2011-05-24T00:12:46 *** McLeopold has joined #aichallenge 2011-05-24T00:14:19 *** bhasker has joined #aichallenge 2011-05-24T00:25:36 *** BAM_ has quit IRC (Ping timeout: 252 seconds) 2011-05-24T00:32:35 *** Mathnerd314 has quit IRC (Ping timeout: 276 seconds) 2011-05-24T00:39:10 *** McLeopold has quit IRC (Ping timeout: 252 seconds) 2011-05-24T00:48:23 *** Palmik has joined #aichallenge 2011-05-24T00:51:56 *** Nebri has quit IRC () 2011-05-24T01:22:51 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2011-05-24T01:28:43 *** McLeopold has joined #aichallenge 2011-05-24T01:55:31 *** FireFly has joined #aichallenge 2011-05-24T03:05:38 *** berak has joined #aichallenge 2011-05-24T03:14:52 *** Ice_Harley has joined #aichallenge 2011-05-24T03:23:05 *** aerique has joined #aichallenge 2011-05-24T03:23:27 *** qacek has quit IRC (Quit: WeeChat 0.3.5) 2011-05-24T03:31:51 *** ltriant has quit IRC (Quit: Computer has gone to sleep) 2011-05-24T03:45:03 re: issue 15, for what it's worth, i just compiled the CL starter bot locally using the sandbox - https://github.com/aichallenge/aichallenge/issues/15 2011-05-24T04:06:48 *** FireFly has quit IRC (Quit: swatted to death) 2011-05-24T04:10:34 *** berak has quit IRC (Ping timeout: 246 seconds) 2011-05-24T04:21:21 *** McLeopold has quit IRC (Read error: Connection reset by peer) 2011-05-24T04:23:28 *** berak has joined #aichallenge 2011-05-24T04:30:28 *** Naktibalda has joined #aichallenge 2011-05-24T04:48:00 *** FireFly has joined #aichallenge 2011-05-24T04:52:39 *** Accoun has quit IRC () 2011-05-24T05:07:25 hey aerique 2011-05-24T05:09:11 *** Accoun has joined #aichallenge 2011-05-24T05:10:10 *** Eruonen has joined #aichallenge 2011-05-24T05:22:12 *** sigh has joined #aichallenge 2011-05-24T05:23:15 mcstar: hiya 2011-05-24T05:31:59 where are you with your visualization? 2011-05-24T05:32:00 *** amstan has quit IRC (Read error: Connection reset by peer) 2011-05-24T05:33:20 i have a working cellular automaton in lispbuilder, and i was wondering you might like to see the code 2011-05-24T05:33:39 *** needsch has joined #aichallenge 2011-05-24T05:41:38 mcstar: sure, give me a link 2011-05-24T05:46:48 *** FireFly has quit IRC (Quit: swatted to death) 2011-05-24T06:15:12 how about dcc? 2011-05-24T06:15:26 im not on the internet lol 2011-05-24T06:16:23 dcc send aerique ~/src/aiall/ca.tar 2011-05-24T06:59:58 *** delt0r___ has joined #aichallenge 2011-05-24T07:00:33 *** delt0r_ has quit IRC (Read error: Operation timed out) 2011-05-24T07:52:14 *** _0x47 <_0x47!~slash@p57B3E967.dip.t-dialin.net> has joined #aichallenge 2011-05-24T07:54:50 What general algorithms you think are going to be employed the most in this challenge? 2011-05-24T07:56:03 *** amstan has joined #aichallenge 2011-05-24T07:56:03 *** ChanServ sets mode: +o amstan 2011-05-24T08:14:56 10 if ai == perfect goto 40 2011-05-24T08:14:59 20 call make_better_ai 2011-05-24T08:15:01 30 goto 10 2011-05-24T08:15:08 40 print "that was easy" 2011-05-24T08:16:00 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-05-24T08:27:43 breadth first search! 2011-05-24T08:28:07 i'm claiming that to be the most used 2011-05-24T08:35:42 antimatroid1: well, i still think mine will be used more often 2011-05-24T08:36:53 (and arguably mine is the more general :D) 2011-05-24T08:46:54 mcstar: are you willing to put your money where your fingers are? 2011-05-24T08:46:57 :p 2011-05-24T08:47:15 :) 2011-05-24T08:56:45 *** berak_ has joined #aichallenge 2011-05-24T08:58:35 *** smiley1983 has joined #aichallenge 2011-05-24T08:58:37 *** berak has quit IRC (Ping timeout: 246 seconds) 2011-05-24T09:01:10 some kind of multiplayer rubix cube game could be cool 2011-05-24T09:01:24 first player to get a face filled with their colour 2011-05-24T09:01:55 not sure how one would do that simultaneously 2011-05-24T09:17:22 *** sigh has quit IRC (Remote host closed the connection) 2011-05-24T09:18:31 this is for you antimatroid1 http://4.bp.blogspot.com/_eXQh6Qsh7Z8/S1PDRRElvGI/AAAAAAAABv0/s5SSbrLf22g/s1600/wtf.jpg 2011-05-24T09:19:34 -.- http://www.youtube.com/watch?v=AAfpq6EPKck 2011-05-24T09:23:39 ive seen almost every video they showed in the clip 2011-05-24T09:24:52 (not sure thats a good or a bad thing, but im not obsessed with cats, in fact i have dogs) 2011-05-24T09:31:52 mcstar: http://i.imgur.com/qp4Cf.png 2011-05-24T09:31:56 i just spent way too long making that 2011-05-24T09:32:00 i want a physical one 2011-05-24T09:33:39 whoa thats your idea of the multiplayer version of th cube? 2011-05-24T09:34:23 no, that was just an alternate idea for a rubix cube 2011-05-24T09:34:49 how dou you intend to color it? 2011-05-24T09:35:44 30 colors is too much 2011-05-24T09:35:50 no it's not 2011-05-24T09:35:55 i know only 5-6 colors 2011-05-24T09:36:06 i'm a mathematician 2011-05-24T09:36:15 as far as i'm concerned there's an infinite number of colours 2011-05-24T09:36:16 :p 2011-05-24T09:37:10 im not talking about colours, im talking about colors 2011-05-24T09:37:17 hehe 2011-05-24T09:37:43 of course you could print the alphabet with number on the faces 2011-05-24T09:37:45 i'm australian :P 2011-05-24T09:37:52 or a number :P 2011-05-24T09:38:05 no, it has to be 1 character 2011-05-24T09:38:18 yeah, each face gets a unique number, 1-30 2011-05-24T09:38:37 you could start adding more cubes on so that you could transform the object into different things as well 2011-05-24T09:38:47 ie. the object isn't symmetric under all rotations 2011-05-24T09:39:34 *** berak_ has quit IRC (Ping timeout: 246 seconds) 2011-05-24T09:39:58 im sure that i cant conceive of the group structure of your cube 2011-05-24T09:40:16 the original is complex enough 2011-05-24T09:40:23 btw can u do it? 2011-05-24T09:40:41 i can make it in 2.5 minutes 2011-05-24T09:40:52 cause my fingers are slow 2011-05-24T09:40:52 nah, never bothered to learn 2011-05-24T09:41:13 one could write a program to get the group structure, no idea if i could run it though 2011-05-24T09:42:06 i was bad at group theory, i studied it only to get through the exam 2011-05-24T09:42:29 i hated those character tables 2011-05-24T09:43:31 group theory is awesome 2011-05-24T09:43:45 i can see why someone likes it 2011-05-24T09:44:02 it just took me too much of an effort to study it 2011-05-24T09:45:07 now back to work, i have to watch the end of SGU 2011-05-24T09:53:17 *** Kingpin13 has joined #aichallenge 2011-05-24T10:11:36 *** _0x47 <_0x47!~slash@p57B3E967.dip.t-dialin.net> has quit IRC (Ping timeout: 250 seconds) 2011-05-24T10:11:39 *** _0x471 <_0x471!~slash@p5B12EA5F.dip.t-dialin.net> has joined #aichallenge 2011-05-24T10:17:29 *** _0x471 <_0x471!~slash@p5B12EA5F.dip.t-dialin.net> has quit IRC (Ping timeout: 260 seconds) 2011-05-24T10:22:22 *** _0x47 <_0x47!~slash@p5B12F548.dip.t-dialin.net> has joined #aichallenge 2011-05-24T10:27:11 *** _0x471 <_0x471!~slash@p5B12F34D.dip.t-dialin.net> has joined #aichallenge 2011-05-24T10:28:09 *** _0x47 <_0x47!~slash@p5B12F548.dip.t-dialin.net> has quit IRC (Ping timeout: 258 seconds) 2011-05-24T10:34:25 *** aerique has quit IRC (Quit: ...) 2011-05-24T10:48:38 *** berak has joined #aichallenge 2011-05-24T10:56:34 *** berak has quit IRC (Ping timeout: 246 seconds) 2011-05-24T10:58:22 *** _0x47 <_0x47!~slash@p57B3E535.dip.t-dialin.net> has joined #aichallenge 2011-05-24T11:00:29 *** _0x471 <_0x471!~slash@p5B12F34D.dip.t-dialin.net> has quit IRC (Ping timeout: 276 seconds) 2011-05-24T11:00:45 *** FireFly has joined #aichallenge 2011-05-24T11:15:42 oh god, talk written 2011-05-24T11:15:52 i don't think people are going to understand wtf i'm saying 2011-05-24T11:19:00 *** Ice_Harley has quit IRC (Ping timeout: 252 seconds) 2011-05-24T11:32:18 *** demerzel has joined #aichallenge 2011-05-24T11:39:20 *** _0x471 <_0x471!~slash@p5B12F231.dip.t-dialin.net> has joined #aichallenge 2011-05-24T11:39:29 *** _0x47 <_0x47!~slash@p57B3E535.dip.t-dialin.net> has quit IRC (Ping timeout: 276 seconds) 2011-05-24T11:51:37 *** Frontier has quit IRC (Quit: Leaving.) 2011-05-24T12:06:04 *** bhasker has quit IRC (Quit: bhasker) 2011-05-24T12:06:27 *** berak has joined #aichallenge 2011-05-24T12:06:32 *** Mathnerd314 has joined #aichallenge 2011-05-24T12:31:39 *** amstan has joined #aichallenge 2011-05-24T12:31:40 *** ChanServ sets mode: +o amstan 2011-05-24T12:33:34 *** _0x47 <_0x47!~slash@p5B12FB04.dip.t-dialin.net> has joined #aichallenge 2011-05-24T12:34:34 *** _0x471 <_0x471!~slash@p5B12F231.dip.t-dialin.net> has quit IRC (Ping timeout: 246 seconds) 2011-05-24T12:42:07 *** _0x47 <_0x47!~slash@p5B12FB04.dip.t-dialin.net> has quit IRC (Ping timeout: 252 seconds) 2011-05-24T12:50:17 *** amstan has quit IRC (Ping timeout: 252 seconds) 2011-05-24T12:52:17 *** bhasker has joined #aichallenge 2011-05-24T12:59:25 *** berak has quit IRC (Ping timeout: 246 seconds) 2011-05-24T13:02:34 *** berak has joined #aichallenge 2011-05-24T13:13:18 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 4.0.1/20110413222027]) 2011-05-24T13:20:17 *** fgump has joined #aichallenge 2011-05-24T13:27:44 *** _0x47 <_0x47!~slash@p5B12F650.dip.t-dialin.net> has joined #aichallenge 2011-05-24T13:32:19 *** fgump has quit IRC (Quit: Leaving) 2011-05-24T13:37:15 *** Cyndre_ is now known as Cyndre 2011-05-24T14:12:36 *** Frontier has joined #aichallenge 2011-05-24T14:25:00 so IE9 is now officially released, I've got a Windows Update notification. I wonder how much I have to change to get the visualizer going in JScript. 2011-05-24T14:28:38 *** Frontier has quit IRC (Read error: Connection reset by peer) 2011-05-24T14:31:44 *** Redgis has joined #aichallenge 2011-05-24T14:33:19 *** Frontier has joined #aichallenge 2011-05-24T14:39:56 IE9 doesn't seem to respect the size constraints, but otherwise it works nicely. It is even quite fast. 2011-05-24T14:40:40 *** Ow has quit IRC (Ping timeout: 252 seconds) 2011-05-24T14:43:14 you wasted 4:41 minutes of your life to reboot windows 2011-05-24T14:43:43 14 2011-05-24T14:44:02 no 2011-05-24T14:44:07 never mind 2011-05-24T14:53:49 :p 2011-05-24T14:54:10 and now it offers me an update to ie that has been released last year 2011-05-24T14:55:26 it's like 'lets release the alpha version from last year, ppl can gradually patch it to the release version' 2011-05-24T14:57:50 anyway IE is a good example of what you can do when you don't have to care about cross-platformness. It outruns Firefox and Chrome in JavaScript benchmarks now! 2011-05-24T14:58:12 (Firefox 4 that is) 2011-05-24T14:59:59 *** delt0r_ has joined #aichallenge 2011-05-24T15:01:30 *** delt0r___ has quit IRC (Ping timeout: 250 seconds) 2011-05-24T15:08:33 *** Naktibalda has joined #aichallenge 2011-05-24T15:09:20 Though to be honest, my visualizer runs with 20 FPS in IE9 and about 60 in Chrome on Linux. So whatever the benchmark included was no good test case :p 2011-05-24T15:13:16 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2011-05-24T15:13:48 *** smiley1983 has quit IRC (Ping timeout: 240 seconds) 2011-05-24T15:15:00 *** smiley1983 has joined #aichallenge 2011-05-24T15:16:23 i guess drawing performance is not related to js execution speed then 2011-05-24T15:24:21 its gotta phone home to microsoft with every image you view 2011-05-24T15:24:38 no idk but i would turn off compatibility mode and safemode (or whatever it's called) and try it 2011-05-24T15:33:07 *** amstan has joined #aichallenge 2011-05-24T15:33:07 *** ChanServ sets mode: +o amstan 2011-05-24T16:03:01 *** Accoun has quit IRC () 2011-05-24T16:04:09 *** sente has quit IRC (Ping timeout: 252 seconds) 2011-05-24T16:19:01 *** tomh has joined #aichallenge 2011-05-24T16:19:12 hey 2011-05-24T16:19:24 is the Ants game ready to develop against? 2011-05-24T16:20:50 develop against? not sure what you mean by that exactly but the game itself exists on github 2011-05-24T16:21:01 to create a bot for it 2011-05-24T16:22:00 yes 2011-05-24T16:22:05 there are starter packages etc 2011-05-24T16:22:26 yeah I saw that, but it said winter 2011, so I thought it might be still under heavy development 2011-05-24T16:22:52 it is 2011-05-24T16:22:55 but the basic game is there 2011-05-24T16:23:02 ok 2011-05-24T16:23:02 the game itself won't change much 2011-05-24T16:23:06 just the site etc is under dev 2011-05-24T16:23:13 and when is the competition deadline? 2011-05-24T16:23:17 idk 2011-05-24T16:23:33 ok 2011-05-24T16:28:09 tomh: it didn't start.. so.. you're good 2011-05-24T16:29:49 amstan: worker still stalled? 2011-05-24T16:30:01 demerzel: yes, bug mcleopold pls 2011-05-24T16:30:19 contestbot: seen mcleopold 2011-05-24T16:30:19 demerzel: mcleopold was last seen in #aichallenge 21 hours, 46 minutes, and 25 seconds ago: amstan: I think my memcache code is crappy and useless 2011-05-24T16:30:25 yea that was the last one 2011-05-24T16:30:34 got a 'giving up' vibe :( 2011-05-24T16:30:45 hopefully not tho 2011-05-24T16:30:59 winter? so there is quite some time before the contest then? 2011-05-24T16:31:07 * berak would like a debate, why it's taking so long for take-off, ppl, you're half a year late now, already ! 2011-05-24T16:31:10 i might be able to participate after all if so 2011-05-24T16:31:33 berak: well, nobody's really into coding the backend 2011-05-24T16:31:37 berak: we need more coders 2011-05-24T16:31:41 i would be if it wasn't all python lol 2011-05-24T16:31:46 i'd tolerate php 2011-05-24T16:31:47 ;) 2011-05-24T16:32:32 python: literally the only language i won't touch with a 10-foot pole 2011-05-24T16:32:43 sorry 2011-05-24T16:32:46 demerzel: like me with c# and java? 2011-05-24T16:32:59 idk you do seem progress-resistant ;) 2011-05-24T16:33:28 since when was c# progress? i see it more as chains 2011-05-24T16:33:49 i'm not even gonna dignify that lol 2011-05-24T16:34:12 i think its funny linux users like python it's like the cmoplete oppoiste of my mindset 2011-05-24T16:34:13 since when is php progressive compared to python? 2011-05-24T16:34:25 i want complete control over my os but not my coding 2011-05-24T16:34:32 it should 'just do it' 2011-05-24T16:35:28 i mean, i'm with you that python is not a nicely progressive language, but php does not stand in contrast to it 2011-05-24T16:35:39 no i was giving him shit about not liking c# / java 2011-05-24T16:36:05 python is a weird association to linux but it's there for some reason 2011-05-24T16:36:07 java is one of the most, err. regressive languages i know. c# is alright at least 2011-05-24T16:36:35 they're syntactically very similar but c# is better implemented because it was planned that way 2011-05-24T16:36:48 otoh java 'theoretically' is open-platform 2011-05-24T16:36:52 but with mono existing it's a moot point imo 2011-05-24T16:36:52 and was actually designed to be better than the languages it competes with 2011-05-24T16:36:58 yea 2011-05-24T16:37:01 the only new thing java introduced to the mainstream was GC 2011-05-24T16:37:18 which of course was already around 2011-05-24T16:37:22 just not in a "fast" language 2011-05-24T16:37:27 amstan: what kind of backend stuff is not done? 2011-05-24T16:37:35 jmcarthur: The JVM is apparently rather nice. 2011-05-24T16:37:43 (It's arguably a byproduct of Java.) 2011-05-24T16:37:45 jbroman: for anything that is not interactive, yes 2011-05-24T16:37:57 the JIT is amazing 2011-05-24T16:38:09 and the GC fairly nice 2011-05-24T16:38:13 the thing that ties the whole run games/upload/move compiled/everything else is constantly broken 2011-05-24T16:38:18 afaik the actual components work fine 2011-05-24T16:38:32 I can develop the bot in any language I want right? 2011-05-24T16:38:39 tomh: a bunch of bugs, and mcleopold and janzert are the only ones familiar with the code 2011-05-24T16:38:46 tomh: so we're at their mercy 2011-05-24T16:38:51 ok 2011-05-24T16:39:16 its written in such way, that no one else could be familirized with the code? 2011-05-24T16:39:26 amstan is lazy 2011-05-24T16:40:15 * berak sees now, what 'we need more coders!' means.. 2011-05-24T16:40:26 500 lines of python which is 'really easy to read'? might as well be greek 2011-05-24T16:40:33 *** Accoun has joined #aichallenge 2011-05-24T16:40:54 demerzel: then what language do you like? :P 2011-05-24T16:41:13 i am giving amstan shit about not debugging it himself since he said python was really easy to read 2011-05-24T16:41:22 anything that starts with the letter c, generally 2011-05-24T16:41:28 except cobol :P 2011-05-24T16:41:40 demerzel: i am, i never wanted to work on this code, see my zeta suggestion 2011-05-24T16:42:03 hehe ok 2011-05-24T16:42:03 demerzel: Cython 2011-05-24T16:42:11 oshit! 2011-05-24T16:42:33 but for the competitions, you could write the bot in any language right? 2011-05-24T16:42:37 lol the c thing for python is funny 2011-05-24T16:42:40 idgi 2011-05-24T16:42:46 tomh: Most common languages, yes. 2011-05-24T16:42:52 haskell?\ 2011-05-24T16:42:54 if you admit c/c++ extensions would run much faster why not just write teh whole thing in c/c++ 2011-05-24T16:42:59 yep 2011-05-24T16:43:05 they even already have a haskell starter package 2011-05-24T16:43:09 ok good 2011-05-24T16:43:33 demerzel: Some things are faster to write in raw assembly than C/C++. Why would you not write the whole thing in assembly? 2011-05-24T16:43:55 i'm pretty sure the variety and amount of those things are much smaller in scope if not microscopic in comparison 2011-05-24T16:44:12 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.86.1 [Firefox 4.0.1/20110504060726]) 2011-05-24T16:44:27 The same argument holds. People who use higher level languages do so because they prefer the convenient facilities that provides, usually at the cost of some measure of performance. 2011-05-24T16:45:06 i thought this debate was settled when fortran/lisp was created 2011-05-24T16:45:12 50 years ago 2011-05-24T16:45:26 haskell is also quite performant 2011-05-24T16:45:50 i want to learn lisp 2011-05-24T16:45:56 maybe i'll try this year 2011-05-24T16:46:01 +1 2011-05-24T16:46:12 if you can type ( and ), you are halfway there 2011-05-24T16:46:34 tomh is being sarcastic 2011-05-24T16:46:56 i want a straight answer guys: 2011-05-24T16:47:03 jbroman: i don't believe that performance must necessarily be sacrificed for a higher level language, although it is definitely not going to be fast for early iterations of the compiler 2011-05-24T16:47:07 is it possible that the contest really will only start in the winter? 2011-05-24T16:48:53 jmcarthur: Well, some facilities due inevitably sacrifice performance (e.g. it's probably impossible to write GC that is never slower than well-written malloc/free code). Certainly some may be achievable without loss of performance (and I know you're not too subtly talking about GHC here). But those facilities are still the reason that people choose languages like Python. 2011-05-24T16:49:09 s/due in/do in/ 2011-05-24T16:49:18 i'm actually not talking about ghc at all 2011-05-24T16:50:25 Hmm, OK. 2011-05-24T16:50:30 jbroman: regions are often faster than malloc/free 2011-05-24T16:50:44 another language war won't make it happen. 2011-05-24T16:50:58 by that i mean faster than is at all possible with malloc/free unless you write a full blown regions library on top of it 2011-05-24T16:51:40 jmcarthur: I don't really care about the speed of the allocation algorithm. I meant that the GC will sometimes do operations that might be avoidable with foreknowledge of the program. And it was only an example. 2011-05-24T16:51:42 region inference isn't well-studied enough to avoid some explicitness in your code though, unfortunately 2011-05-24T16:52:01 * amstan points that speed of execution is not the only thing that matters 2011-05-24T16:52:05 even faster is not having to use malloc/free by careful design 2011-05-24T16:52:14 jbroman: no i understood your point. i just mean there are alternatives to gc 2011-05-24T16:52:19 regions being one 2011-05-24T16:52:20 i believe speed of developing is also important in this contest at least 2011-05-24T16:52:38 its really not though we have like 5 months to develop nothing but pure math logic 2011-05-24T16:52:42 *** smiley1983 has quit IRC (Ping timeout: 248 seconds) 2011-05-24T16:53:06 demerzel: what? 2011-05-24T16:53:13 thsi is one of the ffew applications i don't even take a hit in time by going back to pure c95 2011-05-24T16:53:20 dev time* 2011-05-24T16:53:57 *** smiley1983 has joined #aichallenge 2011-05-24T16:54:09 amstan: stdio (already written) and just pure logic ie math operation and conditions is all that's left 2011-05-24T16:54:11 jbroman: but since you brought up ghc, i feel i need to point out that ghc is *not* anywhere near where i think it could be. there are several optimizations glaringly missing and questionable design decisions making other kinds of optimizations impossible. even some of haskell's design decisions play a part here, most of which superficially make haskell appear *lower* level 2011-05-24T16:54:23 dev time is not really a factor since any language is going ot be similar in time invested 2011-05-24T16:54:37 higher level languages extract concepts like gui / input etc which is useful but not necessary in this 2011-05-24T16:54:46 demerzel: That is debatable. If that were provable true nobody would use high-level languages. 2011-05-24T16:54:52 s/provable/provably/ 2011-05-24T16:55:01 demerzel: i strongly disagree with that 2011-05-24T16:55:26 how much difference can there really be between languages on 'if h(n) < then' 2011-05-24T16:55:34 For instance, a whole class of memory allocation bugs do not occur in some higher-level languages. Thus you do not need to spend time debugging those issues. 2011-05-24T16:55:35 in terms of dev time not syntax 2011-05-24T16:55:39 higher level languages give you a richer vocabulary in general 2011-05-24T16:55:47 demerzel: i'm not just talking about syntax sugar 2011-05-24T16:55:49 jbroman: your dimensions are static 2011-05-24T16:55:56 given to you on turn1 2011-05-24T16:56:02 you don't even have to malloc 2011-05-24T16:56:17 demerzel: you were only talking about syntax, but let's analyze your example there (if then)... 2011-05-24T16:56:18 demerzel: That's only one example. 2011-05-24T16:56:46 demerzel: If you end up writing an identical program switching one syntax for another, sure, you won't notice much of a difference. 2011-05-24T16:56:59 demerzel: other ways to do it are pattern matching, which gives you a more powerful operation in fewer lines of code, allowing branches to have a context sensitivity which if statements simply can't mimick at all 2011-05-24T16:57:22 but you're giong to want to optimize that as much as possible anyway given the large amount of possibilities that exist 2011-05-24T16:57:52 demerzel: if statements also require booleans. with pattern matching, you can get a truth value which also carries its own proof, and then you can use that proof in the branches 2011-05-24T16:58:09 what do you mean by pattern matching? 2011-05-24T16:58:27 well i could give a simple example, but it won't show its full potential 2011-05-24T16:58:27 saying it brings regex to mind which i don't think is what you mean 2011-05-24T16:58:33 no, not regex 2011-05-24T16:58:57 case foo bar of Nothing -> ; Just x -> 2011-05-24T16:59:07 demerzel: that as opposed to checking for null before using x 2011-05-24T16:59:20 * berak just encountered a (python) syntax err in turn 100, a c-compiler would have got that even before startup 2011-05-24T16:59:27 demerzel: in other words, because you obtain x through pattern matching, it's impossible to use x if it's null 2011-05-24T16:59:46 it's not even syntactically expressible to use a null value 2011-05-24T16:59:52 but you could just initialize x with a value that's impossible to arrive at 2011-05-24T17:00:07 guess i don't understand the example 2011-05-24T17:00:09 but that's just a hack 2011-05-24T17:00:23 true but i don't understand the significance of null 2011-05-24T17:00:42 berak: python syntax errors are detected from the start, unless you're doing evals 2011-05-24T17:01:27 okay how about this... open a source file, find the first if statement in the file, and tell me what you are using it for. nine times out of ten, pattern matching and a high level data structure can do it more nicely 2011-05-24T17:01:47 not just more nicely, that is 2011-05-24T17:01:53 more safely 2011-05-24T17:02:18 demerzel: Suppose you have an object, Person, who maintains a pointer to another object, Company. You might try to do "person->company->name" to get the company name, forgetting to check the possibility that the company field is null. 2011-05-24T17:02:19 In Haskell, you cannot do this because the field is of type "Maybe Company", so you are forced to consider the possibility by the language. 2011-05-24T17:03:01 Pattern matching provides a syntax which makes doing this not painful. 2011-05-24T17:03:11 ok i understand that but how is that different than null checking 2011-05-24T17:03:21 it's different in a few ways 2011-05-24T17:03:31 1. The compiler prevents you from accidentally using a null value. 2011-05-24T17:03:54 2. The synax gives you a "denullified" value to use which you are guaranteed is not null. 2011-05-24T17:04:14 3. The syntax makes it impossible to use a denullified value without having first checked it via pattern matching. 2011-05-24T17:04:36 pattern matching is actually much more general than this. this is just a very simple example 2011-05-24T17:05:06 how is that different than initializing to a non-null value that is either that min/max or unreachable by calculation? 2011-05-24T17:05:13 not only is this higher level, but it can also make your code faster because you don't have to perform redundant null checks in different functions 2011-05-24T17:05:15 depending on application 2011-05-24T17:05:41 what is the 'denullified' value, do you define it or is it language-wide 2011-05-24T17:05:53 because your proposal does not prevent you from misusing it and does not give you a syntactically obvious indication that you need to check for null 2011-05-24T17:06:06 the denullified value is x in my example code 2011-05-24T17:06:21 the possibly-null value is "foo bar" 2011-05-24T17:06:48 once you have x you can use it however you want and not have to ever think about the possiblity of it being null 2011-05-24T17:07:15 your "special value" proposal does not have this property 2011-05-24T17:07:26 cool ty gonna look into haskell - i do know of some exchange companies that use it 2011-05-24T17:07:38 well yea i guess it depends on how much you trust the coder haha 2011-05-24T17:08:11 it goes a lot deeper than this though, as i said. overall, haskell's type system makes it much easier to reason about your code. 2011-05-24T17:08:36 jbroman: it also makes it easier for the compiler to reason about your code. that's the direction i was intending to push when i said high level languages don't necessarily have to be slower than low level languages 2011-05-24T17:09:35 jmcarthur: True in theory, but extremely difficult in practice. But point taken. 2011-05-24T17:10:24 dont get wrong i wouldn't even think about using iso c in a gui app or something similar 2011-05-24T17:10:35 it was just super nice to have something like this where i don't have to think at all about user interaction 2011-05-24T17:11:10 seemed an optimal choice because all the stuff that is abstracted in other languages really isn't part of it 2011-05-24T17:11:29 but i only know class-based languages and a handful of others i have never touched haskell/lisp/etc 2011-05-24T17:11:35 i don't think it's that difficult in practice. i think the problem is that so far people have been too afraid to continue purifying (in the sense of purely functional) languages. there are many ways to bring it further. i think turing completeness is vastly overrated, for example. dropping that in certain ways makes crazier optimizations much easier, just as purity makes certain optimizations a 2011-05-24T17:11:35 c-based i meant 2011-05-24T17:11:36 lot easier too 2011-05-24T17:12:28 i guess i was due for my functional languages conversation in #aichallenge 2011-05-24T17:13:15 iso c isn't a functional language? 2011-05-24T17:13:40 demerzel: I wouldn't describe it as one, no. 2011-05-24T17:13:41 tough question to answer honestly. C is not functional in the sense that i am talking about 2011-05-24T17:14:15 since functional is such an ambiguous word, i've been using "denotative" in its place 2011-05-24T17:14:28 a language is denotative if it satisfies the following criteria: 2011-05-24T17:14:44 1. Expressions are made up of subexpressions. 2011-05-24T17:15:13 2. Each subexpression has a well-defined meaning all by itself (a value, mathematical object, function, etc.). 2011-05-24T17:15:29 3. The meaning of an expression depends only on the meanings of its subexpressions. 2011-05-24T17:15:37 C does not satisfy that criteria. 2011-05-24T17:16:17 nor does Haskell, by a very strict interpretation 2011-05-24T17:16:39 i will admit i'm not very knowledgeable at the academic side of coding lol 2011-05-24T17:16:42 jmcarthur: Do you consider the IO monad a violation of that? 2011-05-24T17:16:47 didnt' go to school for cs at least 2011-05-24T17:16:52 the IO monad does not have a well-defined meaning 2011-05-24T17:17:05 *** berak has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2011-05-24T17:17:17 demerzel: even if you did, they probably wouldn't teach you this 2011-05-24T17:19:32 brb coffee time 2011-05-24T17:33:03 *** computerwiz_222 has joined #aichallenge 2011-05-24T17:38:05 well.. now we know where jeff is: http://www.tmx.com/en/news_events/broadcast_centre/events_calendar.html 2011-05-24T17:38:12 * amstan points at blue suit 2011-05-24T17:38:49 yay coffee 2011-05-24T17:39:16 *** _0x47 <_0x47!~slash@p5B12F650.dip.t-dialin.net> has quit IRC (Ping timeout: 244 seconds) 2011-05-24T17:39:52 *** _0x47 <_0x47!~slash@p5B12FF30.dip.t-dialin.net> has joined #aichallenge 2011-05-24T17:41:22 that's really jeff? 2011-05-24T17:41:28 yeah 2011-05-24T17:41:40 and wait, what blue suit? 2011-05-24T17:41:46 the blue shirt you mean? 2011-05-24T17:41:46 in the middle 2011-05-24T17:41:52 Blue shirt. 2011-05-24T17:41:53 he presses something 2011-05-24T17:42:02 that would be a shirt, not a suit :P 2011-05-24T17:42:11 oh, w/e 2011-05-24T17:42:16 tought it was a blazer at first 2011-05-24T17:42:26 heh 2011-05-24T17:48:10 *** ltriant has joined #aichallenge 2011-05-24T17:53:04 if you want to know something about the most powerful computer language, pls watch this promo video, you wont regret it: 2011-05-24T17:53:06 http://www.youtube.com/watch?v=HM1Zb3xmvMc 2011-05-24T17:53:14 I crashed IE9, yay! 2011-05-24T17:54:16 *** _0x47 <_0x47!~slash@p5B12FF30.dip.t-dialin.net> has quit IRC (Quit: Leaving.) 2011-05-24T17:56:08 *** smiley1983 has quit IRC (Ping timeout: 250 seconds) 2011-05-24T17:57:15 *** smiley1983 has joined #aichallenge 2011-05-24T17:57:24 *** McLeopold has joined #aichallenge 2011-05-24T17:57:51 miau..miau..miau...? ^^ 2011-05-24T17:58:26 nyan.cat? 2011-05-24T17:58:30 oh! 2011-05-24T17:58:39 McLeopold: miau! 2011-05-24T17:59:19 btw the game examples is from the book, the clip supposed to promote 2011-05-24T17:59:34 its a nice book teaching you lisp throgh writing simple games 2011-05-24T17:59:49 amstan: hi 2011-05-24T18:01:26 McLeopold: we can has fixed worker? 2011-05-24T18:01:36 mcleopold i think you are obiwan at this point 2011-05-24T18:01:40 re: only hope 2011-05-24T18:01:58 amstan: we can has more free php labor? 2011-05-24T18:02:08 McLeopold: on what? 2011-05-24T18:02:17 everything 2011-05-24T18:02:35 I don't want to be the point of failure 2011-05-24T18:02:40 the website seems to be working, mostly 2011-05-24T18:02:44 the worker is the broken one 2011-05-24T18:04:09 I was fixing up my misguided attempts at improved caching 2011-05-24T18:05:30 where? 2011-05-24T18:07:15 game_list.php and ranking.php 2011-05-24T18:07:22 let me commit, hold on... 2011-05-24T18:07:34 what about the worker? 2011-05-24T18:07:47 the .copy() stuff 2011-05-24T18:08:00 *** orivej has joined #aichallenge 2011-05-24T18:08:09 yeah, that too 2011-05-24T18:13:05