| 01:25.22 | *** join/#brlcad abhi2011 (~chatzilla@117.200.81.140) |
| 01:40.48 | brlcad | apparently cmake has some parallel build race condition bug too... ended up compiling one file to a 0-length .o file, linking it in and naturally then resulting in undefined symbols |
| 01:50.03 | CIA-48 | BRL-CAD: 03brlcad * r47361 10/brlcad/trunk/src/conv/g-vrml.c: remove dead code, #if 0 |
| 01:50.59 | CIA-48 | BRL-CAD: 03brlcad * r47362 10/brlcad/trunk/src/conv/g-x3d.c: restructure bu exception handling into try/catch form, make 'reg' also be static in order to avoid gcc longjmp clobbering |
| 02:01.47 | CIA-48 | BRL-CAD: 03brlcad * r47363 10/brlcad/trunk/src/conv/intaval/regtab.h: quell warning about being unable to inline this constructor. move the init of desc over with the others. |
| 02:02.59 | CIA-48 | BRL-CAD: 03brlcad * r47364 10/brlcad/trunk/src/conv/ (36 files in 10 dirs): ws |
| 03:15.10 | starseeker | O.o |
| 03:15.17 | starseeker | has never encountered that... |
| 04:04.35 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) |
| 04:07.59 | *** join/#brlcad packrat (~packrator@c-98-209-146-133.hsd1.mi.comcast.net) |
| 04:54.53 | *** join/#brlcad abhi2011 (~chatzilla@117.200.84.208) |
| 06:16.33 | abhi2011 | hmm , the contact point approach works with the simple case of sphere-sphere collision |
| 06:16.42 | abhi2011 | forces are getting applied now |
| 06:16.55 | abhi2011 | lets see whats the issue with box shapes then |
| 07:08.17 | CIA-48 | BRL-CAD: 03Abhi2011 07http://brlcad.org * r3190 10/wiki/User:Abhijit: /* Log */ |
| 08:07.39 | *** join/#brlcad forth (~10497E4F9@92.242.118.253) |
| 09:55.16 | *** join/#brlcad packrat (~packrator@c-98-209-146-133.hsd1.mi.comcast.net) |
| 11:37.11 | *** join/#brlcad CIA-109 (~CIA@cia.atheme.org) |
| 12:22.44 | *** join/#brlcad abhi2011 (~chatzilla@117.200.80.21) |
| 12:46.07 | *** join/#brlcad abhi2011 (~chatzilla@117.200.83.165) |
| 14:01.10 | CIA-109 | BRL-CAD: 03brlcad * r47365 10/brlcad/trunk/include/rt/defines.h: not yet used, but add the start of a common definitions file for librt just so this directory isn't empty |
| 14:02.38 | abhi2011 | ok now the box-box collision also works |
| 14:02.46 | abhi2011 | with contact pairs |
| 14:02.54 | abhi2011 | will try to roll a sphere on a plane |
| 14:03.33 | CIA-109 | BRL-CAD: 03brlcad * r47366 10/brlcad/trunk/include/ (CMakeLists.txt Makefile.am): add the new rt/defines.h header for install/dist |
| 14:07.48 | brlcad | abhi2011: so have you figured out what the problem was? |
| 14:08.10 | brlcad | too many contact points? |
| 14:08.18 | brlcad | wrong vector direction? |
| 14:08.21 | abhi2011 | yes, the problem seems to lie on the direction of the normals |
| 14:08.35 | brlcad | that'd 'splain it ;) |
| 14:08.37 | abhi2011 | which have to point from what ever object that bullet considers as object B |
| 14:08.46 | abhi2011 | to whichever points towards object A |
| 14:09.27 | abhi2011 | yes, I think I can try to roll a sphere now and see if I can generate contact points when its touching the ground |
| 14:09.47 | brlcad | awesome |
| 14:09.55 | abhi2011 | I have to use the curvature information of the sphere though, to ensure I dont generate more than 1 contact point |
| 14:10.16 | abhi2011 | as more than 1 point separated by a non zero distance would prevent its roll |
| 14:10.18 | abhi2011 | :) |
| 14:10.34 | abhi2011 | so some logic has to figured out there to work on all cases |
| 14:11.08 | brlcad | after that, a great test case for the ray casting would be to roll an ellipsoid down a bit bowl |
| 14:11.21 | abhi2011 | a bit bowl ? |
| 14:11.24 | abhi2011 | what does the bit mean |
| 14:11.44 | brlcad | a *big* bowl |
| 14:11.53 | abhi2011 | ah ok |
| 14:11.55 | abhi2011 | :P |
| 14:11.58 | brlcad | yeah :) |
| 14:12.03 | abhi2011 | yes |
| 14:12.11 | brlcad | a big concave dish |
| 14:12.17 | abhi2011 | yes, the thing that I have to think about though is |
| 14:12.33 | abhi2011 | that i can only generate contact points when objects interpenetrate |
| 14:13.05 | abhi2011 | which can cause the motion to be bumpy as an ellipsoid rolls along a bowl surface for example |
| 14:13.34 | brlcad | sure, maybe |
| 14:13.52 | brlcad | but if that can be reduced with more sampling points, that might just work fine |
| 14:14.08 | abhi2011 | maybe i can reduce the bumpiness to un-noticeable extents |
| 14:14.13 | brlcad | right |
| 14:14.16 | abhi2011 | ok |
| 14:14.58 | brlcad | by the way, there are lots of folks really excited about your work :) |
| 14:15.18 | brlcad | your quick youtube demo made the rounds amongst many of the expert modelers |
| 14:15.40 | brlcad | they loved it! .. of course wanting to try it right away |
| 14:15.48 | brlcad | told them they'll have to wait a bit still :) |
| 14:17.26 | abhi2011 | hehe, yes I am on it for sure, should have something by tomorrow hopefully :) |
| 14:17.54 | brlcad | well, I told them at least a month, so no worries :) |
| 14:18.03 | abhi2011 | and dont worry I ll keep working on it inspite of socis completing |
| 14:18.10 | abhi2011 | as i had promised :) |
| 14:18.27 | brlcad | that's great news ;) |
| 14:18.33 | brlcad | this is really going to be big |
| 14:19.42 | abhi2011 | cool, I am actually plannning to add this to Bullet also , so that it supports irregular objects, since the raytracing is done within the small overlap region and can be parallelized, it may be near real time |
| 14:31.08 | brlcad | that'd probably make a huge impact in their flexibility |
| 14:31.46 | brlcad | even for polygonal shapes, our ray engine can shoot a mesh prety darn fast |
| 14:32.33 | brlcad | starseeker: CPack Error: Cannot create symlink: .cmake/share/brlcad/7.20.3/html/presentations/en/images/tk-based-gui-for-mged.png--> /home/sean/brlcad/doc/docbook/presentations/en/images/tk-based-gui-for-mged.png |
| 14:56.50 | *** join/#brlcad abhi2011 (~chatzilla@117.200.83.165) |
| 15:25.44 | *** join/#brlcad abhi2011 (~chatzilla@117.200.87.86) |
| 15:27.14 | CIA-109 | BRL-CAD: 03brlcad * r47367 10/brlcad/trunk/NEWS: getting much closer, should be today if this last pass succeeds |
| 15:31.45 | CIA-109 | BRL-CAD: 03brlcad * r47368 10/brlcad/branches/STABLE/ (57 files in 26 dirs): again merging trunk to STABLE from r47343 to HEAD r47367, in prep for retagging release 7.20.4 |
| 15:51.00 | CIA-109 | BRL-CAD: 03brlcad * r47369 10/brlcad/trunk/doc/docbook/Makefile.am: update CatalogManager.properties file for autotools dist, in builddir, not srcdir |
| 16:02.20 | CIA-109 | BRL-CAD: 03brlcad * r47370 10/brlcad/branches/STABLE/doc/docbook/Makefile.am: merging trunk to STABLE from r47367 to HEAD r47369 |
| 18:00.51 | *** join/#brlcad abhi2011 (~chatzilla@117.200.80.50) |
| 19:29.54 | brlcad | finishes implementing support for our whitespace indentation standard into astyle |
| 19:30.23 | brlcad | crappy indentation of continuations, though .. wonder if there's a way to fix that |
| 19:47.06 | brlcad | finishes what is hopefully the last distcheck for retag |
| 21:17.11 | starseeker | erm... is that CPack error consistent? |
| 21:17.53 | starseeker | astyle... cool! |
| 21:18.27 | starseeker | hopes he will accept the patch |
| 21:20.30 | starseeker | wonders why he does not see these CMake errors... |
| 21:45.12 | cvds_ | !seen rkoeppl_printing |
| 21:45.23 | cvds_ | sorry wrong windos |
| 23:54.12 | *** join/#brlcad bhinesley (~bhinesley@adsl-99-52-241-103.dsl.bkfd14.sbcglobal.net) |