2013-03-09T02:33:06 *** srgpqt has quit IRC (Quit: leaving) 2013-03-09T03:33:02 *** UncleVasya has joined #aichallenge 2013-03-09T03:43:50 *** sigh has joined #aichallenge 2013-03-09T04:56:55 *** epicmonkey has joined #aichallenge 2013-03-09T05:11:46 *** rofer has quit IRC (Ping timeout: 272 seconds) 2013-03-09T05:15:26 *** rofer has joined #aichallenge 2013-03-09T05:16:42 *** skynet_ has joined #aichallenge 2013-03-09T05:18:22 *** skynet_ has quit IRC (Client Quit) 2013-03-09T05:33:30 *** kilae has joined #aichallenge 2013-03-09T06:04:47 *** Garf has joined #aichallenge 2013-03-09T06:25:25 *** epicmonkey has quit IRC (Ping timeout: 260 seconds) 2013-03-09T06:29:48 *** foRei has quit IRC (Quit: Bye) 2013-03-09T06:46:39 *** coeus_ has joined #aichallenge 2013-03-09T07:22:50 *** GarfTop has joined #aichallenge 2013-03-09T07:23:15 *** Garf has quit IRC (Ping timeout: 276 seconds) 2013-03-09T07:33:41 *** UncleVasya has quit IRC (Quit: UncleVasya) 2013-03-09T07:34:27 *** UncleVasya has joined #aichallenge 2013-03-09T07:44:53 *** sigh has quit IRC (Remote host closed the connection) 2013-03-09T08:28:38 *** mcstar has joined #aichallenge 2013-03-09T08:32:15 *** GarfTop has quit IRC (Quit: Make a new plan, Stan!) 2013-03-09T08:50:29 *** coius has joined #aichallenge 2013-03-09T08:50:44 *** coius is now known as Guest91383 2013-03-09T08:54:07 *** coeus_ has quit IRC (Ping timeout: 260 seconds) 2013-03-09T09:16:25 *** Devnix has joined #aichallenge 2013-03-09T09:18:13 Hi :) 2013-03-09T09:18:29 Does someone know if there will be another contest like ants soon? 2013-03-09T09:23:23 i am not unsure whether or not there will be any such contest anytime soon with any probability 2013-03-09T09:23:49 *** epicmonkey has joined #aichallenge 2013-03-09T09:24:15 joking aside, we dont know when the next aichallenge contest comes 2013-03-09T09:24:49 *** Scooper has joined #aichallenge 2013-03-09T09:58:51 mcstar: good job at solving that puzzle and _not_ being flashed 2013-03-09T09:59:01 hey 2013-03-09T09:59:07 flashed? 2013-03-09T09:59:19 well, it happened for me for the 2x2 case, sort of 2013-03-09T09:59:24 that's what we say around here. maybe it is bad english 2013-03-09T10:00:11 mleise: i will be more 'flashed' once i solve this haskell type system trick 2013-03-09T10:00:23 we also call mobile phones "handy" as if it was an english word 2013-03-09T10:01:27 mleise: i have a puzzle for you, it is very educational 2013-03-09T10:01:30 are you ready? 2013-03-09T10:01:31 haskell type system trick. i don't know if i want to know more about that 2013-03-09T10:01:34 not a logic puzzle 2013-03-09T10:01:47 ok, shoot 2013-03-09T10:02:13 i think of a positive integer coefficient polynom 2013-03-09T10:02:22 you have to tell me what it is 2013-03-09T10:02:36 wait let me google that 2013-03-09T10:02:36 you can ask me the value of the polynom at 2 points 2013-03-09T10:02:39 nooo 2013-03-09T10:02:45 dont google anything 2013-03-09T10:02:52 if you have questions ask me 2013-03-09T10:03:07 a polynom is this, right? : a+b*x+c*x²... 2013-03-09T10:03:11 yes 2013-03-09T10:03:20 the coefficients are integers and > 0 2013-03-09T10:03:39 by asking the value of it at two x's you have to tell me what the polynom is 2013-03-09T10:03:46 (i.e. tell me all the coefficients) 2013-03-09T10:03:55 that works? 2013-03-09T10:04:00 yes 2013-03-09T10:04:09 alright, x can be negative 2013-03-09T10:04:20 it can 2013-03-09T10:04:46 and there is no limit to the number of elements? 2013-03-09T10:04:53 or coefficients 2013-03-09T10:04:54 no, there isnt 2013-03-09T10:05:00 nope 2013-03-09T10:05:06 so it must be a simple one 2013-03-09T10:05:12 what do you mean? 2013-03-09T10:05:15 it can be anything 2013-03-09T10:05:18 damn 2013-03-09T10:05:26 as many terms as i think of 2013-03-09T10:05:36 what if you think of 100 terms? 2013-03-09T10:05:40 thats ok too 2013-03-09T10:05:48 it would take ages to find that out 2013-03-09T10:06:34 mleise: you can do it from 2 evaluations 2013-03-09T10:07:14 I guess I can get a by asking for x=0 2013-03-09T10:08:01 sadly i didnt figure this out fully, only went halfway with it, but i had a few beers in me when i was asked 2013-03-09T10:08:37 but homebound i fully grasped the solution, and it is quite nice 2013-03-09T10:09:16 mleise: imagine 3 of us in a pub, loudly talking about logic puzzles... 2013-03-09T10:09:32 ok at x>1 the terms should leave a distinguishable signature 2013-03-09T10:10:18 oh I had an awkward situation with a group in a cinema waiting for the start of the movie 2013-03-09T10:11:20 wanna tell me about it? 2013-03-09T10:11:25 no. ok so where was I? 2013-03-09T10:11:30 o.O 2013-03-09T10:11:30 ah x=2 2013-03-09T10:12:00 you tell me there was an awkward, possibly funny situation, but not tell me what was it about.. 2013-03-09T10:12:06 i dont like your style, man.. 2013-03-09T10:12:43 it was funny for us until reality caught us in form an employee coming round and telling us to be quiet because we were heard in one of the cinemas 2013-03-09T10:13:30 i see 2013-03-09T10:13:38 you party crashers 2013-03-09T10:13:55 one of the group worked in a video store and talked about fetish movies and from there people had the craziest ideas and questions and we almost rofled 2013-03-09T10:20:13 does it have to do with prime numbers? 2013-03-09T10:21:15 no 2013-03-09T10:30:31 everything x>1 is a valid "base" in terms of a dual system (x=2) or decimal system (x=10). so it doesn't make sense to look for a specific x>1. any will do 2013-03-09T10:30:47 maybe I need 2 of them. then I'd chose 2 and 3 2013-03-09T10:31:36 maybe you are in the right track, but you need to think this over better 2013-03-09T10:32:17 I'll do 2013-03-09T10:32:24 dual system -> binary number system 2013-03-09T10:33:23 hi. can i bore you with a simplier task? 2013-03-09T10:33:33 simpler 2013-03-09T10:33:53 bore means also to bore a hole... 2013-03-09T10:33:54 catched me :D 2013-03-09T10:34:13 here we go: 2013-03-09T10:34:49 also, does your name come from Uncle Vazya? a play about a russian count or something? 2013-03-09T10:35:10 vanya 2013-03-09T10:35:37 sorry, a professor, not count 2013-03-09T10:35:55 idk how, but i navigated there on wikipedia 2013-03-09T10:36:05 c*3² + d*3³ == (c+3)*3² + (d-1)*3³ => a single x > 1 is not enough. the y value can be the result of multiple coefficient combinations 2013-03-09T10:36:09 i have a 2d grid: Cell boad[BOARD_HEIGHT][BOARD_WIDTH] and a list of coordinates [list is an array to but it's irrelevant]. 2013-03-09T10:36:49 *** Garf has joined #aichallenge 2013-03-09T10:37:04 and i need to return a list that consist of those coordinates plus coordinates of their neighbours 2013-03-09T10:37:44 sounds like ants UncleVasya 2013-03-09T10:38:15 so i do following: for every cell in first list i loop on its neighs and if it wasn't added before i add it to the list 2013-03-09T10:39:00 mleise: some board game, but not ants 2013-03-09T10:39:05 the problem is that if(cell_added[row][col]) takes too much time (as profiler says) 2013-03-09T10:39:08 i think he is into board games these days 2013-03-09T10:39:21 yep, this is for Game of Life. 2013-03-09T10:39:34 well, thats not really a game 2013-03-09T10:39:43 Game Of Life was what I had in mind a second ago ;) ah an performance optimization problem 2013-03-09T10:39:46 stupid name, as far as im concerned 2013-03-09T10:39:59 since you said "not yet in the list" 2013-03-09T10:40:04 that didn't sound like a regular game 2013-03-09T10:40:17 i'm optimizing simulation function 2013-03-09T10:40:45 yeah first of all use D for programming 2013-03-09T10:40:47 UncleVasya: are you just trying to simulate GoL and nothing else? 2013-03-09T10:41:16 mleise: yoiu should just solve the riddle first... you D evangelist! 2013-03-09T10:41:20 XD 2013-03-09T10:41:50 so my question is: do you know some faster way to get cells + neighbours or can you suggest some data structure or another technical trick. 2013-03-09T10:42:15 mcstar: ai for 2-player GoL. 2013-03-09T10:42:17 it is a basic task of stencil computations 2013-03-09T10:42:28 ok next I'd say that negative x doesn't solve the problem of "needing two of x > 1", because x² and x⁴ terms are unaffected and can cause ambiguos results by themselves 2013-03-09T10:42:36 just iterate over the cells, and read out the nbors... 2013-03-09T10:43:08 UncleVasya: i dont understand the performance issue, how large your grid is? 2013-03-09T10:43:17 29x29 2013-03-09T10:43:22 lol thats nathing 2013-03-09T10:43:24 o 2013-03-09T10:43:39 sry, brb in 5-10 mins 2013-03-09T10:43:46 i think i did a 1000x1000 at 60 fps, idk 2013-03-09T10:44:02 and that's in haskell even! 2013-03-09T10:44:38 no 2013-03-09T10:44:41 i think lisp 2013-03-09T10:44:56 It's funny how tasks that were solved on 20Mhz CPUs are now impossible on dual-core 3 Ghz machines ^^ 2013-03-09T10:47:04 UncleVasya: http://hpaste.org/83741 2013-03-09T10:47:51 i think it has the Brian's brain rules, not GoL 2013-03-09T10:47:56 but almost the same thing 2013-03-09T10:51:13 i'm here 2013-03-09T10:51:32 i'm pasting code now 2013-03-09T10:52:13 so, does your name come from Uncle Vanya or not? 2013-03-09T10:52:43 no, it has another meaning 2013-03-09T10:54:06 in Russian uncle Vasya generally means 'some guy whose name I don't even know'. 2013-03-09T10:55:53 does uncle have any relevance to the eglish 'uncle'? 2013-03-09T10:56:02 like: - I know you cheated on me! - wtf, why do you think so? - one my friend told me. - so we're breaking up just because some uncle Vasya told you I'm cheated on you?! 2013-03-09T10:56:28 i see 2013-03-09T10:56:39 ah you are a Hans Otto! 2013-03-09T10:57:40 i am still at loss to find such a hungarian expression 2013-03-09T10:57:47 or - and now we'll learn this cool theorem that was proved in emm... 19xx by... emmm... some uncle Vasya 2013-03-09T10:57:50 i am either too stupid, or there isnt one 2013-03-09T10:58:08 damn how do i find two values of x such that the difference of the factors results in all factors that are not multiples of another? 2013-03-09T10:58:14 *was proven 2013-03-09T10:58:34 mleise: nice puzzle, aint it? 2013-03-09T10:58:42 yes, nice 2013-03-09T10:59:14 in the meantime: "Could not deduce (t ~ (Bool -> t)) 2013-03-09T10:59:16 " 2013-03-09T11:04:46 here's my code: http://ideone.com/45HVjh 2013-03-09T11:05:57 if(!cell_added[row][col]){ is the most time consuming part 2013-03-09T11:07:20 i dont understand why would that be the most time consuming 2013-03-09T11:07:25 except if you call it many times 2013-03-09T11:07:54 what i dont understand also, is what does this cell_added mean 2013-03-09T11:08:01 I tried to look at the bitmaps (or something like this) as a replacement for an array of bools but i think it will optimize only memory in this case and will be slower in terms of time 2013-03-09T11:08:03 the update rule is local 2013-03-09T11:08:13 you dont need this stuff, as far as i am concerned 2013-03-09T11:08:27 yes, many times. i run thousands of simulations 2013-03-09T11:08:41 UncleVasya: do you have 2 boards or one? 2013-03-09T11:11:27 mcstar: i came up with my own algo for simulations and the task 'get list of cells and return list of them + neighbours' is used several times. 2013-03-09T11:12:03 is it a local rule, or not? 2013-03-09T11:13:09 let's call it local for now 2013-03-09T11:14:09 this bool array is needed only for one thing 2013-03-09T11:15:31 basically, int cells_to_process_num; Coords cells_to_process; bool cell_added[BOARD_FULL_HEIGHT][BOARD_FULL_WIDTH] can be seen as a one data structure. 2013-03-09T11:16:32 i should refactor this to a struct i think 2013-03-09T11:19:19 if the code seems okay to you then there is nothing i can do with it to improve speed 2013-03-09T11:20:07 UncleVasya: i just cant get my head around, why dont you iterate over the 1D representation of the first array, and write to the second array 2013-03-09T11:20:32 you can ofc do it with 2D arrays too 2013-03-09T11:21:07 what do you mean by first and second? 2013-03-09T11:21:33 you operate on 2 arrays 2013-03-09T11:21:46 the first contains the state of the simulation at a given time 2013-03-09T11:22:01 and you write to the second 2013-03-09T11:22:11 no, it is not the task 2013-03-09T11:22:27 my task is not to make another step 2013-03-09T11:22:29 the task is to run the GoL simulation 2013-03-09T11:22:52 since it is based on a local update rule, you just go ahead, and decide for each cell 2013-03-09T11:23:06 you dont need to make a global statistics or anything 2013-03-09T11:23:40 i do :) 2013-03-09T11:24:18 i'm writing a bot for 2-player version of GoD. 2013-03-09T11:24:47 and while evaluating moves 2013-03-09T11:25:11 instead of running separate simulations for every position 2013-03-09T11:26:59 i just merging histories of different simulations 2013-03-09T11:27:13 and it goes pretty fast 2013-03-09T11:27:29 this is what i am proud about :) 2013-03-09T11:27:36 great 2013-03-09T11:27:56 so i'm happy with general algo for running simulations 2013-03-09T11:28:39 just some technical stuff like this takes more time than i think it should 2013-03-09T11:29:48 now i have one thought how to improve this part 2013-03-09T11:31:12 no, this idea is useful only if i need neighs in radius >1 2013-03-09T11:50:00 *** epicmonkey has quit IRC (Ping timeout: 264 seconds) 2013-03-09T11:52:03 *** Devnix has quit IRC (Quit: Page closed) 2013-03-09T11:55:35 *** amstan_ has quit IRC (Ping timeout: 260 seconds) 2013-03-09T12:08:39 *** epicmonkey has joined #aichallenge 2013-03-09T12:55:18 mleise: http://hpaste.org/83746 2013-03-09T12:56:09 i gave it an annotation 2013-03-09T12:56:26 if you give a predicate of n truth variables in ghci 2013-03-09T12:56:32 you get back the truthtable of it 2013-03-09T12:57:02 is this for me? 2013-03-09T12:57:12 it's some sort of functional language 2013-03-09T12:57:56 mleise: haskell maybe? 2013-03-09T12:58:14 thats what i was working on, with the help of #haskell 2013-03-09T12:58:48 so you print a truth table from ... well looks like actual source code? 2013-03-09T12:59:16 you reflect upon the boolean expression and build a truth table 2013-03-09T12:59:24 hm? 2013-03-09T12:59:28 no, not from source code 2013-03-09T12:59:33 it is a simple function 2013-03-09T12:59:44 there is absolutely no reflection involved 2013-03-09T12:59:49 ah ok 2013-03-09T13:01:29 so what's the deal with that snippet? 2013-03-09T13:02:12 i think it is kind of cool, that i can provide an n-ary function, and evaluate it at all possible inputs 2013-03-09T13:02:29 so you do reflect upon the parameter count 2013-03-09T13:02:37 can you do that with D without metaprogramming or reflection? 2013-03-09T13:02:41 no 2013-03-09T13:02:59 when you say n-ary I think n is variable 2013-03-09T13:03:10 it is a constant for a given function 2013-03-09T13:03:16 not has arity 1 2013-03-09T13:03:22 && has arity two 2013-03-09T13:03:32 \a b c -> a && b || c has arity 3 2013-03-09T13:03:39 sure 2013-03-09T13:04:03 and somehow the parameters can be ... indexed and iterated over in haskell? 2013-03-09T13:04:09 but you dont tell the function (tt) this beforehand, that evaluates this given predicate 2013-03-09T13:04:15 mleise: no, cant 2013-03-09T13:04:38 then how does it go over all permutations of parameters? 2013-03-09T13:04:56 you just curry it until curry cant curry anymore or what? :p 2013-03-09T13:05:00 because i defined the tt function to do that 2013-03-09T13:05:06 mleise: well, yes 2013-03-09T13:05:06 *** thestinger has joined #aichallenge 2013-03-09T13:05:16 the type of the function changes with each parameter application 2013-03-09T13:05:32 well, when you fill in a parameter it changes, looses a Bool -> 2013-03-09T13:05:49 it is a typelevel recursion, via typeclasses 2013-03-09T13:06:38 it is very simple (looks simple) but since i hold on to all the truth values for each argument to build a truthtable it looks complicated 2013-03-09T13:08:14 mleise: refresh the page, and see the last annotation 2013-03-09T13:08:30 thats the code, that evaluates an n-ary function at True values 2013-03-09T13:09:12 say, you have \a b c -> a && b || c, then tt (\a b c -> a && b || c) = True && True || True = True 2013-03-09T13:10:52 mleise: do you want the solution to the puzzle? 2013-03-09T13:11:08 no I'll keep it around for times when I feel bored :p 2013-03-09T13:11:20 mleise is tortured :P 2013-03-09T13:11:22 you are almost there... 2013-03-09T13:11:27 I know ;) 2013-03-09T13:12:24 tt (\a b c -> a && b || c) = True && True || True = True 2013-03-09T13:12:27 what is this tt ? 2013-03-09T13:13:00 ah truth table 2013-03-09T13:13:02 it evaluates an arbitrary n-ary predicate 2013-03-09T13:13:06 no, not yet 2013-03-09T13:13:15 mleise: i think we should invest in our future and make mcstar go for a PhD in management so he'll be able to make next aichallenge happen 2013-03-09T13:13:18 at least, it doesnt make a truth table in the last annotation 2013-03-09T13:13:26 mleise: thats where i started from http://hpaste.org/83733 2013-03-09T13:13:37 UncleVasya: hehe, maybe in Hungary this time 2013-03-09T13:13:43 it could evaluate 1-5 arity functions :) 2013-03-09T13:14:05 lol 2013-03-09T13:14:24 i hated micro and macroeconomics soo much, i still havent finished my micro course 2013-03-09T13:14:29 lol 2013-03-09T13:14:45 at least, not the official one, i hope i can make them accept this course in place of the official one 2013-03-09T13:17:49 In D I would do the same: Ask on IRC how. :D 2013-03-09T13:18:21 it is an enlightening example 2013-03-09T13:18:34 similar to how printf works 2013-03-09T13:18:40 eurica! 2013-03-09T13:18:53 yeah and printf sucks and isn't type safe :p 2013-03-09T13:19:01 yes, but this is 2013-03-09T13:19:15 it only works oon bools 2013-03-09T13:19:21 i see. and it uses these cool overloads on function parameters 2013-03-09T13:19:33 not really 2013-03-09T13:19:40 it is a proper type 2013-03-09T13:19:52 what is a type? 2013-03-09T13:20:12 function(some amount of boolean parameters) ? 2013-03-09T13:20:23 A 2013-03-09T13:20:29 a function type: 2013-03-09T13:20:32 A -> B 2013-03-09T13:20:43 everything is made fromt he combinations of those 2013-03-09T13:21:00 i should have written lowe case letters, only those are allowed as type variables 2013-03-09T13:21:07 also a type can take arguments 2013-03-09T13:21:12 a = b c 2013-03-09T13:22:22 a = b c means a(b, c); ? 2013-03-09T13:22:25 well, types can have predicates too 2013-03-09T13:22:28 no 2013-03-09T13:22:37 lol, I'll never get it 2013-03-09T13:22:48 'a' is a type variable that can stand in place of 'b c' 2013-03-09T13:22:53 It's good I stay miles away from Haskell 2013-03-09T13:23:00 where b is a type variable that takes a type argument 'c' 2013-03-09T13:23:23 so, 'b' has kind * -> * 2013-03-09T13:23:29 while 'c' has kind * 2013-03-09T13:23:31 a is a wildcard allright 2013-03-09T13:23:38 kind is the type of types 2013-03-09T13:23:58 mcstar: http://habrahabr.ru/post/169671/ 2013-03-09T13:24:03 and as i said, types can have predicates 2013-03-09T13:24:05 like: 2013-03-09T13:24:15 Show a => a 2013-03-09T13:24:41 that says 'a' is a type variable whenever 'a' is an instance of Show 2013-03-09T13:24:49 allright. so wherever _you_ use a, I'd call it a template argument. 2013-03-09T13:25:02 why? 2013-03-09T13:25:09 it is just like a type in D 2013-03-09T13:25:09 so i understand it 2013-03-09T13:25:18 Int -> int 2013-03-09T13:25:23 Bool -> bool 2013-03-09T13:25:24 when a can be different types it's a template argument 2013-03-09T13:25:29 no way around that 2013-03-09T13:25:41 no, something that has type 'a' is a polymorphic constant 2013-03-09T13:25:53 O.o 2013-03-09T13:26:13 template argument it is 2013-03-09T13:26:16 for example, 1 :: Num a => a 2013-03-09T13:26:38 1 is a number that evaluates to a number? 2013-03-09T13:26:45 so, 1 can be any type that if that type is a part of Num 2013-03-09T13:26:52 be of* 2013-03-09T13:27:03 for example 1 :: Int, or 1 :: Integer 2013-03-09T13:27:12 machine vs. arbitrary precision 2013-03-09T13:28:09 UncleVasya: some kind of game? 2013-03-09T13:28:33 does it simulate a cat chasing a mouse? 2013-03-09T13:29:05 mleise: 1 is a value constructor 2013-03-09T13:29:43 data Int = 1 | 2 | 3 | ... ad inifinitum (up to 2^64 e.g.) 2013-03-09T13:29:50 oh man, why does every single sentence introduce a new scientific word? 2013-03-09T13:30:01 sry 2013-03-09T13:30:02 you know what an ivory tower is? 2013-03-09T13:30:06 yes 2013-03-09T13:30:09 :p 2013-03-09T13:30:12 im looking at you from there 2013-03-09T13:30:13 XD 2013-03-09T13:30:19 kidding 2013-03-09T13:30:37 terminology is important, cause we have to think of the same things, when we talk about something 2013-03-09T13:30:48 mcstar: one guy runs FP-contest on his livejournal every month. this a task from prev month. 2013-03-09T13:31:00 well, not every second month. 2013-03-09T13:31:10 mcstar: it's easier to start off from C or something similar 2013-03-09T13:31:46 mcstar: tasks aren't hard for smart people (i think you can solve any of them in less then 40 mins) 2013-03-09T13:31:47 mleise: i think you should read a little LYAH 2013-03-09T13:31:51 'data' looks like a generator now 2013-03-09T13:32:10 it is the keyword that introduces data structures in haskell 2013-03-09T13:32:14 mcstar i think you'll enjoy participiating 2013-03-09T13:32:25 lol, so Int can only be positive ? 2013-03-09T13:32:35 something is wrong 2013-03-09T13:32:38 mleise: thats how i defined it 2013-03-09T13:32:40 XD 2013-03-09T13:32:41 ok 2013-03-09T13:33:01 lets call those positive integers... 2013-03-09T13:33:16 you need only register in lj before next challenge and make a comment with solution 2013-03-09T13:33:28 mleise: it is not defined in ghc like that, since it is a primitive type, but thats how you would define an enumerations 2013-03-09T13:33:29 -s 2013-03-09T13:33:50 data Bool = False | True 2013-03-09T13:33:56 thats how Bool is defined 2013-03-09T13:34:08 True is a value constructor 2013-03-09T13:34:17 UncleVasya: ok, thanks, also, i am not that smart 2013-03-09T13:34:22 mleise: right! 2013-03-09T13:34:29 mcstar: this guy loves haskell too. you even get bonus points if use it instead of other lang 2013-03-09T13:34:33 it constructs a value of type Bool 2013-03-09T13:34:33 but why is it? 2013-03-09T13:34:53 because the language parses True as built-in value ? 2013-03-09T13:35:10 hm? 2013-03-09T13:35:18 or did you just define it as a new "keyword"? 2013-03-09T13:35:23 no, if there was no Bool in the language, you would define it like that 2013-03-09T13:35:32 mcstar: you are talking about stuff i'll never understand + you publish scientific articles => you work for science. you are a hell smart badass :P 2013-03-09T13:35:38 does data Bool = True | False bring those two into existence in the first place? 2013-03-09T13:35:47 mleise: right 2013-03-09T13:36:09 ah ok, and this is how you think about Int as well 2013-03-09T13:36:35 UncleVasya: then keep in mind, that there are several hell smart badass ppl in this channel besides me 2013-03-09T13:36:36 1 is defined to construct ... well what now ? an Int or an Integer? 2013-03-09T13:36:41 maybe both? 2013-03-09T13:36:53 with a little effort you can be one too(if not already) 2013-03-09T13:36:54 ah! it is polymorhpic! 2013-03-09T13:37:02 a polymorphic constant 2013-03-09T13:37:11 mcstar: every other guy here, i know :D 2013-03-09T13:37:48 this bottom up approach works much better 2013-03-09T13:39:01 mleise: here're links to another 17 tasks: http://users.livejournal.com/_darkus_/689264.html 2013-03-09T13:39:36 afk 2013-03-09T13:43:00 mleise: imagine, the numbers 0,1,2,.., are symbols, and they are instance constants 2013-03-09T13:43:18 and there are real representations of machine integers and bigints too 2013-03-09T13:43:34 and these polymorphic constants will be replace by those, depending on the assumed type 2013-03-09T13:44:05 who is the assumed type determined ? 2013-03-09T13:44:14 some program may not be clear about it 2013-03-09T13:44:21 it is always clear 2013-03-09T13:44:32 ill write an example 2013-03-09T13:44:34 e.g. why isn't 1 a floating point number in the end? 2013-03-09T13:44:58 how does the compiler now it can store it in a machine word for example 2013-03-09T13:51:34 and one more question, why doe you have 12 values for a function taking 4 parameter combinations? 2013-03-09T13:52:10 shouldn't there always be 2^parameter_count values ? 2013-03-09T13:53:00 mleise: http://hpaste.org/83754 2013-03-09T13:53:20 where do i have that? 2013-03-09T13:55:02 UncleVasya: i dont read well russian, i dont think i can compete 2013-03-09T13:55:53 mleise: where do i have 12 values for 4 parameters? 2013-03-09T13:56:14 in the annotation 2013-03-09T13:56:36 2 parameters, 12 values 2013-03-09T13:57:25 can you paste it? 2013-03-09T13:57:26 the link 2013-03-09T13:57:43 http://hpaste.org/83746 2013-03-09T13:57:50 under sample output for pred1 2013-03-09T13:58:07 2 parameters, 2^2 4 cases 2013-03-09T13:58:18 pred1 p q = not p && ((p ==> q) == not (q && not p)) 2013-03-09T13:58:22 of course and that's what I expected 2013-03-09T13:58:23 its a 2-ary function 2013-03-09T13:58:33 _4_ cases :p 2013-03-09T13:58:34 then whats the problem? 2013-03-09T13:58:41 why are there 12 values? 2013-03-09T13:58:49 what do you mean? 2013-03-09T13:58:55 do you know what is a truth table? 2013-03-09T13:59:18