2011-12-31T00:07:19 *** CowTipperVirus has joined #aichallenge 2011-12-31T00:22:15 *** Migi32 has quit IRC (Remote host closed the connection) 2011-12-31T00:38:26 *** grc48 has quit IRC (Quit: Leaving) 2011-12-31T00:38:40 *** sigh has quit IRC (Remote host closed the connection) 2011-12-31T00:38:48 *** grc48 has joined #aichallenge 2011-12-31T00:39:07 *** JorgeB has joined #aichallenge 2011-12-31T00:46:21 *** sigh has joined #aichallenge 2011-12-31T00:59:53 *** CowTipperVirus has quit IRC (Quit: Page closed) 2011-12-31T01:13:56 *** sigh has quit IRC (Remote host closed the connection) 2011-12-31T01:15:03 *** pairofdice has quit IRC (Quit: coalescence cascade) 2011-12-31T01:21:49 *** sigh has joined #aichallenge 2011-12-31T01:29:21 *** wraithonchelle is now known as Wraithan 2011-12-31T01:42:46 *** yoden1 has joined #aichallenge 2011-12-31T01:43:29 *** yoden has quit IRC (Ping timeout: 252 seconds) 2011-12-31T01:48:24 *** MagnumPU has joined #aichallenge 2011-12-31T01:48:52 *** MagnumPU has left #aichallenge 2011-12-31T01:53:02 *** sigh has quit IRC (Remote host closed the connection) 2011-12-31T02:10:41 *** treeform has joined #aichallenge 2011-12-31T02:30:57 *** Paradoxial has quit IRC (Quit: :) 2011-12-31T02:44:39 *** smjms has joined #aichallenge 2011-12-31T03:00:15 *** smjms has quit IRC (Quit: Konversation terminated!) 2011-12-31T03:10:09 *** Kurnevsky has joined #aichallenge 2011-12-31T03:11:13 *** epicmonkey has joined #aichallenge 2011-12-31T03:34:20 *** cseyu01 has quit IRC (Ping timeout: 258 seconds) 2011-12-31T03:49:59 *** Jak_o_Shadows1 has quit IRC (Remote host closed the connection) 2011-12-31T04:48:08 *** alc has quit IRC (Quit: Leaving) 2011-12-31T04:48:48 *** replore_ has joined #aichallenge 2011-12-31T04:52:35 *** amstan has quit IRC (Read error: No route to host) 2011-12-31T04:52:56 *** amstan has joined #aichallenge 2011-12-31T04:52:56 *** ChanServ sets mode: +o amstan 2011-12-31T04:59:33 *** grc48 has quit IRC (Quit: This computer has gone to sleep) 2011-12-31T05:20:00 *** replore_ has quit IRC (Remote host closed the connection) 2011-12-31T05:29:48 *** mleise has joined #aichallenge 2011-12-31T05:32:46 *** mj41 has joined #aichallenge 2011-12-31T05:34:56 *** replore_ has joined #aichallenge 2011-12-31T05:41:37 *** foRei has joined #aichallenge 2011-12-31T06:06:18 *** Xunie` has quit IRC (Quit: Leaving) 2011-12-31T06:16:11 *** mcstar has joined #aichallenge 2011-12-31T06:19:39 *** treeform has quit IRC (Remote host closed the connection) 2011-12-31T06:32:25 *** kilae has joined #aichallenge 2011-12-31T06:36:54 *** Anilm3 has joined #aichallenge 2011-12-31T06:37:12 Hi 2011-12-31T06:43:05 *** kire has joined #aichallenge 2011-12-31T06:47:39 *** Regis has joined #aichallenge 2011-12-31T06:48:02 *** yoden has joined #aichallenge 2011-12-31T06:49:53 *** X-Scale has quit IRC (Ping timeout: 268 seconds) 2011-12-31T06:50:41 *** yoden1 has quit IRC (Ping timeout: 248 seconds) 2011-12-31T07:02:32 *** ahow984_ has joined #aichallenge 2011-12-31T07:04:29 *** Anilm3 has quit IRC (Ping timeout: 240 seconds) 2011-12-31T07:10:23 *** Anilm3 has joined #aichallenge 2011-12-31T07:33:40 *** Anilm3 has quit IRC (Quit: leaving) 2011-12-31T07:48:59 *** ikaros has joined #aichallenge 2011-12-31T07:55:02 *** GeorgeJ has quit IRC (Quit: No Ping reply in 180 seconds.) 2011-12-31T07:55:06 *** GeorgeJ_ has joined #aichallenge 2011-12-31T08:06:38 *** amstan has quit IRC (Ping timeout: 240 seconds) 2011-12-31T08:12:35 *** links234 has joined #aichallenge 2011-12-31T08:15:54 *** grc48 has joined #aichallenge 2011-12-31T08:16:15 *** grc48 has left #aichallenge 2011-12-31T08:25:25 *** kire has joined #aichallenge 2011-12-31T08:26:03 *** links234 has quit IRC (Ping timeout: 258 seconds) 2011-12-31T08:41:51 *** dmj111 has joined #aichallenge 2011-12-31T08:46:53 *** Migi32 has joined #aichallenge 2011-12-31T09:31:21 *** cseyu01 has joined #aichallenge 2011-12-31T09:43:16 *** TheLinker has joined #aichallenge 2011-12-31T09:56:16 *** epicmonkey has quit IRC (Ping timeout: 276 seconds) 2011-12-31T09:57:31 contestbot: later tell mcleopold sorry, i haven't been around the last week or two. there are almost certainly faster ways to do what your code does, with varying tradeoffs 2011-12-31T09:57:31 jmcarthur: Ready to serve. 2011-12-31T10:02:17 *** smjms has joined #aichallenge 2011-12-31T10:04:00 *** replore_ has quit IRC (Remote host closed the connection) 2011-12-31T10:10:51 *** jmcarthur has quit IRC (Read error: Connection reset by peer) 2011-12-31T10:11:29 *** replore has joined #aichallenge 2011-12-31T10:11:39 *** jmcarthur has joined #aichallenge 2011-12-31T10:15:07 *** pairofdice has joined #aichallenge 2011-12-31T10:16:21 *** replore has quit IRC (Ping timeout: 260 seconds) 2011-12-31T10:22:38 *** kire has quit IRC (Remote host closed the connection) 2011-12-31T10:31:59 *** kire has joined #aichallenge 2011-12-31T10:32:46 *** replore has joined #aichallenge 2011-12-31T10:37:08 *** replore has quit IRC (Ping timeout: 240 seconds) 2011-12-31T10:42:01 *** mj41 has quit IRC (Ping timeout: 260 seconds) 2011-12-31T10:44:37 *** epicmonkey has joined #aichallenge 2011-12-31T10:51:46 *** epicmonkey has quit IRC (Read error: Operation timed out) 2011-12-31T10:57:54 *** UncleVasya has joined #aichallenge 2011-12-31T11:24:38 *** mleise has quit IRC (Read error: Connection reset by peer) 2011-12-31T11:25:04 *** kire has quit IRC (Remote host closed the connection) 2011-12-31T11:30:54 *** mj41 has joined #aichallenge 2011-12-31T11:39:23 *** nha has joined #aichallenge 2011-12-31T11:50:04 *** srgpqt has joined #aichallenge 2011-12-31T11:50:56 hi srgpqt 2011-12-31T11:51:05 <-- MettaWorldPeace 2011-12-31T11:51:34 cseyu01: hi :) i've been reading the channel logs haha 2011-12-31T11:51:45 lol 2011-12-31T11:51:58 your bot is pretty solid 2011-12-31T11:52:34 i removed its the attack strategy code, it still needs a lot of work 2011-12-31T11:52:43 but other than that i'm pretty happy about it :) 2011-12-31T11:53:05 I'm working on getting my ants to move out without specifically telling them to distance themselves from each other 2011-12-31T11:53:42 *** smjms has quit IRC (Changing host) 2011-12-31T11:53:42 *** smjms has joined #aichallenge 2011-12-31T11:53:50 yeah, that's pretty essential for exploration 2011-12-31T11:54:22 *** dom7b5 has joined #aichallenge 2011-12-31T11:54:30 I spent yesterday working with graphics so I could actually see the maps that i create within the game in a sldeshow 2011-12-31T11:54:56 *** nha has quit IRC (Ping timeout: 260 seconds) 2011-12-31T11:54:57 alot easier than looking thru a long list of co-ordinates 2011-12-31T11:55:14 ctrl-f "(21, 128)" 2011-12-31T11:55:39 yup, that was a very valuable capability for me as well. i dump PPM images of the game. 2011-12-31T11:56:01 would be nice to be able to be able to view the images live, as they are output.. 2011-12-31T11:56:32 the border between my uncharted territory and my visible area with 0's in a dictionary 2011-12-31T11:56:54 *** sigh has joined #aichallenge 2011-12-31T11:56:56 then add 1's to neighboring land then 2's etc 2011-12-31T11:57:41 *** replore has joined #aichallenge 2011-12-31T11:57:48 it cuts off after a certain amount of ms 2011-12-31T11:57:54 *** replore has quit IRC (Read error: Connection reset by peer) 2011-12-31T11:58:21 so when the map is big it doesn't go far enough to reach my central ants 2011-12-31T11:59:34 I can view images as i go as long as I don't look at one when my program is trying to overwrite it 2011-12-31T11:59:51 *** raemde_ has joined #aichallenge 2011-12-31T12:00:05 I dump images into a folder 2011-12-31T12:00:27 then use faststone image viewer to look at them 2011-12-31T12:00:48 i have one of those wheels on my mouse that spins 2011-12-31T12:00:59 so i can whip through them really fast 2011-12-31T12:01:42 *** raemde has quit IRC (Read error: Operation timed out) 2011-12-31T12:02:32 :) 2011-12-31T12:05:26 *** raemde_ has quit IRC (Ping timeout: 260 seconds) 2011-12-31T12:11:43 *** raemde has joined #aichallenge 2011-12-31T12:15:34 *** treeform has joined #aichallenge 2011-12-31T12:15:38 *** UncleVasya has quit IRC (Read error: Connection reset by peer) 2011-12-31T12:17:01 cseyu01: so you're coding it in C/C++ ? 2011-12-31T12:19:02 *** replore has joined #aichallenge 2011-12-31T12:20:42 *** sigh has quit IRC (Remote host closed the connection) 2011-12-31T12:31:43 no, pyhton 2011-12-31T12:31:49 just learnt it 2011-12-31T12:32:18 ah. 2011-12-31T12:32:25 when I started with the tutorial, I couldn't get their java to run 2011-12-31T12:32:51 and I wanted to start out cutting and pasting 2011-12-31T12:33:25 I could probably speed it up by re-writing in c++ 2011-12-31T12:33:41 I think I have a solution for my map problem :) 2011-12-31T12:34:00 in this challenge i've learned that you can get very good speed in most languages with good code 2011-12-31T12:34:05 keep a static dictionary that's updated every 10 or 20 turns 2011-12-31T12:34:28 and a dynamic one that my program can work on every turn 2011-12-31T12:34:52 then just copy the dynamic to the static at regular intervals 2011-12-31T12:35:15 map info will be a little old sometimes but at least it'll be there 2011-12-31T12:35:18 *** nha has joined #aichallenge 2011-12-31T12:35:43 that way my ants always know where the uncharted territory is 2011-12-31T12:36:09 *** dmj111 has quit IRC (Ping timeout: 240 seconds) 2011-12-31T12:36:57 the numbering works really well when the map info is there, very easy to tell them, "go to the lowest numbered neighboring tile" 2011-12-31T12:37:35 fails miserably when all neighbors have no value 2011-12-31T12:38:58 yes, i do something similar in my bot, but i keep values for the entire map, and reuse the values from the previous turn... 2011-12-31T12:39:50 *** mcstar has quit IRC (Quit: WeeChat 0.3.6) 2011-12-31T12:39:59 need to find a happy medium between being current and staying within time constraints 2011-12-31T12:40:00 each turn, i run the map through a bunch of propagation iterations 2011-12-31T12:40:12 same 2011-12-31T12:40:16 *** replore has joined #aichallenge 2011-12-31T12:40:32 on the next turn i just continue where i left off 2011-12-31T12:41:13 for loop that only ends when there are no new entries 2011-12-31T12:41:56 *** cbad has joined #aichallenge 2011-12-31T12:41:58 *** cbad has joined #aichallenge 2011-12-31T12:42:11 *** cbad has left #aichallenge 2011-12-31T12:43:02 i do this for a lot of my strategies, so for example after a few turns all of my ants (anywhere on the map) can go towards a discovered enemy hill just by looking up the neighboring tile map values 2011-12-31T12:43:52 so it can handle 1000 ants almost as easily as it can handle 10 ants 2011-12-31T12:45:09 wow 2011-12-31T12:46:44 I noticed that when I went map based the processing times for the actual decisions went way down, but I've never had 1000 ants 2011-12-31T12:47:07 I only leave 150ms for decisions 2011-12-31T12:47:21 that's plenty of time for 400-500 2011-12-31T12:48:14 i don't even keep track of turn time.. teehee 2011-12-31T12:49:46 i should measure how long each turn takes, just for fun 2011-12-31T12:52:07 ants.turntime - ants.time_remaining() 2011-12-31T12:52:46 I haven't even touched battle resolution yet 2011-12-31T12:52:58 if (his_crew >= my_crew): score = score - PENALTY 2011-12-31T12:53:12 this is my quickfix for battles ^^ 2011-12-31T12:55:59 *** nha has quit IRC (Ping timeout: 252 seconds) 2011-12-31T13:01:33 *** replore has joined #aichallenge 2011-12-31T13:02:35 time for entire first time : 40.2 ms 2011-12-31T13:02:35 *** Israfel has quit IRC (Read error: Operation timed out) 2011-12-31T13:03:41 *** Israfel has joined #aichallenge 2011-12-31T13:04:32 time for entire first turn : 40.2 ms last turn (235 ants): 58.4 ms 2011-12-31T13:05:08 <3 lua 2011-12-31T13:05:15 I think when you've explored alot more of the map it'll go up significantly 2011-12-31T13:05:38 58.4 on a big map? 2011-12-31T13:05:54 134x196 map 2011-12-31T13:06:00 impressive 2011-12-31T13:06:10 fully explored before the last turn 2011-12-31T13:06:59 my food search takes long because I reset the values every turn 2011-12-31T13:07:31 i reset food too, because they don't stay very long anyway, and i only want close ants to reach them 2011-12-31T13:07:39 but I can still find 15 tile paths to food in under 100ms 2011-12-31T13:08:10 do you store all the values seperately or combine scores? 2011-12-31T13:08:40 i keep seperate dictionaries and then convert to a score at decision time 2011-12-31T13:08:59 i think some ppl just say this tile has x score 2011-12-31T13:09:21 i have separate maps for combined food, combined enemy hills 2011-12-31T13:09:21 @later tell amstan I'm ready to shut off the main server except for not knowing what to do with the forums. I'm a little afraid of putting them on the archive server given that it's so small but I don't really have anywhere else to go with them either. 2011-12-31T13:09:21 janzert: Yep. 2011-12-31T13:09:52 i would like to unify the maps but haven't gotten to it (nor found it necessary yet) 2011-12-31T13:10:12 I don't think unifying them would help 2011-12-31T13:10:29 *** Israfel has quit IRC (Ping timeout: 248 seconds) 2011-12-31T13:11:04 better to know what you're assigning the points for, then to just say this tile has x-value 2011-12-31T13:12:31 *** McLeopold has joined #aichallenge 2011-12-31T13:12:46 jmcarthur: ping 2011-12-31T13:13:13 i have a separate map for each of my hills too now (previously i didnt have any).. a map for exploration, a map for which tiles have been seen, a map for which tiles i've walked on before 2011-12-31T13:14:47 i used to use a dictionary tracking how many times my ants have stepped on each tile 2011-12-31T13:15:28 it helps avoid oscillation and get them to explore 2011-12-31T13:15:37 McLeopold: pong, but i'll brb 2011-12-31T13:16:28 but I'd rather use more direct methods 2011-12-31T13:16:35 i wonder what performance penalty you pay for using dictionaries instead of a list (or better yet, a true array from the array module) 2011-12-31T13:17:22 not much difference in time when i loop thru lists dicts or sets 2011-12-31T13:18:22 for storing whether or not a tile has been explored or if it's land or water i use lists 2011-12-31T13:18:59 McLeopold: alright, so what are you willing to change for optimization? functions only, or also data structures? there are tradeoffs, of course 2011-12-31T13:19:11 but if I need to associate a value like the number of steps from a target I have to use a dict 2011-12-31T13:19:18 jmcarthur: what are we talking about? 2011-12-31T13:19:29 McLeopold: that pastebin thing from days ago 2011-12-31T13:19:38 McLeopold: not sure if you even care anymore, but just following up 2011-12-31T13:19:41 oh, haskell? :) 2011-12-31T13:19:45 yes I do 2011-12-31T13:20:03 I've since read that show should only be used to match read 2011-12-31T13:20:11 that's the ideal, yes 2011-12-31T13:20:12 and that I should create render functions 2011-12-31T13:20:56 so, my main question is, if I care about performance, what is the best way to do a matrix? 2011-12-31T13:21:03 aha 2011-12-31T13:21:10 I would want to continually update values in it 2011-12-31T13:21:11 there are a few ways 2011-12-31T13:21:18 a list of lists is certainly the worst 2011-12-31T13:21:28 also, what is the fastest way to print the map 2011-12-31T13:21:32 alright 2011-12-31T13:22:21 *** Israfel has joined #aichallenge 2011-12-31T13:22:43 there are a few different array libraries. if you are willing to do some mind-bending, repa is pretty nice for higher dimensional arrays and if used intelligently will be very efficient 2011-12-31T13:22:58 *** replore has joined #aichallenge 2011-12-31T13:23:05 *** replore has quit IRC (Read error: Connection reset by peer) 2011-12-31T13:23:16 there's also vector, which is for 1D arrays. to get a matrix, you will have to nest them. this isn't as fast as the unboxed representation you can get from repa 2011-12-31T13:23:45 I did stumble upon repa 2011-12-31T13:23:48 on the other hand, using boxed vectors means it's easier to work with custom data types (no crazy type class instances for unboxing), and the elements can be lazy 2011-12-31T13:23:54 I wasn't sure if it was overkill or not 2011-12-31T13:24:06 another way would be to nest a couple IntMaps 2011-12-31T13:24:18 I'm thinking my Tile data type can't be used unboxed? 2011-12-31T13:24:27 which might sound weird, but it's actually pretty nice to work with since it's purely functional and not entirely too slow 2011-12-31T13:24:41 Tile can be unboxed, but it requires some prep work 2011-12-31T13:24:52 a lot of boilerplate to get there 2011-12-31T13:25:02 the Hill constructor is different from the rest, it takes an Int 2011-12-31T13:25:17 another approach would be to use a hash table. unordered-containers has a decent purely functional one 2011-12-31T13:25:23 yeah, that would be fine 2011-12-31T13:25:32 the hash table means you can just has the coord 2011-12-31T13:25:36 *hash 2011-12-31T13:26:25 since there are a limited number of players, you could pack the whole thing into a little Word8 representation 2011-12-31T13:26:45 without even losing the pattern matching and stuff, thanks to the way the unboxing stuff works 2011-12-31T13:27:00 i use 1d numeric vector without nesting (simply append each row linearly in the vector) in my (non-haskell) implementation 2011-12-31T13:27:17 yea, I thought about doing it the way the starter bot did: data Tile = Land | Water | Enemy1 | Enemy2, but I'd rather have the Hill Player 2011-12-31T13:27:21 srgpqt: that's basically what repa does for you :) 2011-12-31T13:27:27 neat 2011-12-31T13:27:41 *** deltree_ has joined #aichallenge 2011-12-31T13:28:12 so if you Try and get map 3 4, it translate to map 3 * cols + 4? 2011-12-31T13:28:35 McLeopold: there's also the array package, which i don't really like much anymore, but it does allow you to index in interesting ways 2011-12-31T13:28:43 yeah 2011-12-31T13:28:48 so, about printing 2011-12-31T13:28:55 the thing with repa, though, is that it's meant to be used in a particular way 2011-12-31T13:28:57 I'm assuming that my method is slow 2011-12-31T13:29:00 you don't really get direct mutation 2011-12-31T13:29:46 instead you are intended to use various folds and accumulations and stuff, and the mutation is performed under the hood. for some algorithms this probably doesn't work, so you'd have to extract the underlying unboxed vector representation and do the mutation yourself 2011-12-31T13:29:50 Yeah, I would be doing a lot of direct mutation, I think. The plan was to rewrite my cell_maze generator as an excersice in learning Haskell 2011-12-31T13:30:22 mutation is not necessary as often as you might think, but it's hard to wrap your head around some of the folds and slices you do instead with repa 2011-12-31T13:31:04 is concatMap show row fast? is there a better way? 2011-12-31T13:31:26 well, Strings in general are pretty slow 2011-12-31T13:31:34 since they are linked lists :( 2011-12-31T13:31:54 so there are two ways to go about this 2011-12-31T13:32:05 I should be able to build it fairly fast, and print it at the same speed? 2011-12-31T13:32:20 you could go with a ByteString representation, or you could stick with String and do your best to optimize the render function 2011-12-31T13:32:27 which would be your preference? 2011-12-31T13:32:40 idk yet 2011-12-31T13:32:44 k 2011-12-31T13:32:59 so, for your String representation... 2011-12-31T13:33:06 I just didn't want to accidently create an O(n^2) algo 2011-12-31T13:34:07 I didn't like my method of nested printing, because I would never just print 1 row only 2011-12-31T13:34:11 concatMap show is not *terrible*. it's not going to be exponential 2011-12-31T13:35:14 I would know my total string length up front 2011-12-31T13:35:18 "foo" ++ "bar" only has to rebuild the "foo" part. concat ["foo", "bar", "baz"] associates (++) to the right, so it won't rebuild the same parts over and over 2011-12-31T13:35:56 so, using ByteString might be equivilant to a stringbuilder class? 2011-12-31T13:36:19 no, ByteString is just a more compact (sane) string type 2011-12-31T13:36:52 blaze-builder would give you something more like a stringbuilder class 2011-12-31T13:38:29 i would probably either use blaze-builder or, if i care *a whole lot* about speed, roll by own direct renderer into a mutable buffer 2011-12-31T13:38:39 blaze-builder would be much more idiomatic though 2011-12-31T13:39:03 i'm recommending ByteString instead of Text since you don't need unicode 2011-12-31T13:40:03 btw, there is a showList method of the Show type class that can be specialized on showing a list of values 2011-12-31T13:40:25 ok 2011-12-31T13:41:02 anyway, i hope i have been helpful. my advice has been rather open-ended 2011-12-31T13:41:12 i have to go now 2011-12-31T13:41:12 it has :) 2011-12-31T13:41:12 tbc... 2011-12-31T13:43:13 *** contestbot_ has joined #aichallenge 2011-12-31T13:43:21 *** JamesMG_ has joined #aichallenge 2011-12-31T13:44:12 *** hkraal_ has quit IRC (Ping timeout: 252 seconds) 2011-12-31T13:44:12 *** cYmen has quit IRC (Ping timeout: 252 seconds) 2011-12-31T13:44:13 *** 31NAACW1K <31NAACW1K!~jamesq@cpc21-rdng21-2-0-cust111.15-3.cable.virginmedia.com> has quit IRC (Ping timeout: 252 seconds) 2011-12-31T13:44:15 *** replore has joined #aichallenge 2011-12-31T13:44:17 *** hkraal_ has joined #aichallenge 2011-12-31T13:44:24 *** replore has quit IRC (Read error: Connection reset by peer) 2011-12-31T13:44:25 *** rb__ has joined #aichallenge 2011-12-31T13:44:40 *** isbric has quit IRC (Ping timeout: 252 seconds) 2011-12-31T13:44:40 *** mawif has quit IRC (Ping timeout: 252 seconds) 2011-12-31T13:44:41 *** krishna has quit IRC (Ping timeout: 252 seconds) 2011-12-31T13:44:47 *** GeorgeJ has joined #aichallenge 2011-12-31T13:44:47 *** GeorgeJ has joined #aichallenge 2011-12-31T13:44:56 *** GeorgeJ_ has quit IRC (Ping timeout: 252 seconds) 2011-12-31T13:44:58 *** delt0r has quit IRC (Ping timeout: 252 seconds) 2011-12-31T13:45:21 *** JamesMG has quit IRC (Ping timeout: 252 seconds) 2011-12-31T13:45:38 *** TheLinker has quit IRC (Ping timeout: 240 seconds) 2011-12-31T13:45:41 *** mattswe has quit IRC (Ping timeout: 252 seconds) 2011-12-31T13:45:59 *** cYmen has joined #aichallenge 2011-12-31T13:46:08 *** mawif has joined #aichallenge 2011-12-31T13:46:14 *** QGazQ has quit IRC (Ping timeout: 252 seconds) 2011-12-31T13:46:25 *** isbric has joined #aichallenge 2011-12-31T13:46:40 *** krishna has joined #aichallenge 2011-12-31T13:46:58 *** andar has joined #aichallenge 2011-12-31T13:47:23 *** delt0r has joined #aichallenge 2011-12-31T13:47:29 *** srgpqt has quit IRC (Ping timeout: 252 seconds) 2011-12-31T13:47:39 *** mattswe has joined #aichallenge 2011-12-31T13:48:07 *** srgpqt has joined #aichallenge 2011-12-31T13:52:52 *** ikaros has quit IRC (Quit: Ex-Chat) 2011-12-31T13:59:08 *** mj41 has quit IRC (Ping timeout: 240 seconds) 2011-12-31T14:05:31 *** replore has joined #aichallenge 2011-12-31T14:09:47 *** replore has quit IRC (Ping timeout: 252 seconds) 2011-12-31T14:10:55 No snow at new years, sad 2011-12-31T14:10:57 *** sevikkk has quit IRC (Ping timeout: 252 seconds) 2011-12-31T14:11:42 *** sevikkk has joined #aichallenge 2011-12-31T14:16:09 *** dom7b5 has quit IRC (Ping timeout: 240 seconds) 2011-12-31T14:16:16 *** McLeopold has quit IRC (Quit: Leaving.) 2011-12-31T14:21:24 *** Accoun has quit IRC () 2011-12-31T14:26:48 *** replore has joined #aichallenge 2011-12-31T14:29:45 *** dom7b5 has joined #aichallenge 2011-12-31T14:32:22 *** mcstar has joined #aichallenge 2011-12-31T14:32:39 rwest: hey, are you there? 2011-12-31T14:33:31 @later tell rwest 356898446 2011-12-31T14:33:31 mcstar: Ready to serve. 2011-12-31T14:33:58 in 1.9 secs 2011-12-31T14:36:10 ok, actually only 0.76 secs, because of symmetry 2011-12-31T14:38:36 *** Accoun has joined #aichallenge 2011-12-31T14:42:26 *** X-Scale has joined #aichallenge 2011-12-31T14:46:30 *** deltree_ has quit IRC (Quit: ChatZilla 0.9.88-rdmsoft [XULRunner 1.9.0.17/2009122204]) 2011-12-31T14:48:04 *** replore has joined #aichallenge 2011-12-31T14:49:57 *** amstan has joined #aichallenge 2011-12-31T14:49:57 *** ChanServ sets mode: +o amstan 2011-12-31T14:54:43 @later tell rwest lol, ok my integer overflowed, i was suspicious, since they said "it will fit a 64bit int", so the answer is 309594543758 2011-12-31T14:54:43 mcstar: Job's done. 2011-12-31T14:54:46 same time 2011-12-31T14:56:29 *** kilae has quit IRC (Ping timeout: 255 seconds) 2011-12-31T15:07:31 *** dmj111 has joined #aichallenge 2011-12-31T15:09:21 *** replore has joined #aichallenge 2011-12-31T15:09:34 *** replore has quit IRC (Read error: Connection reset by peer) 2011-12-31T15:30:42 *** replore has joined #aichallenge 2011-12-31T15:35:57 *** Paradoxial has joined #aichallenge 2011-12-31T15:46:38 *** UncleVasya has joined #aichallenge 2011-12-31T15:46:57 this is rubbish 2011-12-31T15:47:03 there are multiple overflows 2011-12-31T15:50:24 whats this about? 2011-12-31T15:50:32 sounds like a projecteuler problem :P 2011-12-31T15:50:54 @later tell rwest 806844323190414 is the solution, i had an immediate int that truncated my computation 2011-12-31T15:50:54 mcstar: As you wish. 2011-12-31T15:51:24 @later tell rwest now the result in monotinically increasing in the height parameter 2011-12-31T15:51:24 mcstar: OK 2011-12-31T15:51:25 https://mattmccutchen.net/bigint/ 2011-12-31T15:51:37 no bigint, thx 2011-12-31T15:51:42 if you're using C++, that handles big numbers 2011-12-31T15:51:48 i used gmp before 2011-12-31T15:51:59 *** replore has joined #aichallenge 2011-12-31T15:52:01 kk 2011-12-31T15:52:15 *** replore has quit IRC (Read error: Connection reset by peer) 2011-12-31T15:52:34 ahow984_: https://www.dropbox.com/s/89hd7283yx3546i/Block_Puzzle.pdf 2011-12-31T15:54:49 huh, interesting 2011-12-31T15:54:55 I might try it for practise :) 2011-12-31T15:55:01 what time did you do it in? 2011-12-31T15:55:52 0.7 secs 2011-12-31T15:56:14 kk 2011-12-31T15:56:23 did you mean that, or figuring it out? 2011-12-31T15:57:15 ahow984_: ^^ 2011-12-31T15:57:22 :P I meant runtime 2011-12-31T15:57:35 k 2011-12-31T15:57:52 you know, im an android 2011-12-31T15:57:57 so i wasnt sure 2011-12-31T15:59:36 figured 2011-12-31T16:13:18 *** replore has joined #aichallenge 2011-12-31T16:23:42 mcstar: do you know what is 999899874? 2011-12-31T16:23:59 yes 2011-12-31T16:25:38 UncleVasya: talk to me, what is it? 2011-12-31T16:25:47 (other than an integer) 2011-12-31T16:26:07 *** Kurnevsky has quit IRC (Quit: Instantbird 1.0) 2011-12-31T16:27:13 the answer for one fun problem. 2011-12-31T16:27:21 I saw your big number in above and decided you like this stuff. 2011-12-31T16:27:32 duh 2011-12-31T16:28:27 mcstar: is there a mathematical way to solve this problem? 2011-12-31T16:28:46 I have this number in notepad that close to me. 2011-12-31T16:28:50 what is that criterion? mathematical? 2011-12-31T16:29:02 as in not brute-forced 2011-12-31T16:29:30 the brute force solution takes thinking too 2011-12-31T16:29:35 ive got that first 2011-12-31T16:29:49 but for the required problem size, it is useless 2011-12-31T16:30:11 so theres 2 bits of thinking 2011-12-31T16:30:24 the first one to figure it out A solution 2011-12-31T16:30:33 yeah 2011-12-31T16:30:35 mcstar: when you'll need your c++ knowledge? 2011-12-31T16:30:40 job interview or whatever 2011-12-31T16:30:44 yes 2011-12-31T16:30:45 I thought I had it a smart mathematical way but I was way off 2011-12-31T16:31:09 i was thinking about a pure mathematical solution, like a formula 2011-12-31T16:31:20 did you find one? 2011-12-31T16:31:21 but i couldnt bypass an enumeration pahse 2011-12-31T16:31:29 no 2011-12-31T16:31:41 btw, i did it in haskell 2011-12-31T16:31:44 the brute force one 2011-12-31T16:31:50 weeks ago 2011-12-31T16:31:59 when rwest first presented this 2011-12-31T16:32:01 heh, fair enough 2011-12-31T16:32:08 I didn't see it the first time 2011-12-31T16:32:10 just now 2011-12-31T16:32:18 and i tried to make it better today, but i couldnt do it in haskell 2011-12-31T16:32:28 for the last step i needed mutable arrays 2011-12-31T16:32:40 I think I have a (someway) smart way to do it with vectors in C++ 2011-12-31T16:32:47 and its awkward for me to use that in haskell, so i did it in c++ instead 2011-12-31T16:32:55 good choice ;) 2011-12-31T16:33:01 its under 100 lines 2011-12-31T16:33:25 recursive? 2011-12-31T16:34:11 what is? 2011-12-31T16:34:20 your solution 2011-12-31T16:34:26 which part? 2011-12-31T16:34:29 :) 2011-12-31T16:34:33 *** replore has joined #aichallenge 2011-12-31T16:34:42 i dont want to help 2011-12-31T16:34:44 is there any recursion in it lol 2011-12-31T16:34:45 *** replore has quit IRC (Read error: Connection reset by peer) 2011-12-31T16:34:52 its just a general question 2011-12-31T16:34:55 i mean its better to figure it out yourself 2011-12-31T16:35:00 yeah 2011-12-31T16:35:19 never mind my question, it wasn't that I was trying to copy what you had done 2011-12-31T16:35:32 I'm not good at recursion and was wondering if thats how you did it 2011-12-31T16:35:53 i heavily used recursion in the haskell solution 2011-12-31T16:36:05 (the cats out of the bag..) 2011-12-31T16:36:07 ok 2011-12-31T16:36:14 :P hardly a big secret lol 2011-12-31T16:36:26 but it implies that i dont use it anymore 2011-12-31T16:36:48 I don't want to solve it brute forcedly 2011-12-31T16:37:00 I want to try find an "elegant" solution first 2011-12-31T16:37:13 lol 2011-12-31T16:37:16 wrong 2011-12-31T16:37:24 do it the other way around 2011-12-31T16:37:29 have a solution first 2011-12-31T16:37:38 and after try to find something better 2011-12-31T16:37:41 or at least some way that doesn't involve me going through a 48*10 vector a couple of million times 2011-12-31T16:37:56 48? 2011-12-31T16:38:01 theres no 48 2011-12-31T16:38:01 yes 2011-12-31T16:38:08 its 48/3=16 2011-12-31T16:38:12 :) 2011-12-31T16:38:23 How many different ways are there 2011-12-31T16:38:23 for your niece to build a 48”×10” panel? 2011-12-31T16:38:43 I don't understand why you divide it by 3.... 2011-12-31T16:38:50 My niece would break the pieces 2011-12-31T16:39:03 ahow984_: it just means you havent payed much attention 2011-12-31T16:40:02 no, I'm completely lost now.... 2011-12-31T16:40:48 *** UncleVasya has quit IRC () 2011-12-31T16:41:26 yup, nfi what you're talking about now. nowhere can I see anything to suggest that layer 1 should be made entirely of 3" blocks or anything.... 2011-12-31T16:42:32 you have 3x1 and 4.5x1 blocks 2011-12-31T16:42:55 now, your "units" are count of those elements 2011-12-31T16:43:01 counts* 2011-12-31T16:43:12 i have 16 a and 0 b 2011-12-31T16:43:25 then you can have 13 a and 2 b 2011-12-31T16:43:32 then you can have 10 a and 4 b 2011-12-31T16:43:37 then you can have 7 a and 6 b 2011-12-31T16:43:39 .... 2011-12-31T16:43:49 you dont actually use 48 anywhere 2011-12-31T16:44:05 i->you 2011-12-31T16:44:50 wow, I _never_ would have thought of it that way lol 2011-12-31T16:44:58 what do you mean? 2011-12-31T16:45:05 its the first thing that came to mind 2011-12-31T16:45:13 I was going to have a 96*10 vector of bools :P 2011-12-31T16:45:17 :S 2011-12-31T16:45:26 haha, yeah.... 2011-12-31T16:45:29 id never thought about it that way :) 2011-12-31T16:45:37 strange how people think 2011-12-31T16:45:38 *** amstan has quit IRC (Ping timeout: 240 seconds) 2011-12-31T16:45:58 actually, had you sticked with that, you could not have proceed easily 2011-12-31T16:46:13 then have * set to false, set them to true if there's a block end there 2011-12-31T16:46:18 the next step is very naturally comes from my description 2011-12-31T16:46:27 then check +-1 layer for intersections 2011-12-31T16:46:43 yes, that makes it very different 2011-12-31T16:55:53 *** replore has joined #aichallenge 2011-12-31T17:01:27 WOOO, FIREWORKS 2011-12-31T17:03:33 *** cseyu01_ has joined #aichallenge 2011-12-31T17:04:19 *** cseyu01 has quit IRC (Ping timeout: 258 seconds) 2011-12-31T17:06:54 *** TheLinker2 has quit IRC (Remote host closed the connection) 2011-12-31T17:10:49 *** amstan has joined #aichallenge 2011-12-31T17:10:49 *** ChanServ sets mode: +o amstan 2011-12-31T17:22:33 *** mj41 has joined #aichallenge 2011-12-31T18:01:24 *** sigh has joined #aichallenge 2011-12-31T18:07:21 *** flowenol has joined #aichallenge 2011-12-31T18:11:13 *** smjms has quit IRC (Quit: Konversation terminated!) 2011-12-31T18:29:22 happz new zear 2011-12-31T18:30:57 *** cseyu01_ has quit IRC (Ping timeout: 258 seconds) 2011-12-31T18:32:25 It's old already! 2011-12-31T18:35:37 *** flowenol has quit IRC (Ping timeout: 276 seconds) 2011-12-31T18:42:46 *** replore has joined #aichallenge 2011-12-31T18:47:08 *** replore has quit IRC (Ping timeout: 240 seconds) 2011-12-31T18:51:17 *** sigh has quit IRC (Remote host closed the connection) 2011-12-31T18:59:48 three words: Planes, Trains and Automobiles 2011-12-31T19:03:26 *** sigh has joined #aichallenge 2011-12-31T19:04:06 *** replore has joined #aichallenge 2011-12-31T19:05:20 *** mj41 has quit IRC (Ping timeout: 255 seconds) 2011-12-31T19:08:23 *** replore has quit IRC (Ping timeout: 252 seconds) 2011-12-31T19:37:44 *** dom7b5_ has joined #aichallenge 2011-12-31T19:39:29 *** dom7b5 has quit IRC (Ping timeout: 240 seconds) 2011-12-31T19:52:00 *** ahow984_ has quit IRC (Quit: leaving) 2011-12-31T19:54:58 *** dom7b5_ has quit IRC (Read error: Connection timed out) 2011-12-31T19:55:35 *** dom7b5_ has joined #aichallenge 2011-12-31T19:58:34 *** mcstar has quit IRC (Quit: WeeChat 0.3.6) 2011-12-31T20:12:55 *** dom7b5_ has quit IRC (Read error: Connection timed out) 2011-12-31T20:19:29 *** Regis has quit IRC (Ping timeout: 240 seconds) 2011-12-31T20:25:49 *** amstan has quit IRC (Ping timeout: 240 seconds) 2011-12-31T20:26:27 *** grc48 has joined #aichallenge 2011-12-31T20:29:11 *** replore has joined #aichallenge 2011-12-31T20:33:29 *** replore has quit IRC (Ping timeout: 240 seconds) 2011-12-31T21:32:45 *** replore has joined #aichallenge 2011-12-31T21:35:57 *** amstan has joined #aichallenge 2011-12-31T21:35:57 *** ChanServ sets mode: +o amstan 2011-12-31T21:39:59 *** amstan_ has joined #aichallenge 2011-12-31T21:39:59 *** ChanServ sets mode: +o amstan_ 2011-12-31T21:41:10 contestbot_: echo hi 2011-12-31T21:41:10 hi 2011-12-31T21:42:15 hau 2011-12-31T21:43:04 contestbot_: nick contestbot 2011-12-31T21:43:04 *** contestbot_ is now known as contestbot 2011-12-31T21:43:09 hmm 2011-12-31T21:43:54 *** amstan_ has quit IRC (Client Quit) 2011-12-31T21:44:04 *** amstan_ has joined #aichallenge 2011-12-31T21:44:04 *** ChanServ sets mode: +o amstan_ 2011-12-31T21:44:07 *** foRei has quit IRC (Read error: Connection reset by peer) 2011-12-31T21:53:40 *** amstan_ has quit IRC (Quit: Konversation terminated!) 2011-12-31T21:54:00 *** replore has joined #aichallenge 2011-12-31T21:55:03 *** amstan_ has joined #aichallenge 2011-12-31T21:55:03 *** ChanServ sets mode: +o amstan_ 2011-12-31T21:56:47 *** Migi32 has quit IRC (Remote host closed the connection) 2011-12-31T22:00:23 *** treeform has quit IRC (Remote host closed the connection) 2011-12-31T22:15:20 *** replore has joined #aichallenge 2011-12-31T22:29:29 *** amstan_ has quit IRC (Read error: Connection reset by peer) 2011-12-31T22:30:02 *** amstan_ has joined #aichallenge 2011-12-31T22:30:02 *** ChanServ sets mode: +o amstan_ 2011-12-31T22:43:06 *** amstan_ has quit IRC (Remote host closed the connection) 2011-12-31T22:44:20 *** amstan_ has joined #aichallenge 2011-12-31T22:44:20 *** ChanServ sets mode: +o amstan_ 2011-12-31T22:44:58 *** amstan_ has quit IRC (Client Quit) 2011-12-31T22:45:07 *** amstan_ has joined #aichallenge 2011-12-31T22:45:07 *** ChanServ sets mode: +o amstan_ 2011-12-31T22:52:02 *** amstan_ has quit IRC (Quit: Konversation terminated!) 2011-12-31T22:52:37 *** amstan_ has joined #aichallenge 2011-12-31T22:52:37 *** ChanServ sets mode: +o amstan_ 2011-12-31T23:13:08 *** amstan_ has quit IRC (Ping timeout: 240 seconds) 2011-12-31T23:13:38 *** amstan has quit IRC (Ping timeout: 240 seconds) 2011-12-31T23:17:22 *** amstan has joined #aichallenge 2011-12-31T23:17:22 *** ChanServ sets mode: +o amstan 2011-12-31T23:27:52 *** amstan_ has joined #aichallenge 2011-12-31T23:27:52 *** ChanServ sets mode: +o amstan_ 2011-12-31T23:40:52 *** replore has joined #aichallenge 2011-12-31T23:45:32 *** replore has quit IRC (Ping timeout: 248 seconds)