| 00:36.12 | mafm | great :) |
| 00:36.49 | mafm | d-lo: go and smack your system by testing my evil package! https://devel.adenu.ia.uned.es/~mafm/deb/brlcad/ |
| 00:37.01 | mafm | btw, thanks everybody for your patience :) |
| 02:29.03 | *** join/#brlcad _yukonbob (~svs@S010600235a187d92.ok.shawcable.net) | |
| 04:50.26 | *** join/#brlcad IriX64 (~MarioDUli@bas2-sudbury98-1096601417.dsl.bell.ca) | |
| 07:00.46 | *** join/#brlcad merzo (~merzo@smartbussiness.mobicom.net.ua) | |
| 08:19.41 | *** join/#brlcad mafm (~mafm@165.Red-79-159-0.staticIP.rima-tde.net) | |
| 10:56.34 | *** join/#brlcad Zaebos (~irc@pd95b7f5e.dip0.t-ipconnect.de) | |
| 11:08.00 | mafm | d-lo: are you around? |
| 12:02.00 | brlcad | he should be soon |
| 12:03.01 | mafm | I want victims for my evil artifact :þ |
| 12:25.51 | brlcad | :) |
| 14:01.13 | CIA-88 | BRL-CAD: 03starseeker * r40343 10/brlcad/branches/cmake/misc/CMake/FindTERMLIB.cmake: Use the existing macro in CMake to look for tputs, like autotools does with AC_CHECK_LIB |
| 14:09.39 | *** join/#brlcad kanzure (~kanzure@131.252.130.248) | |
| 14:19.12 | CIA-88 | BRL-CAD: 03bob1961 * r40344 10/brlcad/trunk/src/tclscripts/archer/ArcherCore.tcl: Minor tweaks to updateTreeTopWithName. |
| 14:22.05 | CIA-88 | BRL-CAD: 03starseeker * r40345 10/brlcad/branches/cmake/CMakeLists.txt: We were getting this setting from ZLIB - put it in our own toplevel - need it if we're not building our own ZLIB. |
| 14:29.50 | CIA-88 | BRL-CAD: 03bob1961 * r40346 10/brlcad/trunk/src/tclscripts/archer/ (Archer.tcl ArcherCore.tcl): Change updateTree to updateTreeDrawLists. Add a call to updateTreeDrawLists to the end of syncTree. |
| 15:01.38 | CIA-88 | BRL-CAD: 03bob1961 * r40347 10/brlcad/trunk/src/tclscripts/archer/ (Archer.tcl ArcherCore.tcl): Changed refreshTree to rebuildTree and removed the optional _restore parameter. |
| 16:17.09 | *** join/#brlcad Zaebos (~irc@pd95b7f5e.dip0.t-ipconnect.de) | |
| 17:53.50 | CIA-88 | BRL-CAD: 03bob1961 * r40348 10/brlcad/trunk/src/tclscripts/archer/ArcherCore.tcl: Modified selectTreePath to set mSelectedObj and mSelectedObjPath. |
| 17:55.58 | CIA-88 | BRL-CAD: 03bob1961 * r40349 10/brlcad/trunk/src/tclscripts/archer/Archer.tcl: Added buildInvalidObjEditView and initInvalidObjEditView. Arrange to call initInvalidObjEditView from initEdit if the object in question does not exist. |
| 18:07.18 | CIA-88 | BRL-CAD: 03brlcad * r40350 10/brlcad/trunk/sh/ws.sh: 'echo -n' is not compliant with the posix/unix definition for echo. this breaks mac 10.5+ which is now unix-compliant, so use printf instead. |
| 18:13.22 | CIA-88 | BRL-CAD: 03brlcad * r40351 10/brlcad/trunk/src/libbu/ (12 files): testing indent.sh. minor ws indent cleanup. |
| 18:14.08 | brlcad | mafm: oops, he's not online today .. maybe monday or weekend :) |
| 18:17.26 | CIA-88 | BRL-CAD: 03brlcad * r40352 10/brlcad/trunk/src/libbu/ (9 files): more cleanup (via ws.sh) to remove trailing line ws and insert embedded tabs. |
| 18:47.41 | CIA-88 | BRL-CAD: 03brlcad * r40353 10/brlcad/trunk/src/libbn/ (8 files): ws indent update |
| 19:17.02 | CIA-88 | BRL-CAD: 03erikgreenwald * r40354 10/brlcad/trunk/src/libbn/plane.c: rogue backslash in a macro define removed (bad ws.sh) |
| 19:54.06 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) | |
| 20:07.00 | CIA-88 | BRL-CAD: 03brlcad * r40355 10/brlcad/trunk/src/libbn/plane.c: bad ROT_VERT altogether. missing semicolon after VMINMAX. |
| 20:15.36 | *** join/#brlcad Led_Zeppelin (~user@ool-4573f43b.dyn.optonline.net) | |
| 20:50.44 | *** join/#brlcad mafm (~mafm@165.Red-79-159-0.staticIP.rima-tde.net) | |
| 21:17.18 | *** join/#brlcad IriX64 (~MarioDUli@bas2-sudbury98-1096601512.dsl.bell.ca) | |
| 21:25.58 | CIA-88 | BRL-CAD: 03bob1961 * r40356 10/brlcad/trunk/src/tclscripts/archer/Archer.tcl: A few tweaks to get the global_undo to work better when moving/renaming objects. |
| 21:45.55 | CIA-88 | BRL-CAD: 03n_reed * r40357 10/brlcad/trunk/ (6 files in 4 dirs): more changes to boteditor interface plus some minor changes to rt bot routines |
| 23:09.29 | brlcad | woo hoo |
| 23:09.50 | brlcad | significantly optimized bot vertex fusing |
| 23:12.00 | brlcad | was an O(n^2) algorithm (maybe O(n^3)), but now it's roughly O(n) .. |
| 23:12.43 | brlcad | mesh with 150k vertices processed in about 50sec is now 0.7sec |
| 23:18.01 | mafm | bah, that's peanuts |
| 23:18.26 | mafm | any good hacker would have optimized to O(n^1/3) |
| 23:18.32 | brlcad | mesh with 1.7M vertices processed in about an hour is now 4sec |
| 23:18.48 | brlcad | mafm: :P |
| 23:19.09 | mafm | you should have ended processing it before you even thought of starting! |
| 23:19.14 | mafm | congrats ;) |
| 23:19.41 | brlcad | of course, I didn't validate that the new one actually works so it could be all smoke :) |
| 23:19.58 | mafm | did you see my announcement? I hope that my package gets some love... he's so lonely in his directory... |
| 23:20.09 | brlcad | but it definitely processes into bins correctly, sorts, and identifies duplicates |
| 23:20.25 | brlcad | mafm: which announcement? |
| 23:21.09 | mafm | mailing list, about the package available |
| 23:27.27 | brlcad | ah, I've been in a hole coding for the past few hours |
| 23:27.37 | brlcad | will read it up tonight |
| 23:32.37 | CIA-88 | BRL-CAD: 03brlcad * r40358 10/brlcad/trunk/src/librt/primitives/bot/bot.c: (log message trimmed) |
| 23:32.37 | CIA-88 | BRL-CAD: after seeing nick waiting over an hour for a vertex fuse to complete, a quick peek at the code made it quite apparent that there was a possible O(n^3), or maybe O(n^2) + O(n^2), algorithm being used to fuse vertices. |
| 23:32.37 | CIA-88 | BRL-CAD: this implements bin-sort space partitioning where vertices are spatially grouped |
| 23:32.37 | CIA-88 | BRL-CAD: in one dimension into bins so that only nearby vertices need to be compared. |
| 23:32.38 | CIA-88 | BRL-CAD: the new approach is asymptotic O(n) (albeit with a heavy linear constant). the |
| 23:32.38 | CIA-88 | BRL-CAD: savings are proving to be quite substantial with fusing 150k vertices going from |
| 23:32.38 | CIA-88 | BRL-CAD: 50sec to 0.7sec and nick's 1.7M vertex data set going from over an hour to just |
| 23:33.40 | CIA-88 | BRL-CAD: 03brlcad * r40359 10/brlcad/trunk/TODO: yeah, need to more exhaustively verify that bot_vertex_fuse actually still works when there are lots of duplicate vertices.. the optimization is good, but needs more testing. |
| 23:34.23 | mafm | np, you won't test it anyway |
| 23:34.27 | CIA-88 | BRL-CAD: 03brlcad * r40360 10/brlcad/trunk/src/librt/Makefile.am: ws |
| 23:42.55 | brlcad | I had ubuntu running in a vm a few weeks ago .. might still have that image around to test it |
| 23:43.40 | mafm | ah, that's good then |
| 23:43.56 | mafm | but note that I only produced amd64 packages at the moment |
| 23:44.31 | CIA-88 | BRL-CAD: 03brlcad * r40361 10/brlcad/trunk/src/librt/ (63 files in 31 dirs): lots more ws. mostly end-of-line and indent tabbage. |
| 23:52.26 | *** join/#brlcad IriX64 (~MarioDUli@bas2-sudbury98-1096600771.dsl.bell.ca) | |