| 00:14.50 | CIA-21 | BRL-CAD: 03brlcad * r31011 10/brlcad/trunk/ (10 files in 8 dirs): |
| 00:14.50 | CIA-21 | BRL-CAD: this renames the wdb_binary / dbbinary mged command to just 'binary' .. this |
| 00:14.50 | CIA-21 | BRL-CAD: seems to be what it is actually used in archer already, what makes more sense, |
| 00:14.50 | CIA-21 | BRL-CAD: and is more consistent with other commands. happened to fix a mistake in the |
| 00:14.50 | CIA-21 | BRL-CAD: html docs in the process. also updated the command's help so that it's less |
| 00:14.53 | CIA-21 | BRL-CAD: brain dead wrong in response to tom browder's sf bug 1532699 (dbbinary -o help |
| 00:14.55 | CIA-21 | BRL-CAD: is misleading) |
| 01:09.06 | *** join/#brlcad Twingy (n=justin@74.92.144.217) |
| 01:21.05 | yukonbob | waves in.. |
| 01:36.55 | brlcad | howdy yukonbob :) |
| 01:37.02 | brlcad | clears out some bug reports |
| 01:42.07 | starseeker_ | finally gets the docbook toolchain to do his bidding... |
| 01:42.15 | louipc | hooray x2 |
| 01:44.28 | CIA-21 | BRL-CAD: 03brlcad * r31012 10/brlcad/trunk/configure.ac: try to prevent -Lyes/lib64 -Iyes/include etc from happening if the user only provides --with-jdk or --with-java without a path. |
| 01:54.42 | brlcad | woot, only three left before the bugs are back down to just one page! |
| 02:06.09 | CIA-21 | BRL-CAD: 03brlcad * r31013 10/brlcad/trunk/regress/solids.sh: don't need to refer to $1/regress, just use the files |
| 02:06.30 | brlcad | two.. |
| 02:08.45 | brlcad | one more! |
| 02:12.00 | brlcad | woot! |
| 02:28.35 | starseeker_ | wow. |
| 02:28.52 | starseeker_ | checks in disbelief... |
| 02:29.45 | starseeker_ | #define state_of_jaw "dropped" |
| 02:33.27 | starseeker_ | nice job! |
| 02:35.47 | brlcad | has since cleared about a half dozen others .. down to only 42 open bugs woo hoo :-) |
| 02:41.45 | yukonbob | updates svn co, gets ready to do some (unrelated to BRL-CAD) Tcl work |
| 02:52.02 | brlcad | nice to see the bug list actually decrease for once .. didn't realize how many issues got fixed with the latest 7.12 |
| 02:52.23 | brlcad | down to about 40 now, few more that might be fixed but need investigating |
| 02:53.09 | yukonbob | is the dsp still on the list? |
| 02:53.27 | brlcad | thinks he needs to host a BRL-CAD Bug Zapper |
| 02:53.44 | brlcad | yeah :( |
| 02:53.47 | brlcad | well sorta.. |
| 02:54.02 | brlcad | it was never reported to the tracker, just to the BUGS file |
| 02:54.10 | brlcad | these are tracker items I was clearing out |
| 02:54.46 | yukonbob | needs to work-out "issues" w/ Tcl on his system (Tcl wrt NetBSD in general, /me thinks), including getting 8.5 up/running |
| 02:55.02 | yukonbob | ah -- |
| 02:55.57 | yukonbob | it's nice to see bugs from 1997 getting squashed... ;) |
| 03:24.13 | louipc | hmm wow sf.net wants me to run 27 jscript files |
| 03:32.21 | quentusrex | Hello |
| 03:32.25 | quentusrex | Anyone around? |
| 03:32.39 | yukonbob | nobyd -- shh.. |
| 03:32.43 | quentusrex | :) |
| 03:32.44 | yukonbob | *nobody |
| 03:33.28 | quentusrex | Do you know if there has been any progress into taking multiple pictures of a 3D object and translating that into CAD? |
| 03:33.47 | yukonbob | ?by magic |
| 03:34.35 | quentusrex | I'm talking about 3D object detection, and translation into some form of Computer modeling. |
| 03:35.20 | yukonbob | has no idea what project/task quentusrex is talking about, of if this a general "start of the art" question for the whole field of 3d visualization |
| 03:35.30 | yukonbob | *state of the art |
| 03:35.41 | yukonbob | damn latency, again (still) |
| 03:36.44 | quentusrex | Well, I have experience in video and image transcoding(which basically is just dealing with many individual frames) and I had a through. With object detection in images, what about having a video? with many frames? |
| 03:37.12 | quentusrex | If there are several images with the objects 'detected' couldn't those be modeled? |
| 03:37.30 | quentusrex | by taking many 2D 'slices' of the profile of an object? |
| 03:39.01 | quentusrex | wouldn't you think so yukonbob? |
| 03:42.20 | yukonbob | I'll qualify my answer with "I'm not an expert" |
| 03:43.54 | yukonbob | I'd be skeptical -- you'd need "perfect" lighting, perfectly configured, and work w/ objects that conform to formats you "know", I'd think |
| 03:44.25 | yukonbob | after all, video isn't 3d, no matter if you shoot images of the real world |
| 03:44.53 | louipc | har har I opened another bug report. |
| 03:46.13 | yukonbob | so it's an interpolation job... also depednds on what you expect to get out of it -- BRL-CAD specfically uses CSG, which wouldn't work well with your idea, since a _lot_ of what a CSG odel can be is not necessarily visible -- BREP, otoh, might work to some degree. |
| 03:46.16 | louipc | quentusrex: yeah I saw a program that would help do that but it didn't take slices |
| 03:46.57 | louipc | it took points on the image and somehow analysed it |
| 03:47.38 | yukonbob | guesses you'd also need to know objects material/finish, to know how light interracts w/ it... |
| 03:48.27 | yukonbob | because light is the only metric you've got when dealing w/ an image... |
| 03:50.46 | yukonbob | is inclined to find the "issues" w/ questions like this -- I'm sure it'd be a fun project, and results could be impressive. |
| 05:20.48 | pacman87 | http://www.cs.unc.edu/Research/urbanscape/ |
| 05:21.17 | pacman87 | a bit late to the discussion, but this kind of looks like what you're describing |
| 05:21.37 | brlcad | tunes in |
| 05:21.51 | yukonbob | waves |
| 05:22.30 | yukonbob | would is interested if pacman87 or brlcad have comments about the preceeding... |
| 05:22.39 | yukonbob | looks at pacman87's link |
| 05:23.50 | brlcad | i've actually had a research project directly related to what he's suggested "on tap" for several years now |
| 05:23.51 | yukonbob | pacman87: kewl |
| 05:24.11 | brlcad | lots of thoughts and ideas on the subject, it's an entirely non-trivial problem, but intersting nonetheless |
| 05:24.46 | brlcad | there are a couple groups working on model reconstruction from video/photos, pretty much in its infancy right now |
| 05:25.17 | yukonbob | s/would is/is/ |
| 05:26.00 | yukonbob | brlcad: seems interesting, but "messy" to me -- esp. when in the headspace of the precision/correctness of dealing w/ BRL-CAD -type issues |
| 05:26.13 | yukonbob | feh |
| 05:26.16 | brlcad | http://ieeexplore.ieee.org/Xplore/login.jsp?url=/iel5/10468/33215/01565435.pdf?temp=x is another group (based off research from stanford iirc) |
| 05:26.50 | yukonbob | miilitary probably _loves_ this stuff, though, for quick/dirty recon |
| 05:27.35 | yukonbob | send in automaton which sends out signal in RT... review, attack! |
| 05:27.57 | brlcad | it's the ultimate in reverse modeling techniques .. just video tape the thing and *poof*, 3D model on computer |
| 05:28.11 | brlcad | goal-wise at least |
| 05:28.26 | brlcad | it'll happen eventually (at high-res), most of the pieces are there |
| 05:28.26 | pacman87 | re: need to know objects material/finish, to know how light interracts w/ it... |
| 05:28.37 | yukonbob | well... depending on goal... is so dirty, but that may be fine... |
| 05:28.41 | pacman87 | i'd think that's only a problem for very shiny stuff |
| 05:28.53 | brlcad | shiney stuff isn't really that hard |
| 05:28.54 | pacman87 | with reflections that change depending on your viewing angel |
| 05:28.55 | pacman87 | angle |
| 05:29.04 | yukonbob | pacman87: or transparent/translucent, /me imagines |
| 05:29.09 | brlcad | in a single image it is, even with just a couple frames, you can correlate really easily |
| 05:29.25 | brlcad | soft bodies are a bitch |
| 05:29.41 | yukonbob | "soft body"? |
| 05:29.55 | yukonbob | ie: a dog, vs. a building? |
| 05:30.01 | brlcad | yeah |
| 05:30.03 | brlcad | people |
| 05:30.06 | brlcad | plastic bags |
| 05:32.48 | brlcad | just about anything highly deformable (e.g. people), or with ill-defined or hard-to-detect boundaries (e.g., hair, fur), liquids, plastics under a load, something undergoing a deformation .. all not something that can be captured via video/images really well/easily/at all |
| 05:33.16 | brlcad | quentusrex: why do you ask? you interested in working on that? :) |
| 06:26.53 | brlcad | louipc: thanks for the report.. |
| 07:00.09 | *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch) |
| 07:02.10 | louipc | brlcad: sure thing |
| 07:56.01 | *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6) |
| 08:28.24 | quentusrex | brlcad, Well, I'm currently busy on a project that is going corporate. So for the next year or so I'll be really busy. But I'd love to work on something like this. |
| 08:28.47 | quentusrex | The next company I start I hope will deal more directly with AI and robitics issues. |
| 08:30.06 | quentusrex | Real world application for something like this would be(for the military) is flying a drone over a city and then being able to have a 3D map of the city. |
| 08:30.28 | quentusrex | and moving soft body objects like dogs and humans would just be considered 'noise' |
| 09:52.04 | *** join/#brlcad mafm (n=mafm@elnet-111.lip.pt) |
| 09:58.21 | mafm | allo |
| 10:26.19 | clock_ | What I find fascinating if you turn the view in BRL-CAD and do clockwise or counterclockwise circles with the mouse, it rotates around the axis of view |
| 10:26.47 | clock_ | the bigger circles, the larger rotation per completed circle |
| 10:27.07 | clock_ | This must be some property of the algebraic group of rotation, any explanation? |
| 10:54.37 | *** join/#brlcad thing0 (n=ric@123.208.23.133) |
| 11:15.17 | quentusrex | Also, for a project like this I wouldn't expect a perfect rendering of what ever the object is. I would think it would still need a person to 'fix the fine details' but it would give a significant head start on the modeling process. |
| 11:18.13 | clock_ | quentusrex: BRL-CAD produces an almost perfect rendering |
| 11:18.49 | quentusrex | sorry clock_, that comment is in relation to a much earlier conversation |
| 16:03.32 | clock_ | Oh my bugreports are being processed |
| 16:03.35 | clock_ | I appreciate that. |
| 16:12.30 | *** join/#brlcad Elperion (n=Bary@p548744AB.dip.t-dialin.net) |
| 16:56.40 | *** join/#brlcad d_rossberg (n=rossberg@bz.bzflag.bz) |
| 17:22.45 | brlcad | any czech speakers here? |
| 17:31.55 | *** join/#brlcad andre|away (n=chatzill@189.71.62.7) |
| 17:32.23 | andrecastelo | good morning :D |
| 18:19.21 | CIA-21 | BRL-CAD: 03starseeker * r31014 10/brlcad/trunk/src/nirt/ (command.c nirt.1): |
| 18:19.21 | CIA-21 | BRL-CAD: Change behavior of backout command in interactive nirt - backout flag can now be |
| 18:19.21 | CIA-21 | BRL-CAD: activated and deactivated, and repeated invocations of backout will not increase |
| 18:19.21 | CIA-21 | BRL-CAD: the backout distance with each invocation. Also, change of origin due to |
| 18:19.21 | CIA-21 | BRL-CAD: backout use in shoot is now temporary - previous value is restored after shot is |
| 18:19.24 | CIA-21 | BRL-CAD: complete. |
| 18:22.09 | mafm | brlcad: I think clock_ does -- if you can wait for him to log again |
| 18:23.21 | CIA-21 | BRL-CAD: 03starseeker * r31015 10/brlcad/trunk/NEWS: Add NEWS note for nirt change |
| 18:25.11 | brlcad | mafm: yeah, I know he does as do two other guys |
| 18:25.17 | brlcad | but I don't see theem, so figured I'd query ;) |
| 18:27.54 | mafm | it's a bit confusing, since he has swiss ISPs :) |
| 18:28.54 | mafm | 1) Whitespace -- Indents are 4 characters, tabs are 8 characters. |
| 18:29.09 | brlcad | yes? |
| 18:29.44 | mafm | how can you indent to 4 characters in emacs then? |
| 18:30.02 | brlcad | there are emacs local variable footers in all files |
| 18:30.16 | brlcad | it should be automatic |
| 18:30.49 | brlcad | if you create any new files with sh/template.sh, it'll also have the correct footer |
| 18:31.23 | brlcad | otherwise the answer more specifically is you set the indentation to 4 and then just hit tab, emacs will do the right thing |
| 18:31.23 | mafm | lemme check |
| 18:31.48 | brlcad | the footers set up the indentation style |
| 18:32.16 | brlcad | there's also a vi/vim line with similar decls |
| 18:32.55 | mafm | hmm, is the "stroustrup" thing? I thought that the "tab-width: 8" was the one working for indentation |
| 18:33.39 | brlcad | yeah, it used to be explicit |
| 18:33.52 | brlcad | but then the file style was set, which for stroustrup is 4 char indents |
| 18:34.58 | brlcad | e.g. at the bottom of http://brlcad.svn.sourceforge.net/viewvc/brlcad/brlcad/tags/rel-7-8-4/src/libbu/association.c?revision=27207&view=markup |
| 18:35.25 | brlcad | c-basic-offset for c-mode files |
| 18:35.36 | brlcad | but since it's redundant with a style set, it was removed |
| 18:37.21 | brlcad | tab-width basically just describes where the tab stops are at, used during indentation |
| 18:39.27 | mafm | I see |
| 18:46.21 | mafm | int the_answer() { return 42; }; |
| 18:46.33 | mafm | aren't one-line functions forbidden? :) |
| 18:56.55 | brlcad | heh |
| 18:57.17 | brlcad | omfg, clock's perpend actually tessellates |
| 18:57.51 | brlcad | at least with the latest sources, so either my bot fixes fixed the dangling face errors or just better roll of the dice |
| 18:58.14 | brlcad | just took it forever to tessellate .. 15 hours |
| 18:58.25 | mafm | try to buy lottery today, just in the case |
| 19:01.10 | brlcad | :) |
| 19:01.20 | brlcad | mega millions is up to 160M or something |
| 19:03.13 | *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6) |
| 19:07.49 | mafm | I tend to be the luckiest among my family and friends thugh, by not buying it |
| 19:07.58 | mafm | so at least I don't waste money :) |
| 19:16.37 | mafm | well, I go home now |
| 19:16.45 | mafm | or drinking a beer first, maybe :) |
| 19:17.00 | mafm | bye! |
| 19:25.15 | brlcad | png results here, http://brlcad.org/tmp/perpend/ |
| 19:30.31 | brlcad | wow, down to less than 35 bugs |
| 20:24.21 | yukonbob | waves in |
| 20:24.30 | brlcad | 6howdy yukonbob |
| 20:24.43 | yukonbob | 7howdy back, bugkilla |
| 20:45.41 | PrezKennedy | howdy y'all |
| 20:45.46 | starseeker | howdy :-) |
| 20:46.47 | *** join/#brlcad andrecastelo (n=chatzill@189.71.62.7) |
| 20:54.19 | andrecastelo | ``Erik: hey erik |
| 21:09.37 | brlcad | hola andrecastelo |
| 21:09.59 | andrecastelo | hey brlcad :D |
| 21:10.30 | brlcad | how's it going? |
| 21:10.38 | andrecastelo | i'm fine, you ? |
| 21:10.49 | brlcad | peachy :) |
| 21:10.59 | andrecastelo | i've started writing a mlt_def.h, with a few structure definitions |
| 21:11.01 | brlcad | happy that the bug reports were cut in half |
| 21:11.10 | brlcad | cool |
| 21:11.16 | andrecastelo | cool, heard about that |
| 21:11.22 | andrecastelo | but i kinda got a problem |
| 21:11.37 | brlcad | you should post it up as a patch ;) |
| 21:11.38 | andrecastelo | i'm going to use struct point_list, that i created in viewarea.c (in rtarea) |
| 21:11.50 | andrecastelo | where can i add it so that i may use it in mlt as well ? |
| 21:12.34 | brlcad | depends whether you're going the liboptical route or isolated tool |
| 21:13.06 | andrecastelo | i'm going isolated tool ;) |
| 21:13.27 | brlcad | okay, then just put it in your header for now then |
| 21:13.35 | andrecastelo | hm ok, then |
| 21:14.15 | brlcad | or keep it in its own header |
| 21:14.30 | brlcad | so it can be moved around to whereever makes sense |
| 21:14.50 | andrecastelo | ok.. |
| 21:15.24 | andrecastelo | i'll add some stuff and then upload as a patch so you can give me some feedback, can i do that ? |
| 21:15.46 | brlcad | otherwise, raytrace.h normally/presently holds most of the rendering structures |
| 21:15.56 | brlcad | yeah, that'd be great |
| 21:16.14 | andrecastelo | ok, just a minute, i need a reboot |
| 21:16.18 | brlcad | after one or two of those, should be able to get you committing directly |
| 21:16.27 | brlcad | or reboot :P |
| 21:17.10 | brlcad | notes there is an rt_pt_node, linked list of points |
| 21:17.54 | *** join/#brlcad Elperion (n=Bary@p548744AB.dip.t-dialin.net) |
| 21:30.37 | *** join/#brlcad andrecastelo (n=chatzill@189.71.62.7) |
| 21:30.50 | andrecastelo | k, back |
| 21:35.15 | andrecastelo | brlcad: i've updated my working copy but i have some conflicts in misc/win32-msvc8.. my msvc updated some files in there to be compatible to the latest msvc9.. what should i do ? can i simply ignore them ? |
| 21:45.27 | brlcad | andre|afk: if you have conflicts, you can't ignore them -- it puts markers in the files |
| 21:46.01 | brlcad | should probably blow the directory away, svn up again, then cp win32-msvc8 to win32-msvc9 to update |
| 21:46.10 | brlcad | or provide a diff of the updates it made for vc9 |
| 21:46.20 | brlcad | as a patch (presuming it works with 8) |
| 21:47.21 | andre|afk | brlcad: there's no problem in deleting it ? the source code i have is different from the one in the svn, because of the center points patch |
| 21:48.51 | brlcad | you'd have to readd those changes to the build files |
| 21:49.04 | brlcad | you can try to resolve the conflicts, merge the changes in |
| 21:49.17 | brlcad | but if you've not made more than a couple edits, probably easier to just make them again |
| 21:57.41 | andrecastelo | brlcad: is it easier to use a command line svn? i'm using tortoiseSVN, is that a problem ? |