| 15:20.35 | *** join/#brlcad ibot (~ibot@rikers.org) |
| 15:20.35 | *** topic/#brlcad is BRL-CAD Open Source Solid Modeling || http://brlcad.org || logs: http://ibot.rikers.org/%23brlcad/ || Congratulations to our 11 accepted GSoC students!! |
| 15:42.40 | *** join/#brlcad Stattrav (~Stattrav@ns.cmi.ac.in) |
| 15:42.41 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) |
| 16:01.10 | CIA-68 | BRL-CAD: 03brlcad * r52527 10/brlcad/trunk/src/libged/bot_fuse.c: quell may be used uninitialized |
| 16:09.21 | CIA-68 | BRL-CAD: 03carlmoore * r52528 10/brlcad/trunk/misc/CMake/Distcheck.cmake: fix 2 spellings |
| 16:10.49 | CIA-68 | BRL-CAD: 03carlmoore * r52529 10/brlcad/trunk/misc/CMake/distcheck_autotools.cmake.in: fix spelling in a quoted string |
| 16:20.34 | *** join/#brlcad [1]elf11_ (~elf11_@92.80.57.5) |
| 16:27.10 | brlcad | elf11_: here's what was written |
| 16:27.10 | brlcad | 22:30 < brlcad> elf11_: you should totally document everything abhijit has been telling you in the source code |
| 16:27.13 | brlcad | 22:31 < brlcad> like the broadphase/narrowphase information, bounding boxes, how things are organized and processed, etc |
| 16:27.16 | brlcad | 22:31 < brlcad> a comment in one of the source files or even in a text README file is warranted |
| 16:27.19 | brlcad | 22:31 < brlcad> as much detail an information as possible |
| 16:36.42 | elf11_ | Hey, Sean. I was already doing that, documenting for myself what Abhi told me, to organize stuff a little bit in my mind - by writing it down, so I will add a file with it in the source files |
| 16:38.02 | brlcad | great |
| 16:38.55 | elf11_ | and about the mm/m problem that we encountered? |
| 16:39.18 | elf11_ | also I am currently trying to see why the mass of the body is growing during the simulation |
| 16:45.12 | brlcad | elf11_: can you summarize the problem? |
| 16:45.31 | brlcad | limiting working units to mm is fine for now, but it does seem like a peculiar limitation |
| 16:47.02 | elf11_ | In the meter case we concluded that when transfering information from mged to bullet, mged has the units stored as n * 1000mm and not as n meters, so bullet gets n*1000 units, since bullet knows only about meters that's the unit we get |
| 16:47.49 | elf11_ | so scaling when transferring from mged to bullet is an option |
| 16:49.17 | elf11_ | also found out why the cube was penetrating the gp in that simulation, bullet has a tolerance of 0.04m so for collision detection to work properly geometry should be larger than that tolerance |
| 16:49.26 | elf11_ | the geometry for that simulation wasn't |
| 16:59.09 | brlcad | elf11_: brl-cad stores everything as mm and lets you have working units be nearly anything |
| 16:59.59 | brlcad | you shouldn't be "scaling" anything.. should be passing it in as-is |
| 17:01.23 | brlcad | if bullet has a tolerance of 0.04m, then it may make sense to "scale" it when feeding it to bullet, but only to overcome that limitation |
| 17:03.13 | brlcad | makes a note of http://wikihelp.autodesk.com/Alias/enu/2013/Help/0124-Autodesk124/0281-How_to..281/0282-Model282/0363-Modify_g363/0427-Cut_or_t427/0433-Convert_433 |
| 17:04.59 | elf11_ | So you say that brlcad is storing everything in mm, but when we transfer from mged to bullet we should have got meters? Of course if the geometry we created was in meters. |
| 18:45.35 | brlcad | elf11_: I wasn't speaking to what happens when you transfer to bullet -- just what brl-cad does |
| 18:46.14 | brlcad | everything is (always) stored as mm along with the working units so that numerics are stable -- conversion happens on the fly to your working units only when needed |
| 18:47.17 | brlcad | if you created geometry in m, then the mm values getting passed to bullet would be "too big" if bullet treats the values as m |
| 20:30.42 | starseeker | Well, nuts. Why'd Pixar have to go with Microsoft Public License... http://www.opensubdiv.com/?p=197 |
| 22:05.51 | *** join/#brlcad crdueck (~cdk@24-212-219-10.cable.teksavvy.com) |
| 22:29.57 | *** join/#brlcad Yoshi47 (~jan@d24-204-236-81.home4.cgocable.net) |
| 22:46.32 | *** join/#brlcad ChanServ (ChanServ@services.) |
| 22:46.32 | *** mode/#brlcad [+o ChanServ] by kornbluth.freenode.net |
| 22:48.47 | elf11_ | Hmm okay, thanks brlcad for that, then the scaling when transfering from mged to bullet is necessarily and when working in mm there should be some kind of warning about the size of the geometry. |
| 22:55.28 | starseeker | brlcad: hmm, another open font: http://sourceforge.net/projects/sourcecodepro.adobe/ |