2011-08-13T00:17:34 *** sigh has joined #aichallenge 2011-08-13T01:55:47 *** amstan has quit IRC (Ping timeout: 250 seconds) 2011-08-13T03:03:20 *** ibdknox has quit IRC (Remote host closed the connection) 2011-08-13T03:04:15 *** ibdknox has joined #aichallenge 2011-08-13T03:55:04 *** boegel has joined #aichallenge 2011-08-13T03:56:50 *** Palmik has joined #aichallenge 2011-08-13T04:04:53 *** Redgis has joined #aichallenge 2011-08-13T04:30:54 *** janzert has quit IRC (Ping timeout: 240 seconds) 2011-08-13T05:32:24 *** berak has joined #aichallenge 2011-08-13T05:33:28 *** ibdknox has quit IRC (Remote host closed the connection) 2011-08-13T05:42:20 *** JaycobColeman has joined #aichallenge 2011-08-13T05:54:25 *** JaycobColeman has quit IRC (Quit: JaycobColeman) 2011-08-13T06:45:50 *** foRei has joined #aichallenge 2011-08-13T07:01:18 *** FireFly has joined #aichallenge 2011-08-13T07:02:44 *** smellyhippy has quit IRC (Quit: Take your stinking clause off me you damn dirty predicate!) 2011-08-13T07:21:39 *** berak has quit IRC (Ping timeout: 250 seconds) 2011-08-13T07:24:58 *** mceier has joined #aichallenge 2011-08-13T07:40:15 *** Redgis has quit IRC (Ping timeout: 255 seconds) 2011-08-13T07:44:22 *** Redgis has joined #aichallenge 2011-08-13T07:53:43 *** sigh has quit IRC (Ping timeout: 258 seconds) 2011-08-13T08:08:36 *** Redgis has quit IRC (Ping timeout: 255 seconds) 2011-08-13T08:56:14 *** sigh has joined #aichallenge 2011-08-13T09:01:12 *** mleise has joined #aichallenge 2011-08-13T09:33:16 *** boegel has quit IRC (Quit: Leaving) 2011-08-13T09:40:03 *** mcstar has joined #aichallenge 2011-08-13T09:40:09 antimatroid: 2011-08-13T09:41:01 antimatroid: 2011-08-13T09:41:19 i guess hes sleeping 2011-08-13T09:41:37 anyone else knows cpp templates? 2011-08-13T09:49:06 *** Accoun has quit IRC (Ping timeout: 258 seconds) 2011-08-13T09:53:19 *** Accoun has joined #aichallenge 2011-08-13T10:12:43 *** Naktibalda has joined #aichallenge 2011-08-13T10:17:12 mcstar: No, and I think they are a bit difficult now and then. Maybe Google can help you if you want to do something specific 2011-08-13T10:17:40 mleise: hi 2011-08-13T10:17:55 well, i would like to use them to genera cpp code 2011-08-13T10:18:01 generate* 2011-08-13T10:18:47 maybe you can answer this: 2011-08-13T10:18:57 lets say i have a template argument, an int 2011-08-13T10:19:07 and the template has a loop in it 2011-08-13T10:19:31 does the compiler expand that loop if i use the template parameter as the upper limit for exmaple 2011-08-13T10:19:32 ? 2011-08-13T10:27:28 nicht versteht? 2011-08-13T10:28:29 oh you wrote something 2011-08-13T10:29:49 hmm, I don't know for sure, but I'd say that the compiler turns the template into cpp code as you said and it then depends on the compiler and settings if it unrolls the loop. So I assume yes, it does. 2011-08-13T10:31:38 i guess ill use it then ill be more or less sure what it can/cant do 2011-08-13T10:31:51 it just feels sort of wtf 2011-08-13T10:32:19 you know, after learning lisp, it spoiled me, and now im in the dark as to how to use them 2011-08-13T10:32:36 i mean other syntax/languge contructs 2011-08-13T10:32:55 but this is my problem 2011-08-13T10:33:01 so thx 2011-08-13T10:35:48 *** Kingpin13 has joined #aichallenge 2011-08-13T10:39:16 why do you switch to c++ anyway mcstar 2011-08-13T10:41:04 because those infantile people at NVIDIA thought they would choose c/c++ 2011-08-13T10:41:31 i guess i could use fortran too, but im not in the mood... 2011-08-13T10:44:46 *** mathis_ has joined #aichallenge 2011-08-13T10:45:27 If it is C you can probably bind to LISP, no? If it is C++ you're screwed of course. :p 2011-08-13T10:46:11 its complicated 2011-08-13T10:46:30 common lisp has a famous foreign function interface 2011-08-13T10:46:34 so thats not a problem 2011-08-13T10:46:55 but the vuda program is compiled with nvcc, which separates cuda code from c/c++ 2011-08-13T10:47:01 cuda* 2011-08-13T10:47:23 and of course i cant interface programs this way 2011-08-13T10:47:29 but there is a cuda driver api 2011-08-13T10:47:49 which is basically no more than a c api for device initialization/execution 2011-08-13T10:48:04 so its possible to call those functions from lisp 2011-08-13T10:48:50 but i would still need to compile a c-like language code to ptx, which is an assembly laguage for a virtual machine 2011-08-13T10:49:06 and compile that code with the nvidia driver to device machine code 2011-08-13T10:49:14 which again, i could load from lisp 2011-08-13T10:49:24 but this is way too convoluted 2011-08-13T10:49:36 pycuda does something along those lines i believe 2011-08-13T10:50:05 it cant hurt if i pick up some c++ along the way, im not against it 2011-08-13T10:55:19 =-O This sounds complicated, even though I knew how CUDA compiles to gfx card 'assembly' 2011-08-13T10:57:16 what would be nice, is if i could eliminate #define's 2011-08-13T10:57:30 in simulations, we have a lot of parameters 2011-08-13T10:57:57 and there are combinations that would be nice not to compute them by hand every time we change something 2011-08-13T10:58:15 and, still, the compiled program would contain them as calculated values 2011-08-13T10:58:38 like a macro system, that actually a whole language in itslef 2011-08-13T10:58:53 so with lisp macros its really easy, and would fit perfectly 2011-08-13T10:59:06 and thats why im asking about templates 2011-08-13T10:59:47 i want to write down my equations in a way that is both handy for me, and abstract, but the compiled code has no superfluous computation 2011-08-13T11:00:25 mcstar: I see, try "-funroll-all-loops". I don't know if gcc unrolls loops on its own 2011-08-13T11:00:41 it doesnt matter 2011-08-13T11:00:49 the equations are in the device code 2011-08-13T11:00:57 and theres a #pragma unroll 2011-08-13T11:01:01 which works 2011-08-13T11:01:02 kk 2011-08-13T11:01:19 but what i asked previously, was related to using loops for code generation 2011-08-13T11:02:11 So is that template in the gfx card code or in the x86 code? 2011-08-13T11:02:36 i would like it to be on the divice 2011-08-13T11:02:53 my kernel is already templated for its data type float/double 2011-08-13T11:03:04 but i need more 2011-08-13T11:03:23 like, giving them constants through templates 2011-08-13T11:03:48 mykernel<<<>>>(); 2011-08-13T11:03:56 which is totally possible in a modern language :p 2011-08-13T11:04:13 and const1 const2 would be replaced by the values i put there 2011-08-13T11:04:18 so they are not variables 2011-08-13T11:04:19 what does <<<>>> mean??? 2011-08-13T11:04:28 and dont consume memory, only code 2011-08-13T11:04:41 <<<>>> is the execution parametrization 2011-08-13T11:05:26 mykernel<<>>(vars); 2011-08-13T11:05:41 bytes* 2011-08-13T11:05:43 stream* 2011-08-13T11:06:26 Is that a CUDA feature? I've never seen <<<>>> 2011-08-13T11:06:38 yes 2011-08-13T11:07:26 And a C-style macro doesn't work at all? 2011-08-13T11:07:32 For what you want? 2011-08-13T11:07:50 well, its not really nice 2011-08-13T11:07:58 that's right 2011-08-13T11:08:12 if i cant do any other way, i will do it with simple #define's 2011-08-13T11:08:30 but i cant do arithmetic on them 2011-08-13T11:08:47 you see, lets say i have DIMX * DIMY in device code 2011-08-13T11:08:58 it will compute this value every time 2011-08-13T11:09:17 so i use a #define DIMXTIMESDIMY 324234234 2011-08-13T11:09:30 no, don't do that :p 2011-08-13T11:09:32 but now, this isnt flexible, i have to hand-code it if i change DIMX,or DIMY 2011-08-13T11:09:49 mleise: this is the usual way 2011-08-13T11:10:06 in simlation code, we usually have a page of this kind of defines 2011-08-13T11:10:34 #define DIMAREA (DIMX * DIMY) 2011-08-13T11:10:46 that doesnt help 2011-08-13T11:10:53 it gets computed every time 2011-08-13T11:11:14 A macro gets computed by the preprocessor once 2011-08-13T11:11:41 The device wont compute it again every execution step 2011-08-13T11:11:55 but does the preprocessor evaluate the multiplication? 2011-08-13T11:12:16 i highly doubt it 2011-08-13T11:12:24 it has no knowledge of the type system 2011-08-13T11:12:29 hmm, let's find that out 2011-08-13T11:12:35 but templates are typed 2011-08-13T11:12:49 mcstar: still need me? 2011-08-13T11:12:49 the preprocessor maybe only knows about some INT-type 2011-08-13T11:13:01 i doubt that too 2011-08-13T11:13:13 i mean, it doesnt evaluate anything i think 2011-08-13T11:13:17 it has conditionalas and such 2011-08-13T11:13:40 #if VAR < 3 * 4 2011-08-13T11:13:43 does this work? 2011-08-13T11:14:07 (correctly? thats now even correct there) 2011-08-13T11:14:13 antimatroid: Does C++ reduce arithmetics in macros to a single value? 2011-08-13T11:14:24 yeah, he is sleeping 2011-08-13T11:14:25 i have no idea sorry 2011-08-13T11:14:27 wtf 2011-08-13T11:14:35 you werent supposed to answer 2011-08-13T11:14:38 i;m drunk too :P 2011-08-13T11:14:41 i have no idea - wtf (lol) 2011-08-13T11:14:43 don't expect me to be of much use :P 2011-08-13T11:14:56 happy drinking 2011-08-13T11:15:07 drinking is over, i just got home 2011-08-13T11:15:30 did you see my first question? 2011-08-13T11:15:34 it was titled to you 2011-08-13T11:16:02 but im not forcing you in the light of drunkenness 2011-08-13T11:16:56 mleise: are you coding something up right now? 2011-08-13T11:18:21 mcstar: if the macro "#define AREA (DIMX * DIMY)" gets expanded in "int something = x * AREA" it becomes "int something = x * (3 * 2)" which the compiler can optimize to a single value, as both are literal numbers 2011-08-13T11:19:32 yeah, but lets say i want double precicion, or flaot 2011-08-13T11:19:35 I assume you are right about the pre-processor not doing any calculations, so macros just 'expand' 2011-08-13T11:19:37 i have no control over this 2011-08-13T11:19:51 and the device code gets compiled by openccx 2011-08-13T11:20:01 or wtf, what nvcc calls 2011-08-13T11:20:03 T something = x * T(AREA) 2011-08-13T11:20:04 so i cant rely on this 2011-08-13T11:20:56 you cannot rely on the compiler merging 3 * 2 to 6? This compiler is useless -.- 2011-08-13T11:21:23 lets say i have 2 long double's 2011-08-13T11:21:26 now what? 2011-08-13T11:21:40 or, i want to typecast one of them 2011-08-13T11:21:42 or anything 2011-08-13T11:21:50 its not flexible 2011-08-13T11:22:11 2 A compile-time constant is something that can be evaluated at compile-time, such as an integer, or the address of a global 2011-08-13T11:22:14 variable. There are restrictions on what can be a compile-time constant, e.g. floating point values can not be compile-time constants 2011-08-13T11:22:34 i dont like what this sentence tells me though 2011-08-13T11:22:50 its regarding template parameters 2011-08-13T11:23:21 *** onensora has joined #aichallenge 2011-08-13T11:23:53 So you cannot have floats as template parameters? that's bad 2011-08-13T11:24:39 yeah 2011-08-13T11:25:05 mleise: what is your main language? 2011-08-13T11:25:09 javascript? 2011-08-13T11:26:02 Looks like it, doesn't it? Currently I'd say it is D. I've come a long way. 2011-08-13T11:26:50 did you program in cpp for money? 2011-08-13T11:27:01 im just guessing that you cant earn with D 2011-08-13T11:27:01 And in D you can have floats as template parameters of course and compile time function evaluation to define some constants :p 2011-08-13T11:27:39 Oh there was one job offer in Berlin for D1 (or C or C++) 2011-08-13T11:28:15 So far I've earned my money with Delphi and Java. 2011-08-13T11:29:50 ah right, the firm you mention in you profile is delphisomething 2011-08-13T11:32:18 Oh that's actually a forum focussed on Delphi and OpenGL. One of the members actually put together quite a nice engine for a robbery game: http://gael.godlikesoft.de/ 2011-08-13T11:32:51 *** onensora has quit IRC () 2011-08-13T11:33:35 *** onensora has joined #aichallenge 2011-08-13T11:34:04 He went so far as to create a scripting language for it and simple physics. The sound lib is from another forum member, but all in all it is one of these pet projects that have reached an impressive state for a one-man-project 2011-08-13T11:35:49 the problem is that im short on time 2011-08-13T11:35:58 ive got to get these equations in 2011-08-13T11:36:22 Then you better switch to some cpp channel here 2011-08-13T11:36:53 ah, ill read a doc on templates, it will enlighten me, but now i will just code it in 2011-08-13T11:36:55 or just do it with defines until you find a better solution 2011-08-13T11:37:01 yeah 2011-08-13T11:45:11 *** bhasker has joined #aichallenge 2011-08-13T11:46:10 *** bhasker has quit IRC (Client Quit) 2011-08-13T11:49:49 *** gilymerkado has joined #aichallenge 2011-08-13T11:55:56 it makes sense for floating point not to be compile time constants 2011-08-13T11:56:19 the meaning of a floating point computation depends on global settings 2011-08-13T11:56:30 *dynamic* global settings, that is 2011-08-13T11:57:22 it could also destroy cross-compilation, aside from that 2011-08-13T11:57:46 eh, i guess not necessarily the latter 2011-08-13T11:59:32 well, now you have to define "makes sense" 2011-08-13T11:59:43 http://software.intel.com/sites/products/documentation/studio/composer/en-us/2009/compiler_c/fpops/common/fpops_fp_env.htm 2011-08-13T11:59:57 if you look at a technology, and see its limitations, then after thinking about them, they can make sense 2011-08-13T12:00:18 but it doesnt mean i like to put up with something wrongly-thought-out 2011-08-13T12:01:27 i dont see how this is relevant 2011-08-13T12:01:53 i mean the link 2011-08-13T12:02:15 it means that a statically computed floating point value isn't necessarily going to be the same as a dynamically computed floating point value 2011-08-13T12:03:01 well, yeah, pre-runtime calculations have nothing to do with machine-state specific run-time calculations 2011-08-13T12:03:07 exactly 2011-08-13T12:03:09 but we knew this already 2011-08-13T12:03:40 and since floating point computations are context-dependent then performing the computations statically would be a suspicious feature 2011-08-13T12:03:42 but it doesnt explain why templates cant have floating point arguments 2011-08-13T12:04:12 the compiler usually has access to the same hardware it compiles to 2011-08-13T12:04:21 this isn't hardware dependent 2011-08-13T12:04:23 and if it cross-compiles, it could emulate 2011-08-13T12:04:27 this is *dynamically* dependent 2011-08-13T12:04:46 that doesnt matter 2011-08-13T12:04:59 you can run a computation, then change some settings, then run it again. the results will be different 2011-08-13T12:05:03 those parameters can be controlled when compilation occurs 2011-08-13T12:05:04 i mean pure float computations 2011-08-13T12:05:10 *** onensora has quit IRC () 2011-08-13T12:05:45 the standard would only need to define a standard environment and all is done 2011-08-13T12:07:17 to make myslef clear, when you write down two doubles, you multiply them, and get a result 2011-08-13T12:07:34 and you dont have to specify any rounding settings or flsuh to zeros 2011-08-13T12:07:41 its a default setting 2011-08-13T12:07:49 templates could have that too 2011-08-13T12:08:00 or whatever, i dont know them 2011-08-13T12:08:13 i mean im not even sure what i want is possible with them 2011-08-13T12:08:27 i just mean that the reason these are dynamically controllable is because it might matter. making it static means you lose that control 2011-08-13T12:08:32 *** onensora has joined #aichallenge 2011-08-13T12:08:33 what is it that you want? 2011-08-13T12:09:01 jmcarthur: i had a conversation with m.leise, i dont have the time to repeat it 2011-08-13T12:09:07 but if you want you can read it :) 2011-08-13T12:09:29 i read it, but it wasn't clear to me. is this anything more than partial evaluation? 2011-08-13T12:09:46 or do you mean the loop question? 2011-08-13T12:10:01 the first is more important 2011-08-13T12:11:09 i think i would write a separate program that generates code for me. templates only go so far 2011-08-13T12:11:32 but are you sure templates cant do that? 2011-08-13T12:11:44 *** gilymerkado has quit IRC (Quit: Page closed) 2011-08-13T12:11:56 its is referred to as metaprogramming daman it 2011-08-13T12:12:02 what is so meta about it then? 2011-08-13T12:12:08 -a 2011-08-13T12:12:21 it generates code 2011-08-13T12:12:26 that doesn't mean it's very general 2011-08-13T12:14:01 they say its turing complete 2011-08-13T12:14:13 templates are really intended to allow static type parameterization. anything else they enable is just a symptom of being turing complete. unfortunately, they are turing complete, but don't provide facilities for symbolic manipulation 2011-08-13T12:14:37 basically htey suck 2011-08-13T12:14:42 yeah 2011-08-13T12:15:24 i could talk about this with you too 2011-08-13T12:15:30 but im really short on time 2011-08-13T12:15:33 so excuse me 2011-08-13T12:23:36 I look at the floating point state issue from the point that floating point operations are generally 'lossy'. So unless I really depend on rounding mode and precision settings in the FPU I'd go with the default for compile-time evaluation. In the end I'm setting up a constant so I won't expect the value of it to depend on what the FPU state is in the context it is used in. 2011-08-13T12:25:35 an example of lossy integer arithmetic: (int)(2<<32) 2011-08-13T12:26:32 well, i think i wanted to write unsigned int 2011-08-13T12:27:35 but the worst part of it all 2011-08-13T12:27:39 wait for it...... 2011-08-13T12:27:44 That's a totally different kind of loss though in practice. 2011-08-13T12:28:01 is that the reference implementation spits out data in .mat files 2011-08-13T12:28:11 i hate matlab literally 2011-08-13T12:28:13 With floats you lose precision (and that is usually what you influence with FPU flags) and with ints you overflow 2011-08-13T12:28:41 overflow is a lossy for of arithmetic 2011-08-13T12:28:43 I think my sister told me about that program. 2011-08-13T12:28:51 for->form 2011-08-13T12:29:52 floating point numbers are just like integers, the result of a computation is deterministic 2011-08-13T12:30:14 of course the previously mentioned registers need to be kept unchanged 2011-08-13T12:30:26 *** amstan has joined #aichallenge 2011-08-13T12:30:26 *** ChanServ sets mode: +o amstan 2011-08-13T12:30:33 so this is not a loss you can control/arent aware of 2011-08-13T12:30:44 cant* 2011-08-13T12:30:45 hell 2011-08-13T12:31:23 *** bhasker has joined #aichallenge 2011-08-13T12:32:06 Weren't we talking about why FLOATS cannot be computed at compile time? 2011-08-13T12:32:28 yeah, but i dont see any problem why they couldnt be 2011-08-13T12:32:53 *** Redgis has joined #aichallenge 2011-08-13T12:32:58 just as there is a default setting of these, when you use fp arithmetic at runtime 2011-08-13T12:33:08 there could be a similar defualt at compile time 2011-08-13T12:33:41 but its the "who the f cares" category for me now 2011-08-13T12:33:54 it seems scipy can read mat files 2011-08-13T12:33:57 I agree. But you lose precision in the general case with FP math. I most cases that is ok 2011-08-13T12:34:31 not always 2011-08-13T12:34:41 i have a vendetta against floating point 2011-08-13T12:34:56 But your point about INT arithmetic being lossy is a different for me. You don't loose more or less precision keeping the value usable, but instead get something horribly different from an overflow 2011-08-13T12:35:22 mcstar: not always = in the general case 2011-08-13T12:35:23 int arithmetic is just modular arithmetic 2011-08-13T12:35:30 i mean the value may change because there are not representable numbers, but precision loss usually comes when you try to subtract numbers with huge difference in the exponentials 2011-08-13T12:35:43 modular arithmetic is well-defined and easy to reason about 2011-08-13T12:35:57 (at least when the number of bits is known) 2011-08-13T12:37:07 s/well-defined/has a simple definition/ 2011-08-13T12:37:12 the truth is that floating point numbers suck too 2011-08-13T12:37:30 sure let's all hate them for what they are :p 2011-08-13T12:37:31 i will try to use fixed point number in this simulation 2011-08-13T12:37:43 floating point is great when you need to support almost arbitrary ranges in the same number of bits 2011-08-13T12:37:51 because the upper/lower bound of the values of the fields is well known 2011-08-13T12:38:27 excuse my s's 2011-08-13T12:38:43 and is's/are's 2011-08-13T12:38:46 hehe 2011-08-13T12:39:13 it sounds perfectly fine to me 2011-08-13T12:40:12 look closer, i made a few mistakes 2011-08-13T12:43:49 *** pmnathan has joined #aichallenge 2011-08-13T12:44:45 *** pmnathan has left #aichallenge 2011-08-13T12:50:26 *** sigh has quit IRC (Remote host closed the connection) 2011-08-13T12:51:57 *** sigh has joined #aichallenge 2011-08-13T13:00:42 *** demerzel has joined #aichallenge 2011-08-13T13:01:09 amstan: i think ratings are unfairly weighting more games, i'm rank 77 with my busted ass bot from 2 months ago 2011-08-13T13:01:45 it literally just goes in a straight line and usually kills itself if it has more than 3 lol 2011-08-13T13:01:52 no way i should be ranked in the top 100 even 2011-08-13T13:01:58 yeah, something's not cool there 2011-08-13T13:02:24 just got an invoice for the server looks like it's still busy 2011-08-13T13:02:41 also do you have some kind of nocache set up on pages? notice it does a server pull every time i click back 2011-08-13T13:03:05 there shouldn't be 2011-08-13T13:03:13 we have use memcached for some stuff 2011-08-13T13:03:15 well if i go to my profile, then watch a game 2011-08-13T13:03:20 then press back, it repulls the page 2011-08-13T13:03:28 yeah, and makes everything slow as hell 2011-08-13T13:03:32 yep 2011-08-13T13:03:43 memcached? in the viewer you mean? 2011-08-13T13:04:14 gonna have to build one - when does the official contest start? ie when are you moving to cloud, not the actual ranking 2011-08-13T13:04:28 unless you'r eplanning on keeping beta server until the actual competition 2011-08-13T13:04:52 well, i was hoping we were moved by now 2011-08-13T13:04:54 not that i'm trying to hurry you to the cloud 2011-08-13T13:05:08 idc about the server just wondering when i have to get in gear if i'm gonna participate 2011-08-13T13:05:37 i'm not sure 2011-08-13T13:06:29 np just wondering 2011-08-13T13:06:36 been a couple months i know i've already lost out on a lot of time 2011-08-13T13:06:38 *** sigh has quit IRC (Remote host closed the connection) 2011-08-13T13:06:40 workk crazy 2011-08-13T13:11:15 *** Naktibalda has quit IRC (Remote host closed the connection) 2011-08-13T13:14:43 *** mcstar has left #aichallenge ("WeeChat 0.3.5") 2011-08-13T13:22:35 *** onensora has quit IRC (Ping timeout: 252 seconds) 2011-08-13T13:25:33 *** Indecipherable_ has joined #aichallenge 2011-08-13T13:25:48 *** onensora has joined #aichallenge 2011-08-13T13:25:53 hey 2011-08-13T13:28:14 *** Naktibalda has joined #aichallenge 2011-08-13T13:46:48 *** berak has joined #aichallenge 2011-08-13T13:52:42 I need help 2011-08-13T13:56:06 Anyone? 2011-08-13T13:56:58 here is one 2011-08-13T13:58:11 oh 2011-08-13T13:58:19 I ran into a problem 2011-08-13T13:59:09 I'm running Windows, and neither of the lisp starter bots work locally 2011-08-13T13:59:16 They just crash 2011-08-13T13:59:40 oh have you asked that question a few days ago? 2011-08-13T14:00:23 Yep. Nobody answered lol. Goin to join the forum today 2011-08-13T14:00:39 As soon as I get the activation message 2011-08-13T14:00:40 *** mathis_ has quit IRC (Ping timeout: 252 seconds) 2011-08-13T14:00:53 I don't know LISP and don't use Windows, but mcstar knows LISP. He isn't around at the moment, but maybe he could give you a hint. 2011-08-13T14:01:31 I was talking to mcstar yesterday on #lisp i think 2011-08-13T14:01:40 lol. Thanks 2011-08-13T14:01:48 There should be an activation email for the forum, but not for your 'contest login'. (Unless that has been worked on recently) 2011-08-13T14:02:18 No I meant the activation link for the forum user account 2011-08-13T14:06:19 *** ibdknox has joined #aichallenge 2011-08-13T14:07:13 What language are most of the people here using? 2011-08-13T14:11:10 *** janzert has joined #aichallenge 2011-08-13T14:11:34 english 2011-08-13T14:18:10 I meant programming 2011-08-13T14:25:44 *** Indecipherable_ has quit IRC (Quit: Copywight 2007 Elmer Fudd. All wights wesewved.) 2011-08-13T14:41:34 Indecipherable_: You can look at the rankings for that 2011-08-13T14:54:17 Indecipherable_: Perl * 1, Go * 1, Clojure * 1, 2011-08-13T14:54:17 | CoffeeScript | 1 | 2011-08-13T14:54:17 | Scala | 1 | 2011-08-13T14:54:17 | Python3 | 2 | 2011-08-13T14:54:17 | D | 2 | 2011-08-13T14:54:17 | OCaml | 3 | 2011-08-13T14:54:17 | Haskell | 4 | 2011-08-13T14:54:18 | C | 5 | 2011-08-13T14:54:18 | Lisp | 9 | 2011-08-13T14:54:19 | Javascript | 10 | 2011-08-13T14:54:19 | PHP | 12 | 2011-08-13T14:54:20 | Ruby | 13 | 2011-08-13T14:54:20 | C# | 22 | 2011-08-13T14:54:21 | Python | 44 | 2011-08-13T15:03:12 *** boegel has joined #aichallenge 2011-08-13T15:13:03 *** boegel has quit IRC (Ping timeout: 276 seconds) 2011-08-13T15:19:20 *** amstan has quit IRC (Ping timeout: 260 seconds) 2011-08-13T15:33:50 *** dlila has joined #aichallenge 2011-08-13T15:34:34 *** amstan has joined #aichallenge 2011-08-13T15:34:35 *** ChanServ sets mode: +o amstan 2011-08-13T15:38:10 *** boegel has joined #aichallenge 2011-08-13T15:41:59 So here is the statistic on language usage again: http://pastebin.com/vfrbEij9 2011-08-13T15:49:47 *** berak has quit IRC (Quit: ChatZilla 0.9.84 [SeaMonkey 2.0a3/20090223135443]) 2011-08-13T15:58:36 *** Naktibalda has quit IRC (Quit: ChatZilla 0.9.87 [Firefox 5.0/20110622232440]) 2011-08-13T15:59:05 *** Accoun has quit IRC () 2011-08-13T16:12:46 *** Accoun has joined #aichallenge 2011-08-13T16:15:46 *** boegel has quit IRC (Quit: Leaving) 2011-08-13T16:18:24 *** jbrechtel has joined #aichallenge 2011-08-13T16:22:47 This might be a pretty dumb question, but why are the squares of all of the relevant radius' passed in as parameters instead of the radius themselves? 2011-08-13T16:26:58 *** demerzel has quit IRC (Ping timeout: 252 seconds) 2011-08-13T16:35:51 Nevermind 2011-08-13T16:36:06 That is a dumb question. 2011-08-13T16:55:09 *** demerzel has joined #aichallenge 2011-08-13T16:55:17 *** demerzel is now known as demerzel[work] 2011-08-13T16:55:56 *** Palmik has quit IRC (Remote host closed the connection) 2011-08-13T17:00:11 *** foRei has quit IRC (Read error: Connection reset by peer) 2011-08-13T17:00:26 *** foRei has joined #aichallenge 2011-08-13T17:03:51 *** Accoun has quit IRC (Remote host closed the connection) 2011-08-13T17:04:33 i forgot who made the viewer but all the buttons are fuzzy grey 2011-08-13T17:04:38 prob a cross-site thing idk 2011-08-13T17:29:02 demerzel[work]: mleise did 2011-08-13T17:29:39 is that a new name for him? i don't remember that name at all 2011-08-13T17:29:51 Frontier 2011-08-13T17:30:00 @amstan tell him that in firefox 4 with noscript (with hypertriangle enabled) the buttons are greyed out 2011-08-13T17:30:01 demerzel[work]: An error occured while trying to show the previous error. 2011-08-13T17:30:06 oh 2011-08-13T17:30:14 @amstan tell him that in firefox 4 with noscript with hypertriangle enabled the buttons are greyed out 2011-08-13T17:30:15