| 00:00.18 | ``Erik | ooh, not logged in, and still at the office |
| 00:00.42 | brlcad | wonder who that is |
| 00:01.16 | CIA-109 | BRL-CAD: 03Erik 07http://brlcad.org * r3199 10/wiki/Contributor_Quickies: /* EASY: Model new BRL-CAD Logo using BRL-CAD */ Add time guess |
| 00:03.43 | CIA-109 | BRL-CAD: 03Sean 07http://brlcad.org * r3200 10/wiki/Contributor_Quickies: clarify docs |
| 00:04.31 | CIA-109 | BRL-CAD: 03Sean 07http://brlcad.org * r3201 10/wiki/Contributor_Quickies: already had time estimate added, update |
| 00:06.15 | ``Erik | whups, assumed fime would be between body and references |
| 00:09.06 | CIA-109 | BRL-CAD: 03Sean 07http://brlcad.org * r3202 10/wiki/Contributor_Quickies: /* Outreach */ idea on interviewing jordi |
| 00:14.11 | CIA-109 | BRL-CAD: 03Sean 07http://brlcad.org * r3203 10/wiki/Contributor_Quickies: /* Outreach */ another on writing geometry cpp articles |
| 00:26.48 | CIA-109 | BRL-CAD: 03Sean 07http://brlcad.org * r3204 10/wiki/Contributor_Quickies: /* Quality Assurance */ deep unit test, and find bugs in archer |
| 00:32.44 | CIA-109 | BRL-CAD: 03Sean 07http://brlcad.org * r3205 10/wiki/Contributor_Quickies: /* Research */ update the spreadsheet |
| 00:39.42 | *** join/#brlcad juan_man (~quassel@unaffiliated/juanman) | |
| 00:41.04 | CIA-109 | BRL-CAD: 03Sean 07http://brlcad.org * r3206 10/wiki/Contributor_Quickies: /* User Interface */ reorganize mged's menu |
| 00:47.38 | CIA-109 | BRL-CAD: 03Sean 07http://brlcad.org * r3207 10/wiki/Contributor_Quickies: /* Translation */ translate our intro mged docs |
| 00:55.12 | CIA-109 | BRL-CAD: 03Sean 07http://brlcad.org * r3208 10/wiki/Contributor_Quickies: /* Translation */ be specific on the desired languages |
| 00:56.57 | CIA-109 | BRL-CAD: 03Sean 07http://brlcad.org * r3209 10/wiki/Contributor_Quickies: /* Translation */ |
| 01:01.55 | CIA-109 | BRL-CAD: 03Sean 07http://brlcad.org * r3210 10/wiki/Contributor_Quickies: /* Translation */ HACKING |
| 01:04.40 | CIA-109 | BRL-CAD: 03Sean 07http://brlcad.org * r3211 10/wiki/Contributor_Quickies: /* Translation */ our portuguese contingent deserve props for all the attention they've given over the years |
| 01:07.25 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) | |
| 02:09.50 | starseeker | hah cool, didn't know about this project: http://www.helenos.org/ |
| 04:04.36 | *** join/#brlcad abhi2011 (~chatzilla@117.200.89.70) | |
| 04:09.09 | CIA-109 | BRL-CAD: 03Abhi2011 07http://brlcad.org * r3212 10/wiki/Contributor_Quickies: /* EASY: Translate a chapter from the Introduction to MGED to Hindi */ |
| 04:09.21 | abhi2011 | :P |
| 06:16.49 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) | |
| 06:25.15 | abhi2011 | hmm getting a number of erros from a custom build rule in cmakelist.txt : http://bin.cakephp.org/view/1879910803 |
| 07:00.39 | CIA-109 | BRL-CAD: 03abhi2011 * r47389 10/brlcad/trunk/src/libged/simulate/simutils.c: Corrected a bug in the primitive lookup code for a comb |
| 08:36.30 | cvds_ | tgc(thumbPlungerTop1.s): A not perpendicular to B, f=-0.21693 <-- hmmm I did not know this was a requirement -_- |
| 08:36.57 | cvds_ | in thumbPlungerTop1.s rec 0 0 0 0 0 3 0 7.5 0 22.5 -5 0 <-- this is what I more or less want |
| 08:38.29 | cvds_ | (its combined with a in thumbPlungerTop2.s rpp 0 32.5 -7.5 7.5 0 3 thats why its not perpendicular) |
| 08:53.55 | cvds_ | resolved it by orot the rec inside the combination then pushing it |
| 10:03.53 | cvds_ | brlcad: I ordered a lot of thing for the printer so hopefully I can give you those pictures somewhere december (provided I can tune the printer well enough) |
| 10:32.45 | *** join/#brlcad archivist (~archivist@host81-149-189-98.in-addr.btopenworld.com) | |
| 10:32.45 | *** join/#brlcad piksi (piksi@pi-xi.net) | |
| 10:32.45 | *** join/#brlcad brlcad (~sean@BZ.BZFLAG.BZ) | |
| 10:32.45 | *** join/#brlcad alex_joni (~alex_joni@emc/board-of-directors/alexjoni) | |
| 10:33.21 | *** join/#brlcad starseeker (~starseeke@BZ.BZFLAG.BZ) | |
| 10:34.09 | *** join/#brlcad ``Erik (Here@c-69-140-109-104.hsd1.md.comcast.net) | |
| 10:34.18 | *** join/#brlcad cvds_ (~leila@h111030.upc-h.chello.nl) | |
| 10:34.50 | *** join/#brlcad packrat (~packrator@c-98-209-146-133.hsd1.mi.comcast.net) | |
| 10:51.43 | *** join/#brlcad bhinesley (~bhinesley@adsl-99-52-241-103.dsl.bkfd14.sbcglobal.net) | |
| 10:51.43 | *** join/#brlcad yiyus (1242712427@je.je.je) | |
| 10:51.43 | *** join/#brlcad ChanServ (ChanServ@services.) | |
| 10:51.43 | *** mode/#brlcad [+o ChanServ] by calvino.freenode.net | |
| 11:21.43 | CIA-109 | BRL-CAD: 03starseeker * r47390 10/brlcad/trunk/src/libbu/CMakeLists.txt: Whoops, ignoring wrong file |
| 11:53.06 | *** join/#brlcad abhi2011 (~chatzilla@117.200.84.234) | |
| 12:02.05 | *** join/#brlcad abhi2011 (~chatzilla@117.200.88.30) | |
| 12:05.36 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) | |
| 12:45.27 | *** join/#brlcad juanman (~quassel@201.216.198.121) | |
| 12:45.32 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) | |
| 12:51.41 | brlcad | cvds_: if you'd like to generalize the tgc even further, go for it ;) |
| 12:52.02 | brlcad | the intersection calculations get even more hairy if they're not perpendicular |
| 12:52.54 | brlcad | you can get non-perpendicular caps with a subtraction, so it's still an achievable shape -- just not with one tgc |
| 12:53.32 | brlcad | can't wait to see the pics ;) |
| 13:34.12 | CIA-109 | BRL-CAD: 03brlcad * r47391 10/brlcad/trunk/HACKING: freshmeat change their name to freecode |
| 13:47.28 | *** join/#brlcad abhi2011 (~chatzilla@117.200.86.135) | |
| 14:53.45 | cvds_ | brlcad: I see, I sorted it with a normal rec, looks good enough for now |
| 14:55.44 | cvds_ | http://flic.kr/p/aBer23 you can see the result here |
| 14:57.59 | CIA-109 | BRL-CAD: 03brlcad * r47392 10/brlcad/trunk/HACKING: |
| 14:57.59 | CIA-109 | BRL-CAD: add a regex one-liner awesomeness for automatically extracting the latest NEWS |
| 14:57.59 | CIA-109 | BRL-CAD: section into a release notes README-#-#-#.txt file. also fix the release steps |
| 14:57.59 | CIA-109 | BRL-CAD: so that binary platform maintainers are notified before public release |
| 14:57.59 | CIA-109 | BRL-CAD: announcements are posted (so they can have a chance to get started on binary |
| 14:57.59 | CIA-109 | BRL-CAD: builds) |
| 15:07.14 | cvds_ | and for the live of me I dont get solid rotation |
| 15:15.38 | cvds_ | rot takes into account current view angle ? |
| 15:18.39 | cvds_ | hmm arot actually seem to do what I expect |
| 15:30.33 | *** join/#brlcad n_reed (~molto_cre@BZ.BZFLAG.BZ) | |
| 15:36.47 | brlcad | cvds_: primitives rotate around some primitive-specific keypoint, which might not be where you'd expect an origin to be |
| 15:37.01 | brlcad | for a cylinder, for example, it's the center of the base ellipse |
| 15:37.11 | brlcad | for an arb8, it's the first corner |
| 15:37.50 | brlcad | you'd probably expect the object center for both, but to get that behaviour you'll either need to use one of the other rotation commands or set a keypoint explicitly |
| 15:38.50 | cvds_ | brlcad: nope, with rot I was expecting a rotation over 1 primary vertex, but it rotated over more. with arot I specify the vertex explicitly and things are spiffy ;_ |
| 15:38.53 | cvds_ | :) |
| 15:47.38 | CIA-109 | BRL-CAD: 03brlcad * r47393 10/brlcad/trunk/src/util/ (bw-png.c pix-png.c png-bw.c png-pix.c png_info.c): zlib.h needs to be included before png.h in case compression flags are used. also, they're system headers, so use brackets instead of quotes and pull them up into the right section. |
| 15:50.07 | CIA-109 | BRL-CAD: 03brlcad * r47394 10/brlcad/trunk/src/libged/png.c: they're system headers, so use brackets instead of quotes and pull them up into the right section. |
| 15:52.35 | CIA-109 | BRL-CAD: 03brlcad * r47395 10/brlcad/trunk/src/fb/ (fb-png.c png-fb.c): more header cleanup. png/zlib are system headers. use bin.h instead of directly including winsock.h |
| 16:11.34 | *** join/#brlcad abhi2011 (~chatzilla@117.200.83.152) | |
| 16:31.16 | CIA-109 | BRL-CAD: 03abhi2011 * r47396 10/brlcad/trunk/src/libged/simulate/simrt.c: Shooting rays in y direction now and analyzing the normals generated. |
| 16:40.11 | brlcad | ``Erik: you see the new gcc farm server? |
| 16:40.25 | brlcad | 64-proc power7 .. frickin awesome :) |
| 16:40.52 | brlcad | rather, 64-core |
| 16:47.08 | *** join/#brlcad n_reed (~molto_cre@BZ.BZFLAG.BZ) | |
| 17:04.46 | brlcad | starseeker: http://paste.debian.net/142105/ |
| 17:05.39 | brlcad | that was a default "cmake .." build |
| 17:48.52 | cvds_ | http://flic.kr/p/aBfMH8 <-- fun making these shapes |
| 17:56.59 | CIA-109 | BRL-CAD: 03abhi2011 * r47397 10/brlcad/trunk/src/libged/simulate/ (simrt.c simrt.h): Added code for shooting z rays and analyzing normals. |
| 18:07.46 | brlcad | cvds_: nice :) |
| 18:07.59 | brlcad | some sort of switch? |
| 18:08.08 | brlcad | electric contact switch ? |
| 18:12.14 | *** join/#brlcad Forth (~Forth@92.242.118.253) | |
| 18:53.45 | CIA-109 | BRL-CAD: 03128.63.32.62 07http://brlcad.org * r3213 10/wiki/Early_Raytracing_History: Stub out page for organizing early raytracing historical reports |
| 18:55.31 | starseeker | brlcad: are the opengl headers present? |
| 18:56.01 | starseeker | what platform? |
| 18:58.29 | brlcad | I don't see any opengl headers |
| 18:58.36 | brlcad | it's a linux |
| 18:59.06 | brlcad | looks like it's Fedora 16 |
| 19:01.26 | starseeker | hmm. Yeah, that's not going to be a widely tested setup |
| 19:01.30 | brlcad | obviously I could probably disable opengl and get past this, but implies the cmake logic isn't right if the default case doesn't properly autodetect and disable |
| 19:01.52 | brlcad | fedora is basically a future RHEL |
| 19:01.52 | starseeker | it's supposed to turn off opengl if it's not there, but I'm not surprised it's not quite right |
| 19:02.10 | starseeker | sans-opengl machines are a rarity these days |
| 19:02.40 | brlcad | this is a brand new system, so not really -- just a different type of configuration |
| 19:03.04 | brlcad | brand new ibm power series |
| 19:03.08 | starseeker | correction - they *should* be a rarity these days, even if they aren't |
| 19:03.17 | brlcad | server platform |
| 19:03.21 | starseeker | still |
| 19:03.25 | brlcad | servers rarely have graphics cards :P |
| 19:03.32 | starseeker | people run opengl apps on servers |
| 19:03.56 | brlcad | not necessarily on compute servers |
| 19:05.36 | brlcad | regardless, it's a bonefide build system bug so it should get fixed... |
| 19:23.19 | *** join/#brlcad Forth (~Forth@92.242.118.253) | |
| 19:47.18 | CIA-109 | BRL-CAD: 03starseeker * r47398 10/brlcad/trunk/ (3 files in 3 dirs): Tweak the OpenGL find routines to care more if the headers are found. |
| 20:00.17 | cvds_ | brlcad: thumb controlled switch. Magnetic / optical |
| 20:01.45 | brlcad | ~starseeker++ |
| 20:01.52 | brlcad | that seems to have done the trick, trying the build now |
| 20:02.31 | brlcad | giggles at make -j100 |
| 20:02.57 | cvds_ | -j100 ... thats many many cores |
| 20:03.07 | cvds_ | I run -j9 *3 |
| 20:09.22 | brlcad | it's a 64 core server |
| 20:09.37 | brlcad | so it actually should be able to juggle that many efficiently :) |
| 20:10.03 | brlcad | lesse how long this build takes.. :) |
| 20:10.18 | starseeker | winces... that's some heavy duty parallelism |
| 20:11.05 | starseeker | haven't actually tried a non-opengl build for months |
| 20:22.25 | CIA-109 | BRL-CAD: 03abhi2011 * r47399 10/brlcad/trunk/src/libged/simulate/simrt.c: |
| 20:22.25 | CIA-109 | BRL-CAD: Some bug fixes to ray shots, now the normals for rigid body B are summed |
| 20:22.25 | CIA-109 | BRL-CAD: correctly when there are overlapping objects : rigid body A and rigid body B. |
| 20:22.25 | CIA-109 | BRL-CAD: Next is shooting a bunch of rays in the direction opposite to this normal, to |
| 20:22.25 | CIA-109 | BRL-CAD: measure the depth of penetration of B into A and finally calculate the bunch of |
| 20:22.26 | CIA-109 | BRL-CAD: contact points on the surface of B which lies inside A(due to the penetration). |
| 20:22.27 | CIA-109 | BRL-CAD: Then we have all the required info to inject into Bullet |
| 20:22.33 | CIA-109 | BRL-CAD: 03starseeker * r47400 10/brlcad/trunk/src/other/CMakeLists.txt: If we turn off opengl, turn off togl too |
| 20:32.11 | brlcad | starseeker: hehe, we have a new winner! |
| 20:32.12 | brlcad | Elapsed compilation time: 41 seconds |
| 20:32.13 | brlcad | Elapsed time since configuration: 1 minute 19 seconds |
| 20:32.21 | starseeker | O.O |
| 20:32.37 | starseeker | it succeeded? |
| 20:32.38 | brlcad | attempts autotools for comparison |
| 20:32.39 | brlcad | yep |
| 20:32.53 | starseeker | if you're doing autotools compare, make sure you turn off the static libs |
| 20:33.01 | starseeker | (unless you enabled them in CMake) |
| 20:33.40 | brlcad | they're default enabled in cmake, so it's fair |
| 20:33.50 | starseeker | only for release config |
| 20:34.01 | starseeker | unless something changed, I have them off in Debug mode |
| 20:35.15 | brlcad | hm, hm |
| 20:35.40 | starseeker | (cept for opennurbs - need to fix that) |
| 20:36.34 | brlcad | is the summary not written out anywhere? |
| 20:36.43 | starseeker | you mean to a file? |
| 20:36.51 | brlcad | I see nothing in the CMakeOutput.log where I'd expect it.. |
| 20:37.02 | starseeker | no - it's just on the console |
| 20:37.22 | brlcad | mm, that's nfg .. there's no way to see the summary ? |
| 20:37.48 | brlcad | seeing AUTO in the cache tells me nothing :) |
| 20:37.59 | starseeker | ah |
| 20:38.26 | starseeker | can probably write it to a file easy enough |
| 20:38.47 | brlcad | ideally the entire cmake output |
| 20:39.34 | brlcad | but minimally the summary is super helpeful for debugging, can tell people to just send you the log file and see what they saw |
| 20:39.57 | starseeker | I don't know of any way offhand to capture just the on-screen output of CMake |
| 20:39.58 | brlcad | use config.log that way all the time |
| 20:40.14 | brlcad | it doesn't have to be just the output |
| 20:40.16 | starseeker | I can teach my summary logic to make a CMakeSummary.log file |
| 20:40.22 | brlcad | config.log is a superset, for example |
| 20:40.42 | brlcad | better to write to the existing CMakeOutput.log |
| 20:41.11 | brlcad | if you need someone to send you a log, you really only want to have to explain how to find/send one file |
| 20:44.53 | starseeker | right - I should be able to append to that file |
| 20:45.51 | brlcad | three times I've been bitten by the bundled/system/auto trio |
| 20:46.29 | starseeker | bitten how? |
| 20:46.35 | brlcad | most of the vars are on/off/auto, but the dep build flags aren't |
| 20:46.53 | CIA-109 | BRL-CAD: 03bob1961 * r47401 10/brlcad/trunk/src/tclscripts/mged/openw.tcl: Check the display manager type before setting the zbuffer state in "proc gui" |
| 20:47.15 | brlcad | I configure with BRLCAD_BUNDLED_LIBS as "ON" .. and don't get them all on, defaults back to auto because I didn't enter "BUNDLED" ;) |
| 21:24.00 | CIA-109 | BRL-CAD: 03brlcad * r47402 10/brlcad/trunk/src/ (12 files in 6 dirs): gcc continues to get smarter. apply fixes for issues detected by gcc 2.6.1, vars set but unused, unsigned 'char' that need to be int, and other type corrections. |
| 21:24.24 | starseeker | 2.6.1? |
| 21:24.28 | starseeker | yow |
| 21:26.34 | *** join/#brlcad merzo (~merzo@205-134-133-95.pool.ukrtel.net) | |
| 21:27.41 | CIA-109 | BRL-CAD: 03abhi2011 * r47403 10/brlcad/trunk/src/libged/simulate/ (simrt.c simrt.h simutils.c): Need to keep track of normals encountered so far, for a ray passing through rigid_body B, otherwise the same normals added twice will skew the resultant normal direction. |
| 21:28.37 | brlcad | with static: Elapsed compilation time: 57 seconds |
| 21:29.01 | brlcad | though not as controlled as the first build |
| 21:29.22 | brlcad | libpng is provoking some ld bug that I had to work around |
| 21:30.37 | CIA-109 | BRL-CAD: 03abhi2011 * r47404 10/brlcad/trunk/src/libged/simulate/simrt.c: Corrected the initialization of the number of normals. |
| 21:31.16 | starseeker | still - pretty darn impressive |
| 21:31.43 | starseeker | (or possibly impressive - perhaps autotools will do as well) |
| 21:33.36 | starseeker | brlcad: does that include docbook? |
| 21:34.59 | brlcad | if I had the summary in a log file, I could tell you ;) |
| 21:35.10 | starseeker | heh |
| 21:35.30 | starseeker | getting close |
| 21:35.47 | brlcad | autotools failed in the docbook section after 3min, so presuming it's compiling the same stuff, cmake is considerably faster |
| 21:36.53 | brlcad | though that's really a comparison of recursive make to non-recursive make, it's still a huge gain |
| 21:39.47 | CIA-109 | BRL-CAD: 03starseeker * r47405 10/brlcad/trunk/ (3 files in 2 dirs): |
| 21:39.47 | CIA-109 | BRL-CAD: Try a cute trick - override the message command to enhance logging. |
| 21:39.47 | CIA-109 | BRL-CAD: CMakeFiles/CMakeOutput.log should now have almost all messages from the cmake |
| 21:39.47 | CIA-109 | BRL-CAD: configure process - possible exceptions are those written out without using the |
| 21:39.47 | CIA-109 | BRL-CAD: MESSAGE command. Also make a stab at accepting ON and OFF for the |
| 21:39.48 | CIA-109 | BRL-CAD: AUTO/BUNDLED/SYSTEM vars |
| 21:39.48 | brlcad | there we go, so yeah .. about 3min20sec for autotools, enable-all, debug, with docs (no pdf) |
| 21:40.03 | brlcad | that's my usual compilation benchmark |
| 21:40.12 | starseeker | nifty :-) |
| 21:40.33 | starseeker | that's configure + build? |
| 21:40.54 | brlcad | configure was faster than cmake.. :) |
| 21:41.05 | starseeker | humph - figures |
| 21:41.33 | brlcad | and that's not even considering that I have to run cmake three times if run via ccmake to get an enable-all build |
| 21:41.51 | starseeker | nods - command line cmake ftw |
| 21:41.57 | brlcad | the system is using ccache, so I'll have to rerun to make sure there's not some object file caching going on too |
| 21:43.25 | brlcad | if I could get a list of our variables from cmake (e.g., cmake --help-variables) like configure, it'd be more feasible to run via command-line |
| 21:43.36 | brlcad | don't have the vars memorized |
| 21:43.48 | starseeker | nods - I need to finish the configure.cmake script |
| 21:44.20 | starseeker | I've got enable-all in there, but not the static libs flag |
| 21:44.52 | brlcad | in where? |
| 21:45.01 | brlcad | oooh, wrapper |
| 21:45.05 | starseeker | yeah |
| 21:45.21 | brlcad | bleh, that's cheating :) |
| 21:45.27 | starseeker | have expected you to take that and finish it so you wouldn't have to worry about the CMake options anymore |
| 21:45.33 | starseeker | :-P |
| 21:46.07 | brlcad | only after the core system is already near-optimal |
| 21:46.16 | brlcad | hacking on top of hacks is bad practice |
| 21:46.57 | brlcad | that goes for absolutely any code, it's only closed API that you have to put up with that |
| 21:47.14 | starseeker | hopes to $deity that we don't have much further to go, but suppose he knows better... |
| 21:47.28 | brlcad | I give it about two years |
| 21:52.38 | brlcad | doesn't persist an "off" setting, otherwise looks like it's closer |
| 21:55.33 | brlcad | there, more controlled rebuild was 1min38sec |
| 21:55.49 | brlcad | cmake, enable all (cept png), debug, with docs |
| 21:55.55 | brlcad | not too shabby |
| 21:56.08 | brlcad | double that time for the three config passes |
| 21:57.23 | starseeker | the second and third config passes should be considerably shorter... |
| 21:59.27 | cvds_ | about 50% according to the previous statement ? |
| 22:00.51 | brlcad | there, about 3min30sec for a repeat autotools build |
| 22:01.10 | brlcad | so roughly cutting the time in half or a third |
| 22:01.46 | cvds_ | heh... using combination and oed you can cut back on a lot of solids |
| 22:04.47 | CIA-109 | BRL-CAD: 03starseeker * r47406 10/brlcad/trunk/misc/CMake/ (BRLCAD_Util.cmake ThirdParty_TCL.cmake): Check for on and off in the optname, not the default |
| 22:38.06 | *** part/#brlcad n_reed (~molto_cre@BZ.BZFLAG.BZ) | |