want to know is "will I run out of memory" 2011-06-15T02:23:15 you don't care about the difference between allocating 1000 small segments vs 1 large segement... what about the performance impact of performing a realloc... are the consequences to allocation and reallocating many times? 2011-06-15T02:23:52 mleise, and for a functional language I can say "All you want to know is the result of the function" 2011-06-15T02:24:00 *** Ice_Harley has joined #aichallenge 2011-06-15T02:24:05 point is that when it comes to optimisation, you have to know the details 2011-06-15T02:24:11 that's not unique to functional languages 2011-06-15T02:24:24 hmm once you understand the "continuous block" thing you can deduct the implications for realloc and ... yeah malloc is *slow*. 2011-06-15T02:26:33 sigh: maybe I would understand the implementation details of Haskel just as well if I used it for years. I'm just alienated by all the different concepts now. 2011-06-15T02:27:02 that was my point :P 2011-06-15T02:27:27 but for starters don't worry about the implementation details, just worry about correct code :) 2011-06-15T02:27:37 good luck with your ants then :) 2011-06-15T02:28:27 eh? it's the same way for any language... first you have to have a working program before you can optimise 2011-06-15T02:32:04 We have only 500ms. Performance is very important now. An algorithm must scale well with 1000 ants. 2011-06-15T02:32:42 with 1000 ants your battle algorithm is probably sucking hard 2011-06-15T02:32:45 O(n²) is practically out :p 2011-06-15T02:32:52 you just could do a russian style attack 2011-06-15T02:32:55 lol rabidus 2011-06-15T02:33:17 sure, but all you are doing is saying you don't want to work with a language/paradigm you are not familiar with... you aren't really making an argument against any particular language 2011-06-15T02:33:37 which is fair enough, but I was never disagreeing with that 2011-06-15T02:34:47 ok maybe with a language that is close to the system level it is easier to estimate the cpu time. is that an argument for using C? 2011-06-15T02:35:25 why not assembly? 2011-06-15T02:36:49 and even then you have don't know what magic system calls are doing 2011-06-15T02:37:05 sigh: I just remember something that's hard to do. With no threading we are assumed to keep track of time and return from any executing function if we are running out of time. Can a functional language help here by 'automatically' adding time awareness to all the code? 2011-06-15T02:39:10 mleise, depends on the language how easy that would be... I'm not proficient enough to say 2011-06-15T02:39:11 or is it more straight forward, because you would just add a "time remaining" function as a parameter to every other function? in a OOP language that would be kinda paradigm breaking and ugly. 2011-06-15T02:39:49 mleise, no you would not add a parameter to every function, that's just as ugly as in a procedural language 2011-06-15T02:40:17 hang on 2011-06-15T02:40:46 *** antimatroid has joined #aichallenge 2011-06-15T02:40:53 ay 2011-06-15T02:47:15 *** mcstar has joined #aichallenge 2011-06-15T02:47:59 mleise, sorry I'm not sure... haskell has System.Timeout for wraping functions, but I don't know if that works without threads or anything 2011-06-15T02:52:40 not to self, ice is REALLY slippery on a bike going downhill 2011-06-15T02:54:36 ice tends to be slippery regardless of the slope of the hill and your mode of transportation 2011-06-15T02:55:22 sigh: yes, well, i slid about 15 metres on the road this morning :P 2011-06-15T02:55:56 luckily I just had time to realise what was happening and it actually didn't hurt, but i wasn't far off sliding out onto another road at the bottom of the hill which could have been quite bad 2011-06-15T02:56:19 you didn't realise you were sliding? 2011-06-15T02:57:05 i did realise 2011-06-15T02:57:21 i knew i was on ice, but i was moving, best I could do was put a foot down and try to slide the wheels and my feet 2011-06-15T02:57:31 but i ended up down on my ass anyway, but could have been way worse 2011-06-15T02:57:37 yeah 2011-06-15T02:58:04 well, that'll learn you to put ice on your roads 2011-06-15T02:59:04 d; 2011-06-15T03:00:31 http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=Bath+St,+Battery+Point+TAS,+Australia&aq=0&sll=-42.902313,147.336617&sspn=0.024237,0.066047&ie=UTF8&hq=&hnear=Bath+St,+Battery+Point+Tasmania,+Australia&ll=-42.894159,147.333403&spn=0.00303,0.008256&t=h&z=18&layer=c&cbll=-42.894167,147.333555&panoid=CACaXubvKKvmnqVhyk7Pkw&cbp=12,294.41,,0,15.85 2011-06-15T03:00:59 i fell at the end of the hill the car is turning into 2011-06-15T03:01:52 *** berak has joined #aichallenge 2011-06-15T03:01:58 that street doesn't look very safe 2011-06-15T03:02:20 it's alright when it's not ice at the bottom :P 2011-06-15T03:02:39 wow... looking at australian roads looks weird now that I've been driving in the states 2011-06-15T03:02:48 the cars are all on the other side! 2011-06-15T03:02:57 :p 2011-06-15T03:03:05 i would struggle to drive on the rhs 2011-06-15T03:03:37 I keep turning on my wipers when I am trying to indicate :( 2011-06-15T03:04:48 *** UncleVasya has joined #aichallenge 2011-06-15T03:06:10 sigh: System.Timeout probably uses a thread, but what I like is this asynchroneous exception stuff, where you can throw exceptions at other threads any time :p 2011-06-15T03:06:54 So if pathfinding is slow you can throw an exception at it to finish faster, hehe 2011-06-15T03:07:10 australia and ice ? 2011-06-15T03:08:15 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-06-15T03:08:56 more like tasmania :) 2011-06-15T03:10:01 mleise: which is australia! :P 2011-06-15T03:14:32 *** aerique has joined #aichallenge 2011-06-15T03:15:50 i dont see no ice on that road... 2011-06-15T03:16:10 lol 2011-06-15T03:16:11 :D 2011-06-15T03:16:21 This is my place in december: http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=Burgstra%C3%9Fe,+Hannover,+Deutschland&aq=0&sll=-42.894167,147.333553&sspn=0.002114,0.00456&g=Bath+St,+Battery+Point+Tasmania,+Australia&ie=UTF8&hq=&hnear=Burgstra%C3%9Fe,+30159+Hannover,+Region+Hannover,+Niedersachsen,+Germany&ll=52.371879,9.732363&spn=0.000874,0.00228&t=h&z=19&layer=c&cbll=52.371764,9.732507&cbp=12,147.04,,0,0&photoid=po-4 2011-06-15T03:17:30 lol, it gives me an all black screen 2011-06-15T03:17:56 yeah, I'm in Nirvana already. They have internet here 2011-06-15T03:17:57 seet bike! 2011-06-15T03:17:59 sweet* 2011-06-15T03:18:00 ah, me too 2011-06-15T03:18:15 black screen for me too 2011-06-15T03:18:41 hell, my irc client only copies the first line of the link, i cant be bothered to copy it again line by line 2011-06-15T03:18:44 move the little guy around a little bit 2011-06-15T03:18:46 fixes it up then 2011-06-15T03:19:07 mcstar: pidgin! 2011-06-15T03:19:15 thats not console 2011-06-15T03:19:26 tissue? :P 2011-06-15T03:19:40 you mean...? 2011-06-15T03:19:50 dw :P 2011-06-15T03:19:54 hehe 2011-06-15T03:20:35 sigh: i realized why did you need the rule that ants trying to move to the same square must die 2011-06-15T03:20:56 mcstar, why is that? 2011-06-15T03:21:13 well, otherwise you couldnt update the game-board 2011-06-15T03:21:20 at least not that easily 2011-06-15T03:21:28 you could, but yeah... not easily :P 2011-06-15T03:21:28 the moves would be interdependent 2011-06-15T03:21:46 a few other options were discussed 2011-06-15T03:21:59 but we settled on this 2011-06-15T03:22:15 its simple and effective 2011-06-15T03:22:23 *** olexs has joined #aichallenge 2011-06-15T03:22:58 i'm a big fan of not having ant id's etc. 2011-06-15T03:23:12 also mcleopold said something to me about combining iterative closest and damage battle options :| 2011-06-15T03:23:17 that sounds awful complexity wise 2011-06-15T03:23:28 and for intuitive understanding 2011-06-15T03:23:36 oh btw to everyone on linux: try the freetype2-infinality font rendering, its great 2011-06-15T03:23:49 i wont be happy until unicode has greek 2011-06-15T03:23:59 so i can use it in my source for programs 2011-06-15T03:24:03 and mathematical relations etc. 2011-06-15T03:24:17 ωτf αrε γσμ ταlκιηg αbσμτ? 2011-06-15T03:24:19 its not font-faces 2011-06-15T03:24:20 and on ma keyboard 2011-06-15T03:24:23 :P 2011-06-15T03:24:24 its rendering 2011-06-15T03:24:37 best antialiasing and hinting ive ever seen, even in gtk 2011-06-15T03:25:47 antimatroid: you really dont use Mathematica? 2011-06-15T03:26:02 mcstar: no 2011-06-15T03:26:09 i mean yes 2011-06-15T03:26:12 i really don't :P 2011-06-15T03:26:25 interesting, i thought every mathematician on the planet has it 2011-06-15T03:26:27 :D 2011-06-15T03:26:40 meh, i'm a big fan of open source 2011-06-15T03:26:49 ok, thats understandable 2011-06-15T03:26:49 i have wxMaxima 2011-06-15T03:26:50 mainly because i don't have to pay anything or worry about pirating it 2011-06-15T03:26:56 sage 2011-06-15T03:27:02 im trying to switch 2011-06-15T03:27:16 meh, i'm happy with normal programming stuff 2011-06-15T03:27:21 sage is in python anyway yeah? 2011-06-15T03:27:25 not a huge fan of python :P 2011-06-15T03:27:33 not the whole thing 2011-06-15T03:27:36 *** olexs has quit IRC (Quit: Leaving.) 2011-06-15T03:27:48 it interfaces my subprograms, and yes, some part are in python 2011-06-15T03:27:48 once i bother learning haskell well enough, i don't expect people will turn me back to anything but c++ and haskell 2011-06-15T03:27:54 my->many 2011-06-15T03:28:25 c++ is so close to my perfect language :( 2011-06-15T03:28:40 it cant be the choice for everything 2011-06-15T03:29:03 mathematica is very good for prototyping, i mean really very good 2011-06-15T03:29:03 "use the right tool for the problem"? never! 2011-06-15T03:29:09 i have fairly acute interests that require programming :P 2011-06-15T03:29:17 they mostly require squeezing the most out computationally 2011-06-15T03:29:39 mcstar: I like JavaScript and the HTML canvas for prototyping :p 2011-06-15T03:29:59 i mean more like scientific prototyping 2011-06-15T03:30:01 *** UncleVasya has quit IRC (Ping timeout: 260 seconds) 2011-06-15T03:30:35 i do theory stuff other than ai and my games stuff 2011-06-15T03:30:56 theory sucks :p 2011-06-15T03:31:00 ergh :P 2011-06-15T03:31:02 applied sucks! 2011-06-15T03:31:04 hey, im not trying to converty anyone or wtf, just saying that those CAS has its many applications, and im surprised at you antimatroid that you dont use one 2011-06-15T03:31:19 applied is too messy for my liking, and I understand stuff better if I build from the bottom up 2011-06-15T03:31:25 rather than top to bottom 2011-06-15T03:31:45 mcstar: yeah, sorry, I'm just saying why I haven't used it :P 2011-06-15T03:31:54 i'm sure there'll come a time where I probably find it useful 2011-06-15T03:32:21 one upside for whatever work you do, is the notebook frontend 2011-06-15T03:32:32 either in mma or in sage 2011-06-15T03:32:44 in sage i can inline latex code 2011-06-15T03:33:10 or any other scripting i would like for that matter 2011-06-15T03:33:27 lets say python/bash/whatever interface they made, even mathematica in sage 2011-06-15T03:34:03 thats very handy, you can see the input with the output, and the whole notebook is saved wo explicit saving 2011-06-15T03:34:09 you can publish it... 2011-06-15T03:34:43 *** smiley1983 has joined #aichallenge 2011-06-15T03:34:52 i guess you have to try it to understand how convenient that is 2011-06-15T03:36:54 yeah 2011-06-15T03:37:03 i've added LaTeX functions to my games library 2011-06-15T03:37:15 so i don't have to go through the horror of manually formatting games 2011-06-15T03:37:51 hey, can i see your house? 2011-06-15T03:38:51 http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=Bath+St,+Battery+Point+TAS,+Australia&aq=0&sll=-42.902313,147.336617&sspn=0.024237,0.066047&ie=UTF8&hq=&hnear=Bath+St,+Battery+Point+Tasmania,+Australia&ll=-42.894159,147.333403&spn=0.00303,0.008256&t=h&z=18&layer=c&cbll=-42.894167,147.333555&panoid=CACaXubvKKvmnqVhyk7Pkw&cbp=12,294.41,,0,15.85 2011-06-15T03:38:59 i still live with my father, so not my house :P 2011-06-15T03:39:07 the one on the left 2011-06-15T03:39:37 in the corner? 2011-06-15T03:40:01 nah, the one with the hedge 2011-06-15T03:40:08 the hedge blocked it more from front on 2011-06-15T03:40:28 near the beach, sweet 2011-06-15T03:40:33 oh wait 2011-06-15T03:40:39 i don't think i copied the link properly :P 2011-06-15T03:40:55 mcstar: I'm using these statements too muffle warnings: https://github.com/aichallenge/aichallenge/commit/ffdf138f6cf3d8c54e81cfd04689ff9f5bc47c17 2011-06-15T03:40:58 lol, coming down from up the street, theres a cyclist, driving in the intersection 2011-06-15T03:41:02 is that you? 2011-06-15T03:41:10 http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=Tower+Rd,+New+Town+TAS,+Australia&aq=0&sll=-42.903665,147.327465&sspn=0.006091,0.016512&ie=UTF8&hq=&hnear=Tower+Rd,+New+Town+Tasmania+7008,+Australia&ll=-42.852986,147.302628&spn=0.003048,0.008256&t=h&z=18&layer=c&cbll=-42.85301,147.302472&panoid=EkaFXp2hCfJZ-BbvVyrqCA&cbp=12,149.53,,0,13.16 2011-06-15T03:41:11 mcstar: ran into the same problems as you when I started working on a real bot 2011-06-15T03:42:38 too bad, you have some variables that get assigned functions at runtime? 2011-06-15T03:42:48 mcstar: my house wasn't in the first link, i failed to copy it :P 2011-06-15T03:43:15 *** qacek has quit IRC (Quit: WeeChat 0.3.5) 2011-06-15T03:43:15 i was returned a lambda with closure, and thats why sbcl was giving me warnings 2011-06-15T03:43:52 returned->returning 2011-06-15T03:45:55 *** UncleVasya has joined #aichallenge 2011-06-15T03:47:37 mcstar: no, not yet anyway 2011-06-15T03:47:51 then why the warnings? 2011-06-15T03:48:00 mcstar: SBCL could really use a more uniform (and actually working) way of controlling these warnings 2011-06-15T03:48:38 mcstar: optimisation declartions in some functions and a 3rd-party lib 2011-06-15T03:49:39 oh, yeah sweet optimization :D 2011-06-15T03:49:42 antimatroid: cool, you've got a shag bay nearby :) 2011-06-15T03:49:54 haha yeah, you can go camping there 2011-06-15T03:49:59 but there's way cooler places up the east coast 2011-06-15T03:50:03 those can really give a couple of page warnings 2011-06-15T03:50:28 110 swanston road? 2011-06-15T03:50:49 me? 2011-06-15T03:50:53 nope 2011-06-15T03:50:57 argh 2011-06-15T03:51:02 i'm in tower road :P 2011-06-15T03:51:09 http://www.google.com.au/search?q=wineglass+bay&hl=en&biw=1920&bih=897&prmd=ivnscm&source=lnms&tbm=isch&ei=SGT4TbfBH4_EvgP-hZ2DDA&sa=X&oi=mode_link&ct=mode&cd=2&sqi=2&ved=0CBsQ_AUoAQ that's up the east coast of tassie 2011-06-15T03:51:23 you have to walk for like an hour from any parking over the saddle of a mountain 2011-06-15T03:51:45 but there's always tourists there, so cooler places to go if you want to go camping/drinking 2011-06-15T03:53:00 the water looks nice, but even on a hot day it's damn cold! 2011-06-15T03:53:56 oh, a marina 2011-06-15T03:54:15 does it come straight from antarctica? 2011-06-15T03:54:16 *** sigh has quit IRC (Quit: This computer has gone to sleep) 2011-06-15T03:54:55 http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=Wineglass+Bay,+Freycinet,+Tasmania,+Australia&aq=2&sll=-42.881903,147.323815&sspn=0.775854,2.113495&ie=UTF8&hq=&hnear=Freycinet+Tasmania,+Australia&ll=-42.212754,148.332596&spn=0.196053,0.528374&t=h&z=12 2011-06-15T03:55:09 it goes out to the ocean, but it's not the south end 2011-06-15T03:57:16 its looks just like miami 2011-06-15T03:57:19 -s 2011-06-15T03:57:34 (i dont know why i make so many spelling mistakes) 2011-06-15T03:58:22 tassie is way cooler :P 2011-06-15T03:58:45 *** locutus2 has joined #aichallenge 2011-06-15T03:59:32 in the literal sense of the word 2011-06-15T04:00:02 well, looks very nice i give you that 2011-06-15T04:00:38 the only beauty where i live is a nice sunset, and a small pine forest 2011-06-15T04:04:42 *** Naktibalda has joined #aichallenge 2011-06-15T04:13:13 i one downside of comedey central is: the comedist only says "and here we are on beautiful time square", and she gets applauded, like she told the most hilarious joke in the century 2011-06-15T04:20:55 *** sir_macelon has joined #aichallenge 2011-06-15T04:21:13 hi 2011-06-15T04:21:24 hi 2011-06-15T04:23:33 *** onensora has joined #aichallenge 2011-06-15T04:23:41 *** UncleVasya has quit IRC (Ping timeout: 260 seconds) 2011-06-15T04:27:33 *** boegel has joined #aichallenge 2011-06-15T04:35:18 *** UncleVasya has joined #aichallenge 2011-06-15T04:48:12 *** nux67 has joined #aichallenge 2011-06-15T05:01:17 *** guyverthree has joined #aichallenge 2011-06-15T05:01:24 alrighty guys 2011-06-15T05:06:50 *** Ice_Harley has quit IRC (Ping timeout: 252 seconds) 2011-06-15T05:07:26 The Java and CL runtimes are really in eachother's way at times. Is it known yet how much memory the official servers will have? I don't think we should ever have maps with more than 8 players. 2011-06-15T05:08:07 we're saying you will have at least 1gb of memory each 2011-06-15T05:08:11 (i think) 2011-06-15T05:08:24 that's quite a lot 2011-06-15T05:08:38 although i'm a bit worried about our ability to limit a bots memory usagr 2011-06-15T05:08:54 otherwise you hope you get to start up first then take up all the memory and don't release it :P 2011-06-15T05:08:59 well the jail's doing a pretty good job at times :) 2011-06-15T05:09:12 i have no idea about what the jail even is :P 2011-06-15T05:09:27 cool, seems like you guys are close to a new open contest 2011-06-15T05:10:20 boegel: welcome to the party :P 2011-06-15T05:10:25 have you seen the beta server? 2011-06-15T05:17:02 I have 2011-06-15T05:17:17 don't have time right now to give it a shot, but I'll try and find some time 2011-06-15T05:17:34 I will be using Haskell, so I might need to look into a Haskell starters package first 2011-06-15T05:22:19 *** UncleVasya has quit IRC () 2011-06-15T05:22:25 boring ... basic bfs makes me timeout when i have more than 10 ants in python ... 2011-06-15T05:22:45 dr-: what are you doing? 2011-06-15T05:23:05 also, python is going to be pretty slow for this :P 2011-06-15T05:23:25 yeah I know, but I'm pretty good with python :/ 2011-06-15T05:24:07 the c++ bot is pretty awesome though :P 2011-06-15T05:24:20 you built it ? :p 2011-06-15T05:24:24 yes 2011-06-15T05:24:32 that's why ... ;) 2011-06-15T05:24:36 :P 2011-06-15T05:24:58 last time, just to compare c++ and python, I made a loop from 1 to 1000000 (maybe more zeros) 2011-06-15T05:25:17 c++ took less than 1 sec and python 12sec 2011-06-15T05:25:18 antimatroid: C++ is boring to advertise. Don't you have a more exotic language? 2011-06-15T05:25:30 D 2011-06-15T05:25:43 :D definitly my way 2011-06-15T05:25:51 i thought you might like that :P 2011-06-15T05:25:55 I'm just bored because I don't want not to use python just because it's slow 2011-06-15T05:26:14 dr-: take a look at the c++ bot, it's pretty solid 2011-06-15T05:26:24 it's built on the structure of my pw bot 2011-06-15T05:27:00 okay, i'll take a look 2011-06-15T05:27:04 :) 2011-06-15T05:27:10 converting people one at a time 2011-06-15T05:27:21 I need to do something about that. 2011-06-15T05:28:26 Oh look at this: http://www.digitalmars.com/d/2.0/cpptod.html 2011-06-15T05:31:17 mleise: does it let me do any of the following, and if so which... overload anything as an operator, overload operator precedence, have template functions that are declared in the .h equivalent and defined in the .cc equivalent, treat functions as parameters, a type system more like haskells, with type classes etc. I would love type variables in c++ too 2011-06-15T05:31:29 those are my complaints about c++ 2011-06-15T05:31:43 NUMPY 2011-06-15T05:31:57 dr-: cant be 12 sec sorry 2011-06-15T05:33:59 antimatroid: that all can be done in cl as fas as i know 2011-06-15T05:34:08 fas->far 2011-06-15T05:34:18 how gross is the syntax? 2011-06-15T05:34:25 no syntax 2011-06-15T05:34:26 i'm a big fan of most of the c++ syntax 2011-06-15T05:34:30 (function arguments) 2011-06-15T05:34:42 even -,+ is a functions 2011-06-15T05:34:45 -s 2011-06-15T05:34:47 no operators 2011-06-15T05:34:58 operators are functions silly :P 2011-06-15T05:35:01 no 2011-06-15T05:35:09 mathematically speaking :P 2011-06-15T05:35:11 operators are built-ins 2011-06-15T05:35:31 mathematically speaking, not strictly 2011-06-15T05:35:38 pretty much 2011-06-15T05:35:40 antimatroid: overloading anything as an operator? - no. that's horrible to check for the lexer. The language is supposed to be easily parsable. that also answers the operator precedence question *g* 2011-06-15T05:35:59 ok, antimatroid do you consider if a function? 2011-06-15T05:36:03 mleise: meh, i don't care about compile time :P 2011-06-15T05:36:04 antimatroid: there is no preprocessor or .h files. 2011-06-15T05:36:23 hmmm, okay 2011-06-15T05:36:34 antimatroid: It is also meant to be easier for tools like IDEs to figure out the syntax. 2011-06-15T05:36:57 meh, i don't care that much about syntax highlighting, it's nice but doesn't need to be perfect 2011-06-15T05:37:07 i'm much more interested in having flexibility with my code 2011-06-15T05:37:57 antimatroid: You get very nice template functionality which is very flexible. Everything you imagine could be calculated at compile time can be done so I think. 2011-06-15T05:38:00 mleise: I'd be willing to have a few exceptions like . for access, but c++ is really restrictive about which operators you can overload 2011-06-15T05:38:25 hehe, lisp has accessor functions 2011-06-15T05:38:29 no need for . 2011-06-15T05:38:47 functions as parameters: definitly yes! also lazily evaluated arguments in general, which is nice for logging complex expressions for example 2011-06-15T05:38:53 i think i need to look at lisp 2011-06-15T05:39:15 good idea, pls do, ill find a nice article i read 2011-06-15T05:39:16 not sure what a type class is :-( 2011-06-15T05:40:27 mleise: it's a set of types 2011-06-15T05:40:47 like in haskell they have Ord, which is the set of all types which can be ordered (totally?) 2011-06-15T05:41:10 and then you can write functions that are somewhat like templates using type classes, so you enforce some kind of structure on the "generic" type 2011-06-15T05:41:22 jmcarthur would probably kill me for some of that wording :P 2011-06-15T05:42:29 i found it: http://abhishek.geek.nz/docs/features-of-common-lisp/ 2011-06-15T05:42:32 mcstar > try this (a loop from 0 to 100000000) with python (without psyco) and c++ 2011-06-15T05:42:53 time command give me 9.7s and 0.2s 2011-06-15T05:43:03 you wrote 1e6 before 2011-06-15T05:43:29 I wrote 12sec, not far from 9sec ;) 2011-06-15T05:43:42 but this is 1e8 2011-06-15T05:43:49 thats a factor of 100 2011-06-15T05:43:52 antimatroid: I don't think this exists in D. You can check a template argument with isIntegral(T) so you can limit your template to that class of types, yes. 2011-06-15T05:44:11 dr-: but ill try, im interested 2011-06-15T05:44:30 haskell has by far the best type system of the very few languages i have ever looked at 2011-06-15T05:45:30 Say one had a template T, I would like T to be like a variable, so i could call something like if(T == string) 2011-06-15T05:45:58 I think both Delphi and Java understand Classes as objects of their own right and you can limit gemeric types in Delphi to descendants of a certain class. 2011-06-15T05:46:16 Java can go and die :) 2011-06-15T05:46:49 lisp has multiple dispatch, can specialize any arguments, has a sophicticated precedence ordering, that you can tweak 2011-06-15T05:46:57 there are before, after, around methods 2011-06-15T05:47:56 In D you have in and out blocks to check the input and output of a function, too. 2011-06-15T05:48:23 no, thats not it 2011-06-15T05:48:43 it changes the precedence of type dispatch 2011-06-15T05:49:02 but of course i dont know any D 2011-06-15T05:51:42 and I don't know what the precedence of type dispatch is :p 2011-06-15T05:53:56 you can have a method, that specializes its arguments for classes 2011-06-15T05:54:09 the system must have a way, to call the correct method 2011-06-15T05:54:29 you see, theres an ambiguity, that must be resolved 2011-06-15T05:54:31 mcstar: are you talking overloaded methods? 2011-06-15T05:54:34 thats dispatch precedence 2011-06-15T05:54:44 i dont talk overloaded 2011-06-15T05:54:49 hehe, thats cpp 2011-06-15T05:54:57 yes :p 2011-06-15T05:55:03 my world is too small 2011-06-15T05:55:05 but some like it, yes 2011-06-15T05:56:13 is it one class with ambiguous methods or one method that works on many classes? 2011-06-15T05:57:14 you see, in lisp, classes dont own methods, a function that you want specialize in its types, is called a method 2011-06-15T05:57:23 *** UncleVasya has joined #aichallenge 2011-06-15T05:57:27 so naturally when you deal with classes you also deal with methods 2011-06-15T05:57:52 and methods can have multiple parameters, in fact a whole lambda list 2011-06-15T05:58:13 ah lisp has dynamic types so it is not necessary to have a strict binding of methods and classes 2011-06-15T05:58:18 so the correct way is to say, many methods on many classes 2011-06-15T05:58:45 well, the class is itself a type 2011-06-15T05:58:57 and every object in lisp belongs to a class 2011-06-15T05:59:03 the topmost superclass is t 2011-06-15T05:59:11 the true value 2011-06-15T05:59:32 wait... the top CLASS is 'true' ?! 2011-06-15T05:59:47 what, please tell me D and lisp allow overloading of function names? 2011-06-15T05:59:50 i would die without that 2011-06-15T05:59:57 dynamic typing only means that a certain lexical veriable can take any kind of value of any types 2011-06-15T06:00:13 its much more than simple overlaoding 2011-06-15T06:00:23 this is what im trying to tell 2011-06-15T06:00:36 pls antimatroid take a look at the link, its enlightening 2011-06-15T06:00:47 antimatroid: Sure, overloading is supported 2011-06-15T06:00:53 mcstar: i got part way through then distracted by reddit :P 2011-06-15T06:01:01 damn reddit 2011-06-15T06:02:37 mcstar: I never used any other languages than the procedural / OOP types. Except for JavaScript, but even then I write mostly OOP. I wonder why it never happened that I beheld the glory of functional languages :p. 2011-06-15T06:03:39 im writing my second project(this game and the engine) in oop style 2011-06-15T06:03:44 actually oop+functional 2011-06-15T06:04:04 the first one was an opengl mandelbrot viewer in cpp 2011-06-15T06:04:19 so, im really a novice in oop 2011-06-15T06:04:52 *** Zaphus has joined #aichallenge 2011-06-15T06:05:16 and lisp is not functional 2011-06-15T06:05:22 CRCs are something I often recall for OOP: http://en.wikipedia.org/wiki/Class-responsibility-collaboration_card 2011-06-15T06:05:23 its multi-paradigm 2011-06-15T06:05:33 choose what fits your problem 2011-06-15T06:05:55 Basically you think of a class in terms of "what is it responsible for" and "who does it collaberate with" 2011-06-15T06:06:14 *** UncleVasya has quit IRC (Ping timeout: 252 seconds) 2011-06-15T06:06:45 well, i always find it useful to draw the structure of anything im dealing with 2011-06-15T06:06:54 hmm, but lisp has its focus on functional, like D has a focus on imerative even if there are functional parts 2011-06-15T06:07:07 no 2011-06-15T06:07:08 *** MuTaLiSk has joined #aichallenge 2011-06-15T06:07:15 it is advised, though 2011-06-15T06:07:23 but not functional per say 2011-06-15T06:07:37 you have powerful global variablews 2011-06-15T06:07:41 ok, its just that you see lisp on the wikipedia page for functional languages 2011-06-15T06:08:01 the global variables are called special, because they are really dynamic variables 2011-06-15T06:08:03 not lexical 2011-06-15T06:08:49 *** MuTa-ofd00m has quit IRC (Read error: Operation timed out) 2011-06-15T06:08:52 since functions are first class objects, yes, its more functional than C 2011-06-15T06:09:55 now I will have a look at this dispatch thing you talked about. may the web enlighten me 2011-06-15T06:10:25 search for CLOS common lisp object system 2011-06-15T06:11:01 tracing paths with circular brushes for mapgen with rounded ends: http://ants.zeroviz.us/exvYM/ 2011-06-15T06:13:57 *** UncleVasya has joined #aichallenge 2011-06-15T06:14:33 Looks like it's set up for some blocking fun :-) 2011-06-15T06:14:58 one of the most annyoing things is blocking 2011-06-15T06:15:07 working on it :) 2011-06-15T06:15:17 i hate when nubs block me in UrbanTerror 2011-06-15T06:16:42 mcstar: This made it clear to me: http://en.wikipedia.org/wiki/Multiple_dispatch#Common_Lisp 2011-06-15T06:17:10 It is also a good example where you know you could not have done that better in OOP 2011-06-15T06:18:20 *** UncleVasya has quit IRC (Ping timeout: 252 seconds) 2011-06-15T06:18:39 i dont know what other languages have this concept too 2011-06-15T06:18:39 mcstar: Some FPS have soft collisions for teammates, so you get pushed out of them slowly when you are standing in them, but you never get blocked 2011-06-15T06:19:22 Well with templates you can emulate it. But it doesn't feel at home then. 2011-06-15T06:19:28 blocking in the sense we use is when a nub tries to rape you, so by doing this blocks the way you otherwise want to go 2011-06-15T06:20:02 for example im running away bleeding from the enemy, and the nub comes right at you, killing us both 2011-06-15T06:20:34 or, im sniping and i have to dodge the enemy bullet reall fast, and a nub decides to snipe from right next to me, blockiing my dodge 2011-06-15T06:22:31 but the nice thing in lisp, is that you dont have to modify the compiler to add your own things 2011-06-15T06:22:42 yeah I know these situations. and you don't see what happens around you in scoped mode 2011-06-15T06:22:52 yes, exactly 2011-06-15T06:23:34 for example if i wanted to add oop to the language, i could do so 2011-06-15T06:23:36 mcstar: lisp is not scheme. what additions do you mean 2011-06-15T06:23:53 there are real macros 2011-06-15T06:24:03 i dont have to rewrite the compiler, or interpreter 2011-06-15T06:24:16 the languge is designed to modify code 2011-06-15T06:24:30 ah the 'code is data' concept, yes 2011-06-15T06:24:37 what do you mean lisp is not scheme? 2011-06-15T06:24:48 i dont know scheme, but cant be that different 2011-06-15T06:25:07 oh, and in lisp a symbol can have a property list 2011-06-15T06:25:32 for example you can have a function value or a common value assigned to it 2011-06-15T06:26:01 and as the plist suggests, other thigs too(but im not versed in that yet) 2011-06-15T06:27:39 i once watched a video where someone showed how in scheme you basically build up your own language 2011-06-15T06:27:58 well thats at least as true for common lisp 2011-06-15T06:28:17 cl derived from an early version of scheme 2011-06-15T06:28:46 and they defined production standard functionality for it in an ansi standard 2011-06-15T06:29:01 scheme remained academic somewhat 2011-06-15T06:31:17 I know that Halo 1 used LISP for the AI :-) That was the first time I heard of it 2011-06-15T06:31:50 never played halo 2011-06-15T06:31:59 And people went like "I don't know why they used that script language in a game. It is more something for scientific tasks" 2011-06-15T06:32:06 hehe 2011-06-15T06:32:19 lisp is NOT scripting language 2011-06-15T06:32:26 it compiles to machine code 2011-06-15T06:32:30 every time 2011-06-15T06:32:35 at least sbcl does this 2011-06-15T06:33:03 They probably interpreted lisp through a c++ library, but I don't know 2011-06-15T06:33:08 btw im sure u are aware that lisp was associated with ai from their beginnings 2011-06-15T06:34:29 i started watching sicp from mit, it is rally educational 2011-06-15T06:34:42 yes, I'm somewhat aware of this. haven't heard of it in other contexts 2011-06-15T06:34:43 they do amazing things with so few constructs 2011-06-15T06:35:28 there are big commercial applications written in lisp, or at least were written in it 2011-06-15T06:35:51 but i can only say that you have to try it, then see it for yourself 2011-06-15T06:35:52 mcstar: Do you think the execution speed of a lisp program can compare to c++? 2011-06-15T06:36:02 they say definitely yes 2011-06-15T06:36:22 depends on the problem and it'll take you some work though 2011-06-15T06:36:31 it can 2011-06-15T06:36:40 i wonder how. you can redefine the language have dynamic types and resulting dispatches all over the place... 2011-06-15T06:36:46 you only have to add this to your functions (declare (type sometype something)) 2011-06-15T06:36:52 then optimize 2011-06-15T06:37:01 depending on the codding style anyhting badly coded can be coded better in a SLOWER lanugage 2011-06-15T06:37:02 and if your lucky youre done :D 2011-06-15T06:37:25 the resulting code will probably start to look like c++ as well 2011-06-15T06:37:34 lol 2011-06-15T06:37:54 you profile, see whats slow, and potimize that 2011-06-15T06:38:09 btw you can instantly disassemble anthing to machine instructions 2011-06-15T06:38:28 but i guess thats implementation dependent 2011-06-15T06:38:35 you can profile lisp? are there special profilers for it or do you use gprof & co.? 2011-06-15T06:38:45 *** onensora has quit IRC () 2011-06-15T06:38:48 implementation provides it 2011-06-15T06:38:54 mleise: depends on implementation again, but we're usually talking SBCL 2011-06-15T06:39:01 *** onensora has joined #aichallenge 2011-06-15T06:39:16 what i see, is when speed matters use sbcl 2011-06-15T06:39:25 or commercial implementations 2011-06-15T06:39:30 there are a few 2011-06-15T06:39:36 mleise: http://www.sbcl.org/manual/#Statistical-Profiler 2011-06-15T06:40:23 what is very different from cpp is the development process 2011-06-15T06:40:51 you can have a running instance of your program when you redefine something 2011-06-15T06:41:11 couple this to slime, and you have the perfect environment 2011-06-15T06:41:28 if only other languages could do this 2011-06-15T06:41:29 mcstar: i promised mcleopold to make a video of this for ants 2011-06-15T06:41:34 hehe 2011-06-15T06:41:36 I like that very much. Recompiling C++ is relatively slow 2011-06-15T06:41:37 would be nice 2011-06-15T06:41:39 smiley1983: nice, i'd pick a radius that doesn't have that single tip 2011-06-15T06:41:41 if that makes sense 2011-06-15T06:42:15 a single out-standing square in each compass direction, yes 2011-06-15T06:42:50 *** FireFly has joined #aichallenge 2011-06-15T06:42:55 time to add some pathfinding to my bot 2011-06-15T06:43:03 I'm writing a visual debugger this time that a client can connect to and draw stuff into 2011-06-15T06:43:10 oh, and if you have problems, theres a nice guy called Xach on #lisp who answers your questions 2011-06-15T06:43:36 aerique: ist he sweet? 2011-06-15T06:43:37 dr-: how's my bot? :P 2011-06-15T06:43:54 I won't use this bracket-mess-language for good ^^. But I want to know on what I'm missing out. 2011-06-15T06:44:15 mcstar: yeah, he'll be happy to be promoted like that ;-) 2011-06-15T06:44:31 (parentheses) 2011-06-15T06:45:04 mleise: your missing on the nicest syntax ever 2011-06-15T06:45:48 of course not everything is made of sugarcake 2011-06-15T06:46:01 antimatroid: yes, it makes sense :) currently the radius is set automatically based on the size of the maze cells, but it could be reduced a little to make the result rounder. 2011-06-15T06:46:10 for example i still have problems with debugging 2011-06-15T06:47:11 aerique: pathfinding is the next on my todo list 2011-06-15T06:47:38 who can I copy the code from when you are both done? :p 2011-06-15T06:48:04 well, im doing it with collaboration with the military 2011-06-15T06:48:13 in col... 2011-06-15T06:48:45 that means what? 2011-06-15T06:48:59 in collaboration instead of with 2011-06-15T06:49:13 means nothing im kidding, as if my code was top secret 2011-06-15T06:49:47 *** onensora has quit IRC () 2011-06-15T06:49:57 ah, language barriers I guess. some jokes just don't end up right on my end :p 2011-06-15T06:50:03 *** onensora has joined #aichallenge 2011-06-15T06:50:05 im not planning on doing any sophisticated thing, i think you can come up with your own before me anyway 2011-06-15T06:50:48 ill have them delivered by UPS if thats better 2011-06-15T06:51:07 i wonder if the paths we calculate can be reused somehow, but caching paths from all squares to all squares is definitly not possible 2011-06-15T06:51:08 *** onensora has quit IRC (Client Quit) 2011-06-15T06:51:43 i cant code and talk 2011-06-15T06:51:51 i must say goodbye now 2011-06-15T06:51:59 bb 2011-06-15T06:52:04 *** mcstar has left #aichallenge ("WeeChat 0.3.5") 2011-06-15T06:52:05 damn 2011-06-15T06:52:07 rows:=256, cols:=256 -> 2^32 entries 2011-06-15T06:52:12 were you talking about ants again? 2011-06-15T06:52:25 sorry, is that wrong? 2011-06-15T06:52:35 mcstar left right as i was about to get into the convo 2011-06-15T06:52:53 aerique: have you seen my a* example with multiple targets? 2011-06-15T06:53:54 http://pastebin.com/6Dh9Jdxg the numbers are the order in which locations will be taken off the search queue, any .'s adjacent to a number are going to be currently in the search queue 2011-06-15T06:53:56 *** onensora has joined #aichallenge 2011-06-15T06:55:07 it's going to be pretty painfully inefficient with LOTS of targets though 2011-06-15T06:55:57 ergh, you can also probably swap which direction you search to minimise the number of "targets", as h(x) is going to be like o(n) 2011-06-15T06:56:08 excuse my terrible understanding of complexity crap :P 2011-06-15T06:57:16 antimatroid: rounding the ends meant chaging one character: http://ants.zeroviz.us/SZW16/ :) 2011-06-15T06:57:36 much nicer ;) 2011-06-15T06:57:55 that's a sweet map 2011-06-15T06:58:03 middle battle ground, then tunnels on the edges 2011-06-15T06:58:23 however, I think the maze generator should be ditched entirely in favour of one based entirely on random walks 2011-06-15T06:58:40 why not have both? 2011-06-15T06:58:53 i'm a big fan of variation with maps 2011-06-15T06:58:55 the maps I'm posting are then exceptions, not the rules among the mazes :) 2011-06-15T06:58:56 provided they're symmetric 2011-06-15T06:59:04 *the 2011-06-15T07:00:09 that is my favourite map i've seen so far 2011-06-15T07:00:11 it's perfect 2011-06-15T07:00:17 there are several unfortunate things about the maze generator for actual games: the algorithm it's based on tends toward long boring dead ends and a complete absence of loops 2011-06-15T07:00:24 could maybe be bigger :P 2011-06-15T07:00:29 adding some random walks can help with this 2011-06-15T07:00:36 that's a 120x120 map, I think 2011-06-15T07:00:47 i made maze maps by accident when i was making maps 2011-06-15T07:00:49 with a grain size of 12, so based on a 10x10 maze 2011-06-15T07:01:02 i switched some paramaters and tada, i accidentally had myself maze maps 2011-06-15T07:01:04 it was quite funny 2011-06-15T07:01:28 it made sense once one thought about it, but it was still a surprise at the time :P 2011-06-15T07:01:40 :) yes, mazes are quite beautifully simple and yet there's so much variation :) 2011-06-15T07:01:45 i guess it doesn't need to be larger 2011-06-15T07:01:54 smaller even 2011-06-15T07:02:00 120x120 isn't going to be feasible 2011-06-15T07:02:10 there's another problem with the maze maps in symmetric rotation: 2011-06-15T07:02:44 it has a very strong tendency to create maps which look like a symbol with a bad and politically sensitive past 2011-06-15T07:02:57 meh? 2011-06-15T07:02:57 lol, I didn't want to mention it 2011-06-15T07:03:17 yes, I've been carefully filtering it out of any maps I post 2011-06-15T07:03:20 I was imagining this map gen in auto run mode on the server and me waiting for the s... 2011-06-15T07:03:28 but the generator makes them, mostly 2011-06-15T07:03:40 can i see one? 2011-06-15T07:03:42 still too soon ? ~75 years 2011-06-15T07:03:50 so it cannot be used 2011-06-15T07:03:52 if you run the code yourself :) 2011-06-15T07:03:53 not to mention the symbol has been used for other things 2011-06-15T07:04:28 http://failblog.org/2008/08/17/bbq-design-fail/ 2011-06-15T07:04:36 for bbqs? 2011-06-15T07:04:41 it's a problem with symmetric rotation generally, but maps based on random walks will reduce the tendency 2011-06-15T07:04:41 you guys have seen the 2012 olympic logo right? 2011-06-15T07:04:56 smiley1983: have you done these maps with 8 ants yet? 2011-06-15T07:05:33 http://whatisthepyramid.com/wp-content/uploads/2009/08/2012_logo_white_385x450.jpg 2011-06-15T07:05:49 yes, but it breaks food generation even with food=random 2011-06-15T07:06:18 antimatroid: there is neither a swastica nor an SS tag -.- 2011-06-15T07:06:42 I like that they seem to have got a map of Australia in there (although Tassie has gone wandering) 2011-06-15T07:06:54 mleise: it looks like lisa simpson giving a blowjob 2011-06-15T07:06:57 cannot unsee 2011-06-15T07:07:26 antimatroid: oh no, now i see it too 2011-06-15T07:07:43 it's perfect 2011-06-15T07:08:17 you can fold up the image of the queen on our $5 notes to make it look like a blowjob too 2011-06-15T07:08:23 Zaphus: you would know this? 2011-06-15T07:09:44 http://fc05.deviantart.net/fs24/f/2007/314/b/1/Secret_Penis_On_Aussie_Money_by_odin123.jpg 2011-06-15T07:09:57 if you fold it over further down it looks more like a mouth giving it rather than a fish 2011-06-15T07:13:34 mleise: my friend had the best comeback "its actually a lot harder to see '2012' in that" 2011-06-15T07:20:04 antimatroid: sorry, was away for lunch. no, hadn't seen your a* stuff before. i'm playing with h(x) weights now 2011-06-15T07:20:35 aerique: the h(x) function there is a big costly, h(x) = min_T dist(x, T) where T are targets 2011-06-15T07:20:40 but it's admissible 2011-06-15T07:20:47 bit* 2011-06-15T07:21:17 and the search queue minimises h if there is a draw for min f 2011-06-15T07:22:41 oh wait 2011-06-15T07:22:52 derp, i can generate h in one sweep at the beginning 2011-06-15T07:22:57 that makes a* way better 2011-06-15T07:23:26 oh, no, nevermind 2011-06-15T07:23:41 i reduce the number of targets each iteration, damn! 2011-06-15T07:23:55 well, it'd still be admissible i guess, but not as accurate 2011-06-15T07:24:21 and try to find food where there really isn't any 2011-06-15T07:24:22 hey guys i'm havign a problem i have undated my tools and now run one game doesn't work 2011-06-15T07:25:46 antimatroid: glad i could help 8-D 2011-06-15T07:27:22 no worries fiured it out now :) 2011-06-15T07:28:50 aerique: are you using a* with multiple targets/sources? 2011-06-15T07:29:01 multiple targets is the hard part 2011-06-15T07:29:51 antimatroid: no, i only just added a* code to my bot 2011-06-15T07:30:26 i haven't programmed anything for that 2011-06-15T07:30:41 i've been trying to work my stuff out before programming properly this time 2011-06-15T07:30:58 antimatroid: played a little with the h(x) weight which is already interesting 2011-06-15T07:31:04 I always skip the "coding properly" step, seems like too much work :-) 2011-06-15T07:31:18 antimatroid: i'm back to real-life work now though :-| 2011-06-15T07:34:04 *** george1914 has quit IRC (Quit: Page closed) 2011-06-15T07:36:35 Zaphus: it's not worth the messiness and bugs 2011-06-15T07:37:32 *** nux67 has quit IRC (Ping timeout: 252 seconds) 2011-06-15T07:41:49 *** delt0r_ has joined #aichallenge 2011-06-15T07:43:40 there is one simple thing airports could do that would really make everyone's life better, and i don't understand why they don't 2011-06-15T07:44:12 there should be a big line painted around baggage claim things with big signs saying stand outside of the line unless you are actually stepping in to pick up your luggage 2011-06-15T07:44:40 i have never seen a lack of space being a problem, the crowd could then swallow people after they've picked stuff up and the world would be a better place 2011-06-15T07:45:19 I've been to airports that have a line... makes little or no difference :-( 2011-06-15T07:52:06 bah, people! 2011-06-15T07:52:23 i hate everything about participating in queue's 2011-06-15T07:54:54 a1k0n_: I think I might have just got what you meant the other day 2011-06-15T07:55:26 was it to paint out the h values with a bfs from all targets, then when i select a target do a kind of inverse search from that to update the information/ 2011-06-15T08:03:00 http://pastebin.com/xhHzs5gP 2011-06-15T08:03:20 the numbers there are different for each example, but it shows the difference in the number of locations one extracts from the search queue 2011-06-15T08:03:55 i have assumed no wrapping with bfs too :P 2011-06-15T08:06:07 what's that A* 2011-06-15T08:06:11 a* 2011-06-15T08:06:25 http://en.wikipedia.org/wiki/A*_search_algorithm 2011-06-15T08:06:59 i have h(x) := min_T manhattendist(x, T) 2011-06-15T08:07:03 oh 2011-06-15T08:07:32 *** delt0r_ has quit IRC (Read error: Operation timed out) 2011-06-15T08:07:35 antimatroid: thats what I was thinking would be a good way of doing it, should be relatively cheap to do 2011-06-15T08:07:56 not that I have any A* implemented in my bot 2011-06-15T08:08:09 Zaphus: except h(x) is expensive to calculate for every square added to the search queue if you have lots of target squares 2011-06-15T08:08:17 i haven't implemented any of this 2011-06-15T08:08:41 which IDE c++ developpers usually use with linux ? 2011-06-15T08:08:43 i'm trying to work out a way to update a grid of h values if i extract a target 2011-06-15T08:08:55 dr-: code::blocks with mingw compiler! 2011-06-15T08:09:09 *** delt0r_ has joined #aichallenge 2011-06-15T08:09:18 you can download that as one thing 2011-06-15T08:09:30 mingw is for window, no ? 2011-06-15T08:09:41 pretty sure i have it on linux, lemme check 2011-06-15T08:10:14 maybe not, well i just installed code::blocks alright and it's wonderful :) 2011-06-15T08:10:23 code::blocks is cross platform at least 2011-06-15T08:10:35 antimatroid: mingw is the Minimalistic GNU for Windows - on linux you don't need a minimalistic one :) 2011-06-15T08:10:56 smiley1983: that would explain why i didn't have problems with my timer on windows :P 2011-06-15T08:12:09 8 player map based on random walks: http://ants.zeroviz.us/hzK3K/ - after a bit of experimentation, I think these may still need some manual filtering, but there is a far better ratio of useable maps 2011-06-15T08:12:15 dr-: probably doesn't need to be said, but create a project for your bot and you can compile it to .exe with like ctrl f11 and not have to worry about linking etc. 2011-06-15T08:12:31 I used once code blocks, and it was just horrible (for me) compared to eclipse 2011-06-15T08:12:37 :O 2011-06-15T08:13:12 smiley1983: that's pretty cool, but not what i meant :P 2011-06-15T08:14:00 I used it on windows, dunno if it was windows or code blocks that made that feeling 2011-06-15T08:14:12 ah, you were thinking of mirroring octants? 2011-06-15T08:14:23 rotating, rather? 2011-06-15T08:14:35 smiley1983: http://pastebin.com/f72B0r8H 2011-06-15T08:15:10 even if shortcuts exist for compiling without doing anything, you'll need to write a make for the official server 2011-06-15T08:15:18 if you have loc = (a,b), then you have [a][b], [rows-a-1][b],... 2011-06-15T08:15:31 s/make/Makefile 2011-06-15T08:15:41 no make files for the official server 2011-06-15T08:15:43 i don't think 2011-06-15T08:15:55 it works out all the linking stuff for you, i've no idea how to do makefiles 2011-06-15T08:16:06 i would struggle to link c++ files properly 2011-06-15T08:17:12 it's written in the c++ starter package that if you include a makefile, the server would use it 2011-06-15T08:17:25 they were planning to 2011-06-15T08:17:30 but i don't think it's materialised 2011-06-15T08:18:06 i mostly wrote that bot at like the start of the year 2011-06-15T08:19:31 okay ... so i don't have to worry about the linking stuff 2011-06-15T08:19:45 antimatroid: I'm afraid I don't follow that notation... do the a and b refer to the letters in the ascii diagram? 2011-06-15T08:22:13 smiley, nah , a and b are numbers specifying a location 2011-06-15T08:22:52 if you pick out a location in the top left quadrant i think that'd work, you end up with 8 symmetric locations to place ants/food/water/whatever 2011-06-15T08:22:53 okay... I'm failing to understand how it can be made symmetrical on the grid 2011-06-15T08:23:44 is this making 8 rotationally symmetrical map segments, or finding starting locations for 8 players on the existing maps? 2011-06-15T08:25:37 you pick a single location to place ants, say (0, 2) (so f = a), then you have {(0,2), (dim-1, 2), (0,dim-1-2), (dim-1,dim-1-2), (2,0), (2, dim-1), (dim-1-2, 0), (dim-1-2, dim-1)} 2011-06-15T08:25:52 those are your ants starting locations 2011-06-15T08:26:08 then do a similar process when placing water/food/land 2011-06-15T08:27:50 yeah, i don't think you even need to pick (a,b) to be in the top left quadrant, should come out alright 2011-06-15T08:28:18 you can't pick (a,a) for ants locations, otherwise you'll double up locations 2011-06-15T08:28:33 and some other rules too i think 2011-06-15T08:28:54 Okay, forgive me if I'm being incredibly dense, but won't this result in starting positions which are dissimilar, unless the mapgen uses the same logic for its tile placement? 2011-06-15T08:29:12 tile placement? 2011-06-15T08:29:29 the idea is to make the map generator add water with that symmetry in mind 2011-06-15T08:29:31 well, water or land 2011-06-15T08:29:40 erm 2011-06-15T08:29:41 right, yes, now I see 2011-06-15T08:29:45 i screwed up the dimensions 2011-06-15T08:29:54 I know, but I get the idea :) 2011-06-15T08:29:59 i meant to have the same number of rows and cols, but does it actually matter? 2011-06-15T08:30:04 i don't think it does in the end 2011-06-15T08:30:49 I could write the generator to use this sort of symmetry... 2011-06-15T08:31:12 at some point, though, the question will arise about whether it would be better to have all of this in Python 2011-06-15T08:31:29 yeah, it should be in python :P 2011-06-15T08:31:38 i will get onto maps again after next thursday 2011-06-15T08:31:38 My enthusiasm for map generation is beginning to wane, and I'm pretty much ready to go back to my own program 2011-06-15T08:31:54 bah, maps are one of the funnest problems :P 2011-06-15T08:32:03 but fair enough 2011-06-15T08:32:14 so I think I'll probably leave my efforts about where they are until/unless there's some call for it to be updated 2011-06-15T08:32:43 I might return to it at some point, because you are right, map generation is fun 2011-06-15T08:33:17 It's more fun in roguelike games (another subject entirely...) because a wider variation of map types is tactically interesting for the player 2011-06-15T08:33:18 i understand getting sick of it though 2011-06-15T08:33:30 I think this game favours mostly-open maps for good games 2011-06-15T08:33:37 yep 2011-06-15T08:33:48 but i think some kind of "layout" for pathfinding is good 2011-06-15T08:33:55 makes it more interesting to watch at least 2011-06-15T08:35:12 I might return and write some different map generation algorithms in Python once that code is more developed, and can do several types of rotation (if that actually happens) :) 2011-06-15T08:35:35 it will :) 2011-06-15T08:36:59 okay, time for me to get up and stretch and not sit in front of the computer for a while :) 2011-06-15T08:44:39 *** smiley1983 has quit IRC (Quit: Lost terminal) 2011-06-15T08:47:41 *** Accoun has quit IRC (Ping timeout: 250 seconds) 2011-06-15T08:47:45 *** Accoun has joined #aichallenge 2011-06-15T08:57:07 *** olexs1 has joined #aichallenge 2011-06-15T08:59:20 *** delt0r_ has quit IRC (Ping timeout: 252 seconds) 2011-06-15T09:01:51 *** Zaphus has quit IRC (Quit: Page closed) 2011-06-15T09:03:32 *** guyverthree has quit IRC (Quit: Page closed) 2011-06-15T09:08:56 *** olexs1 has quit IRC (Quit: Leaving.) 2011-06-15T09:13:24 *** george1914 has joined #aichallenge 2011-06-15T09:30:19 *** aerique has quit IRC (Quit: ...) 2011-06-15T10:30:59 *** carlop has joined #aichallenge 2011-06-15T11:14:44 *** sigh has joined #aichallenge 2011-06-15T11:18:38 *** george1914 has quit IRC (Ping timeout: 252 seconds) 2011-06-15T11:18:38 *** astevenson has quit IRC (Ping timeout: 252 seconds) 2011-06-15T11:19:22 *** sir_macelon has quit IRC (Ping timeout: 252 seconds) 2011-06-15T11:22:20 *** greghaynes has quit IRC (Ping timeout: 252 seconds) 2011-06-15T11:22:55 *** FireFly has quit IRC (Ping timeout: 260 seconds) 2011-06-15T11:24:17 *** BaronTrozo has joined #aichallenge 2011-06-15T11:24:43 *** FireFly has joined #aichallenge 2011-06-15T11:24:58 *** greghaynes has joined #aichallenge 2011-06-15T11:27:15 *** BaronTrozo has quit IRC (Client Quit) 2011-06-15T11:33:45 *** boegel has quit IRC (Ping timeout: 264 seconds) 2011-06-15T11:44:47 *** sigh has quit IRC (Quit: This computer has gone to sleep) 2011-06-15T11:51:06 *** davidd has quit IRC (Ping timeout: 260 seconds) 2011-06-15T11:56:48 *** nux67 has joined #aichallenge 2011-06-15T12:00:41 *** locutus2 has quit IRC (Ping timeout: 240 seconds) 2011-06-15T12:03:15 *** bhasker has quit IRC (Quit: bhasker) 2011-06-15T12:27:45 *** davidd has joined #aichallenge 2011-06-15T12:28:53 *** smiley1983 has joined #aichallenge 2011-06-15T12:34:16 *** sigh has joined #aichallenge 2011-06-15T12:41:19 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 4.0.1/20110413222027]) 2011-06-15T12:48:36 *** sigh has quit IRC (Quit: This computer has gone to sleep) 2011-06-15T12:49:30 *** sigh has joined #aichallenge 2011-06-15T12:51:18 *** bhasker has joined #aichallenge 2011-06-15T12:53:01 *** guyverthree has joined #aichallenge 2011-06-15T12:53:28 hey all anyone else having a problem with making moves ??? 2011-06-15T12:53:43 my bot issues orders but it doesn't move my ants 2011-06-15T12:53:45 *** sigh has quit IRC (Read error: Connection reset by peer) 2011-06-15T12:54:06 are you checking for errors ? 2011-06-15T12:54:57 would that not result in a timeout or crash ??? 2011-06-15T12:55:35 that would be informed 2011-06-15T12:55:48 i thought that that would be the case 2011-06-15T12:57:53 playgame.py -l visualizer -R -S -I -O -E -v [more options] should give you 0.bot1.error in the visualizer-dir 2011-06-15T12:59:31 whcih one generates the error file ???? 2011-06-15T12:59:41 -E 2011-06-15T12:59:55 together with -l dir 2011-06-15T13:00:45 *** Kingpin13 has joined #aichallenge 2011-06-15T13:00:49 *** carlop has quit IRC (Read error: Connection reset by peer) 2011-06-15T13:04:55 i'm confused because my run command looks nothing like the one you say i should be using 2011-06-15T13:05:14 playgame.py --player_seed 42 --end_wait=0.25 --verbose --log_dir game_logs --turns 100 --map_file maps\symmetric_maps\symmetric_10.map [array of bots after this] 2011-06-15T13:06:44 just insert -E -l visualizer after playgame.py 2011-06-15T13:07:31 *** davidd has quit IRC (Ping timeout: 260 seconds) 2011-06-15T13:08:08 thanks i don't gert a log file out 2011-06-15T13:10:35 playgame.py -E -I -O --player_seed 42 --end_wait=0.25 --verbose --log_dir game_logs --turns 100 --map_file maps\symmetric_maps\symmetric_10.map [array of bots after this] // sorry, there was a log folder already in your command 2011-06-15T13:13:24 *** davidd has joined #aichallenge 2011-06-15T13:13:29 *** amstan has joined #aichallenge 2011-06-15T13:13:29 *** ChanServ sets mode: +o amstan 2011-06-15T13:13:40 and the array of bots starts at 0 yeha so if my bot is first in the arrray my bot is number one 2011-06-15T13:13:57 number 0 i mean 2011-06-15T13:22:32 the game logs don't match up with the debug i am getting i'm really confused 2011-06-15T13:22:48 it has different turns in both for the moves 2011-06-15T13:22:52 at least you got them ;( 2011-06-15T13:23:04 very true 2011-06-15T13:23:36 even if i go back to a realy simple bot I.E. more stupid than the first one i get stupid results 2011-06-15T13:28:58 in fact it's since i upgraded the tool yesterday 2011-06-15T13:29:46 tool only, or pulled in the whole repo ? 2011-06-15T13:33:59 i downloaded the new tools yesterday and now my bots doesn't want to work properly 2011-06-15T13:34:36 I'll just get the tools again and start from scratch 2011-06-15T13:34:47 better pull from http://github.com/aichallenge/aichallenge 2011-06-15T13:34:49 and the bot it won't take long ot port the code. 2011-06-15T13:35:33 I have to go now anyway lunch is over:( 2011-06-15T13:35:36 *** guyverthree has quit IRC (Quit: Page closed) 2011-06-15T13:35:52 good lu.. 2011-06-15T13:42:21 hell, BaronTrozo showed up again.. 2011-06-15T14:07:03 *** amstan has quit IRC (Ping timeout: 250 seconds) 2011-06-15T14:26:34 *** UncleVasya has joined #aichallenge 2011-06-15T14:39:13 *** Naktibalda has joined #aichallenge 2011-06-15T14:49:28 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.87 [Firefox 4.0.1/20110518052719]) 2011-06-15T14:50:45 *** Naktibalda has joined #aichallenge 2011-06-15T15:01:55 i'm staring at the appendix of 'the complete strategyst' (RAND_CB113-1.pdf), yeah, 3 pages of random numbers printed on paper ! 2011-06-15T15:03:04 what's puzzling me is, we all know, how to throw dice, but how would you generate a +2 digit random number without any computers ? 2011-06-15T15:06:52 *** eruonen has joined #aichallenge 2011-06-15T15:07:07 *** onensora has quit IRC () 2011-06-15T15:09:53 *** Naktibalda has quit IRC (Ping timeout: 250 seconds) 2011-06-15T15:13:34 *** berak has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2011-06-15T15:16:52 *** amstan has joined #aichallenge 2011-06-15T15:16:52 *** ChanServ sets mode: +o amstan 2011-06-15T15:16:54 *** Naktibalda has joined #aichallenge 2011-06-15T15:28:24 *** smiley1983 has quit IRC (Quit: leaving) 2011-06-15T15:31:53 *** george1914 has joined #aichallenge 2011-06-15T15:34:31 *** boegel has joined #aichallenge 2011-06-15T15:35:38 *** Accoun has quit IRC () 2011-06-15T15:46:37 *** eruonen has quit IRC () 2011-06-15T15:49:00 *** mceier has joined #aichallenge 2011-06-15T15:49:30 *** Accoun has joined #aichallenge 2011-06-15T16:00:27 *** bhasker has quit IRC (Read error: Operation timed out) 2011-06-15T16:01:27 *** UVs has joined #aichallenge 2011-06-15T16:04:58 *** mcstar has joined #aichallenge 2011-06-15T16:09:06 *** bhasker has joined #aichallenge 2011-06-15T16:15:45 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.87 [Firefox 4.0.1/20110518052719]) 2011-06-15T16:34:24 *** CIA-79 has quit IRC () 2011-06-15T16:34:27 *** boegel has quit IRC (Quit: Leaving) 2011-06-15T16:37:34 *** sigh has joined #aichallenge 2011-06-15T16:42:52 *** bhasker has quit IRC (Remote host closed the connection) 2011-06-15T16:43:19 *** CIA-48 has joined #aichallenge 2011-06-15T16:43:42 *** bhasker has joined #aichallenge 2011-06-15T16:59:59 *** UVs has quit IRC (Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org) 2011-06-15T17:04:56 *** UncleVasya has quit IRC (Ping timeout: 260 seconds) 2011-06-15T17:07:06 *** olexs has joined #aichallenge 2011-06-15T17:15:17 *** aumfer has joined #aichallenge 2011-06-15T17:15:32 *** sigh has quit IRC (Quit: This computer has gone to sleep) 2011-06-15T17:25:15 *** sigh has joined #aichallenge 2011-06-15T17:26:16 *** sigh has joined #aichallenge 2011-06-15T17:26:54 *** sigh has quit IRC (Client Quit) 2011-06-15T17:28:43 *** Palmik has quit IRC (Remote host closed the connection) 2011-06-15T17:29:28 *** sigh has joined #aichallenge 2011-06-15T17:30:14 i was just reminded: ICFP contest is this weekend 2011-06-15T17:30:14 http://www.icfpcontest.org/ 2011-06-15T17:31:20 *** sigh has quit IRC (Client Quit) 2011-06-15T17:32:25 *** bhasker has quit IRC (Quit: bhasker) 2011-06-15T17:36:52 *** bhasker has joined #aichallenge 2011-06-15T17:43:33 *** sigh has joined #aichallenge 2011-06-15T17:46:48 a1k0n_, what do you have to do for that contest... I can't find much details regarding that on the site 2011-06-15T17:46:58 yep 2011-06-15T17:47:12 only announcements 2011-06-15T17:47:29 functional programmin, but programming what? 2011-06-15T17:47:45 FP is functional programming? 2011-06-15T17:47:57 its written in full 2011-06-15T17:48:04 international conference on functional programming contest 2011-06-15T17:48:16 but then it is a functional programming programming contest (according to the heading) :P 2011-06-15T17:48:23 sigh: it's usually a pretty interesting problem which you don't necessarily need to use a functional language to solve 2011-06-15T17:48:52 why do you need to run a specific system? 2011-06-15T17:48:57 you can use whatever. previous years had you piloting a mars rover, solving a series of puzzles on a virtual machine (the puzzles themselves being very computer-sciencey ones) 2011-06-15T17:49:10 *** nann has quit IRC (Read error: Connection reset by peer) 2011-06-15T17:49:10 is it more like engineering or computer science? 2011-06-15T17:49:18 it's more like computer science. definitely. 2011-06-15T17:49:31 but it's organized by completely different people every year, so it varies a lot 2011-06-15T17:49:37 *** nann has joined #aichallenge 2011-06-15T17:49:43 i see 2011-06-15T17:49:46 did you participate in it b4? 2011-06-15T17:49:50 they need a better landing page 2011-06-15T17:49:50 and i don't know what the deal with the linux-on-kvm is 2011-06-15T17:49:59 many times. i forgot about it last year. 2011-06-15T17:50:11 mega1/bocsimacko also 2011-06-15T17:50:23 how many teams usually register? 2011-06-15T17:50:30 ...lots? 2011-06-15T17:50:37 *** Zaphus has joined #aichallenge 2011-06-15T17:50:43 like this one? 2011-06-15T17:50:47 in the thousands? 2011-06-15T17:50:56 here's 2006: http://www.boundvariable.org/scoreboard.shtml 2011-06-15T17:51:28 were you part of a team? 2011-06-15T17:51:33 2008: We received 140 lightning round submissions and 336 regular round submissions. 2011-06-15T17:52:03 i think every time i've worked alone, and then sometimes joined up with one other guy 2011-06-15T17:53:34 well, its going to be too hard for me 2011-06-15T17:55:27 do they release the codes after the contest is over? 2011-06-15T17:55:45 or is this binary only? 2011-06-15T17:56:11 up to the organizers, i don't know 2011-06-15T17:56:28 but if you're in the running for top 10 or so, then you need to submit your source 2011-06-15T17:56:38 and i don't think they release them for you 2011-06-15T17:57:34 did you research the field the problem was connected to, or did you start coding right away? 2011-06-15T17:57:50 im asking about the complexity of the problem 2011-06-15T17:58:28 a couple years ago we had to essentially compute orbits and i had to do a bit of research in celestial mechanics (but that wasn't absolutely necessary) 2011-06-15T17:58:46 well, we had to compute thrust vectors to obtain a particular orbit, i should say 2011-06-15T17:58:51 which is surprisingly nontrivial 2011-06-15T17:59:31 guess numerical integrators showed up 2011-06-15T18:00:35 well, the funny thing was you were essentially provided with the "machine" that numerically integrated the ship's path. the first task was just to write an interpreter for the virtual machine that computed the orbit, so that everyone had the same exact program to test against 2011-06-15T18:01:07 the last several years have had virtual machines of some sort. 2007 was especially interesting as the machine code was DNA and the machine operations were string operations 2011-06-15T18:01:33 (not real DNA, just some silly thing) 2011-06-15T18:01:55 whoa, lets not get ahead of ourselfs 2011-06-15T18:02:14 u had a virtual machine that computed the path given inital conditions? 2011-06-15T18:02:33 and it had an instruction set? 2011-06-15T18:02:51 and the problem to be solved was to write in that language? 2011-06-15T18:03:58 or maybe i should understand this, as that you were given a black box, that contained the numerics, and you had to use that for your course planner 2011-06-15T18:04:33 and the problem was to do what? wite an interface? 2011-06-15T18:11:13 *** Tilon has joined #aichallenge 2011-06-15T18:15:30 a1k0n_: so whats the punch line? to code in the machine language of a hypothetical machine? 2011-06-15T18:15:52 a couple of guys at my uni love that contest 2011-06-15T18:15:56 *** amstan_ has joined #aichallenge 2011-06-15T18:15:56 *** ChanServ sets mode: +o amstan_ 2011-06-15T18:16:12 no, the problem was to provide inputs to the machine however you want 2011-06-15T18:16:20 but yes it had a small instruction set and was pretty easy to implement 2011-06-15T18:16:42 the machine was just there to validate your solution to a given problem 2011-06-15T18:16:55 without, say, doing a roundtrip to the contest webserver 2011-06-15T18:17:03 hm 2011-06-15T18:17:22 so a given solution is totally independent of this virtual environment 2011-06-15T18:17:28 exactly 2011-06-15T18:17:40 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-06-15T18:17:59 and independent of the surrounding envirionment. why they're doing a linux virtualbox this contest i couldn't say 2011-06-15T18:18:09 though the mars rover contest did that too. 2011-06-15T18:18:22 security reasons i guess 2011-06-15T18:18:29 like the sanboxing here 2011-06-15T18:18:32 right 2011-06-15T18:18:46 actually they say this on the link you gave 2011-06-15T18:18:52 if they're actually going to execute your program instead of you just providing a bunch of input 2011-06-15T18:18:55 at least i remember something liek that :D 2011-06-15T18:18:57 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-06-15T18:19:52 the money sounds appealing 2011-06-15T18:19:56 *** Tilon has quit IRC (Ping timeout: 252 seconds) 2011-06-15T18:20:11 but since they only reward the first 2, its unlikey to get it 2011-06-15T18:20:12 haha guess what 2011-06-15T18:20:16 icfp 2004's theme was ants 2011-06-15T18:20:19 https://alliance.seas.upenn.edu/~plclub/cgi-bin/contest/ 2011-06-15T18:20:47 totally different problem i can see that right away 2011-06-15T18:20:57 its on hexagonal lattice :D 2011-06-15T18:21:16 hexagonal would be better for this contest too 2011-06-15T18:21:30 this is more interesting 2011-06-15T18:21:48 i like when symmetry is broken 2011-06-15T18:21:51 *** sigh has quit IRC (Quit: This computer has gone to sleep) 2011-06-15T18:24:49 eeek they use windows 2011-06-15T18:24:51 oh, its colony based and u had to submit an FSM 2011-06-15T18:25:01 *** antimatroid has quit IRC (Ping timeout: 240 seconds) 2011-06-15T18:25:06 *** antimatroid has joined #aichallenge 2011-06-15T18:25:14 now, im content, this contest is really different 2011-06-15T18:25:15 yep 2011-06-15T18:25:22 Why do they have the same constest that we run? 2011-06-15T18:25:29 they dont 2011-06-15T18:25:34 and they didnt 2011-06-15T18:25:36 oh lol just read your comment 2011-06-15T18:25:42 also that was 7 years ago 2011-06-15T18:25:50 it was in 2004 2011-06-15T18:25:52 yep 2011-06-15T18:26:00 *** Kingpin13 has quit IRC (Quit: this used to be the life, but I don't need another one) 2011-06-15T18:26:17 lol these comments from kingpin, so enlightening 2011-06-15T18:26:33 a1k0n_: hexagonal wouldn't be very nice for number of moves :P 2011-06-15T18:26:53 i'm not a fan of asymmetry with ai either 2011-06-15T18:27:07 true 2011-06-15T18:27:07 well, i am playing around, but not for a contest 2011-06-15T18:28:57 *** sigh has joined #aichallenge 2011-06-15T18:46:42 *** amstan_ has quit IRC (Ping timeout: 255 seconds) 2011-06-15T18:52:02 *** Zaphus has quit IRC (Quit: Page closed) 2011-06-15T19:04:58 *** sigh has quit IRC (Quit: This computer has gone to sleep) 2011-06-15T19:09:32 *** sigh has joined #aichallenge 2011-06-15T19:21:56 *** sigh has quit IRC (Quit: This computer has gone to sleep) 2011-06-15T19:22:57 *** mcstar has quit IRC (Quit: WeeChat 0.3.5) 2011-06-15T19:26:26 *** fmeyer has joined #aichallenge 2011-06-15T19:28:15 *** filter has quit IRC (Ping timeout: 260 seconds) 2011-06-15T19:32:23 *** filter has joined #aichallenge 2011-06-15T19:39:34 *** sigh_ has joined #aichallenge 2011-06-15T19:41:59 *** nux67 has quit IRC (Quit: Page closed) 2011-06-15T19:45:53 *** amstan_ has joined #aichallenge 2011-06-15T19:45:53 *** ChanServ sets mode: +o amstan_ 2011-06-15T19:48:53 *** sigh_ has quit IRC (Quit: This computer has gone to sleep) 2011-06-15T19:53:37 *** aumfer has quit IRC (*.net *.split) 2011-06-15T19:53:37 *** mleise has quit IRC (*.net *.split) 2011-06-15T19:53:38 *** AlliedEnvy has quit IRC (*.net *.split) 2011-06-15T19:53:38 *** acieroid has quit IRC (*.net *.split) 2011-06-15T20:00:00 *** AlliedEnvy has joined #aichallenge 2011-06-15T20:00:00 *** mleise has joined #aichallenge 2011-06-15T20:00:00 *** aumfer has joined #aichallenge 2011-06-15T20:00:00 *** acieroid has joined #aichallenge 2011-06-15T20:03:28 *** znutar has joined #aichallenge 2011-06-15T20:04:33 *** AlliedEnvy has quit IRC (Ping timeout: 255 seconds) 2011-06-15T20:17:08 *** AlliedEnvy has joined #aichallenge 2011-06-15T20:20:47 *** mleise has quit IRC (Quit: Leaving.) 2011-06-15T20:21:10 *** mleise has joined #aichallenge 2011-06-15T20:27:59 *** sigh has joined #aichallenge 2011-06-15T20:37:37 *** bhasker has quit IRC (Remote host closed the connection) 2011-06-15T20:38:25 *** bhasker has joined #aichallenge 2011-06-15T20:46:12 *** MuTa-ofd00m has joined #aichallenge 2011-06-15T20:46:51 *** javagamer has quit IRC (Ping timeout: 240 seconds) 2011-06-15T20:47:07 *** javagamer has joined #aichallenge 2011-06-15T20:48:20 *** sigh has quit IRC (Quit: This computer has gone to sleep) 2011-06-15T20:48:27 *** MuTaLiSk has quit IRC (Ping timeout: 240 seconds) 2011-06-15T20:48:51 *** moondust has quit IRC (Ping timeout: 240 seconds) 2011-06-15T20:48:55 *** moondust has joined #aichallenge 2011-06-15T20:54:31 *** bhasker has quit IRC (Ping timeout: 240 seconds) 2011-06-15T20:56:12 *** sigh has joined #aichallenge 2011-06-15T21:03:23 *** FireFly has quit IRC (Quit: swatted to death) 2011-06-15T21:18:25 *** sigh has quit IRC (Quit: This computer has gone to sleep) 2011-06-15T21:22:41 *** sigh has joined #aichallenge 2011-06-15T21:27:42 *** InsaneMalkavian has quit IRC (Remote host closed the connection) 2011-06-15T21:35:26 *** sigh has quit IRC (Quit: This computer has gone to sleep) 2011-06-15T21:52:41 *** amstan_ has quit IRC (Ping timeout: 240 seconds) 2011-06-15T21:53:00 *** sigh has joined #aichallenge 2011-06-15T21:56:32 *** amstan_ has joined #aichallenge 2011-06-15T21:56:32 *** ChanServ sets mode: +o amstan_ 2011-06-15T22:01:07 *** amstan_ has quit IRC (Ping timeout: 250 seconds) 2011-06-15T22:12:12 *** sigh has quit IRC (Quit: This computer has gone to sleep) 2011-06-15T22:15:07 *** sigh has joined #aichallenge 2011-06-15T22:16:04 *** sigh has quit IRC (Client Quit) 2011-06-15T22:16:44 *** iShurf has joined #aichallenge 2011-06-15T22:27:16 Hello, everybody! 2011-06-15T22:27:48 *** amstan_ has joined #aichallenge 2011-06-15T22:27:48 *** ChanServ sets mode: +o amstan_ 2011-06-15T22:27:56 I registered in ants beta about 16 hours ago but still haven't received the confirmation letter. Can someone help me? What should I do? 2011-06-15T22:29:52 iShurf: register again and click on the activation link on the next page ;) 2011-06-15T22:31:21 mleise: but how can I return my nickname back? :( 2011-06-15T22:32:02 hmm, then you may have to ask amstan_ 2011-06-15T22:32:37 he'll know who can activate your account 2011-06-15T22:32:52 mleise: you have root, mysql in, update user set status=1 where status=0; 2011-06-15T22:33:03 mysql -u contest 2011-06-15T22:33:09 kk 2011-06-15T22:33:49 *** sigh has joined #aichallenge 2011-06-15T22:35:41 iShurf: your username also doesn't really matter during beta 2011-06-15T22:36:49 ok, it is activated 2011-06-15T22:36:56 it works! Thank you for your help:) 2011-06-15T22:37:05 that was quick :P 2011-06-15T22:37:20 amstan_: your sql was totally wrong, but I found the right db and field in the end it :p 2011-06-15T22:38:26 *** mceier has quit IRC (Quit: leaving) 2011-06-15T22:38:44 I like having root access. It makes me feel important :) 2011-06-15T22:41:09 *** amstan_ has quit IRC (Ping timeout: 255 seconds) 2011-06-15T22:44:41 mleise, yes people with root access tend to feel that way 2011-06-15T22:45:17 over here they have shirts that say "I have root @ google" 2011-06-15T22:45:54 hehe 2011-06-15T22:46:27 (they meaning the people who actaully have root) 2011-06-15T22:47:30 hehe, so they could stop google.com from working. impressive in deed 2011-06-15T22:57:02 *** amstan has joined #aichallenge 2011-06-15T22:57:02 *** ChanServ sets mode: +o amstan 2011-06-15T23:03:23 *** qacek has joined #aichallenge 2011-06-15T23:34:25 *** olexs has quit IRC (Quit: Leaving.)