| 00:22.59 | ``Erik | fucking goddamnit |
| 00:23.13 | ``Erik | I can deal with the heat, but I need to get the a/c fixed, some redneck is blasting lynyrd skynyrd |
| 00:23.30 | ``Erik | it's ok in small doses, but not a whole album when I'm not in the mood for it |
| 00:23.56 | ``Erik | at least the song going now is a good one... simple man... *sigh* |
| 00:26.04 | pacman87 | ``Erik: http://xkcd.com/368/ |
| 00:27.05 | ``Erik | heh |
| 00:28.05 | ``Erik | <-- has thought about using a collander as a near parabolic director, a spark plug at roughly the focal point, and a nice assload of high capacity fast capacitors... |
| 00:28.20 | ``Erik | uhm, not a collander, a metal uh, like, grease sieve thing |
| 00:28.41 | ``Erik | the metal screen thing :) EM cannon style, yo |
| 00:29.32 | ``Erik | (unfortunately, cars seem to be their own faraday cages to protect the target electronics) |
| 00:34.50 | ``Erik | heh |
| 00:34.52 | ``Erik | http://img144.imageshack.us/img144/1855/01721azf4.jpg |
| 01:29.48 | *** join/#brlcad Twingy (n=justin@74.92.144.217) |
| 03:52.30 | *** join/#brlcad andre|away (n=chatzill@189.71.49.194) |
| 03:52.50 | andrecastelo | hey guys |
| 04:23.28 | CIA-22 | BRL-CAD: 03andrecastelo * r31364 10/brlcad/trunk/src/rt/viewmlt.c: |
| 04:23.28 | CIA-22 | BRL-CAD: Removed struct path_list, mlt_app now has a direct access to a struct |
| 04:23.28 | CIA-22 | BRL-CAD: point_list. Added a point list of light sources. Added initialization of mlt_app |
| 04:23.28 | CIA-22 | BRL-CAD: lists (lightsources and path). Added hit point calculation and hit point storage |
| 04:23.28 | CIA-22 | BRL-CAD: in mlt_app->path, in the rayhit() function. |
| 04:30.15 | CIA-22 | BRL-CAD: 03andrecastelo * r31365 10/brlcad/trunk/src/rt/viewmlt.c: Fixed structure handling bugs (ap->a_ray->r_pt ---> ap->a_ray.r_pt, for example). |
| 04:46.28 | *** join/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1177872280.dsl.bell.ca) |
| 04:46.38 | *** part/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1177872280.dsl.bell.ca) |
| 04:47.45 | *** join/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1177872280.dsl.bell.ca) |
| 05:42.15 | *** join/#brlcad clock_ (n=clock@77-56-95-209.dclient.hispeed.ch) |
| 09:32.03 | *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch) |
| 10:02.49 | *** join/#brlcad mafm (n=mafm@elnet-111.lip.pt) |
| 10:06.39 | mafm | howdy folks |
| 10:28.25 | *** join/#brlcad elite01 (n=elite01@dslb-088-071-032-104.pools.arcor-ip.net) |
| 10:58.49 | *** join/#brlcad d_rossberg (n=rossberg@bz.bzflag.bz) |
| 11:25.40 | *** join/#brlcad thing0 (n=ric@58.171.197.15) |
| 11:51.09 | brlcad | howdy |
| 12:19.45 | *** join/#brlcad andrecastelo (n=chatzill@189.71.49.194) |
| 12:21.14 | andrecastelo | good morning guys |
| 12:22.40 | *** join/#brlcad clock__ (n=clock@zux221-122-143.adsl.green.ch) |
| 12:36.10 | mafm | andrecastelo: I wasn't watching the football match, no |
| 12:37.37 | mafm | the guys at the lab set up the conference room with the projector to watch them together, though |
| 12:42.06 | mafm | brlcad: so what about the console design? |
| 13:10.19 | CIA-22 | BRL-CAD: 03erikgreenwald * r31366 10/brlcad/trunk/src/rt/viewmlt.c: irrelevant formating changes |
| 13:19.15 | mafm | hmm |
| 13:19.34 | mafm | do you guys think that listeners should go in separate classes |
| 13:19.46 | mafm | or maybe add that capability to a bigger class too? |
| 13:28.31 | CIA-22 | BRL-CAD: 03mafm * r31367 10/rt^3/trunk/src/g3d/ (GuiConsole.cxx GuiConsole.h): Implemented properly the GuiConsole window auto-resizing when OGRE RenderWindow (i.e. the window that appears in your desktop) is resized, and cleaned up some of the other listeners/callbacks. |
| 13:37.02 | *** join/#brlcad docelic (n=docelic@78.134.196.163) |
| 13:39.05 | CIA-22 | BRL-CAD: 03mafm * r31368 10/rt^3/trunk/src/g3d/GuiConsole.cxx: Commenting out debugging cout's. |
| 14:06.25 | *** join/#brlcad prasad_ (n=psilva@h-72-245-122-226.mclnva23.covad.net) |
| 14:28.29 | *** join/#brlcad homovulgaris (n=homovulg@202.63.233.61) |
| 14:55.24 | CIA-22 | BRL-CAD: 03mafm * r31369 10/rt^3/trunk/src/g3d/ (Application.cxx GuiConsole.cxx): Setting the console text zone to wrap lines instead of cutting them, and setting OIS to not disable keyboard auto-repeat for all the applications in the same DISPLAY -- that's *very* annoying. |
| 15:44.12 | poolio | brlcad: how might I go about calling rt_xxx_tess from a C++ file? I don't see any header with that defined...just the rt table. Should I go through there? |
| 16:09.29 | *** join/#brlcad homovulgaris (n=homovulg@202.63.233.61) |
| 16:52.12 | homovulgaris | hi all :) |
| 16:52.32 | homovulgaris | brlcad: i was writing definitions for constraintnetwork which is basically a graph |
| 16:52.40 | homovulgaris | can we use the boost library ? |
| 16:53.00 | homovulgaris | i think their headers are pretty rigorous and generic |
| 16:53.24 | homovulgaris | 2 hours to first commit hopefully :) |
| 16:54.21 | homovulgaris | can't we split up the existing doxygen librt page. Am i the only one facing difficulty in loading that page |
| 17:01.55 | mafm | never used that page |
| 17:02.11 | homovulgaris | :) it is a pretty "huge" page |
| 17:02.48 | homovulgaris | mafm: gui images look nice |
| 17:05.28 | mafm | that's default theme's fault :) |
| 17:15.12 | homovulgaris | :D indeed |
| 17:15.32 | homovulgaris | i have never used ogre :) |
| 17:16.54 | mafm | actually it's not OGRE, it's a bit more complex than that |
| 17:17.16 | mafm | OGRE folks wanted to specialyze in the 3D part only, and being so portable as to be usable in consoles |
| 17:17.28 | mafm | then they "externalized" support for other things like input and GUI |
| 17:17.37 | homovulgaris | oh :) |
| 17:17.46 | mafm | so actually you have to use several different libraries to get to this point |
| 17:18.24 | mafm | in this case it's the new RBGui, instead of the most widely known CEGUI... but it feels much better when using it, I think |
| 17:19.02 | mafm | I guess that I should try to focus soon in getting the rest of the infrastructure usable so everybody can test :) |
| 17:19.26 | homovulgaris | i tried rt^3 trunk |
| 17:20.02 | homovulgaris | but didnt seem to do anything :) |
| 17:20.17 | homovulgaris | g3d presently doesnt install ? |
| 17:20.32 | mafm | nope |
| 17:20.45 | mafm | otherwise I think that rt^3 is mostly empty |
| 17:20.53 | homovulgaris | any idea how magic.h entries are created :) |
| 17:21.28 | mafm | the problem is that, for compiling g3d, you need all these different libraries working, and some of them need patches |
| 17:21.43 | homovulgaris | yeah it produces a few binaries rt^3 rt^3d etc. which just prints timer started stopped etc. |
| 17:22.04 | mafm | no idea, let me check... |
| 17:24.36 | mafm | magic.c is in SVN, so I guess that it's not autogenerated -- so I guess that the entries are added by hand, if that's what you mean |
| 17:26.01 | homovulgaris | yeah magic.c just has an identify_magic which returns the name corresponding to magic value |
| 17:26.49 | homovulgaris | but there must be some system of generating the hexadecimal magic number right :) right now i just wrote a number i liked .. but not at all a good way i suppose :) |
| 17:28.47 | mafm | oh I see |
| 17:29.05 | mafm | well, you might want to wait for the gurus to come by ;) |
| 17:29.16 | homovulgaris | :) |
| 17:29.28 | homovulgaris | u are from portugal ? |
| 17:29.45 | mafm | usually nobody is around while I'm coding, except for a couple of hours or so |
| 17:29.48 | mafm | yep |
| 17:30.31 | mafm | you must be from America or Canada I guess :P |
| 17:33.02 | homovulgaris | India :P |
| 17:33.49 | mafm | hmm |
| 17:33.57 | mafm | so what time of the day is it, at the moment? |
| 17:34.27 | homovulgaris | 11:11 pm |
| 17:34.42 | homovulgaris | what time at portugal 5:30 6 ? |
| 17:35.11 | mafm | oh, at night.. I though that you were joining because you just got up in the morning or something :D |
| 17:35.19 | mafm | here it's 18h35 |
| 17:35.28 | homovulgaris | i just got up though :) |
| 17:36.06 | mafm | :D |
| 17:53.45 | homovulgaris | did u see pacman's hyperboloid ? awesome right :) ? |
| 17:56.05 | mafm | only in pictures that he posted here |
| 17:56.12 | mafm | but it's nice yep |
| 18:00.37 | mafm | what's yout development platform? |
| 18:05.25 | homovulgaris | u mean language ? C for integration with librt and other existing libraries.. constraint solving etc. in C++ |
| 18:05.54 | homovulgaris | saw this one ? http://brlcad.org/gallery/s/renderings/primitives/hyp.png.html |
| 18:06.15 | mafm | I mean also the OS |
| 18:06.40 | homovulgaris | debian unstable :) amd64 2.6.24-1 |
| 18:06.47 | mafm | heh, nice |
| 18:06.56 | mafm | oh, another Debianite? |
| 18:07.10 | homovulgaris | :) et tu ? |
| 18:07.13 | mafm | int random { return 4; } :P |
| 18:07.17 | mafm | me too |
| 18:07.26 | mafm | they make a lot of fun of me around here :PPP |
| 18:16.10 | mafm | (not in BRL-CAD, I mean in Portugal some of my friends -- it's not a very popular distro here) |
| 18:18.54 | homovulgaris | :) BRL-CAD is a very "secular" organization i am sure :) |
| 18:20.21 | louipc | mafm: whats popular? |
| 18:20.46 | homovulgaris | I cant imagine debian not being popular anywhere :) |
| 18:20.50 | pacman87 | homovulgaris: all my hyp images are here: https://webspace.utexas.edu/trv82/www/ |
| 18:21.15 | mafm | louipc: mostly fedora |
| 18:21.27 | homovulgaris | pacman87: makes me want to write a primitive myself :) |
| 18:21.36 | louipc | ah |
| 18:21.37 | mafm | and now Ubuntu of course |
| 18:21.56 | louipc | yeah I would have guessed ubuntu |
| 18:22.07 | pacman87 | homovulgaris: most of it was pretty easy, but i'm still fighting to get tess() to do what i want |
| 18:22.16 | homovulgaris | how is the sweep thing going to be implemented ? |
| 18:22.31 | homovulgaris | i mean how does hyp help revolve and sweep ? |
| 18:22.55 | pacman87 | i started it as my gsoc app patch requirement |
| 18:23.14 | pacman87 | and it helps me figure out what all a primitive needs |
| 18:23.20 | homovulgaris | method of creating new geometry primitives is pretty well strucutred as i understood from the comments in librt.. but yeah i think the raytracing part must be pretty involved |
| 18:24.07 | pacman87 | there's a fairly standard method for finding intersections of conic-type shapes |
| 18:25.02 | brlcad | poolio: why would you be calling rt_xxx_tess()? .. it doesn't do anything |
| 18:25.59 | pacman87 | brlcad: i think he wanted to call each primitive's tess(), not rt_xxx_tess() specifically |
| 18:26.01 | brlcad | if instead of xxx you mean some given primitive, then you can either call the function directly (just declare the function) or go through rt_functab |
| 18:27.06 | brlcad | homovulgaris: you're welcome to break up librt into doxygen groups -- I'm not sure who all uses it, that's a work in progress |
| 18:27.08 | homovulgaris | sweep would basically involve a profile right ? and for the profile u would be using sketch ? |
| 18:27.23 | homovulgaris | ok.. and magic number generation ? |
| 18:27.36 | pacman87 | homovulgaris: that's the current plan |
| 18:27.43 | ``Erik | hey, burley... should g2asc/asc2g per-primitive functionality be moved out of conv/ and into src/librt/whatver/g_*.c ? |
| 18:28.34 | poolio | brlcad: I meant rt_arbn_tess, declaring it did the trick :) |
| 18:28.36 | ``Erik | (if I have to use linux, debian is the one I like to go with) |
| 18:29.03 | starseeker | would have thought Gentoo for ``Erik |
| 18:29.11 | brlcad | as for boost, it really depends on which parts of boost -- the portions that are going to be part of c++0x are probably fair game, the others on case-by-case |
| 18:29.29 | ``Erik | uhm, I gave up on linux before gentoo came along, I believe |
| 18:29.50 | homovulgaris | :o gave up on linux :) |
| 18:30.22 | ``Erik | (also; gentoo is for ricers) |
| 18:30.38 | starseeker | Oddly enough, I don't performance tune my system much |
| 18:30.46 | pacman87 | uses slackware |
| 18:31.05 | brlcad | homovulgaris: the magic numbers are defined by the maker of the structure that needed them -- they're usually the ascii encoding of some related text string (see magic.h) |
| 18:31.15 | mafm | slackware r:-) |
| 18:31.35 | ``Erik | in the old days, I tuned some, I don't anymore... |
| 18:31.41 | ``Erik | heh |
| 18:31.51 | ``Erik | I still have floppies with slackware disk set files on them |
| 18:32.01 | homovulgaris | basically i wanted to define PC_CONSTRAINT_MAGIC_INTERNAL so i just selected a hex number and added it to magic.h and magic.c .. should it be ok / |
| 18:32.05 | pacman87 | ``Erik: 5.25" or 3.5"? |
| 18:32.13 | ``Erik | 3.5's |
| 18:32.18 | pacman87 | awww :( |
| 18:32.21 | ``Erik | I moved from 5.25's to 3.5's with my commodore |
| 18:32.28 | brlcad | mafm: hadn't had a chance to say it, that is great progress on the gui -- want to walk through a build (maybe tomorrow) to get things up and running myself |
| 18:32.36 | ``Erik | I used to have 4 track data tapes O.o |
| 18:32.53 | pacman87 | i've got carmen sandiego on a 5.25" |
| 18:33.00 | ``Erik | heh, I used to |
| 18:33.27 | ``Erik | I had BUCK ROGERS on cassette tape :D and used to buy the 120 minute ones because you could cram more bytes on 'em and they usually worked :D |
| 18:33.36 | brlcad | bets he still has carmen discs somewhere |
| 18:34.10 | brlcad | pacman87: cool! https://webspace.utexas.edu/trv82/www/hyp_rt16.png .. so is that recent that you got it aligned? |
| 18:34.37 | pacman87 | brlcad: yeah, it should work now |
| 18:34.41 | brlcad | awesome |
| 18:35.28 | pacman87 | i was off before because the rec uses y and -x as maj/min axis, and i use x and y |
| 18:35.29 | mafm | brlcad: nice. maybe I should sort out the building process so everybody could do it... it's a bit complicate though, and I already contacted RBGui guys to fix some of the issues instead of rolling private patches |
| 18:36.06 | ``Erik | brlcad: two cylinders and a hyp in the middle? |
| 18:36.28 | mafm | other than that, I have still the problem of not knowing exactly what to build in the console, etc; and what about the logging -- if I have to use the one in libbu or what |
| 18:36.33 | ``Erik | or pacman, rather? heh :D |
| 18:36.34 | brlcad | ``Erik: yup, anywhere there is per-primitive tables .. that belongs in the new dirs |
| 18:36.57 | brlcad | (asc2g, g2asc, in, mirror, soledit, *thinks* .. ) |
| 18:37.07 | pacman87 | ``Erik: yes, the hyp creation specifies the base dimensions, and neck ratio, so it's easier to match up to cyls |
| 18:37.29 | brlcad | mafm: yeah, especially for bob :) |
| 18:37.43 | ``Erik | g2asc.c has per primitive operations in it... for everything... asc2g.c, too :( |
| 18:41.05 | mafm | if you want something quickly though, I could pack the libraries (already with patches), the media (still not sure about the license) and the simple Makefile+ogre.cfgs |
| 18:43.13 | ``Erik | hehehe --fenable-racing-stripes |
| 18:43.26 | ``Erik | --fuck-upstream |
| 18:43.32 | ``Erik | I love reading about gentoo, it's funny stuff :D |
| 18:46.29 | ``Erik | ahhh, it's moved |
| 18:46.31 | ``Erik | http://www.funroll-loops.info/ |
| 18:47.57 | CIA-22 | BRL-CAD: 03mafm * r31370 10/rt^3/trunk/src/g3d/ (GuiConsole.cxx GuiConsole.h): Adding pointer to listener to be able to delete it, and thus not leak memory (even if it should be only disabled when closing the application...) |
| 18:47.59 | mafm | mmm |
| 18:48.26 | mafm | sf.net won't accept my new ssh key anymore, and asks me for password all the time -- any idea of the reason? |
| 18:48.35 | mafm | maybe another special brl-cad hook? |
| 18:59.51 | CIA-22 | BRL-CAD: 03mafm * r31371 10/rt^3/trunk/src/g3d/ (Application.cxx Application.h): Some cleanups and small code reorganizations, and adding Doxygen documentation in some parts. |
| 19:00.33 | CIA-22 | BRL-CAD: 03erikgreenwald * r31372 10/brlcad/trunk/src/adrt/libtie/tie_kdtree.c: If bu_realloc() could possibly have a 0 size, test and call bu_free() instead. PR 1989853. |
| 19:09.44 | mafm | brlcad: so about my question before about logs and the console functionalities? |
| 19:12.02 | brlcad | mafm: yeah, that's actually our regular modus operandi to include deps (with our srcs) |
| 19:12.34 | brlcad | I'd say commit their sources, them make commit(s) for any/all patches needed |
| 19:12.46 | brlcad | that way we can reapply the patches if/when we need to update the sources |
| 19:13.03 | CIA-22 | BRL-CAD: 03mafm * r31373 10/rt^3/trunk/src/g3d/ (Application.cxx Application.h): Restoring missing character when closing comment, it was causing havok when compiling :) |
| 19:13.36 | mafm | do you mean update their sources to our repository? |
| 19:13.37 | brlcad | mafm: and no regarding ssh keys -- that's entirely outside of the commit/hook infrastructure |
| 19:13.58 | brlcad | you should be able to log into an interactive node and manually verify your ~/.ssh files |
| 19:14.07 | mafm | all RBGui (which needs Mocha as util library), OGRE and OIS? |
| 19:14.26 | *** join/#brlcad WARLOCK (n=Paychris@wsip-24-120-246-34.lv.lv.cox.net) |
| 19:14.27 | brlcad | if you *just* uploaded new keys, I'd give it a day -- sometimes takes a while to sync them to all the svn hosts |
| 19:14.49 | brlcad | mafm: yep, the whole shebang -- how big is their stuff? |
| 19:14.51 | mafm | I can successfully submit to other repositories (gna), but not this |
| 19:14.51 | WARLOCK | anyone know a good place to discuss autocad & get help |
| 19:15.12 | mafm | maybe it's because of pulling SVN from https instead of svn+ssh? |
| 19:15.13 | brlcad | WARLOCK: sure, try an autocad support forum :) |
| 19:15.27 | brlcad | (which this is not) |
| 19:15.28 | WARLOCK | well, i was curious if there was an irc chat room |
| 19:15.31 | WARLOCK | for cad users |
| 19:15.36 | WARLOCK | it would be nice |
| 19:15.36 | brlcad | not that I'm aware of |
| 19:15.50 | WARLOCK | we should have one! |
| 19:15.53 | brlcad | there is #cad with like one dude |
| 19:16.15 | brlcad | "we" ? that wouldn't benefit us given we're developing a different cad system |
| 19:17.03 | ``Erik | O.o |
| 19:17.21 | WARLOCK | really |
| 19:17.27 | WARLOCK | let me navigate |
| 19:17.28 | WARLOCK | to shi site |
| 19:17.34 | WARLOCK | to this site* |
| 19:17.35 | ``Erik | joins #gentoo to talk about the latest vista service patch and the issues he's having |
| 19:17.40 | ``Erik | :D |
| 19:18.20 | *** part/#brlcad WARLOCK (n=Paychris@wsip-24-120-246-34.lv.lv.cox.net) |
| 19:18.43 | homovulgaris | brlcad: is our pkgconfig system working properly ? |
| 19:18.55 | mafm | brlcad: took a while to run: http://rafb.net/p/6UGAlD81.html |
| 19:19.10 | homovulgaris | i mean in the .pc files shouldnt the datarootdir declaration be one line above ? |
| 19:20.53 | mafm | I'd say that 300k sloc from OGRE is a bit... too much |
| 19:21.52 | mafm | as for patches and so on, for me OGRE and OIS stand pretty much by themselves and are quite easy to install (regular build systems, etc) |
| 19:22.15 | mafm | but RBGui and Mocha don't even compile out of the box, even if the patches required are pretty small |
| 19:27.47 | mafm | well, have to run now |
| 19:28.28 | mafm | hopefully I'll come back tomorrow -- tonight is the day of patron of the city, big party \o/ |
| 19:28.36 | mafm | take care folks :) |
| 20:01.51 | homovulgaris | hey erik, shouldnt line 918 of tie_kdtree.c be ((tie_geom_t *)(tie->kdtree->data))->tri_list rather than ((tie->kdtree->data))->tri_list |
| 20:02.00 | homovulgaris | i was getting a compile error |
| 20:04.12 | ``Erik | hrm, yeah, probably |
| 20:05.25 | ``Erik | thanks for catching that |
| 20:05.35 | CIA-22 | BRL-CAD: 03erikgreenwald * r31374 10/brlcad/trunk/src/adrt/libtie/tie_kdtree.c: pheer my pointy hat (fix a void* deref). |
| 20:06.47 | homovulgaris | :) i have honestly no idea what libtie does :) |
| 20:07.06 | ``Erik | uhm, really really fast triangle raytracing, used by adrt |
| 20:07.08 | homovulgaris | :should check out more parts of the source.. 1.5 mil lines is too much :) |
| 20:07.19 | ``Erik | some day, I'll rip out bot stuff in librt and shove tie in there |
| 20:07.24 | ``Erik | 'triangle intersection engine' |
| 20:07.37 | pacman87 | not 'twin ion engine'? |
| 20:07.44 | homovulgaris | oh ok :) bag of triangles never made much sense to me either :P |
| 20:08.20 | ``Erik | a long time ago, in a galaxy far away, the notion of modeling or rendering something just using triangles seemed stupid :) |
| 20:08.44 | ``Erik | I mean, how are you going to get a reasonable model to fit in the massive 4kb of memory your big high powered supercomputer has? |
| 20:09.02 | ``Erik | :D |
| 20:09.05 | homovulgaris | :D |
| 20:10.12 | ``Erik | wonders if his patch actually works |
| 20:15.52 | *** join/#brlcad clock_ (n=clock@77-56-78-196.dclient.hispeed.ch) |
| 20:17.04 | ``Erik | brilliant :D http://bugs.gentoo.org/show_bug.cgi?id=35890 |
| 20:42.12 | homovulgaris | :D |
| 20:42.44 | homovulgaris | no wonder u like #gentoo |
| 20:43.13 | louipc | lol |
| 20:48.27 | homovulgaris | :) btw why do we still attribute copyright to US army ? |
| 20:50.26 | andrecastelo | good evening guys |
| 20:51.36 | ``Erik | <-- personally thinks that donated work should remain under content creators content, the license protects the US gov't, as well as all other consumers O.o but *shrug* not my call |
| 20:51.46 | ``Erik | howdy, andré |
| 20:52.24 | andrecastelo | howdy ``Erik |
| 20:52.43 | andrecastelo | how's the weather ? :D |
| 20:52.55 | ``Erik | day-star too bright *grunt* |
| 20:59.09 | homovulgaris | :( no mged -- ogl_open: couldn't create glXContext. |
| 20:59.29 | CIA-22 | BRL-CAD: 03homovulgaris * r31375 10/brlcad/trunk/ (60 files in 43 dirs): |
| 20:59.29 | CIA-22 | BRL-CAD: Added basic structure of new Parametrics and constraints library in src/libbpc, |
| 20:59.30 | CIA-22 | BRL-CAD: pc_constraint.c file in librt for constraint object import export functionality, |
| 20:59.30 | CIA-22 | BRL-CAD: modification of primitives and table.c to provide rt_primitive_params |
| 20:59.30 | CIA-22 | BRL-CAD: functionality via functab structure, minor change to doxygen_structure |
| 21:18.02 | ``Erik | sobs at the every increasing instance count of "#ifdef __linux__ /* EXTRA special */" in this file |
| 21:18.13 | ``Erik | s,y,, |
| 21:28.51 | pacman87 | ``Erik: which file? |
| 21:33.20 | ``Erik | different project |
| 21:33.43 | pacman87 | briefly contemplates rewriting hyp's tess() to take advantage of the fact that a hyberboloid is a doubly-ruled surface |
| 21:34.07 | ``Erik | compiler chunk that outputs at&t style x86 assembly, there's the linux way and the rest of the known universe way :) |
| 21:35.48 | ``Erik | oh poo, mal isn't here to raz on :D |
| 21:48.14 | pacman87 | success! |
| 21:49.42 | pacman87 | https://webspace.utexas.edu/trv82/www/hyp_rt_tess.png |
| 21:50.26 | pacman87 | still not all that efficient, but at least it finally works |
| 21:57.47 | *** join/#brlcad docelic (n=docelic@77.237.111.4) |
| 22:06.49 | homovulgaris | :) |
| 22:20.43 | andrecastelo | so ``Erik, what do you think about the implementation idea at http://andrecastelo.wordpress.com/ ? |
| 22:58.25 | CIA-22 | BRL-CAD: 03pacman87 * r31376 10/brlcad/trunk/src/librt/primitives/ (hyp/hyp.c rhc/rhc.c): tess() should finally work, but still needs effeciency optimizations |
| 23:01.10 | ``Erik | hrmmm, you are intending on doing a full sample set for each pixel, not building a light map, correct? |
| 23:05.48 | andrecastelo | yup, but i'm open to suggestions |
| 23:07.39 | ``Erik | no, that works for me |
| 23:08.04 | ``Erik | it seems to me that you should shoot a primary ray and begin your pass at the first hitpoint? |
| 23:08.16 | andrecastelo | yes |
| 23:08.24 | ``Erik | opposed to shooting a random direction from the viewpoint as your list indicates? |
| 23:08.50 | andrecastelo | that's what i meant |
| 23:09.00 | ``Erik | okie |
| 23:09.16 | ``Erik | storing the path in mlt_app might be trickier than you might think |
| 23:09.25 | andrecastelo | why ? |
| 23:09.33 | ``Erik | we're going to have N workers going |
| 23:09.48 | ``Erik | you probably have 1 by default, I have 8 by default and will be testing at 128+ |
| 23:10.07 | andrecastelo | i see, if i restrict it to one point list, it will slow down considerably |
| 23:10.20 | ``Erik | in fact, we're getting a new sun t2 machine with 128 ALU's, I'm excited to see this piece of hw :) |
| 23:10.28 | ``Erik | and I have machines with, uh, 4096 cpu's at my disposal |
| 23:10.41 | andrecastelo | O.O |
| 23:10.51 | andrecastelo | holy.. |
| 23:10.53 | ``Erik | well, if you have one point list, the usual occurance is for data to be over-written or confused |
| 23:11.19 | pacman87 | is jealous |
| 23:12.05 | ``Erik | I probably won't use the big one, it requires queuing a batch job and people doing "important" things need time on it |
| 23:12.24 | ``Erik | but I will pretty much be operating solely in a fairly threaded and parallel mode |
| 23:12.27 | CIA-22 | BRL-CAD: 03andrecastelo * r31377 10/brlcad/trunk/src/rt/viewmlt.c: Added memory freeing block for mlt_app->path. (Subject to change to mlt_app->path_list) |
| 23:12.37 | andrecastelo | i see |
| 23:13.05 | andrecastelo | also, should i give more priority to view_pixel() over rayhit() ? |
| 23:13.09 | ``Erik | most of the rt's operating by having each worker responsible for an entire scanline |
| 23:13.21 | ``Erik | priority? how do you mean? |
| 23:13.48 | ``Erik | they do different things, last I looked O.o |
| 23:14.29 | andrecastelo | i know, i meant if i should work first on view_pixel() |
| 23:14.40 | ``Erik | um |
| 23:14.56 | ``Erik | well, if you do view_pixel first, you can stub like a flat shader to see if you're casting right |
| 23:15.08 | andrecastelo | yup, that's what i thought |
| 23:15.10 | ``Erik | I think that would give you a more visual result sooner |
| 23:15.33 | ``Erik | which is always good, not only can you course correct sooner, but it's awesome seeing it and gets ya all enthusiastic :D |
| 23:15.55 | andrecastelo | yep, i've been thinking that should i do view_pixel() later, i could need to rewrite some stuff |
| 23:16.16 | andrecastelo | yup, i'm kind of jealous of pacman87 and his nifty images |
| 23:16.23 | ``Erik | *nod* |
| 23:16.35 | pacman87 | :D |
| 23:16.50 | andrecastelo | :D |
| 23:17.15 | ``Erik | I sure got a huge thrill out of my metaball image |
| 23:17.19 | ``Erik | http://brlcad.org/gallery/s/renderings/primitives/niceballs.png.html |
| 23:17.33 | ``Erik | ask brlcad, I was running around the office yelling "check out my balls! look at my balls!" |
| 23:18.21 | andrecastelo | rofl! |
| 23:19.23 | andrecastelo | rtmlt's view_pixel() would be something like rt's view_pixel(), right? |
| 23:19.34 | ``Erik | I think so |
| 23:19.54 | ``Erik | generate the primary cast, "go do it", pack the result |
| 23:21.33 | andrecastelo | primary cast is the primary rays? |
| 23:21.38 | ``Erik | yes |
| 23:22.04 | ``Erik | ray caster vs ray tracer... :) |
| 23:23.47 | andrecastelo | ah ok ok :D |
| 23:23.47 | ``Erik | hrmmmmmm, gpl or bsd *think* |
| 23:24.40 | ``Erik | leans towards the 'new' bsd license (two clause, not 3) |
| 23:24.53 | ``Erik | side project, unrelated to BRL-CAD or graphics :D |
| 23:25.04 | andrecastelo | was just going to ask that :D |
| 23:25.20 | andrecastelo | what's so bad about gpl? (or what's so good about bsd) |
| 23:27.16 | ``Erik | bsd allows use in closed source projects (with attribution) |
| 23:28.08 | ``Erik | I've released some stuff under the bsd license that is use in, uh, shoutcast and stuff |
| 23:28.08 | ``Erik | so my name gets in their readme or whatever :) |
| 23:28.35 | andrecastelo | is impressed how open source ``Erik is |
| 23:28.43 | andrecastelo | hehehe |
| 23:29.06 | ``Erik | *shrug8 |
| 23:29.18 | ``Erik | in the late 80's, I was releasing code on BBS's under pretty much public domain |
| 23:29.59 | andrecastelo | must be awesome to have someone use your code, isn't it ? |
| 23:30.07 | ``Erik | yeah |
| 23:30.12 | CIA-22 | BRL-CAD: 03homovulgaris * r31378 10/brlcad/trunk/ (27 files in 15 dirs): changes in comments to conform to doxygen, update of misc/Doxyfile |
| 23:30.16 | ``Erik | http://www.brianlee.org/digital-picture-frame/ <-- especially neat stuff like that |
| 23:30.59 | andrecastelo | opens site |
| 23:34.03 | ``Erik | neat, my name is on http://www.centova.com/pages/icescc too |
| 23:35.23 | andrecastelo | awesome, ``Erik.. that digital picture frame looks really nice |
| 23:36.01 | ``Erik | I thought so, I got an email from brian with a couple questions and we got to talking, he did some really awesome stuff, I'm glad he chose my software :D |
| 23:36.52 | andrecastelo | neat, neat, i can imagine the feeling :D |
| 23:37.41 | andrecastelo | ices-cc is the shoutcast thing you talked about? or did you work on more related stuff ? |
| 23:38.02 | ``Erik | uhm, it's a fork of the thing my stuff was used for, I think |
| 23:38.18 | ``Erik | I didn't work on it, I just had some open source software out there under a bsd license, and people used it |
| 23:38.36 | ``Erik | sometimes I get lucky and they tell me what they're using my stuff for :) |
| 23:40.14 | ``Erik | xiph icecast is the one I was thinking |
| 23:41.30 | andrecastelo | wants to write good code too :b |
| 23:41.47 | ``Erik | oh, whoa, wait up there, I never claimed I wrote GOOD code ;) |
| 23:43.14 | andrecastelo | hehehe, true, but i assume that if people want to use it, it must be somewhat good |
| 23:43.38 | pacman87 | andrecastelo: that just means it's useful ;) |
| 23:44.05 | andrecastelo | still, i think that both concepts overlap :) |
| 23:44.41 | ``Erik | hum, some neat renderings from gsi O.o |
| 23:52.35 | CIA-22 | BRL-CAD: 03andrecastelo * r31379 10/brlcad/trunk/src/rt/viewmlt.c: Fixed some issues in view_end(): structure freed changed from mlt_app to point_list; moved declarations to the beginning of the block. |
| 23:55.45 | andrecastelo | hey ``Erik : can you give me some tips on understanding view_pixel() ?? what is the core of the function?? (i'm thinking it must be one of the switch cases) |
