| 00:11.38 | CIA-62 | BRL-CAD: 03starseeker * r46549 10/brlcad/trunk/src/libdm/CMakeLists.txt: Apparently we don't need the termlib include dir in libdm after all. |
| 00:49.03 | *** join/#brlcad nsd_ (~nicholas@c-24-0-153-224.hsd1.pa.comcast.net) |
| 02:43.11 | CIA-62 | BRL-CAD: 03starseeker * r46550 10/brlcad/trunk/ (5 files in 5 dirs): |
| 02:43.11 | CIA-62 | BRL-CAD: Do for other files what we do for .dist - fatal error if we're trying to ignore |
| 02:43.11 | CIA-62 | BRL-CAD: something that doesn't exist, unless it's a generated file (which will specify |
| 02:43.11 | CIA-62 | BRL-CAD: itself in a src list with a full path, and thus can be recognized.) Actually |
| 02:43.11 | CIA-62 | BRL-CAD: caught a number of bugs in the CMake logic, which are also fixed in this commit. |
| 08:18.16 | *** join/#brlcad juan_man (~quassel@unaffiliated/juanman) |
| 08:50.36 | plaes | uhh.. r46548 -- passing 20 compile jobs for build is a bit too much |
| 09:51.59 | ``Erik | plaes: yeah... I'd argue it should be single threaded unless otherwise requested... I'll alter that once I get into the office, mebbe about an hour from now... and roll up a newspaper to swat starseeker some :D |
| 09:57.42 | plaes | ;) |
| 09:58.21 | plaes | \o/ |
| 10:55.42 | CIA-62 | BRL-CAD: 03erikgreenwald * r46551 10/brlcad/trunk/CMakeLists.txt: Remove -j20. Parallel builds should be at the builders explicit request to avoid accidentally hammering the machine. |
| 11:16.03 | *** join/#brlcad merzo (~merzo@193.254.217.44) |
| 11:19.33 | *** join/#brlcad d_rossberg (~rossberg@BZ.BZFLAG.BZ) |
| 12:24.26 | brlcad | meh, that was specifically for distcheck builds which only affects devs -- make is running the build so the dev user doesn't have an immediate means to specify parallel |
| 12:25.02 | brlcad | it should sense the number of cpus and use that by default |
| 12:32.32 | ``Erik | cmake doesn't carry MAKE_FLAGS for you? |
| 12:35.23 | brlcad | not that I'm aware of |
| 12:35.26 | brlcad | maybe something like that could be added |
| 12:36.32 | brlcad | maybe a poor-mans version like: http://www.cmake.org/pipermail/cmake/2010-October/040122.html |
| 12:50.30 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) |
| 12:51.54 | ``Erik | I'm not keen on the automatically parallel thing, *shrug* if I set up a cron to do a distcheck on, say, bz, any parallel would cause a bit of service degredation. mebbe set NPROCS=1 and do "make NPROC=`sysctl -n hw.ncpu` distcheck" or something? |
| 12:53.19 | ``Erik | ('cept with the var names matching and all that) |
| 12:55.02 | ``Erik | hm, survice is putting on another BRL-CAD training course on sept 19, this time in eglin (ft walton beach, florida) O.o |
| 12:59.22 | CIA-62 | BRL-CAD: 03d_rossberg * r46552 10/brlcad/trunk/CMakeLists.txt: |
| 12:59.22 | CIA-62 | BRL-CAD: WIN32 is an add_executable() flag => sort it out from the source file names too |
| 12:59.22 | CIA-62 | BRL-CAD: because it's a CMake variable it has to be prefixed by a "x" e.g. |
| 13:01.00 | brlcad | ``Erik: mebbe, it just needs *some* way to go parallel or nobody will perform a distcheck very often -- it'll take way too long |
| 13:01.42 | brlcad | better to be useful and dangerous than relatively useless and safe |
| 13:03.11 | ``Erik | *shrug* or mebbe go parallel by default and have some way to force it to be low(ered) impact? |
| 13:03.28 | brlcad | if it can come from the initial command-line, even better, then it's both safe and useful but the goal should definitely be towards getting everyone running distcheck more frequently |
| 13:04.09 | ``Erik | I'm thinking more towards having it run automatically, continuous integration style :) |
| 13:04.43 | brlcad | we should *also* be doing that :) |
| 13:05.24 | brlcad | shouldn't be a crutch, imho though |
| 13:05.58 | brlcad | i should be able to invoke a distcheck on-demand without hesitation and get a result within 2x-3x of a regular build time |
| 13:06.14 | brlcad | if it's any longer or has more steps, most won't do it |
| 13:06.51 | ``Erik | regular build is single threaded, though... I do "nice make -j17 build install" on the 16 core machines |
| 13:07.17 | ``Erik | I wouldn't mind having to type "nice make -j17 distcheck", if that'd work on the subdir build |
| 13:07.19 | brlcad | I can't even do the on-demand part right now because there are extra steps (got to have a pristine checkout), but we talked about that one last night and it should be doable to get it going on any checkout |
| 13:07.46 | brlcad | yeah, that'd make sense except how the make distcheck rule is implemented |
| 13:07.53 | brlcad | make -jXX distcheck won't work |
| 13:08.11 | ``Erik | yeah, back to the MAKE_FLAGS ... :D |
| 13:08.30 | brlcad | just calls a distcheck rule that reinvokes a cmake build rule -- so you can distcheck from msvc or xcode, etc |
| 13:09.02 | brlcad | the cmake build rule knows nothing of the make -j flag that might have been set (though there might be some way to capture that) |
| 13:09.30 | brlcad | finds this disturbing, and highly suspect: https://sourceforge.net/tracker/?func=detail&atid=640803&aid=3402908&group_id=105292 |
| 13:10.56 | ``Erik | saw that, dunno what 'comodo' is, where the exe came from, if it's a false positive, mebbe a marketing move by that comodo company, ... :/ |
| 13:12.09 | ``Erik | comodo seems to be a chinese company that sells services which involve them having remote admin access to your winderz boxen |
| 13:15.28 | ``Erik | http://en.wikipedia.org/wiki/Comodo_Group hm |
| 13:16.32 | ``Erik | anim_cascade.exe is the only one marked |
| 13:19.33 | CIA-62 | BRL-CAD: 03d_rossberg * r46553 10/brlcad/trunk/include/CMakeLists.txt: |
| 13:19.33 | CIA-62 | BRL-CAD: there is no brlcad_config.h.in in the repository |
| 13:19.33 | CIA-62 | BRL-CAD: it appears to be CMake generated and lives in the CMake binaries' directory |
| 13:35.06 | brlcad | yeah, I saw the pic -- incredibly unlikely but possible |
| 13:35.35 | brlcad | who made the 7.20.0 binary? |
| 14:37.49 | *** join/#brlcad n_reed (~nreed1@c-68-55-142-136.hsd1.md.comcast.net) |
| 14:43.35 | brlcad | n_reed: 0.0005 is what it's "supposed" to be .. there are just a LOT of places that value is inconsistent |
| 14:44.24 | brlcad | so you had it right ;) |
| 14:44.50 | n_reed | brlcad: I'll change it back then. |
| 14:45.39 | n_reed | brlcad: Richard told me yesterday that it was supposed to be .005. |
| 14:46.27 | n_reed | brlcad: That is the tolerance that prints if you open a db model and type tol in mged. |
| 14:51.37 | brlcad | yeah, that's part of the inconsistency |
| 14:52.45 | brlcad | changing tolerance values requires a fair bit of retesting to make sure it's backwards-compatible |
| 15:34.29 | *** join/#brlcad packrat (~packrator@99-67-225-40.lightspeed.livnmi.sbcglobal.net) |
| 15:39.18 | CIA-62 | BRL-CAD: 03n_reed * r46554 10/brlcad/trunk/src/conv/obj-g_new.c: Reverted to previous revision. .0005mm is the correct distance tolerance. |
| 15:41.06 | brlcad | abhi2011: a matter of nomenclature -- the word "region" has a very specific meaning within BRL-CAD |
| 15:41.45 | brlcad | a region is a database combination with a flag set that indicates the combination occupies physical space |
| 15:42.43 | brlcad | all regions are combinations, but not all combinations are regions .. and no primitive is a region |
| 15:43.08 | brlcad | "groups" (aka assemblies) is a combination that contains one or more regions |
| 15:45.12 | abhi2011 | brlcad: ok |
| 15:46.18 | brlcad | simulate is pretty much agnostic to all of that, just working with "geometry objects" |
| 15:46.35 | brlcad | which can be regions, groups, combs, prims, etc :) |
| 15:57.31 | CIA-62 | BRL-CAD: 03brlcad * r46555 10/brlcad/trunk/src/tclscripts/mged/ (pattern.tcl tclIndex): use the new 'exists' command instead of rolling our own |
| 16:21.32 | CIA-62 | BRL-CAD: 03brlcad * r46556 10/brlcad/trunk/src/tclscripts/ (6 files in 2 dirs): more cases where the new 'exists' command can be put to use. use exists instead of db get or db get_type to test whether a database object already exists. |
| 16:54.27 | CIA-62 | BRL-CAD: 03starseeker * r46557 10/brlcad/trunk/CMakeLists.txt: Get a bit fancier with the package name and version for RPMs - commented out by default, but available if desired. |
| 16:59.31 | brlcad | conforming to HACKING or different? we need to be consistent for all our releases (for a whole variety of reasons) |
| 16:59.57 | brlcad | otherwise, it's basically "wrong" and should be fixed |
| 17:02.28 | brlcad | the format as-is "should" fit most any distribution system I'm aware of including support for optional notes and annotations |
| 17:09.47 | CIA-62 | BRL-CAD: 03abhi2011 * r46558 10/brlcad/trunk/src/libged/simulate/ (simphysics.cpp simulate.c simulate.h): Modified simulate to look through all regions in the current DB and add them as rigid bodies to the sim, objects now fallto the ground correctly |
| 17:31.13 | *** join/#brlcad n_reed (~nreed1@c-68-55-142-136.hsd1.md.comcast.net) |
| 19:00.17 | CIA-62 | BRL-CAD: 03erikgreenwald * r46559 10/brlcad/trunk/src/libgcv/bottess.c: simplify some bit mangling |
| 19:23.47 | CIA-62 | BRL-CAD: 03starseeker * r46560 10/brlcad/trunk/doc/docbook/system/mann/en/ (CMakeLists.txt exists.xml): Add preliminary man page for new version of the exists command. |
| 19:24.18 | starseeker | brlcad: figured I'd let you have a go at that before I start getting TOO deep into the C code |
| 19:31.43 | CIA-62 | BRL-CAD: 03starseeker * r46561 10/brlcad/trunk/doc/docbook/system/mann/en/exists.xml: whoops, extra char - stay consistent, 3 letters for those options |
| 19:37.40 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) |
| 19:58.24 | CIA-62 | BRL-CAD: 03starseeker * r46562 10/brlcad/trunk/src/libged/exists.c: Start roughing out the test.c->exists.c translation. Not doing any major rewiring yet until we've got the options more solidly pinned down, but it's a start. |
| 20:04.08 | abhi2011 | here is the obligatory bricks toppling over video : http://www.youtube.com/watch?v=TIgm0tNYseM |
| 20:04.21 | abhi2011 | made entirely with mged and rt |
| 20:06.48 | starseeker | cool! |
| 20:07.03 | starseeker | how come the block at the right on the end doesn't end up flat? (just curious) |
| 20:07.20 | starseeker | doesn't look like things quite go to equilibrium... |
| 20:11.30 | abhi2011 | yes there is some penetration of the ground plane , probably due to a error with positioning it or the default collision detection not working as expected :P |
| 20:11.56 | ``Erik | bullet ftw, awesome stuff |
| 20:13.11 | abhi2011 | so whats the easiest way to convert the image sequence output by rt into a mpeg movie in linux |
| 20:14.09 | abhi2011 | tried imagemagick, but its requires ffmpeg delegates or something similar |
| 20:14.15 | ``Erik | I used ffmpeg a while back |
| 20:14.22 | ``Erik | I hear the linux mplayer does well, too |
| 20:17.18 | ``Erik | making a driving game? O.o |
| 20:19.12 | abhi2011 | hehe yeah was trying to integrate bullet raycast vehicle into a moon buggy kinda thing |
| 20:19.36 | abhi2011 | was quite smooth with a basic engine model |
| 20:41.27 | *** join/#brlcad b0ef (~b0ef@166.195.251.212.customer.cdi.no) |
| 21:49.15 | *** join/#brlcad alex_joni (~alex_joni@emc/board-of-directors/alexjoni) |
| 22:02.28 | ``Erik | huzzah, truck is fixed |
| 22:02.44 | starseeker | awesome |
| 22:04.37 | ``Erik | idjit who rotated the tires didn't do a very good job of tightening the lug nuts, that was all |
| 22:15.07 | *** join/#brlcad IriX64 (~kvirc@bas2-sudbury98-1096600997.dsl.bell.ca) |
| 22:32.45 | *** join/#brlcad abhi2011_ (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) |
| 23:07.07 | brlcad | abhi2011: hah, that's f'n awesome :) |
| 23:07.24 | brlcad | starseeker: will do (reviewing the exists usage) |
| 23:08.42 | brlcad | abhi2011: so I see the view keeps changing on you as the simulation progresses -- that's due to something getting redrawn |
| 23:09.07 | brlcad | for animation purposes, you can either set up a view script and just keep re-rendering that script for each frame |
| 23:09.43 | brlcad | or you can capture the view size before and restore it after the command is updated (or don't erase and redraw, just redraw) |
| 23:10.40 | brlcad | abhi2011: I'd love to showcase that to the website if you can pull together a better video |
| 23:10.54 | brlcad | suggest a 2x zoom |
| 23:11.29 | brlcad | and maybe a different starting configuration if only to avoid the block that penetrates the surface |
| 23:48.01 | ``Erik | <PROTECTED> |
| 23:48.18 | ``Erik | I think the ogl demos that come with bullet do it that way :/ not sure |