| 00:49.28 | silvap_ | justin |
| 00:49.29 | silvap_ | u there? |
| 01:00.20 | Twingy | yah |
| 02:59.29 | heina | justin. ola |
| 02:59.56 | Twingy | hi |
| 02:59.57 | heina | i met u at SIGGRAPH. after the rtrt day. |
| 03:00.02 | Twingy | ah, ok |
| 03:00.06 | heina | remember now. |
| 03:00.08 | heina | kewl |
| 03:00.18 | Twingy | you were talking to gordon with me? |
| 03:00.47 | heina | nodnod |
| 03:00.52 | heina | you mentioned 700 lines of code. |
| 03:00.56 | Twingy | yep |
| 03:00.57 | heina | poking around, i assume you mean in the rt dir |
| 03:01.01 | Twingy | nope |
| 03:01.06 | heina | whereat? |
| 03:01.09 | Twingy | that's brl-cad's raytracer |
| 03:01.15 | Twingy | brlcad/src/adrt/libtie/tie.c |
| 03:01.31 | Twingy | that's the source for the engine |
| 03:01.33 | heina | ah! |
| 03:01.37 | heina | ok. wouldn't have found that. |
| 03:02.10 | Twingy | we were poking at the assembly output of the engine today |
| 03:02.20 | Twingy | I'm starting on some of the SSE intrinsics next week |
| 03:02.41 | heina | so what's in the tie engine exactly? |
| 03:02.46 | Twingy | look at the header |
| 03:03.09 | Twingy | and the top of tie.c |
| 03:03.17 | Twingy | tie.c has doxygen documentation |
| 03:03.28 | heina | ah |
| 03:03.36 | heina | i have ctags anyway. |
| 03:04.25 | heina | so that's what you're using mostly to "throw together" some rt'ing? |
| 03:04.27 | Twingy | in its current state you'll get about 250k - 400k rays on 250k triangles on a 3ghz p4 |
| 03:04.40 | Twingy | it's the backend to the two applications, "RISE" and "ISST" |
| 03:04.50 | Twingy | ISST is the interactive application, RISE is for producing realistic images |
| 03:05.09 | Twingy | like this |
| 03:05.15 | Twingy | www.rt.cx/~justin/images/stryker_slat.png |
| 03:05.26 | heina | so, basically it's the non-shader part of the rt, which does the kd-tree and ray calculations on arbitrary meshes. |
| 03:05.35 | Twingy | right |
| 03:05.47 | Twingy | one source file and 2 headers |
| 03:05.52 | heina | love it |
| 03:05.53 | heina | ;) |
| 03:06.07 | Twingy | turn-key |
| 03:06.10 | heina | btw, i picked up that physically based rendering book that they recommended. haven't gotten too into it, but it's a definite great book |
| 03:06.16 | heina | ooh...pretty pictures there. |
| 03:06.20 | heina | how're you shading that? |
| 03:06.24 | Twingy | path tracing |
| 03:06.26 | Twingy | 8 trillion rays |
| 03:06.36 | heina | lol. ah, that's the non-interactive one. |
| 03:07.04 | Twingy | right |
| 03:07.15 | Twingy | that's with depth of field |
| 03:07.21 | heina | was just noticing that. |
| 03:07.54 | heina | the ambient occlusion looks so good on rt images. (global illumination) |
| 03:08.02 | Twingy | once I implement energy redistributed path tracing it'll take 1/10th the rays |
| 03:08.20 | Twingy | yah, AO is not bad, but it's still a dirty hack imho :) |
| 03:08.25 | heina | nice. what's the overhead on that. |
| 03:08.27 | heina | AO rocks! |
| 03:08.28 | Twingy | for interactive stuff it's the way to go |
| 03:08.29 | heina | :) for offline |
| 03:08.34 | heina | doing bent normals as well. |
| 03:08.42 | Twingy | AO or instant radiosity |
| 03:08.44 | heina | going to play around with switching it all over to PRT though |
| 03:08.52 | heina | heh |
| 03:09.00 | Twingy | there is a lib there called libtienet |
| 03:09.06 | Twingy | so if you want to distribute you code with tie |
| 03:09.10 | Twingy | you can use libtienet as well |
| 03:09.12 | heina | we've got a mental ray plugin that generates AO/BN and bakes it to textures. |
| 03:09.29 | heina | doh. i see it's copyright army. |
| 03:09.39 | Twingy | GPL |
| 03:09.41 | heina | is it GPL? |
| 03:09.46 | heina | great |
| 03:09.53 | Twingy | LGPL actually |
| 03:09.59 | Twingy | <PROTECTED> |
| 03:09.59 | Twingy | <PROTECTED> |
| 03:09.59 | Twingy | <PROTECTED> |
| 03:10.08 | heina | i can't imagine i'd distribute it for anything, but definitely for offline tools and such. |
| 03:10.14 | heina | nodnod <-- knows LGPL and GPL |
| 03:10.29 | Twingy | welp, it'd be nice to see people other than us using it |
| 03:10.34 | Twingy | then I can get some feedback for improvements and whatnot |
| 03:10.37 | heina | ok kewl. i'm glad i caught you on here, cuz i was poking around in the rt directory and it didn't look right. |
| 03:10.45 | Twingy | yah, that's all CSG stuff |
| 03:10.48 | heina | what i would use it for is the following... |
| 03:10.51 | Twingy | ~25k - 30k rays/sec |
| 03:11.09 | heina | ...i'd use it for generating AO/BN or PRT information in plugins to DCC tools (like maya) |
| 03:11.23 | heina | i might also play around with a rt on the Cell, but you didn't hear that from me. hehe |
| 03:11.46 | Twingy | PRT is photon mapping? |
| 03:11.52 | silvap_ | cell..? |
| 03:11.55 | heina | the rt is ~25-30k cuz of CSG? the tie is doing more like 400k/sec? |
| 03:12.02 | Twingy | right |
| 03:12.05 | heina | cell - new IBM processor in the PlayStation 3. |
| 03:12.07 | Twingy | TIE is over an order of magnitude faster |
| 03:12.27 | Twingy | librt was written by Mike Muuss, the guy that invented "ping" |
| 03:12.31 | heina | PRT --> pre-computed radiance transfer for relighting and getting low-frequency lighting (soft shadows, AO, BN) at interactive rates |
| 03:12.32 | silvap_ | thought so :P |
| 03:12.36 | heina | ping! |
| 03:12.42 | heina | rock on. (old network programmer here ) |
| 03:12.49 | Twingy | it was the first real-time raytracer |
| 03:12.49 | silvap_ | are u on the ps3 dev team? |
| 03:12.57 | Twingy | ~1995 |
| 03:13.09 | heina | no, i'm with a 3rd party developer |
| 03:13.20 | heina | there was a guy...Pat....H-something who was at SIGGRAPH who is. |
| 03:13.22 | silvap_ | game dev? |
| 03:13.29 | Twingy | well, hopefully that number will jump from ~400k to ~800k over the next month |
| 03:13.33 | heina | Take Two/Visual Concepts |
| 03:13.36 | Twingy | Pat Hanrahan? |
| 03:13.40 | heina | yes |
| 03:13.41 | silvap_ | whoa cool |
| 03:13.45 | heina | damn, that's SSE for ya |
| 03:13.49 | heina | grin, thx |
| 03:14.12 | Twingy | if I can get the MLRTA in I should be around 1.6 mil or so |
| 03:14.22 | heina | my last game was for THQ/Pandemic studios....if you guys are military, you might know it. Full Spectrum Warrior. i did all the networking for that game |
| 03:14.29 | heina | MLRTA? |
| 03:14.35 | silvap_ | lol sweet |
| 03:14.39 | Twingy | multi level ray tracing algorithm |
| 03:14.49 | heina | didn't catch that one |
| 03:14.57 | heina | fuck. 1.6 million tris/sec. |
| 03:15.06 | Twingy | it's a frustum culling method |
| 03:15.16 | Twingy | primarily for optical rendering |
| 03:15.21 | silvap_ | so whats up with the interest in rt? trying to get real time ray tracing on the ps3? |
| 03:15.31 | heina | grin |
| 03:15.31 | Twingy | we do alot of ballistic analysis that involves firing rays all the way through geometry building a hitlist |
| 03:15.44 | heina | oh sweet. that's kewl stuff. |
| 03:15.50 | Twingy | as far as I know this is the fastest open-source ray-tracing engine |
| 03:15.55 | Twingy | the faster ones are all closed source |
| 03:16.00 | heina | damn. go army |
| 03:16.01 | heina | :) |
| 03:16.09 | Twingy | so, hopefully I'll begin closing the gap |
| 03:16.16 | heina | silvap_, we are currently using RT for generating information offline. |
| 03:16.38 | heina | silvap_, we can't do RT cuz it looks like crap, but probably you'll see some demos soon. next-gen (5 years from now) we might. |
| 03:16.55 | heina | nice. how much with the MLRTA complicate the code? |
| 03:16.59 | silvap_ | nice |
| 03:17.04 | Twingy | saarland is doing some work with cell processors and rt |
| 03:17.08 | Twingy | quite a bit |
| 03:17.11 | silvap_ | oh, didnt know |
| 03:17.15 | heina | yeah, that's big news |
| 03:17.17 | Twingy | the frustum culling stuff complicates things |
| 03:17.28 | silvap_ | how's development on the ps3? |
| 03:17.29 | heina | if they get that FPGA implementation onto an ASIC or if Nvidia or ATI pick that up...then we'll have vun |
| 03:17.36 | silvap_ | i assume u guys already have a reference platform? |
| 03:17.37 | Twingy | the engine code itself is too cryptic to understand without the comments as it is |
| 03:17.44 | heina | lol |
| 03:17.48 | Twingy | I asked them how their asic progress was coming |
| 03:17.49 | heina | that sucks. make it longer and clearer. |
| 03:17.52 | heina | yeah? |
| 03:17.56 | heina | i hope they get traction on that. |
| 03:17.59 | Twingy | I bought an fpga for testing some rt code at work, but haven't had time yet |
| 03:18.03 | heina | silvap_, it's interesting. still very early. |
| 03:18.09 | Twingy | I think they should shove their transistors into an nvidia card |
| 03:18.11 | heina | haha |
| 03:18.22 | heina | you bought an FPGA. nice |
| 03:18.35 | silvap_ | what's the bottom line? 2 baziilion flops hype is true? =) |
| 03:18.37 | heina | yeah, i agree with that. there's a fair bit of shared approach, but it's really the memory bus that is key |
| 03:18.45 | heina | silvap_, not supposed to say. |
| 03:18.50 | silvap_ | ohh |
| 03:18.56 | heina | silvap_, let's just say hype is hype. |
| 03:19.10 | Twingy | heina, I think all you need is about a dozen lines of code to actually write a test app with the engine stand-alone |
| 03:19.31 | Twingy | just remember to use all the optimizations |
| 03:19.34 | heina | there are some awesome things you can do with the Cell you can't do on other proc's just cuz it is so setup for massive SSE stuff, but it's weak in other areas. extremely complicated to code for, but we'll get it figured out |
| 03:19.42 | Twingy | not just -O3 |
| 03:19.57 | heina | Twingy, you should put together a "hello world" style rt app. something that renders a triangle. |
| 03:20.04 | Twingy | I am, tommorrow morning |
| 03:20.12 | heina | send me that. |
| 03:20.15 | heina | you have my email? |
| 03:20.20 | Twingy | nope |
| 03:20.25 | Twingy | send me a msg, justin@arl.army.mil |
| 03:21.03 | heina | sent |
| 03:21.07 | Twingy | k |
| 03:21.21 | Twingy | I'll be interested to see how my engine would fair on a cell processor with such limited cache |
| 03:21.28 | silvap_ | ps4 with 48 cells |
| 03:21.30 | heina | sweet. ok, i gotta get back to this shadow code. doing shadow map implementation |
| 03:21.37 | Twingy | okie |
| 03:21.38 | silvap_ | justin, 60 fps for stryker |
| 03:21.45 | heina | when i get a cell under my desk, i'll set it up and let you know |
| 03:21.55 | Twingy | hehe, k |
| 03:21.57 | heina | i want to play around with some rt stuff on there. |
| 03:22.05 | Twingy | all we got is 2,308 cpu opteron cluster |
| 03:22.14 | silvap_ | lies |
| 03:22.17 | Twingy | err 2,380 rather |
| 03:22.20 | heina | the trick is the SPU cache is only 200k, so going to either need tiny models instanced, or a streaming solution. |
| 03:22.23 | heina | fuck |
| 03:22.33 | Twingy | 6GB per node |
| 03:22.50 | heina | heh. so what do you need an optimized RT for again? |
| 03:22.51 | heina | :) |
| 03:23.00 | Twingy | beats me :) |
| 03:23.01 | silvap_ | he's anal like that |
| 03:23.17 | Twingy | those machines are batch queue'd |
| 03:23.25 | Twingy | so getting interactive time on them is not really possible |
| 03:23.25 | heina | ah |
| 03:23.30 | heina | good point. |
| 03:23.32 | silvap_ | im telling u man; lock everyone out |
| 03:23.35 | Twingy | for RISE it would be nice |
| 03:23.45 | Twingy | path tracing a few hundred trillion rays |
| 03:23.46 | heina | you could tile render like crazy to that |
| 03:23.56 | Twingy | the network i/o would kill you |
| 03:24.01 | Twingy | it's myrinet |
| 03:24.02 | heina | 8 pixels to each cpu |
| 03:24.08 | Twingy | so I'd have to use MPI to get any benefit |
| 03:24.19 | Twingy | unless you used the myrinet it would not be worth it |
| 03:24.28 | heina | push all the geometry out. all you're pushing is camera matrix and getting pixel colors back. |
| 03:24.41 | Twingy | correct, but like I said, the latency is what kills you |
| 03:24.45 | heina | nodnod |
| 03:24.49 | heina | well, the Cell is kinda like that |
| 03:24.57 | heina | you do all the work out on the SPU's but it's got a wicked fast bus |
| 03:25.05 | heina | the bus can write directly to the VRAM too....sexy |
| 03:25.08 | heina | rather, the SPU's |
| 03:25.09 | Twingy | shoulda gone to siggraph prasad |
| 03:25.20 | heina | omg SIGGRAPH ftw |
| 03:25.54 | *** join/#brlcad starseeker (n=starseek@ip68-230-220-222.rd.hr.cox.net) |
| 03:25.57 | heina | if you get the proceedings on dvd, check out the PRT half-day. from monday. you'll find that really interesting. all kinds of kewl linear algebra. lotsa Spherical Harmonics and such. kewl. |
| 03:26.00 | heina | nice |
| 03:26.06 | Twingy | k |
| 03:26.15 | Twingy | how'd that sketch go? |
| 03:26.25 | heina | k. back to work. gonna idle on here. be around. |
| 03:26.34 | Twingy | okie, I'm usually on here during work day |
| 03:26.40 | heina | kewl. c u round |
| 03:27.13 | starseeker | is this the channel for the BRL computer aided design package? |
| 03:27.27 | Twingy | yes it is |
| 03:27.32 | starseeker | Cool |
| 03:27.41 | Twingy | we've been discovered, flee! |
| 03:27.45 | starseeker | gonna take a run at 7.4.0 on Gentoo |
| 03:27.49 | Twingy | k |
| 03:28.00 | Twingy | probly take you 15 - 30 mins to compile |
| 03:28.10 | silvap_ | stop torturing me with siggraph! |
| 03:28.20 | Twingy | :) |
| 03:28.22 | silvap_ | i was a fool! |
| 03:28.31 | Twingy | next year prasad, boston |
| 03:28.40 | silvap_ | yay |
| 03:28.45 | silvap_ | no jet lag |
| 03:28.49 | Twingy | you could even... ride your bike |
| 03:28.49 | starseeker | Twingy: my main concern is avoiding library name conflicts - last time I did this I hosed my system but good |
| 03:29.03 | Twingy | well set your configure prefix to something like |
| 03:29.10 | Twingy | /home/username/local |
| 03:29.24 | Twingy | configure --enable-optimized --prefix=/home/username/local |
| 03:29.28 | Twingy | then add ~/local/bin to your path |
| 03:29.44 | Twingy | then you're set |
| 03:29.44 | starseeker | Twingy: That would work for a local install, but I'm trying to create/debug a gentoo ebuild |
| 03:29.50 | Twingy | ah, k |
| 03:29.57 | Twingy | then maybe put in /usr/local/brlcad |
| 03:30.00 | silvap_ | Twingy, u have a gcc x87 instr reference handy? |
| 03:30.06 | silvap_ | online |
| 03:30.07 | Twingy | silva, negative |
| 03:30.22 | silvap_ | how am i going to trace this code then.. |
| 03:30.30 | Twingy | wield the mighty google |
| 03:30.35 | silvap_ | negative |
| 03:31.34 | Twingy | http://66.102.7.104/search?q=cache:food28db1d0J:security-protocols.com/programming-tutorials/Assembly/fpuopcode.html+x87+instruction+list&hl=en |
| 03:32.43 | Twingy | I should put that onto my objectives for next year |
| 03:33.04 | silvap_ | who has hot friends |
| 03:33.29 | Twingy | me |
| 03:33.30 | Twingy | :) |
| 03:34.08 | silvap_ | who has hot female friends |
| 03:34.31 | silvap_ | look, there are plenty of x87 references online |
| 03:34.37 | silvap_ | none with gcc specific stuff |
| 03:34.42 | silvap_ | ie: flds |
| 03:35.24 | Twingy | http://www.rt.cx/~justin/images/disney_2005/DSCN0747.JPG |
| 03:35.28 | starseeker | Out of curosity, has anybody here ever successfully figured out SALOME? |
| 03:35.41 | Twingy | http://www.rt.cx/~justin/SIGGRAPH_2005/images/DSCN0917.jpg |
| 03:37.09 | starseeker | http://www.salome-platform.org/home/presentation/overview/ |
| 03:37.47 | silvap_ | and who dumped his gf for a ray tracer? |
| 03:37.49 | silvap_ | huh? HUH? |
| 03:37.51 | silvap_ | =) |
| 03:38.20 | Twingy | is that a FEM engine? |
| 03:38.54 | Twingy | hrm, dad is buying sushi tommorrow |
| 03:39.01 | starseeker | It seems to be for numerical simulation on models, but I've never run it to see |
| 03:39.06 | Twingy | ah |
| 03:39.08 | Twingy | looks interesting |
| 03:39.08 | starseeker | dependency list is a killer |
| 03:39.39 | starseeker | Seems to use open cascade, that's why I'm curious |
| 03:40.04 | starseeker | might be worth an ebuild (or 10) if it's powerful enough to be useful - Linux is a tad short on CAD options at the moment :-( |
| 03:43.07 | starseeker | building... building... |
| 03:43.42 | Twingy | got 2 cpus? |
| 03:43.44 | Twingy | or HT? |
| 03:43.59 | starseeker | Nah, just one - an old AMD AthlonXP |
| 03:44.03 | Twingy | k |
| 03:51.31 | silvap_ | justin |
| 03:51.39 | Twingy | you rang? |
| 03:51.46 | silvap_ | are u doing the sse in asm or using the c funcs/macros |
| 03:51.49 | silvap_ | whatever they are |
| 03:51.53 | Twingy | SSE intrinsincs |
| 03:52.02 | Twingy | google for that, check out the MSDN site |
| 03:52.08 | Twingy | 70 instructions |
| 03:53.26 | silvap_ | well they are exposed through functions, no? |
| 03:53.37 | Twingy | they're... intrinsics |
| 03:53.59 | Twingy | the compiler treats them in a special way |
| 03:54.05 | silvap_ | u and ur buzzwords |
| 03:55.23 | Twingy | buzzzzzz |
| 03:55.28 | Twingy | werds |
| 03:56.19 | starseeker | Curosity question - could the java parts of brlcad conceivably work with gcj, or do they need sun/something else mature? |
| 03:57.10 | Twingy | no clue, I'm a java troglodyte |
| 03:57.17 | silvap_ | so am i |
| 03:57.24 | Twingy | I took a 4 month course on it, but purged it from my head |
| 03:57.29 | Twingy | apparently I got an A in it |
| 03:57.38 | silvap_ | smart man |
| 03:58.02 | Twingy | hrm, these bananas are a tad green |
| 03:58.05 | silvap_ | ive lost my roots since m3 |
| 03:58.14 | Twingy | I think erik has too |
| 03:58.25 | starseeker | ah, libbu one of many sources of name troubles :-). |
| 03:58.28 | Twingy | feels good to dive into some assembly no? |
| 03:58.42 | Twingy | starseeker, you really wanna talk to sean |
| 03:58.46 | Twingy | he'll be on tommorrow |
| 03:58.50 | starseeker | Cool. |
| 03:59.09 | starseeker | He may remember me - I'm the one who was on here a long while back mucking with Gentoo |
| 03:59.16 | Twingy | k |
| 03:59.39 | starseeker | I don't know cad worth squat, but I figure doing my part to bring BRLCAD to a larger audience is the least I can do... |
| 04:00.41 | starseeker | Twingy: Are you a core coder for BRL-CAD? |
| 04:01.03 | Twingy | erm, define brl-cad :) |
| 04:01.14 | Twingy | I know very little about brl-cad, I implemented photon mapping into it, but that's it |
| 04:01.17 | starseeker | The huge thingy I just downloaded ;-) |
| 04:01.19 | silvap_ | me like assembly |
| 04:01.21 | Twingy | well |
| 04:01.30 | Twingy | I also did the high performance ray-tracer in "adrt" under src/ |
| 04:01.37 | Twingy | but it's not exactly "part" of brl-cad |
| 04:01.42 | Twingy | it's just sitting there for open source reasons |
| 04:01.47 | starseeker | :-) |
| 04:01.57 | silvap_ | starseeker, the core coder is stuck on a train somewhere in the midwest at this very moment |
| 04:02.03 | Twingy | yah |
| 04:02.05 | Twingy | so if it explodes |
| 04:02.07 | Twingy | we're all doomed |
| 04:02.09 | starseeker | Ooo, that's not fun |
| 04:02.27 | silvap_ | seriously, where is sean? |
| 04:03.50 | Twingy | hiding under my bed |
| 04:04.01 | Twingy | don't tell lee |
| 04:04.03 | silvap_ | from tranese? |
| 04:04.09 | silvap_ | she's in a bad mood |
| 04:04.14 | Twingy | from darth vadress |
| 04:04.17 | silvap_ | thanks to mister owens |
| 04:04.23 | Twingy | ah |
| 04:04.25 | Twingy | well |
| 04:04.30 | Twingy | stir up the bee hive and bad shit happens |
| 04:05.20 | silvap_ | stirring up hives don't usually end up with the queen leaving |
| 04:05.27 | Twingy | heh |
| 04:05.36 | Twingy | they do when you do her sister |
| 04:05.38 | Twingy | err |
| 04:05.40 | Twingy | :) |
| 04:05.48 | Twingy | wrong topic |
| 04:06.44 | Twingy | I think nasa should send a few monkey's to mars |
| 04:07.05 | Twingy | rig up some electrodes to their head with some 'ai' software to build a colony |
| 04:07.59 | Twingy | silvap_, you ever listen to XRM radio? |
| 04:08.36 | silvap_ | no |
| 04:08.49 | Twingy | try it sometime, shoutcast.com, xrm electronic |
| 04:09.18 | Twingy | I really need to finish my nose cone for the rocket |
| 04:09.44 | Twingy | maybe I'll just do something lame and make a linear cone |
| 04:11.53 | Twingy | we need vulnerability algorithms that use SSE intrinsics |
| 04:13.38 | Twingy | prasad, check this out |
| 04:14.02 | Twingy | rt.cx/~justin/inspiron_700m.pdf |
| 04:15.22 | starseeker | bah, more weirdness |
| 04:15.36 | starseeker | Too tired to mess with this tonight, I'll probably try tomorrow |
| 04:15.39 | starseeker | thanks all! |
| 04:15.46 | Twingy | k |
| 04:15.58 | silvap_ | did u put in the orders yet? |
| 04:16.09 | Twingy | yah, but nothings been ordered yet |
| 04:16.13 | Twingy | why? |
| 04:16.23 | silvap_ | im supposed to get a mac lappy |
| 04:16.33 | Twingy | I put you down for a 15" w/ 2GB |
| 04:16.52 | silvap_ | ok |
| 04:17.04 | Twingy | or maybe it was a 8" |
| 04:17.07 | Twingy | I can't remember |
| 04:17.08 | Twingy | ;) |
| 04:17.13 | silvap_ | wondering if gettting a sparc10 lappy is more useful |
| 04:17.17 | Twingy | heh |
| 04:17.25 | Twingy | why on earth would you want that |
| 04:17.29 | silvap_ | that's right |
| 04:17.31 | silvap_ | no more m3 |
| 04:18.20 | Twingy | say hello to m3.5 |
| 04:19.47 | silvap_ | nM |
| 04:20.03 | silvap_ | nano m***** |
| 04:20.07 | Twingy | heh |
| 04:20.42 | silvap_ | did u say u wanted to do the procedural textures? |
| 04:20.47 | silvap_ | or was that something else |
| 04:21.48 | Twingy | I have a couple already |
| 04:21.53 | Twingy | but more are welcome |
| 04:22.01 | silvap_ | couple what? |
| 04:22.01 | Twingy | I'd like a sparkle texture |
| 04:22.06 | Twingy | procedural textures |
| 04:22.11 | silvap_ | oh |
| 04:22.13 | Twingy | like glitter/sparkle |
| 04:22.30 | Twingy | then we could do glitter stryker |
| 04:22.31 | silvap_ | i meant the material core |
| 04:22.38 | Twingy | or sparkle stryker |
| 04:22.42 | Twingy | that's what I mean |
| 04:22.47 | Twingy | texture/material |
| 04:22.58 | Twingy | I'm using interchangeably |
| 04:23.02 | silvap_ | ahso |
| 04:23.19 | Twingy | hrm |
| 04:23.23 | Twingy | only $94 to fly out to reno |
| 04:23.33 | silvap_ | this was in post processing yes? |
| 04:23.42 | silvap_ | or is this aggregation per interaction |
| 04:23.47 | Twingy | no |
| 04:23.51 | Twingy | ray his surface |
| 04:23.56 | Twingy | surface generates color for surface |
| 04:24.08 | Twingy | (and/or normal) |
| 04:24.14 | Twingy | and gets processed by rendering method |
| 04:24.21 | Twingy | *hits |
| 04:24.53 | Twingy | what I really need is clouds |
| 04:24.55 | Twingy | more than anything |
| 04:25.05 | Twingy | some kickass volumetric cell clouds |
| 04:25.09 | silvap_ | volumetric |
| 04:25.12 | silvap_ | jinx |
| 04:25.19 | Twingy | moo |
| 04:25.31 | silvap_ | MoooOOoOOoOOoOOOO |
| 04:25.56 | Twingy | you see that bovine programming language? |
| 04:25.59 | silvap_ | @.@ Oroooooo!? |
| 04:26.06 | silvap_ | i believe i did |
| 04:26.15 | Twingy | I think muves should get written in that |
| 04:27.05 | silvap_ | u think gcc CFLAGS_SUPER_OPT produces the fastest code possible? |
| 04:27.19 | silvap_ | or u think i can find some opcodes to prune out |
| 04:27.35 | Twingy | I think I'm going to come up with a benchmark app tommorrow and you guys can help me test ideas :) |
| 04:27.59 | silvap_ | 'write the whole thing in assembly' |
| 04:28.01 | silvap_ | that's my idea |
| 04:28.04 | Twingy | bad idea :) |
| 04:28.18 | silvap_ | it's not really that hard |
| 04:28.20 | Twingy | though the engine it's pretty much defined well enough you could |
| 04:28.38 | silvap_ | im talkng about _work |
| 04:28.57 | Twingy | yah |
| 04:29.26 | silvap_ | send me the asm listing u printed out today thru email |
| 04:29.44 | silvap_ | i need to.. uh.. substitute a cpl of stack-offsets with variable names |
| 04:29.46 | silvap_ | for my own sanity |
| 04:30.06 | Twingy | we can do it tommorrow, I'm heading to bed soon :) |
| 04:30.11 | silvap_ | same |
| 04:30.16 | silvap_ | gntie |
| 04:30.18 | Twingy | I have like elevendy billion things to get done in the next 2 weeks |
| 04:30.19 | silvap_ | gnite |
| 04:30.22 | Twingy | later |
| 05:31.57 | *** join/#brlcad mahesh (n=mahesh@12-217-239-24.client.mchsi.com) |
| 07:49.19 | *** join/#brlcad mahesh (n=mahesh@12-217-239-24.client.mchsi.com) |
| 11:11.39 | *** join/#brlcad archivist_ (n=djc@host217-35-76-52.in-addr.btopenworld.com) |
| 14:12.51 | *** join/#brlcad Twingy_ (n=justin@pcp0011647505pcs.aberdn01.md.comcast.net) |
| 21:22.46 | *** join/#brlcad CIA-5 (n=CIA@flapjack.navi.cx) |
| 22:29.13 | CIA-5 | BRL-CAD: 03lbutler * 10brlcad/src/libbn/ (29 files): Doxygen changes |
| 22:30.01 | CIA-5 | BRL-CAD: 03lbutler * 10brlcad/src/libbu/ (50 files): Doxygen changes |
| 22:30.50 | CIA-5 | BRL-CAD: 03lbutler * 10brlcad/src/libfb/ (46 files): Doxygen changes |
| 22:31.47 | CIA-5 | BRL-CAD: 03lbutler * 10brlcad/src/librt/ (151 files): Doxygen changes |
| 22:41.11 | CIA-5 | BRL-CAD: 03lbutler * 10brlcad/Doxyfile: Doxygen incorporation |
| 22:42.36 | CIA-5 | BRL-CAD: 03lbutler * 10brlcad/src/gtools/ (g_qa.1 g_qa.c): Fixed paralell bug |
| 23:38.36 | Twingy | moo |
