| 00:26.06 | CIA-62 | BRL-CAD: 03kunigami * r46239 10/brlcad/trunk/src/liboptical/ (render_svc.cpp render_svc.h): the osl that I commited on svn defines a few more pure virtual functions. added definitons for these functions copying from /osl/src/testshade/ |
| 01:33.06 | abhi2011 | brlcad: regarding the bb function in librt, there was a issue with wdb_put_internal() always freeing the passed rt_db_internal * |
| 01:33.28 | abhi2011 | well making the rt_db_internal * parameter constant makes no difference |
| 01:34.39 | abhi2011 | the compiler discards the const qualifier and throws a warning : warning: passing argument 3 of âwdb_put_internalâ discards qualifiers from pointer target type , which is due to the same parameter not being declarared as const in wdb_put_internal() |
| 01:35.45 | abhi2011 | so I currently allow it to be freed and look it up again afterwards using db_lookup() and rt_db_get_internal() |
| 01:37.28 | abhi2011 | however after this, when I try to walk the tree for a combination, using db_functree(dbip, dp, comb_func, leaf_func, &rt_uniresource, NULL) the comb_func gets called as expected when the combination is detected |
| 01:37.54 | abhi2011 | but the leaf_func for the leaves of the combination, is not called |
| 01:40.16 | abhi2011 | instead db_lookup() reports missing primitives, so I don't think the primitives' information is still available for adding to the in-mem dbip , using db_functree() |
| 01:43.04 | abhi2011 | here is the modified code : http://bin.cakephp.org/view/1374634731 not committed yet as it does not work yet |
| 02:11.55 | bhinesley | abhi2011: it doesn't have to work to commit, it just has to build strict |
| 02:29.05 | bhinesley | abhi2011: there must not be any solids in your combination |
| 02:35.17 | bhinesley | redacts his last statement |
| 02:35.47 | bhinesley | I don't know what's happening, so personally, I would examine the tree in gdb |
| 04:07.36 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) | |
| 06:56.10 | *** join/#brlcad Yoshi47 (~jan@64.235.102.210) | |
| 07:44.25 | *** join/#brlcad kunigami (~kunigami@201.53.206.27) | |
| 09:51.45 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 10:34.10 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 10:40.29 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 11:41.44 | *** join/#brlcad merzo (~merzo@193.254.217.44) | |
| 11:44.45 | *** join/#brlcad juanman (~quassel@201.255.22.2) | |
| 11:44.51 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) | |
| 11:57.49 | kunigami | is it possible to a script permanently set an environment variable? after building osl I'd like brl-cad to know where it is installed. any other option to do that? (maybe a config file...) |
| 11:59.35 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) | |
| 12:07.34 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 12:09.39 | starseeker | kunigami: which part of BRL-CAD needs to know where osl is? |
| 12:10.35 | kunigami | starseeker: the cmake at liboptical and at other/osl/shaders |
| 12:15.27 | CIA-62 | BRL-CAD: 03kunigami * r46240 10/osl/trunk/compile.sh: added option to compile each library alone |
| 12:15.43 | *** join/#brlcad Yoshi47 (~jan@64.235.102.210) | |
| 12:49.09 | starseeker | kunigami: I'd suggest looking at what we do for the src/other libraries |
| 12:49.44 | starseeker | however, if it's always going to be installed first, you probably want find_library |
| 12:50.01 | starseeker | or write a FindOSL.cmake file |
| 12:56.44 | *** join/#brlcad Yoshi47 (~jan@64.235.102.210) | |
| 12:56.46 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) | |
| 13:09.00 | CIA-62 | BRL-CAD: 03tbrowder2 * r46241 10/brlcad/trunk/sh/conversion.sh: added an OPATH (object path); added 's' to seconds displays; wrapped a long line for ease of checking output format |
| 13:09.38 | CIA-62 | BRL-CAD: 03tbrowder2 * r46242 10/brlcad/trunk/NEWS: updated for change to conversion.sh |
| 13:10.18 | CIA-62 | BRL-CAD: 03tbrowder2 * r46243 10/brlcad/trunk/NEWS: corrected my last entry |
| 13:11.07 | CIA-62 | BRL-CAD: 03tbrowder2 * r46244 10/brlcad/trunk/NEWS: wrapped overflowing entry |
| 13:22.58 | kunigami | starseeker: but src/other libraries are build together with brl-cad, so it can set cmake variables that brl-cad sees. osl is build separately and is not necessarily build in any default place or a fixed place relative to brl-cad, so unless osl itself informs cmake, I don't see how find_library or FindOSL would be able to find it |
| 13:41.25 | CIA-62 | BRL-CAD: 03kunigami * r46245 10/osl/trunk/openexr/configure: missing pthread flag on tests |
| 13:42.51 | CIA-62 | BRL-CAD: 03kunigami * r46246 10/osl/trunk/openexr/exrenvmap/main.cpp: missing string.h library |
| 13:43.25 | CIA-62 | BRL-CAD: 03kunigami * r46247 10/osl/trunk/openexr/exrmaketiled/main.cpp: missing string.h library |
| 13:44.56 | CIA-62 | BRL-CAD: 03kunigami * r46248 10/osl/trunk/oiio/src/ico.imageio/icooutput.cpp: missing zlib.h header |
| 14:15.15 | *** join/#brlcad Yoshi47 (~jan@64.235.102.210) | |
| 15:06.46 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 15:06.54 | *** join/#brlcad Yoshi47 (~jan@64.235.102.210) | |
| 15:13.45 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) | |
| 15:24.24 | starseeker | FindOSL would work like any other find_package file |
| 15:24.40 | starseeker | assuming you're installing osl somewhere standard |
| 15:31.31 | dli | starseeker, what does this mean? http://pastebin.com/hHMAsVYd |
| 15:45.07 | starseeker | dli: I can't see that site - can you use another pastebin? (mozilla's should work) |
| 15:45.37 | dli | starseeker, never mind, I guess it's due to gentoo cmake-utils eclass changes, fixed already |
| 15:45.51 | starseeker | cool |
| 15:46.29 | dli | starseeker, pushed to overlay, when will 7.20.4 be released? |
| 15:46.52 | dli | starseeker, either I should try to make cmake work for 7.20.2 or just wait for 7.20.4 |
| 16:02.36 | starseeker | would recommend waiting for 7.20.5 |
| 16:02.40 | starseeker | er 7.20.4 |
| 16:03.03 | CIA-62 | BRL-CAD: 03starseeker * r46249 10/brlcad/trunk/src/librt/primitives/ (part/part.c table.c): Add a bbox routine for part |
| 16:06.30 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 16:06.47 | CIA-62 | BRL-CAD: 03starseeker * r46250 10/brlcad/trunk/NEWS: Tightened the bounding boxes for rpc primitives, particularly axis aligned rpcs - should slightly speed up rpc raytracing. |
| 16:10.31 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 16:16.58 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 16:20.17 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 16:22.46 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 16:30.23 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 17:28.58 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 17:33.13 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 17:42.05 | CIA-62 | BRL-CAD: 03kunigami * r46251 10/osl/trunk/osl/src/cmake/externalpackages.cmake: on ubuntu, llvm is installed in /usr/lib/llvm-2.8/include - how to specify this path to cmake such that it works for any library version? |
| 17:42.47 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 17:52.18 | CIA-62 | BRL-CAD: 03starseeker * r46252 10/brlcad/trunk/ (3 files in 3 dirs): epa bbox routine - also improved bounding boxes for epa, similar arrangement to the rpc. |
| 18:02.18 | CIA-62 | BRL-CAD: 03starseeker * r46253 10/brlcad/trunk/ (3 files in 3 dirs): Add improved bounding box logic for ehy, based on epa improvements. |
| 18:08.08 | kunigami | I'd like to use svn external do bring libz and libpng together with osl. Is it "svn propset svn:external https://brlcad.svn.sourceforge.net/svnroot/brlcad/brlcad/trunk/src/other/libz osl" ? |
| 18:18.20 | kunigami | command that worked: svn propset svn:externals 'https://brlcad.svn.sourceforge.net/svnroot/brlcad/brlcad/trunk/src/other/libz libz' . |
| 18:22.39 | CIA-62 | BRL-CAD: 03starseeker * r46254 10/brlcad/trunk/src/librt/primitives/ (eto/eto.c table.c): Add bbox routine for eto. |
| 18:24.36 | CIA-62 | BRL-CAD: 03n_reed * r46255 10/brlcad/trunk/src/libgcv/wfobj/ (obj_grammar.yy obj_parser.cpp obj_parser_state.h): storing lemon parser-handle in combined state object |
| 18:30.50 | CIA-62 | BRL-CAD: 03kunigami * r46256 10/osl/trunk/: adding libz and libpng as external dependence of osl/trunk |
| 18:41.52 | CIA-62 | BRL-CAD: 03kunigami * r46257 10/osl/trunk/compile.sh: added compile targets for libz and libpng |
| 18:47.35 | CIA-62 | BRL-CAD: 03starseeker * r46258 10/brlcad/trunk/src/librt/primitives/ (hf/hf.c table.c): add an hf bbox routine - not too worried about this one as hf is on its way out. |
| 18:48.47 | kunigami | hmmm osl is seg faulting with the local builds :( maybe it didn't like some of the versions. |
| 18:57.43 | CIA-62 | BRL-CAD: 03kunigami * r46259 10/osl/trunk/: svn propset svn:externals is not additive. have to specify all external dependences at once |
| 18:59.44 | CIA-62 | BRL-CAD: 03starseeker * r46260 10/brlcad/trunk/src/librt/primitives/ (dsp/dsp.c table.c): dsp bbox routine - this is another one that has to dupliate a lot of prep, so don't double up on the work - leave the prep calculations as-is. |
| 19:08.04 | CIA-62 | BRL-CAD: 03kunigami * r46261 10/brlcad/trunk/src/liboptical/ (liboslrend.cpp liboslrend.h osl_rt.cpp sh_osl.cpp): removed unnused code, debug messages and the old AddShader, which is not to be used |
| 19:11.47 | CIA-62 | BRL-CAD: 03bhinesley * r46262 10/brlcad/trunk/doc/docbook/system/mann/en/edit.xml: Started manual; more specifics on the argument style are needed. WIP |
| 19:11.52 | CIA-62 | BRL-CAD: 03kunigami * r46263 10/brlcad/trunk/src/liboptical/ (CMakeLists.txt Makefile.am osl_rt.cpp): deleting the stand-alone osl rt. It was already out-dated and it has no use since sh_osl is functional |
| 19:12.47 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 19:16.33 | CIA-62 | BRL-CAD: 03bob1961 * r46264 10/brlcad/trunk/src/libged/erase.c: |
| 19:16.34 | CIA-62 | BRL-CAD: This reverts the fix for r45409 (i.e. ged_splitGDL was no longer splitting |
| 19:16.34 | CIA-62 | BRL-CAD: things correctly. This caused a noticeable performance issue as well as the |
| 19:16.34 | CIA-62 | BRL-CAD: results of 'who' being wrong). The segmentation fault problem was resolved in |
| 19:16.34 | CIA-62 | BRL-CAD: _ged_eraseFirstSubpath. |
| 19:16.35 | plaes | kunigami: it could be running against system library |
| 19:16.58 | plaes | ldd is your friend ;) |
| 19:17.58 | kunigami | yup, ldd tells me it is using brlcad version. but that was the objective. I'm testing in a system with minimal system libraries to check if the libraries we're providing are enough |
| 19:33.36 | CIA-62 | BRL-CAD: 03starseeker * r46265 10/brlcad/trunk/src/librt/primitives/extrude/extrude.c: gah - extrude bbox logic is pretty close to worst case - don't see any immedate way to do a bbox better, so just try to do a 'self-contained' version that doesn't use stp structures and cleans up after itself. |
| 19:34.40 | CIA-62 | BRL-CAD: 03starseeker * r46266 10/brlcad/trunk/src/librt/primitives/table.c: whoops, add to table.c |
| 19:36.03 | CIA-62 | BRL-CAD: 03kunigami * r46267 10/osl/trunk/jpeg-8c/Makefile.in: missing file... |
| 19:39.07 | starseeker | humph - the way submodel is set up, I can't do a bbox routine for it without adding an rtip parameter |
| 19:41.51 | CIA-62 | BRL-CAD: 03kunigami * r46268 10/osl/trunk/tiff-3.9.5/Makefile.in: doh! missing file... |
| 19:48.09 | CIA-62 | BRL-CAD: 03kunigami * r46269 10/osl/trunk/tiff-3.9.5/ (23 files in 23 dirs): hmm! actually there are a lot more Makefile.in |
| 20:37.05 | CIA-62 | BRL-CAD: 03kunigami * r46270 10/osl/trunk/compile.sh: set linker path for linux env and make oiio use tbb |
| 20:54.37 | kunigami | here's the backtrace of the segfault: http://paste.ubuntu.com/669543/ -- the program does not even finished to load... it seems to be a runtime linking error |
| 21:00.31 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) | |
| 21:31.43 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) | |
| 21:37.47 | bhinesley | starseeker: what handles the generation of html files from docbook format? Using synopfragment tags would be really nice for my uses, but they're not getting formatted correctly. |
| 21:38.15 | bhinesley | I'll take a look at it if you can point me in the right direction |
| 21:39.12 | starseeker | um... the docbook stylesheets probably control that |
| 21:39.46 | starseeker | we haven't customized any of our output from docbook yet - that's a bit of a job |
| 21:40.02 | bhinesley | seems odd that they wouldn't format their own tag correctly |
| 21:41.07 | bhinesley | investigates |
| 21:46.27 | CIA-62 | BRL-CAD: 03starseeker * r46271 10/brlcad/trunk/src/librt/primitives/ (cline/cline.c table.c): add bbox routine for cline |
| 21:51.02 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) | |
| 21:56.41 | CIA-62 | BRL-CAD: 03starseeker * r46272 10/brlcad/trunk/src/librt/primitives/ (superell/superell.c table.c): break out bbox logic for superell |
| 22:16.57 | CIA-62 | BRL-CAD: 03starseeker * r46273 10/brlcad/trunk/src/librt/primitives/ (4 files in 2 dirs): metaballs already had most of the bbox functionality broken out - make it conform to the new functab setup and have it call the sphere routine itself to make it standalone. |
| 22:45.23 | CIA-62 | BRL-CAD: 03starseeker * r46274 10/brlcad/trunk/src/librt/primitives/ (brep/brep.cpp table.c): |
| 22:45.24 | CIA-62 | BRL-CAD: Do the simple thing with rt_brep_bbox and call the openNURBS routine - a quick |
| 22:45.24 | CIA-62 | BRL-CAD: check results in a pretty good bbox, although we'll stick with the bvh result |
| 22:45.24 | CIA-62 | BRL-CAD: for prep. The fact that bi->brep is NULL'ed by prep is a bit worrisome, leading |
| 22:45.24 | CIA-62 | BRL-CAD: to questions about whether this routine will work after raytrace - why are we |
| 22:45.24 | CIA-62 | BRL-CAD: NUll'ing the bi->brep during prep? |
| 23:12.47 | CIA-62 | BRL-CAD: 03starseeker * r46275 10/brlcad/trunk/src/librt/primitives/ (hyp/hyp.c table.c): Go with the epa/ehy approach for hyp too. Might actually be slightly larger surface area for some views when hyp is rotated, but MUCH better than sphere for axis aligned cases. |
| 23:40.33 | CIA-62 | BRL-CAD: 03starseeker * r46276 10/brlcad/trunk/src/librt/primitives/ (revolve/revolve.c table.c): Make a stab at implementing a stand-alone bbox function for revolve - again, too much extra logic to want to have prep call it. |
| 23:44.08 | CIA-62 | BRL-CAD: 03bhinesley * r46277 10/brlcad/trunk/ (3 files in 2 dirs): |
| 23:44.09 | CIA-62 | BRL-CAD: Wrote synopsis for 'edit translate'. It looks fine in brlman, but there is a |
| 23:44.09 | CIA-62 | BRL-CAD: mysterious line break in the html manual browser, and no line breaks in several |
| 23:44.09 | CIA-62 | BRL-CAD: places where there should be. First priority is to fix the issue. Failing that, |
| 23:44.09 | CIA-62 | BRL-CAD: I can simply remove the synopfragment tags (which are what is breaking the |
| 23:44.09 | CIA-62 | BRL-CAD: formatting) and rearrange things a bit. Added synopsis for 'edit', and cleaned |
| 23:44.10 | CIA-62 | BRL-CAD: up indentation. |
| 23:46.45 | *** join/#brlcad betta_y_omega (~betta_y_o@90.166.231.220) | |
| 23:54.00 | CIA-62 | BRL-CAD: 03starseeker * r46278 10/brlcad/trunk/src/librt/primitives/ (pnts/pnts.c table.c): Points apparently don't have any prep, but go ahead and add a bounding box function. |