2012-04-19T00:01:01 *** alehorst has quit IRC (Ping timeout: 272 seconds) 2012-04-19T00:10:37 *** amstan has quit IRC (Ping timeout: 252 seconds) 2012-04-19T00:42:17 *** thestinger has quit IRC (Quit: WeeChat 0.3.7) 2012-04-19T01:07:02 *** Areks has joined #aichallenge 2012-04-19T01:28:33 *** kilae has joined #aichallenge 2012-04-19T01:35:45 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2012-04-19T02:07:53 *** NoxiaZ^ has joined #aichallenge 2012-04-19T02:08:49 *** NoxiaZ^ has quit IRC (Client Quit) 2012-04-19T02:20:53 *** antimatroid has quit IRC (Ping timeout: 245 seconds) 2012-04-19T02:28:41 *** antimatroid has joined #aichallenge 2012-04-19T02:49:41 *** replore_ has quit IRC (Remote host closed the connection) 2012-04-19T03:11:13 *** antimatroid has quit IRC (Ping timeout: 252 seconds) 2012-04-19T03:13:08 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2012-04-19T03:22:54 *** Garf has joined #aichallenge 2012-04-19T04:17:49 *** replore has joined #aichallenge 2012-04-19T04:29:23 *** mcstar has joined #aichallenge 2012-04-19T04:36:04 *** antimatroid has joined #aichallenge 2012-04-19T05:06:21 *** mceier has joined #aichallenge 2012-04-19T05:30:53 *** foRei has joined #aichallenge 2012-04-19T05:31:10 *** replore has quit IRC (Remote host closed the connection) 2012-04-19T05:36:24 *** mcstar has joined #aichallenge 2012-04-19T05:42:27 *** mviel_ is now known as mviel 2012-04-19T05:54:41 *** _dr <_dr!~johannes@faui32o.informatik.uni-erlangen.de> has quit IRC (Remote host closed the connection) 2012-04-19T06:41:14 *** replore_ has joined #aichallenge 2012-04-19T06:46:01 *** delt0r_ has quit IRC (Ping timeout: 248 seconds) 2012-04-19T06:55:13 *** heinrich5991 has quit IRC (Ping timeout: 272 seconds) 2012-04-19T06:58:29 *** delt0r_ has joined #aichallenge 2012-04-19T07:03:13 *** heinrich5991 has joined #aichallenge 2012-04-19T07:18:17 *** antimatroid has quit IRC (Ping timeout: 260 seconds) 2012-04-19T07:23:14 *** smiley1983 has quit IRC (Ping timeout: 264 seconds) 2012-04-19T07:23:28 *** smiley1983 has joined #aichallenge 2012-04-19T07:43:32 *** sigh has joined #aichallenge 2012-04-19T07:50:15 *** antimatroid has joined #aichallenge 2012-04-19T07:56:48 *** smiley1983 has quit IRC (Ping timeout: 265 seconds) 2012-04-19T07:58:13 *** smiley1983 has joined #aichallenge 2012-04-19T08:02:06 *** antimatroid has quit IRC (Ping timeout: 265 seconds) 2012-04-19T08:03:08 *** antimatroid has joined #aichallenge 2012-04-19T08:04:32 *** alehorst has joined #aichallenge 2012-04-19T08:16:32 *** mleise has joined #aichallenge 2012-04-19T09:03:33 *** amstan has joined #aichallenge 2012-04-19T09:03:33 *** ChanServ sets mode: +o amstan 2012-04-19T09:22:37 mleise: http://minus.com/ml6Gvc55x/ look at that 2012-04-19T09:22:55 it works thanks to this code http://hpaste.org/67267 2012-04-19T09:25:33 nice physics :) 2012-04-19T09:26:29 if 2 balls collide, they exchange speed, if more than 2, they disappear 2012-04-19T09:26:55 this all sucks, cause balls arent point particles, so sometimes the collision looks awkward 2012-04-19T09:36:03 why do they dissapear? 2012-04-19T09:38:47 amstan: because 2012-04-19T09:39:10 i was lazy, and it was easier, than writing down the equations and deal with it 2012-04-19T09:40:32 btw, it is kind of hard to imagine 3 point particles colliding anyway 2012-04-19T09:40:36 mcstar: do balls in real life exchange speeds like that? 2012-04-19T09:40:45 in 1 dimension 2012-04-19T09:41:02 don't they each go with the speeds they were originally going in the opposite direction? 2012-04-19T09:41:05 as if they're hitting a wall 2012-04-19T09:41:07 but since they have finite size, they will behave differently 2012-04-19T09:41:27 oh, nvm 2012-04-19T09:41:31 the direction changes 2012-04-19T09:41:46 * amstan needs more physics classes 2012-04-19T09:42:13 btw this was meant to be something like an asteroids clone, so objects when collide, whey should disappear anyway 2012-04-19T09:42:37 i just wanted to make it more spectacular for mleise 2012-04-19T09:42:55 amstan: he is going to rewrite it in D 2012-04-19T09:43:02 tel 2012-04-19T09:43:33 *** Accoun has quit IRC () 2012-04-19T09:44:59 amstan: the apropo of the code is, that i use rendering to detect collisions 2012-04-19T09:45:11 i guess you guys do differently 2012-04-19T09:45:21 you, erique and smiley 2012-04-19T09:45:28 a 2012-04-19T09:45:40 *** Accoun has joined #aichallenge 2012-04-19T09:45:48 *** replore has joined #aichallenge 2012-04-19T09:51:08 hm? 2012-04-19T09:51:43 hi 2012-04-19T09:51:48 nothing 2012-04-19T09:52:35 maybe something 2012-04-19T09:52:43 couldn't you have written that in lisp instead of line noise? ;-) 2012-04-19T09:52:48 pixel scanning, horrible 2012-04-19T09:52:50 what was the evolution of the asteroids code? 2012-04-19T09:53:21 mleise: i only scan pixels in the intersection of bounding rectangles 2012-04-19T09:53:35 aerique: unfortunately no 2012-04-19T09:53:54 another lost soul :-( 2012-04-19T09:54:00 yeah, sorry 2012-04-19T09:54:16 there are many reasons im more into haskell than lisp 2012-04-19T09:54:44 anyway, smiley's asteroids is the current engine afaik 2012-04-19T09:55:13 he didn't fork it from me, not quite sure where he got it from 2012-04-19T09:55:30 well, it started with you, you did it in i dont know what, then i saw amstan did something in python, and after that smiley in html5 2012-04-19T09:55:54 smiley's engine is in python 2012-04-19T09:56:02 amstan and me started both at the same time without knowing about it 2012-04-19T09:56:07 but amstan used pyqt4 2012-04-19T09:56:27 aerique: and what did you write it in? 2012-04-19T09:56:29 smiley came later but his is the most mature 2012-04-19T09:56:44 mcstar: i adapted the ants engine so in python 2012-04-19T09:56:57 and how did you visualize? 2012-04-19T09:57:01 not that i had much experience with it 2012-04-19T09:57:51 the engine wrote out the game state and i had a script that converted each round to a png. all pngs got converted to an avi 2012-04-19T09:57:57 kinda convoluted ;-) 2012-04-19T09:59:01 aerique: do you have the links to the video? 2012-04-19T09:59:25 -s 2012-04-19T10:00:31 *** iglo has joined #aichallenge 2012-04-19T10:00:49 mcstar: https://github.com/aerique/aichallenge/tree/asteroids/asteroids has got the links 2012-04-19T10:01:17 thanks 2012-04-19T10:02:17 *** sigh has quit IRC (Remote host closed the connection) 2012-04-19T10:05:19 do you think it has some merit, to use and exact collision detection (regarding objects shapes), or just treat everything as circles? 2012-04-19T10:05:26 and->an 2012-04-19T10:07:24 SHAPES! 2012-04-19T10:07:53 just grab one of the bazillion box2d libraries and use that 2012-04-19T10:07:55 *** kurnevsky has joined #aichallenge 2012-04-19T10:08:22 it could actually handle all of the physics 2012-04-19T10:09:01 http://code.google.com/p/pybox2d/ 2012-04-19T10:09:46 rwest: so 'fun' is equal to 'import' in your world 2012-04-19T10:10:12 physics and collision detection are not fun 2012-04-19T10:10:13 EVER 2012-04-19T10:10:41 making the game is fun though 2012-04-19T10:11:14 * mcstar has thrown a bag of stinky disagreement in your way 2012-04-19T10:11:27 you sir, are a freak 2012-04-19T10:11:54 * mcstar likes to be called sir 2012-04-19T10:11:55 I have gone down the route of I'll write the physics from scratch 2012-04-19T10:12:02 2 weeks later 2012-04-19T10:12:11 I had imported a nice box2d library 2012-04-19T10:14:10 rwest: asteroids doesnt need a full physics engine 2012-04-19T10:14:43 better change that then! 2012-04-19T10:16:28 mcstar: how does one test the performance of the Asteroids program? 2012-04-19T10:16:46 * mcstar shrugs 2012-04-19T10:16:50 It's full of Cairo drawing and renders at a fixed frame rate, lol 2012-04-19T10:17:09 you up the rate 2012-04-19T10:17:25 but obviously the test field changes :p 2012-04-19T10:17:26 if it becomes sluggish, then it means you are demanding too much 2012-04-19T10:17:35 you averae 2012-04-19T10:17:39 average* 2012-04-19T10:17:48 the ships and rocks vanish after few milliseconds 2012-04-19T10:18:05 how can you reliably measure anything under these circumstances 2012-04-19T10:18:16 you tell them to pass each other 2012-04-19T10:18:28 ok? 2012-04-19T10:18:30 ok 2012-04-19T10:18:40 are you talking about my code? 2012-04-19T10:18:52 or did you finish the rewrite? 2012-04-19T10:19:02 sure, so simply removing the last step (removing dead objects from the list) should do it 2012-04-19T10:19:15 I finished ... well I have a rewrite 2012-04-19T10:19:30 now i need to compile gtkd for GDC 2012-04-19T10:19:37 there is a part, when i take a set difference 2012-04-19T10:19:50 if you dont do that it will be ok 2012-04-19T10:19:53 so i can run it with the outdated GCC implementation of D 2012-04-19T10:19:53 i think 2012-04-19T10:20:14 i dont know what you are talking about 2012-04-19T10:20:24 D doesnt have its own compiler? 2012-04-19T10:20:32 i mean to machine code 2012-04-19T10:20:33 Yes, but it sucks 2012-04-19T10:20:49 It is the reference, but the backend isn't good at optimization 2012-04-19T10:21:35 So the frontend was glued onto GCC which works much better, but I have to rebuild the static libraries for it now. 2012-04-19T10:21:46 since the runtime is different between both compilers 2012-04-19T10:21:55 its like MSVC or GCC on Windows 2012-04-19T10:22:13 mleise: if you comment out the writeIORef line, it will do the work, but will leave objects on screen 2012-04-19T10:22:16 Haskell only has one compiler that is widely used? 2012-04-19T10:22:29 mcstar: thx Ill do that 2012-04-19T10:22:48 far the most used one is GHC 2012-04-19T10:22:56 and if i set the timer to 1 ms I'll have some fun 2012-04-19T10:23:13 ghc has llvm backend, and asm 2012-04-19T10:23:19 im not sure about C 2012-04-19T10:23:35 it could generate code via C, maybe it still does 2012-04-19T10:23:49 yeah these projects are fun 2012-04-19T10:23:56 there is a D to JavaScript compiler as well 2012-04-19T10:25:44 mleise: 1ms looks fine here 2012-04-19T10:25:51 Oo 2012-04-19T10:26:17 also, Xorg eats 1 core up 2012-04-19T10:26:25 there is something wrong with my X 2012-04-19T10:26:25 /tmp/cc4nrJUx.o:(.data+0x640): undefined reference to `___t.8' ... WTF? when did I screw my compiler? 2012-04-19T10:27:06 mcstar: maybe your copy from the Drawable to the Pixbuf slows X down 2012-04-19T10:27:12 no 2012-04-19T10:27:21 i hibernate all the time 2012-04-19T10:27:29 and after a while, X becomes slow 2012-04-19T10:27:41 that sucks 2012-04-19T10:27:53 (not that i would notice, but glxgears for example has like 3-5 fps fullscreen) 2012-04-19T10:27:54 I reboot now and then 2012-04-19T10:28:02 rofl 2012-04-19T10:28:18 make that 1/2 2012-04-19T10:28:18 you are not using an nVidia gfx card 2012-04-19T10:28:28 nvidia integrated 2012-04-19T10:28:32 but it is alright 2012-04-19T10:30:07 i think ill have that reboot now 2012-04-19T10:30:07 *** mcstar has quit IRC (Quit: mcstar) 2012-04-19T10:33:32 *** mcstar has joined #aichallenge 2012-04-19T10:34:41 1ms really stresses X, not that surprising 2012-04-19T10:34:53 but the 3d issue i think i really cased by updates 2012-04-19T10:35:03 is* 2012-04-19T10:35:09 do you use dekstiop effects? 2012-04-19T10:36:16 no 2012-04-19T10:36:21 i use a tiling wm 2012-04-19T10:36:25 which is Awesome 2012-04-19T10:37:12 oh, by updates i mean, probably poor app couldnt load the gl libs, cause the old kernel module was loaded, or something simlar 2012-04-19T10:37:32 *** Areks has quit IRC (Ping timeout: 272 seconds) 2012-04-19T10:37:56 wtf, still low fps, i have to investigate it 2012-04-19T10:40:39 first stop Xorg.0.log 2012-04-19T10:40:50 glxinfo, driinfo 2012-04-19T10:41:01 it says i have direct rendering 2012-04-19T10:41:30 I did that a few times. Just ressurected an old IBM to connect it to the receiver and TV 2012-04-19T10:41:42 so what is up with your OpenGL driver? 2012-04-19T10:42:11 DRI means the drivers have kernel support for some ... fast stuff I don't really understand anyway 2012-04-19T10:42:40 So there is kernel/DRI, the Xorg driver and the OpenGL driver as far as I understand it 2012-04-19T10:43:09 all with their own hardware specialized contribution to the gfx card 2012-04-19T10:43:42 do you use binary drivers from nVidia? 2012-04-19T10:43:54 they are usually painless 2012-04-19T10:44:26 ofc 2012-04-19T10:44:31 everything is loaded, no errors 2012-04-19T10:45:24 no errors? ok. then your card is *really* slow :p 2012-04-19T10:46:08 If it's only OpenGL then maybe the wrong driver is loaded 2012-04-19T10:46:16 something definitely happened to my 3d accel 2012-04-19T10:46:20 no 2012-04-19T10:46:21 i can read 2012-04-19T10:46:28 although ... even a software driver is faster 2012-04-19T10:46:32 mleise: dont try to help me, it just makes me angry 2012-04-19T10:46:56 when hardware fs me in the butt i go for a run 2012-04-19T10:46:56 lol, you are an uneasy person 2012-04-19T10:47:11 ill be back in an hour 2012-04-19T10:47:39 eselect opengl list 2012-04-19T10:47:39 Available OpenGL implementations: 2012-04-19T10:47:39 [1] nvidia * 2012-04-19T10:47:39 [2] xorg-x11 2012-04-19T10:48:36 i have an idea 2012-04-19T10:48:43 i switched to gdm from xdm recently 2012-04-19T10:48:46 *** mcstar has quit IRC (Quit: mcstar) 2012-04-19T10:53:06 *** mcstar has joined #aichallenge 2012-04-19T10:53:15 ill be back soon 2012-04-19T10:57:48 gotta use Mate! 2012-04-19T11:01:42 *** thestinger has joined #aichallenge 2012-04-19T11:06:58 *** kurnevsky has quit IRC (Read error: Connection reset by peer) 2012-04-19T11:09:54 *** kurnevsky has joined #aichallenge 2012-04-19T11:22:28 *** Accoun has quit IRC (Ping timeout: 240 seconds) 2012-04-19T11:25:17 *** pairofdice has joined #aichallenge 2012-04-19T11:36:20 *** kilae has quit IRC (Quit: ChatZilla 0.9.88.2 [Firefox 11.0/20120312181643]) 2012-04-19T11:38:35 *** Accoun has joined #aichallenge 2012-04-19T11:45:23 *** kurnevsky has quit IRC (Read error: Connection reset by peer) 2012-04-19T11:57:07 *** kurnevsky has joined #aichallenge 2012-04-19T12:28:20 pairofdice: done the cs387 challenge question yet? I just decided to do it 2012-04-19T12:28:36 * thestinger reused the ideas for solving a similar problem in crypto-class 2012-04-19T12:28:37 doing 262 2012-04-19T12:29:26 pairofdice: do they have challenge questions? 2012-04-19T12:29:32 or just regular hw 2012-04-19T12:29:53 I guess I'll sign up for all 4 like you :P 2012-04-19T12:31:40 some regexen and nondeterministic finite state machines... 2012-04-19T12:32:05 ie. follow a graph 2012-04-19T12:32:39 *** kurnevsky has quit IRC (Read error: Connection reset by peer) 2012-04-19T12:32:44 as programming assignments 2012-04-19T12:33:18 pairofdice: the coursera crypto class has number theory stuff this week :P 2012-04-19T12:33:23 * thestinger is falling asleep 2012-04-19T12:33:32 Sounds fun! 2012-04-19T12:33:51 ;p 2012-04-19T12:33:52 well this stuff is probably useful if I ever decided to do project euler again 2012-04-19T12:34:19 I just watched 5 minutes and then realized I wasn't paying attention 2012-04-19T12:34:21 and had to rewatch it :P 2012-04-19T12:34:30 web application course feels like it will be usefl 2012-04-19T12:34:40 pairofdice: you don't want to watch 2 hours of number theory lectures? 2012-04-19T12:34:42 :P 2012-04-19T12:34:48 If I ever want to become a billionaire 2012-04-19T12:34:54 No. 2012-04-19T12:35:10 *** kurnevsky has joined #aichallenge 2012-04-19T12:35:53 pairofdice: btw a neat trick from the first crypto-class assignment is that XORing with a space flips the case of regular ascii characters 2012-04-19T12:36:08 oh, cool 2012-04-19T12:36:13 useful for verifying your findings from other methods 2012-04-19T12:36:40 what does xoring a space with space do 2012-04-19T12:36:55 well, the value of space is 32 2012-04-19T12:37:02 so it flips the 32nd bit 2012-04-19T12:37:12 which is the difference between lowercase ascii chars and uppercase ones 2012-04-19T12:37:33 'a' ^ ' ' == 'A' 2012-04-19T12:37:37 'A' ^ ' ' == 'a' 2012-04-19T12:37:37 ye 2012-04-19T12:37:39 etc. 2012-04-19T12:47:10 pairofdice: xor of a value against its own value is always zero 2012-04-19T12:47:20 or where you trying to be funny? 2012-04-19T12:47:26 Right, stupid me 2012-04-19T12:47:35 Nah, tiring day 2012-04-19T12:51:11 the 5th bit 2012-04-19T13:01:56 Zannick: yeah, that :P 2012-04-19T13:04:02 or the bit that represents 32 2012-04-19T13:06:15 *** kilae has joined #aichallenge 2012-04-19T13:06:16 *** Palmik has joined #aichallenge 2012-04-19T13:06:18 bit log(32)/log(2) 2012-04-19T13:06:37 *** replor___ has joined #aichallenge 2012-04-19T13:07:53 *** replore has quit IRC (Ping timeout: 245 seconds) 2012-04-19T13:55:03 *** kilae has quit IRC (Quit: ChatZilla 0.9.88.2 [Firefox 11.0/20120312181643]) 2012-04-19T14:05:50 *** mviel_ has joined #aichallenge 2012-04-19T14:06:49 *** mviel has quit IRC (Ping timeout: 265 seconds) 2012-04-19T14:12:33 *** mviel_ has quit IRC (Ping timeout: 246 seconds) 2012-04-19T14:16:13 *** kilae has joined #aichallenge 2012-04-19T14:24:02 mleise: im back 2012-04-19T14:24:11 Oh, cs387 challenge is neat 2012-04-19T14:24:12 this took a bit longer than i anticipated 2012-04-19T14:24:20 *** mviel_ has joined #aichallenge 2012-04-19T14:24:25 You took the long route? 2012-04-19T14:24:36 mleise: you know potatoes? 2012-04-19T14:24:58 they nub 2012-04-19T14:25:07 i think thats the word 2012-04-19T14:25:15 anyway, they needed to be removed 2012-04-19T14:26:12 Sounds back-breaking, though I sit so much that anything is back-breaking 2012-04-19T14:26:40 pairofdice: you and me both! 2012-04-19T14:26:59 mleise: maybe they are called sprouts 2012-04-19T14:27:21 mleise: anyway, can i see the code? 2012-04-19T14:29:09 yes, in spring i cannot eat my potatoes as fast as they grow these ... things 2012-04-19T14:29:31 potatoes keep pretty well 2012-04-19T14:29:43 mcstar: I just had to disable composite effects and work around the pixbufs not beeing freed, sucking up my memory :p 2012-04-19T14:29:47 *** iglo has quit IRC (Remote host closed the connection) 2012-04-19T14:30:01 oh/ 2012-04-19T14:30:04 ? 2012-04-19T14:30:10 is D no garbage collected? 2012-04-19T14:30:12 not 2012-04-19T14:30:44 composite shouldnt affect 2d performance,should it? 2012-04-19T14:31:06 mleise: will i be able to compile the code? 2012-04-19T14:32:58 I don't think so. I had my trouble getting yours to compile even with cabal 2012-04-19T14:33:39 You'd need to download gtkD and a compiler, but to have good results, you should use GDC. Is that available in your distribution? 2012-04-19T14:34:28 https://en.wikipedia.org/wiki/Letter_frequency#Relative_frequencies_of_letters_in_the_English_language this might help in cs387 challenge... 2012-04-19T14:34:28 i dont know what gdc is 2012-04-19T14:34:38 pairofdice: space! 2012-04-19T14:34:51 gdc is the D compiler tacked onto GCC 2012-04-19T14:34:58 oh, right 2012-04-19T14:35:29 mleise: i have gdc in aur 2012-04-19T14:35:34 probably it wil lbuild 2012-04-19T14:36:18 good luck! I built gtkD with "DC=gdc make", gdc must be on the path 2012-04-19T14:41:06 mleise: help me out 2012-04-19T14:41:31 http://dlang.org/ 2012-04-19T14:41:44 theres some code example, does this supposed to work? 2012-04-19T14:43:09 http://codepad.org/uyCMQ0HC 2012-04-19T14:43:28 that is supposed to work only with the reference compiler DMD 2012-04-19T14:43:46 it comes with a build tool called rdmd that is used in the script 2012-04-19T14:43:49 *** foRei has quit IRC (Read error: Connection reset by peer) 2012-04-19T14:44:11 it should still compile with gdc if you write: "gdc main.d" 2012-04-19T14:44:37 *** foRei has joined #aichallenge 2012-04-19T14:45:16 mleise: i just try to compile something simpler 2012-04-19T14:45:19 but i get errors 2012-04-19T14:45:29 that was a simple program 2012-04-19T14:45:40 http://sprunge.us/PaNP 2012-04-19T14:46:06 looks like ... what does gdc -v print? 2012-04-19T14:46:18 http://sprunge.us/bWOd 2012-04-19T14:46:20 this is the code 2012-04-19T14:46:30 *** delt0r_ has quit IRC (Ping timeout: 276 seconds) 2012-04-19T14:47:14 http://sprunge.us/LWei 2012-04-19T14:47:21 I can compile that with no error. Make sure you use a gdc that uses DMD 2.075 2012-04-19T14:47:25 hmmh, why is my charcount's len() 105... 2012-04-19T14:47:37 what am I doing 2012-04-19T14:47:41 1.020 ... is waaaaaay too old :p 2012-04-19T14:47:51 k 2012-04-19T14:48:04 mleise: i missed it like, 10 years? 2012-04-19T14:48:13 round about, yes 2012-04-19T14:48:20 XD 2012-04-19T14:48:23 oh, damnit... 2012-04-19T14:49:09 mleise: but is the gdc version ok? 2012-04-19T14:50:05 mleise: btw, from a quick glance, your code looks neat 2012-04-19T14:50:08 well, depends how you put it. your gdc is too old to compile any code today, because it was built using an ancient compiler frontend 2012-04-19T14:50:37 really? i felt it was close to becoming ugly 2012-04-19T14:51:14 *** alehorst has quit IRC (Remote host closed the connection) 2012-04-19T14:51:28 ill try gdc-git, that should be up to date 2012-04-19T14:52:41 haha, i tried to compile it from source and ended up with a compiler that produces linker errors. i finally used these sources: hg clone -u 00b19ad7e81f https://bitbucket.org/goshawk/gdc 2012-04-19T14:52:46 that's the version i use 2012-04-19T14:54:57 here are the build instructions: https://bitbucket.org/goshawk/gdc/wiki/Home#!building 2012-04-19T14:55:26 mleise: if the AUR version wont build by itself, im afraid i wont bother with it 2012-04-19T14:55:40 i have to clean up some c++ tonight 2012-04-19T14:55:54 no problem, I was srprised you'd even go as far as to build from sources at all 2012-04-19T14:56:21 i can send you a binary though. it prints a number to the console ever time it processed a frame 2012-04-19T14:56:58 mleise: dcc? 2012-04-19T14:57:20 Pidgin ... I don't know. never tried it or portforwarded anything 2012-04-19T14:57:32 would email be ok? 2012-04-19T14:57:33 *** amstan has quit IRC (Read error: Connection reset by peer) 2012-04-19T14:57:38 k 2012-04-19T14:57:44 *** amstan has joined #aichallenge 2012-04-19T14:57:44 *** ChanServ sets mode: +o amstan 2012-04-19T14:59:02 *** delt0r_ has joined #aichallenge 2012-04-19T15:02:40 oh your code has the same memory leak that mine had 2012-04-19T15:03:02 it's the pixbuf from drawable that isn't freed 2012-04-19T15:04:47 by the way, your haskell code is a great deal faster. congratulations :) 2012-04-19T15:07:01 *** pairofdice has quit IRC (Quit: It is time for you to leave) 2012-04-19T15:16:33 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has joined #aichallenge 2012-04-19T15:16:50 *** Accoun has quit IRC () 2012-04-19T15:17:01 mleise: noway 2012-04-19T15:18:39 mleise: that haskell code has absolutely no optimizations 2012-04-19T15:18:52 i think that was clear to you 2012-04-19T15:19:31 mleise: how do you noticed the memory leak? 2012-04-19T15:19:35 well, it looked nice and straight forward once i undesrtood most of it 2012-04-19T15:19:57 i set the timer to 0ms and let it run for 30 seconds 2012-04-19T15:20:00 *bang* 2012-04-19T15:20:05 HAHA 2012-04-19T15:20:08 slide show 2012-04-19T15:20:41 actually i have that small tray icon system monitor thingy from gnome always running 2012-04-19T15:21:00 it shows me memory use, cpu and network traffic. wouldn't want to live without that 2012-04-19T15:21:20 pix <- pixbufGetFromDrawable pix $ conv rect 2012-04-19T15:21:26 do you mean this pixbuf? 2012-04-19T15:21:31 yep 2012-04-19T15:21:34 or th ebig one? 2012-04-19T15:21:36 ah 2012-04-19T15:21:42 no the 10000000000 small ones 2012-04-19T15:21:45 but these are really tiny ones 2012-04-19T15:22:03 they add up eventually 2012-04-19T15:22:03 they aren't so cute when they clutter up your ram 2012-04-19T15:22:43 mleise: i dont think i got an answer, is D gc'd or not? 2012-04-19T15:22:52 yes, it is 2012-04-19T15:23:28 maybe the gtk runtime holds onto the pixbufs 2012-04-19T15:24:01 mleise: it wouldnt be hard to do without them 2012-04-19T15:24:06 maybe, but don't expect a GC to see that 'foreign' memory 2012-04-19T15:24:41 it may believe that it doesn't need to collect any garbage when there is a lot of external memory in use through wrapper objects 2012-04-19T15:24:54 without pixbufs? 2012-04-19T15:25:04 yeah 2012-04-19T15:25:10 no, cairo offers the functions to do intersection tests 2012-04-19T15:25:11 it was just convenience 2012-04-19T15:25:20 really? 2012-04-19T15:25:30 i hope so :p 2012-04-19T15:25:31 i wrote it in a hurry 2012-04-19T15:25:49 I used quite a lot of convenience functions in my DebugPainter for ants 2012-04-19T15:25:56 i just meant, that all those little pixbufs could be eliminated, by reading from only 1 pixbuf, that holds the whole screen 2012-04-19T15:25:59 like measuring the bounding box of a path 2012-04-19T15:26:03 mcstar: I think it's more a resource leak than a simple memory leak 2012-04-19T15:26:33 thestinger: explain 2012-04-19T15:27:01 mcstar: well, GC takes care of memory but not things like file descriptors and other resources 2012-04-19T15:27:16 I'm sure there's a way to deal with it automatically in haskell though 2012-04-19T15:27:22 but there is no way to 'close' a pixbuf 2012-04-19T15:27:31 oh 2012-04-19T15:28:10 mcstar: do you have to delete them from wherever you had them render? (I have no idea how cairo works) 2012-04-19T15:28:32 *** Accoun has joined #aichallenge 2012-04-19T15:28:40 i dont even render them 2012-04-19T15:28:52 it is a copy of a rendered region of a drawablke 2012-04-19T15:28:56 drawable 2012-04-19T15:29:04 (meaning something you can paint on) 2012-04-19T15:29:22 but there is no connection between them, other than an initial copy 2012-04-19T15:29:26 g_object_unref(pix); worked like a charm for me 2012-04-19T15:29:52 mleise: ill try to kill my code 2012-04-19T15:29:53 that was resorting to pure C though 2012-04-19T15:30:31 I did try to destroy the wrapper object before, but it didn't work 2012-04-19T15:30:44 so it was back to the roots for me 2012-04-19T15:32:02 mleise: haha, X had 4.8GB resource memory 2012-04-19T15:32:21 xrestop? 2012-04-19T15:32:44 hm? 2012-04-19T15:32:45 mcstar: one more thing, it may be unimportant, but on exit your code sometimes complains. maybe the timer is firing when the window is already closed? 2012-04-19T15:33:09 asteroids: user error (widgetGetDrawWindow: no DrawWindow available (the widget is probably not realized) 2012-04-19T15:33:12 this ^^? 2012-04-19T15:33:18 yes! 2012-04-19T15:33:28 i can say this: whatever 2012-04-19T15:33:44 sry, i dont care about gtk stuff much 2012-04-19T15:33:55 that part of the code is from a gtk2hs tutorial 2012-04-19T15:34:03 (which makes the window, and quits it) 2012-04-19T15:39:32 rofl, 25% time spent in nvidia_drv.so 2012-04-19T15:40:13 the rest in xorg, pixman, libc... 2012-04-19T15:40:57 even my network driver has done more than my program 2012-04-19T15:41:40 i guess the profiler has a bug 2012-04-19T15:42:42 :) 2012-04-19T15:43:10 mleise: seems, that D is efficient after all 2012-04-19T15:43:33 in doing nothing? 2012-04-19T15:45:58 The top entry: 0.0903% main.d:262 bool main.checkBuffer(ref const(main.RGB), ref const(main.RGB), gtkc.gdktypes.GdkPixbuf*) 2012-04-19T15:47:21 I'll sleep over this and maybe learn me some cairo and gtk on the way 2012-04-19T15:48:33 maybe.. 2012-04-19T15:49:03 in the meantime, this gdc still builds 2012-04-19T15:54:23 i know :) 2012-04-19T15:54:37 well it is a good sign, isn't it? 2012-04-19T15:56:18 couldnt be better 2012-04-19T15:56:28 well, except having a binary package... 2012-04-19T15:56:48 i'm torn 2012-04-19T15:57:01 whats up? 2012-04-19T15:57:20 idk what i should use for my interface thingy, pyside+qwt(somehow), pyqt+qwt, pyside+matplotlib, or pyside+something of my own built 2012-04-19T15:57:24 i need graphs 2012-04-19T15:57:36 what kind? 2012-04-19T15:57:42 i have a simple kind 2012-04-19T15:58:06 but thats in c++, i always forget you do python 2012-04-19T15:58:38 i read somewhere theres not much difference i pyside and pyqt api 2012-04-19T15:58:44 it'll be complicated 2012-04-19T15:58:51 and it might be commercial, so pyqt might be out 2012-04-19T15:58:53 amstan: matplotlib is awesome 2012-04-19T15:59:00 it is slow 2012-04-19T15:59:03 thestinger: one thing i noticed though: it's slow 2012-04-19T15:59:10 if i want any animation i get like 4fps 2012-04-19T15:59:20 i want to display data as i get it 2012-04-19T15:59:34 oh, yeah 2012-04-19T15:59:38 it sucks for real-time graphs 2012-04-19T15:59:39 thestinger: i do like that i can integrate it in pyqt/pyside though 2012-04-19T16:00:00 might be able to fix that though 2012-04-19T16:00:00 and pyside+qwt doesn't really exist, there's a project for it but it looks abandoned 2012-04-19T16:00:06 by default it makes fancy vector graphics 2012-04-19T16:00:14 but i want the fancy vector stuff too 2012-04-19T16:00:36 i implemented my own graph for pyqt in another project and it's pretty and fast at the same time 2012-04-19T16:00:40 amstan: i just wrote my plotting widget 2012-04-19T16:00:52 it isnt hard, am not sure what do you need 2012-04-19T16:00:53 not to mention that i can make it qt interactive too 2012-04-19T16:01:15 mcstar: i'm building a UI for an electrophoresis device 2012-04-19T16:01:37 it'll have 96 cells, each cell having like 3 graphs(live) each 2012-04-19T16:01:47 so it's a pretty big project, not something i can just throw together 2012-04-19T16:02:03 96x3 real time graphs? 2012-04-19T16:02:10 not on the same screen 2012-04-19T16:02:12 all on one screen? 2012-04-19T16:02:16 i see 2012-04-19T16:02:23 still, it is helluvalot 2012-04-19T16:02:33 anyone try qwt? 2012-04-19T16:05:29 amstan: have you considered c++ and pure qt? 2012-04-19T16:05:40 that sounds kinda scary 2012-04-19T16:05:49 python is scary! 2012-04-19T16:05:59 i also want to do serial interaction with my device 2012-04-19T16:06:16 amstan: we dropped serial in favor of libusb 2012-04-19T16:06:25 it is better 2012-04-19T16:06:27 imho 2012-04-19T16:06:32 mcstar: except my device is serial 2012-04-19T16:06:39 for now 2012-04-19T16:06:55 really serial, or just emulates it? 2012-04-19T16:07:06 for now really serial 2012-04-19T16:07:47 i wrote a wrapper class on top of boost asio, now it is in the wastebin 2012-04-19T16:08:46 i really liked how easy was to make the python stuff cross platform 2012-04-19T16:09:04 amstan: you do asynchron io, right? 2012-04-19T16:09:08 just checking 2012-04-19T16:09:17 yes 2012-04-19T16:09:55 amstan: i had no problem building boost and my project on win7 2012-04-19T16:10:31 i don't feel like c++ though, i want to easily develop this UI thing and even use it during debugging for the device 2012-04-19T16:11:22 and even if i do switch to c++, i'll still have the same problem 2012-04-19T16:11:28 what should i do about graphing stuff 2012-04-19T16:13:11 you need to decide between qwt or roll your own 2012-04-19T16:13:49 amstan: use an html5 library 2012-04-19T16:13:55 bleh 2012-04-19T16:14:07 rwest: and how should i access the serial device? 2012-04-19T16:14:07 can make nice live graphs that run easily 2012-04-19T16:14:16 rwest: not everyone is web developer 2012-04-19T16:14:33 rwest: he has speed issues with matplotlib, canvas will be worse 2012-04-19T16:14:41 you don't have to use a web language to get them 2012-04-19T16:14:54 thestinger: oh, didn't read the whole thing 2012-04-19T16:15:39 he wants to do real-time graphing 2012-04-19T16:15:43 rwest: so then i need to make a python or c++ aberation that includes a html5 rendering engine 2012-04-19T16:15:53 qtwebkit 2012-04-19T16:15:57 thestinger: well.. javascript 2012-04-19T16:16:08 anyway, this is nonsense, excuse me 2012-04-19T16:16:13 just pick qwt and be happy 2012-04-19T16:16:18 amstan: see this is where the nice part would come in, you find one that takes the data in json or xml and you just output it 2012-04-19T16:16:22 except no pyside qwt bindings 2012-04-19T16:16:25 that i could compile 2012-04-19T16:16:35 what is wrong with pyqt? 2012-04-19T16:16:44 but if you have speed troubles natively, it won't help heh 2012-04-19T16:17:09 mcstar: the whole non-commercial thing I think he said 2012-04-19T16:17:16 mcstar: gpl only, pyside is lgpl 2012-04-19T16:17:23 mcstar: it's a commercial thing 2012-04-19T16:17:33 fuck cling 2012-04-19T16:17:36 I can never get it to build 2012-04-19T16:17:52 svn: E155007: '/home/strcat/packages/cling-svn/src/llvm-build/tools/cling/lib/UserInterface' is not a working copy 2012-04-19T16:17:54 CMake Error at tools/cling/lib/UserInterface/CMakeLists.txt:26 (MESSAGE): 2012-04-19T16:17:56 Svn command returned error, is it installed? 2012-04-19T16:17:57 amstan: you would need to free the source? 2012-04-19T16:17:58 mcstar: also, the pyqwt stuff is outdated(uses qwt5 not qwt6) 2012-04-19T16:18:03 mcstar: yep 2012-04-19T16:18:20 whats the licence of the real qwt? 2012-04-19T16:18:22 which i'm not opposed to, but my employer might be 2012-04-19T16:18:26 i think thers even commercial edition 2012-04-19T16:18:41 there is some plotting lib, that uses qwt 2012-04-19T16:18:53 and it is commercial 2012-04-19T16:19:11 amstan: i suggest you do it in c++ with qwt 2012-04-19T16:19:20 feel free to ignore me XD 2012-04-19T16:19:37 yes, but do i even need qwt? that's the question 2012-04-19T16:19:39 mleise: gdc is built! 2012-04-19T16:19:54 is it fun to use? i only need a bunch of graphs 2012-04-19T16:19:56 if you are unsure that you can make those graphs 2012-04-19T16:20:21 i dont think you will be writing more than a couple hundred lines per graph 2012-04-19T16:20:24 i can make them, but i'm affraid the code will get ugly by making my own stuff 2012-04-19T16:20:26 200 2012-04-19T16:20:59 k, time to go home 2012-04-19T16:21:01 oh 2012-04-19T16:21:27 *** racko has joined #aichallenge 2012-04-19T16:21:34 mcstar: gdc -v ? 2012-04-19T16:22:27 mleise: doesnt mention any dmd 2012-04-19T16:22:58 *** amstan has quit IRC (Quit: Konversation terminated!) 2012-04-19T16:23:04 heh, I think it's actually building correctly 2012-04-19T16:23:14 nooope 2012-04-19T16:23:17 then good luck and good night 2012-04-19T16:23:28 ah yeah its late 2012-04-19T16:23:34 good nite everyone 2012-04-19T16:24:11 *** mcstar has quit IRC (Quit: mcstar) 2012-04-19T16:24:54 *** kurnevsky has quit IRC (Read error: Connection reset by peer) 2012-04-19T16:25:09 *** kurnevsky has joined #aichallenge 2012-04-19T16:25:32 *** kurnevsky has quit IRC (Client Quit) 2012-04-19T16:28:24 *** mleise has quit IRC (Ping timeout: 272 seconds) 2012-04-19T16:33:20 *** alehorst has joined #aichallenge 2012-04-19T16:49:12 *** choas has joined #aichallenge 2012-04-19T16:51:26 *** racko has quit IRC (Quit: Page closed) 2012-04-19T16:52:38 *** NoxiaZ^ has joined #aichallenge 2012-04-19T16:52:59 *** kilae has quit IRC (Quit: ChatZilla 0.9.88.2 [Firefox 11.0/20120312181643]) 2012-04-19T17:12:41 *** janzert has quit IRC (Read error: Connection reset by peer) 2012-04-19T17:15:36 *** janzert has joined #aichallenge 2012-04-19T17:36:07 *** Cyndre_ is now known as Cyndre 2012-04-19T17:47:45 *** antimatroid has quit IRC (Ping timeout: 252 seconds) 2012-04-19T17:56:13 *** choas has quit IRC (Ping timeout: 250 seconds) 2012-04-19T17:58:48 *** mceier has quit IRC (Quit: leaving) 2012-04-19T17:59:22 *** antimatroid has joined #aichallenge 2012-04-19T18:02:02 *** amstan has joined #aichallenge 2012-04-19T18:02:02 *** ChanServ sets mode: +o amstan 2012-04-19T18:02:45 *** foRei has quit IRC (Read error: Connection reset by peer) 2012-04-19T18:03:49 *** foRei has joined #aichallenge 2012-04-19T18:39:58 *** NoxiaZ^ has quit IRC (Ping timeout: 246 seconds) 2012-04-19T19:05:52 *** smiley1983 has quit IRC (Ping timeout: 276 seconds) 2012-04-19T19:06:19 *** smiley1983 has joined #aichallenge 2012-04-19T19:06:35 *** replore_ has joined #aichallenge 2012-04-19T19:08:12 *** replor___ has quit IRC (Ping timeout: 260 seconds) 2012-04-19T19:18:19 *** Palmik has quit IRC () 2012-04-19T19:20:02 *** smiley1983 has quit IRC (Ping timeout: 272 seconds) 2012-04-19T19:20:42 *** smiley1983 has joined #aichallenge 2012-04-19T19:21:54 *** Garf has quit IRC (Quit: Make a new plan, Stan!) 2012-04-19T19:24:39 *** replore_ has quit IRC (Remote host closed the connection) 2012-04-19T20:07:09 *** amstan has quit IRC (Read error: Connection reset by peer) 2012-04-19T20:07:29 *** amstan has joined #aichallenge 2012-04-19T20:07:29 *** ChanServ sets mode: +o amstan 2012-04-19T20:08:39 *** foRei has quit IRC (Quit: Bye) 2012-04-19T20:19:00 *** _flag <_flag!~flag@69-165-173-172.dsl.teksavvy.com> has quit IRC (Quit: Lost terminal) 2012-04-19T22:46:23 *** delt0r_ has quit IRC (Ping timeout: 260 seconds) 2012-04-19T22:58:45 *** delt0r_ has joined #aichallenge