| 00:00.25 | CIA-59 | BRL-CAD: 03brlcad * r47748 10/brlcad/trunk/ (include/bu.h src/libbu/tcl.c src/mged/edsol.c): do the same with bu_tcl_structparse_argv(). it was only used in one place, so eliminate it by just making the caller code call bu_structparse_argv() directly. |
| 00:06.22 | starseeker | hah - cool! |
| 00:24.27 | starseeker | doesn't look like it was the ambient occlusion patch |
| 00:30.10 | starseeker | whadya bet it's compile flag related... |
| 00:30.20 | starseeker | fires up autotools |
| 00:53.23 | starseeker | ok, failure is in autotools too... |
| 00:58.55 | starseeker | 47656 is OK... |
| 00:59.55 | starseeker | 47659 is bad |
| 01:05.21 | *** join/#brlcad Technicus (~Technicus@DSLPool-net208-2.wctc.net) | |
| 01:21.23 | CIA-59 | BRL-CAD: 03starseeker * r47749 10/brlcad/trunk/src/libbu/parse.c: Try to undo r47659 without ruining the subsequent commits... this was breaking the solids shader regression test, so needs more checking into. |
| 01:21.41 | brlcad | ah, right -- nick changed how shaders are parsed so tcl syntax works |
| 01:21.47 | brlcad | mater command |
| 01:27.39 | CIA-59 | BRL-CAD: 03brlcad * r47750 10/brlcad/trunk/sh/template.sh: don't just delete the header/footer file before aborting. we might have been running adding a header/footer to a file that already existed. instead restore the backup. |
| 01:28.14 | CIA-59 | BRL-CAD: 03brlcad * r47751 10/brlcad/trunk/sh/ (footer.sh header.sh): add support for adding header/footers to .cmake files, using sh-mode |
| 01:28.39 | CIA-59 | BRL-CAD: 03brlcad * r47752 10/brlcad/trunk/TODO.cmake: added the deprecation warning to the old build system |
| 01:29.38 | CIA-59 | BRL-CAD: 03brlcad * r47753 10/brlcad/trunk/misc/CMake/ThirdParty.cmake: a bug? do what the comment says it is supposed to be doing. |
| 01:34.02 | CIA-59 | BRL-CAD: 03starseeker * r47754 10/brlcad/trunk/ (CMakeLists.txt misc/CMake/BRLCAD_Util.cmake): Handle setting of default option in BRLCAD_OPTION |
| 01:40.15 | CIA-59 | BRL-CAD: 03brlcad * r47755 10/brlcad/trunk/ (9 files in 7 dirs): |
| 01:40.15 | CIA-59 | BRL-CAD: removed bu_badmagic_tcl() and all of the tcl-specific badmagic macros like |
| 01:40.15 | CIA-59 | BRL-CAD: BU_CKMAG_TCL and the various bn and rt macros that wrapped it (like |
| 01:40.15 | CIA-59 | BRL-CAD: RT_CK_AP_TCL, RT_CK_RTI_TCL, BN_CK_VLIST, etc). one step closer to decoupling |
| 01:40.16 | CIA-59 | BRL-CAD: libbu from tcl |
| 02:56.29 | CIA-59 | BRL-CAD: 03brlcad * r47756 10/brlcad/trunk/src/libbu/observer.c: _bu_ prefix was a bad idea |
| 03:14.29 | *** join/#brlcad merzo (~merzo@48-62-132-95.pool.ukrtel.net) | |
| 03:24.43 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) | |
| 03:30.42 | CIA-59 | BRL-CAD: 03brlcad * r47757 10/brlcad/trunk/src/libbu/cmdhist.c: _bu_ prefix was a bad idea, use the file name as the prefix |
| 03:35.05 | CIA-59 | BRL-CAD: 03brlcad * r47758 10/brlcad/trunk/ (include/bu.h src/libbu/tcl.c): change Bu_Init() to take a void* since dynamic loading should still work just fine without knowing the type. helps decouple the header from needing tcl.h |
| 03:42.54 | CIA-59 | BRL-CAD: 03brlcad * r47759 10/brlcad/trunk/src/libbu/cmdhist_obj.c: unfortunately not so simple to remove/move the tcl side of the command history implementation from libbu since it involves non-deprecated tclscripts api. |
| 03:56.06 | CIA-59 | BRL-CAD: 03brlcad * r47760 10/brlcad/trunk/ (10 files in 3 dirs): |
| 03:56.06 | CIA-59 | BRL-CAD: move command history tcl objects from libbu to libtclcad, changing them from |
| 03:56.06 | CIA-59 | BRL-CAD: init during loading of libbu to loading of libtclcad. this should keep things |
| 03:56.06 | CIA-59 | BRL-CAD: working for archer and get libbu one step closer to tcliberation. |
| 04:16.04 | CIA-59 | BRL-CAD: 03brlcad * r47761 10/brlcad/trunk/TODO: break cyclic dependency between libdm and libged |
| 04:16.51 | CIA-59 | BRL-CAD: 03brlcad * r47762 10/brlcad/trunk/doc/deprecation.txt: ws and 3 months should be necessary, not sufficient |
| 14:53.10 | CIA-59 | BRL-CAD: 03erikgreenwald * r47763 10/brlcad/trunk/src/libtclcad/cmdhist_obj.c: Move Cho_Init() to the end to avoid undefined function errors (prototypes were removed). |
| 16:12.04 | n_reed | is looking into the solids-regress failure |
| 16:29.06 | n_reed | I assert that r47659 is correct. I believe the test is what needs to be fixed. |
| 16:29.53 | n_reed | r47659 produces different shader strings for the mater calls in solids.sh which wrap plastic shader parameters in curly braces |
| 16:30.04 | ``Erik | reference and output pix's are different in brightness, dude |
| 16:30.25 | n_reed | that's correct |
| 16:32.55 | ``Erik | heh, long standing bug, even O.o |
| 16:34.03 | ``Erik | brlcad: any special approach to altering reference data? (quorum, s2 signoff, anything?) |
| 16:34.57 | ``Erik | this, the m35 torus intersecting geom for tires edge hit difference, mebbe re-normalizing vgr's to fix sphflake? |
| 16:43.52 | n_reed | so everyone's on the same page: in the old code, plastic {di .9 sp .5} becomes plastic {{di .9 sp .5} } |
| 16:44.08 | n_reed | this doesn't parse correctly, and is the same as applying default plastic shader |
| 16:44.47 | n_reed | r47659 fixed this, causing the parameters to actually be applied, resulting in different/brighter/correct output |
| 16:49.56 | ``Erik | out of curiosity, have you tried putting in a bu_log to display the actual values used before and after the change? |
| 17:09.32 | CIA-59 | BRL-CAD: 03bob1961 * r47764 10/brlcad/trunk/src/libged/polyclip.cpp: Fix ged_export_polygon() and ged_import_polygon() to handle multiple contours. |
| 17:11.40 | CIA-59 | BRL-CAD: 03bob1961 * r47765 10/brlcad/trunk/src/libtclcad/tclcad_obj.c: Update to_data_polygons() import sub-command to set the polygon color, line-width and line-style. |
| 17:27.17 | *** join/#brlcad milamber_ (414f0e1c@gateway/web/freenode/ip.65.79.14.28) | |
| 17:27.26 | *** part/#brlcad milamber_ (414f0e1c@gateway/web/freenode/ip.65.79.14.28) | |
| 17:44.12 | n_reed | erik: I have now. Logs show that phong shader gets correct parameters before revert, and defaults after revert. |
| 17:45.48 | n_reed | furthermore, rt log from post-revert shows error messages indicating the parsing failure and change to defaults |
| 17:46.55 | n_reed | like: "ERROR mlib_setup(/all.g/rec.r) failed. material='plastic', parameters='{sp .4 di .9 re .1}'. Changing /all.g/rec.r material to default and retrying." |
| 17:50.07 | n_reed | it's actually interesting that the shaders-regress didn't fail |
| 17:51.15 | n_reed | that test never uses the tcl-list syntax so it was unaffected, but maybe now it should |
| 20:18.47 | ``Erik | iiiiinteresting: http://stevehanov.ca/blog/index.php?id=130 http://pnylab.com/pny/papers/vptree/vptree/ |
| 20:45.51 | starseeker | ``Erik: is that useful for your bot work? |
| 20:55.32 | ``Erik | nah, there's too much invested in kd trees with tie... but it might be a good approach for the nurbs trimming, among other things |
| 20:56.10 | ``Erik | it's neat stuff, though... always neat learning new data structures and all |
| 21:05.06 | *** join/#brlcad merzo (~merzo@48-62-132-95.pool.ukrtel.net) | |
| 22:39.54 | CIA-59 | BRL-CAD: 03n_reed * r47766 10/brlcad/trunk/ (6 files in 3 dirs): initial cmake build logic for perplex |