| 00:04.06 | *** join/#brlcad gurwinder (75c76ccb@gateway/web/freenode/ip.117.199.108.203) |
| 00:31.54 | Notify | 03BRL-CAD:starseeker * 62520 brlcad/trunk/doc/docbook/system/mann/en/search.xml: Add a -path example to the search man page. |
| 00:46.40 | Notify | 03BRL-CAD:n_reed * 62521 brlcad/branches/brep-debug/src/libbrep/intersect.cpp: quell warning about var possibly being used uninitialized |
| 02:07.10 | Notify | 03BRL-CAD:n_reed * 62522 brlcad/branches/brep-debug/src/libbrep/boolean.cpp: copy input loops |
| 02:32.01 | brlcad | downloaded .. though it would be nice to get the original point cloud data (those pdf files are tessellations from geomagic) |
| 02:43.50 | Stragus | brlcad, you mentioned something related to Ball Pivoting from a starseeker commit, are you really implementing that technique? (This is Alexis from SURVICE, I wrote some code for that same technique, and... ewww!) |
| 02:44.50 | Stragus | It absolutely isn't designed to create watertight geometry, which is probably desired for the kind of stuff you guys do. Also, it's not robust by design, inconsistent. |
| 02:46.07 | brlcad | Stragus: hey alexis ... no, it's your code |
| 02:47.02 | Stragus | Ah right, didn't know it had found its way into BRL-CAD |
| 02:47.27 | Stragus | A different algorithm would have been far more appropriate |
| 02:48.48 | brlcad | funny you say that as that's certainly not how it's been promoted .. ;) |
| 02:49.20 | brlcad | i.e., a next generation method for getting robust tessellations |
| 02:49.26 | Stragus | Darn :) |
| 02:49.47 | Stragus | The mesh decimation code is pretty neat, but the ball pivoting was not the right algorithm |
| 02:51.51 | Stragus | Ball pivoting creates inconsistent results depending where "pivots" come from, leading to conflicts in creating the surface geometry |
| 02:52.09 | Stragus | And more than half the code is just "how to resolve conflicts" |
| 02:54.06 | Stragus | I believe it makes 16 attempts to rebuild conflicts with four different approaches, before giving up. So surface artifacts can be rare... but there's still no guarantees whatsoever about watertight geometry |
| 02:54.18 | brlcad | nods |
| 03:54.27 | Stragus | The normal itself could even be computed from the Nurbs equation I suppose, while ray intersection is dealing with triangles |
| 03:54.30 | Stragus | Oops |
| 04:33.38 | Stragus | For example, if you wanted to optimize the BRL-CAD CSG raytracer... It might involve reorganizing the whole memory layout, how all primitives are stored in memory for SSE processing, and much more |
| 04:33.41 | Stragus | Oops again |
| 07:48.49 | *** join/#brlcad ries (~ries@D979EA84.cm-3-2d.dynamic.ziggo.nl) |
| 08:43.48 | *** join/#brlcad gurwinder (caa43575@gateway/web/freenode/ip.202.164.53.117) |
| 08:45.08 | *** join/#brlcad gurwinder_ (caa43575@gateway/web/freenode/ip.202.164.53.117) |
| 12:28.07 | *** join/#brlcad kanzure (~kanzure@131.252.130.248) |
| 12:46.19 | *** join/#brlcad luca79 (~luca@ge-19-107-51.service.infuturo.it) |
| 12:47.24 | *** join/#brlcad kanzure (~kanzure@131.252.130.248) |
| 12:51.52 | Notify | 03BRL-CAD:indianlarry * 62523 brlcad/trunk/include/brep.h: Initialize m_start and m_end to ON_3dPoint::UnsetPoint and check before using in GetBBox() |
| 12:52.55 | Notify | 03BRL-CAD:indianlarry * 62524 brlcad/trunk/src/libbrep/opennurbs_ext.cpp: Initialize min,max to ON_3dPoint::UnsetPoint before using trims to get a tighter bounding box. Then check to make sure both are set before adjusting the surface bounds. |
| 13:02.03 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) |
| 13:07.19 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) |
| 15:13.57 | Notify | 03BRL-CAD Wiki:Murray4332 * 0 /wiki/User:Murray4332: |
| 15:27.01 | *** join/#brlcad raj12lnm (uid35020@gateway/web/irccloud.com/x-skytawdzxbfofmmp) |
| 15:43.47 | Notify | 03BRL-CAD:carlmoore * 62525 brlcad/trunk/doc/docbook/system/man1/en/pixshrink.xml: fixes to pixshrink man page; remove -h; fix an incorrect reference to -w (should have been -f) |
| 16:04.55 | *** join/#brlcad clock (~clock@77-58-143-135.dclient.hispeed.ch) |
| 16:50.10 | Notify | 03BRL-CAD:starseeker * 62526 (brlcad/branches/dm-work/include/ged.h brlcad/branches/dm-work/src/libged/CMakeLists.txt and 5 others): Start migrating the display_list manipulation logic to something that doesn't explicitly involve gedp. |
| 17:04.51 | Notify | 03BRL-CAD:carlmoore * 62527 (brlcad/trunk/doc/docbook/system/man1/en/pixinterp2x.xml brlcad/trunk/src/util/pixinterp2x.c): fixes to pix2interp2x |
| 17:24.09 | Notify | 03BRL-CAD:starseeker * 62528 (brlcad/branches/dm-work/include/ged.h brlcad/branches/dm-work/src/libged/bigE.c and 7 others): More gedp display list refactoring |
| 17:34.36 | Notify | 03BRL-CAD:starseeker * 62529 (brlcad/branches/dm-work/src/libged/erase.c brlcad/branches/dm-work/src/libged/ged_private.h brlcad/branches/dm-work/src/libged/remove.c): get the rest of the erase.c functions |
| 17:41.09 | Notify | 03BRL-CAD:starseeker * 62530 (brlcad/branches/dm-work/src/libged/CMakeLists.txt brlcad/branches/dm-work/src/libged/erase.c brlcad/branches/dm-work/src/libged/ged_private.h): move some logic around to start grouping display_list only functions |
| 18:06.21 | Notify | 03BRL-CAD:n_reed * 62531 brlcad/branches/brep-debug/src/libbrep/boolean.cpp: fix copy-paste error |
| 18:45.10 | *** join/#brlcad clock (~clock@77-58-143-135.dclient.hispeed.ch) |
| 19:13.54 | *** join/#brlcad clock (~clock@77-58-143-135.dclient.hispeed.ch) |
| 19:37.25 | ``Erik | is fisheye supposed to be cooking that much cpu? |
| 19:41.38 | Notify | 03BRL-CAD:starseeker * 62532 (brlcad/branches/dm-work/include/CMakeLists.txt brlcad/branches/dm-work/include/ged.h and 46 others): Ugly, but get dg.h and friends out of the BRL-CAD libraries. To keep mged working, stuck a hacked up wdb_obj in with src/mged to define the Tcl commands. This will need a fair bit of cleanup to be tolerable, but it serves the immediate need of not wanting to retrofit any display list |
| 19:41.40 | Notify | bits in the deprecated dg.h logic. |
| 19:57.41 | *** join/#brlcad FreezingCold (~FreezingC@CPE602ad06bea2a-CM602ad06bea27.cpe.net.cable.rogers.com) |
| 20:05.28 | *** join/#brlcad FreezingCold (~FreezingC@CPE602ad06bea2a-CM602ad06bea27.cpe.net.cable.rogers.com) |
| 20:17.28 | *** join/#brlcad FreezingCold (~FreezingC@CPE602ad06bea2a-CM602ad06bea27.cpe.net.cable.rogers.com) |
| 20:59.12 | *** join/#brlcad FreezingCold (~FreezingC@CPE602ad06bea2a-CM602ad06bea27.cpe.net.cable.rogers.com) |
| 21:10.06 | Notify | 03BRL-CAD:starseeker * 62533 (brlcad/branches/dm-work/src/libged/autoview.c brlcad/branches/dm-work/src/libged/display_list.c brlcad/branches/dm-work/src/libged/ged_private.h): refactor dl bounding sphere logic into a function and out of autoview.c |
| 21:25.02 | *** join/#brlcad FreezingCold (~FreezingC@CPE602ad06bea2a-CM602ad06bea27.cpe.net.cable.rogers.com) |
| 21:50.48 | Notify | 03BRL-CAD:brlcad * 62534 brlcad/trunk/include/raytrace.h: provide a static initializer for resource structs |
| 21:50.53 | brlcad | ``Erik: dunno, but I killed it |
| 22:03.08 | Notify | 03BRL-CAD:starseeker * 62535 (brlcad/branches/dm-work/src/libged/bot_dump.c brlcad/branches/dm-work/src/libged/display_list.c brlcad/branches/dm-work/src/libged/ged_private.h): A fair bit of work for marginal type hiding, but it does get the explicit dl_headSolid out of bot_dump.c |
| 22:14.47 | *** join/#brlcad Stragus_ (~alexis@modemcable008.161-160-184.mc.videotron.ca) |
| 22:16.10 | Notify | 03BRL-CAD Wiki:BNRAlberto * 0 /wiki/User:BNRAlberto: |
| 22:35.21 | Notify | 03BRL-CAD:carlmoore * 62536 brlcad/trunk/doc/docbook/system/man1/en/plot3-hpgl.xml: remove a word from a man page |
| 22:38.11 | *** join/#brlcad FreezingCold (~FreezingC@CPE602ad06bea2a-CM602ad06bea27.cpe.net.cable.rogers.com) |
| 23:25.50 | Notify | 03BRL-CAD:brlcad * 62537 brlcad/trunk/src/librt/globals.c: zero-initialize the rt_uniresource |
| 23:39.54 | Notify | 03BRL-CAD:brlcad * 62538 brlcad/trunk/src/librt/prep.c: clean up rt_init_resource() to validate all arguments and not set the magic until done initializing (todo: don't conditionally init the lists and ptbls nor allocate them memory until they need it.) |
| 23:44.47 | Notify | 03BRL-CAD:brlcad * 62539 brlcad/trunk/include/raytrace.h: mention that resource structs are are per thread of execution and that you usually want an array of them |
| 23:57.19 | Notify | 03BRL-CAD:brlcad * 62540 (brlcad/trunk/include/raytrace.h brlcad/trunk/src/librt/prep.c): document the rt_init_resource interface |
| 23:58.47 | Notify | 03BRL-CAD:brlcad * 62541 brlcad/trunk/include/raytrace.h: stray end-comment |