| 00:13.52 | Ralith | <3 flyspell-prog-mode |
| 00:15.14 | starseeker | indianlarry: here's an odd thing - the odd hits in that vertical line on the right of the image seem to be due to a bounding box problem of some sort |
| 00:15.45 | starseeker | setting ae to 270 0 0 |
| 00:15.48 | Ralith | tries svn again |
| 00:15.51 | Ralith | yay! |
| 00:15.57 | starseeker | you see a vertical line all the way up and down the object |
| 00:16.13 | CIA-28 | BRL-CAD: 03ralith * r34711 10/rt^3/trunk/src/g3d/CMakeLists.txt: Added Qt4 + OpenGL to dependencies cmake is aware of |
| 00:16.44 | starseeker | if I've got this right, gdb reports five child leaf nodes. All five return hits |
| 00:18.34 | starseeker | of course, it might be that we need to return two hits somewhere and aren't... |
| 00:22.50 | CIA-28 | BRL-CAD: 03homovulgaris * r34712 10/brlcad/trunk/src/libpc/ (pcMathGrammar.h pcMathLF.h): Math expression grammer defined (4/4) . lazy function wrapper implementation of address_of |
| 00:27.52 | ``Erik | svn traps sigterm to try to 'clean up' gracefully |
| 00:28.12 | ``Erik | if you -9 it, you might need to use "svn clean" |
| 00:30.27 | Ralith | thanks for the notice |
| 00:30.32 | Ralith | seemed to ci cleanly, but I'll be sure |
| 00:31.09 | ``Erik | it'll bitch if it needs it :) |
| 00:31.15 | Ralith | okay then |
| 00:31.48 | Ralith | oh cleanup? |
| 00:31.50 | Ralith | yeah I had to do that |
| 00:32.45 | starseeker | hmm - weird |
| 00:33.04 | ``Erik | yes, you are quite weird, cliff :D |
| 00:33.07 | ``Erik | *duck* |
| 00:33.08 | starseeker | the problem area is definitely grazing on a trimming curve... |
| 00:33.21 | starseeker | and proud of it, too! |
| 00:33.33 | Ralith | hm, cmake screwed up the linking :/ |
| 00:36.01 | Ralith | asks #cmake |
| 00:40.55 | Ralith | idea! |
| 00:43.22 | Ralith | ...weird |
| 00:43.41 | *** join/#brlcad samrose (n=samrose@c-24-11-214-181.hsd1.mi.comcast.net) |
| 00:43.42 | Ralith | there's a /lib/rt-2.9.so on my system |
| 00:43.44 | Ralith | wonder what that is. |
| 00:43.48 | Ralith | librt-2.9.so* |
| 00:46.37 | CIA-28 | BRL-CAD: 03ralith * r34713 10/rt^3/trunk/src/g3d/CMakeLists.txt: Fixed Qt linking |
| 00:48.55 | starseeker | ah, so that's it - utah_newton_4corner returns one hit rather than two, which it should be returning two or zero |
| 00:49.30 | starseeker | or rather, utah_brep_intersect_test is |
| 00:52.04 | starseeker | yeah, it's utah_newton_4corner |
| 00:54.28 | starseeker | wonders if we should multiply the u or v parameter of ray and uv space/curves by a factor so the numbers are less small... |
| 00:54.31 | starseeker | hmm... |
| 01:13.33 | Ralith | oh wow |
| 01:13.40 | Ralith | when did sf.net completely redo their layout |
| 01:24.22 | madant | Ralith: didn't you get the mail :) |
| 01:27.24 | *** join/#brlcad Axman6_ (n=Axman6@pdpc/supporter/student/Axman6) |
| 01:29.47 | *** part/#brlcad Axman6_ (n=Axman6@pdpc/supporter/student/Axman6) |
| 01:30.56 | starseeker | indianlarry: if it helps any, here are two specific cases that are provoking errors - the first one is that long vertical miss due to getting one hit point when two are needed: |
| 01:31.06 | starseeker | Origin (x y z) = (-3.21022700 -49.92187372 6.84440400) |
| 01:31.09 | ``Erik | some leenewxes have a librt as an interface to real-time facilities |
| 01:31.18 | starseeker | Direction (x y z) = (0.00000000 1.00000000 0.00000000) |
| 01:31.33 | starseeker | second one is a trimming curve related snaffu: |
| 01:31.43 | starseeker | xyz -1.296691 -0.623331 -0.940507 |
| 01:31.50 | starseeker | dir -0.7424 -0.5198 -0.4226 |
| 01:32.07 | starseeker | (should turn on backout for both of these: backout 1) |
| 01:32.59 | starseeker | if I'm not mistaken, on the second one it's reporting four hit points but the very first one is getting trimmed away - when I look at the uv values this makes sense so I'm not quite sure what's going on there |
| 01:33.35 | starseeker | it's only very close to the trimming curves that this happens |
| 01:34.21 | starseeker | when I plotted the trimming curves over the edges I didn't see any obvious issue where trimming curves were mapping into different lines than the edges were |
| 01:49.22 | starseeker | if I had to guess, it's something to do with not getting a hit point in the box on the surface above where this trimmed hit was reported |
| 01:56.04 | starseeker | indianlarry: is there a possibility that when iterating to a solution very close to the edge of a bounding box, the haulting process is too aggressive? I.e., perhaps it should be "if it goes outside and doesn't attempt to come back in on the next iteration?" |
| 01:56.52 | starseeker | I have a feeling that could be what's going on here, since the trimmed point passed both the linear AND the curved "closest point" on trimming curve based tests |
| 01:56.58 | starseeker | i.e. it wasn't even "close" |
| 02:04.38 | starseeker | or, alternatly, it could be not getting all the boxes needed... |
| 02:09.56 | *** join/#brlcad IriX64 (n=IriX64@bas2-sudbury98-1128564968.dsl.bell.ca) |
| 02:15.43 | werneck | I'm new to brlcad, I just finished a design, is there any tool to get a 2D print of the blocks, or some CNC output? |
| 02:19.13 | starseeker | no cnc - if you want a line rendering there's rtedge |
| 02:19.47 | starseeker | if you need cnc, you can look for export options into formats that can be used to generate cnc paths |
| 02:20.12 | starseeker | indianlarry: nevermind, allowing iteration outside didn't do anything |
| 02:20.18 | indianlarry | starseeker: hey just looking over your comments |
| 02:20.43 | indianlarry | starseeker: i generated some nice images of the black widow |
| 02:21.11 | starseeker | coool :-) |
| 02:22.18 | indianlarry | starseeker: need to undefine (//#define) the KTANGENTBREAK in opennurbs_ext.cpp |
| 02:23.08 | indianlarry | starseeker: otherwise locks in prep(looks to not be converging in the bin iter will look at tomorow' |
| 02:24.28 | werneck | starseeker: ok, thanks |
| 02:24.45 | indianlarry | starseeker: still thinking about trying to iterate uphile in oppisite direction |
| 02:25.05 | indianlarry | starseeker: definitely need something better than the four corner approach |
| 02:26.24 | indianlarry | starseeker: what dotg are you refering to the nurbs_shape1.g ? |
| 02:26.29 | starseeker | I can't help thinking that iterating to a "max" point and using that somehow would help |
| 02:26.33 | starseeker | yes shape1.s |
| 02:26.59 | indianlarry | starseeker: i'll walk it through the debugger and let you know |
| 02:29.42 | starseeker | should try to dope out a routine to draw the 3d bounding boxes of every leaf that intersects a ray, just so we can be sure we're getting the right leaves to start with... |
| 02:29.53 | starseeker | should also get home before he gets in worse trouble... |
| 02:32.00 | indianlarry | starseeker: why buy a house when you can buy a cheap cot and stay there |
| 02:32.41 | starseeker | hehe |
| 02:33.21 | starseeker | can think of few ways of achieving SDDSO (Sudden Death Due to Significant Other) |
| 02:33.47 | starseeker | few better ways rather |
| 02:34.22 | indianlarry | gotta keep the significant other happy |
| 02:34.36 | indianlarry | catch up with you tomorow |
| 02:34.50 | starseeker | sounds good |
| 02:34.55 | starseeker | runs for it |
| 02:43.06 | brlcad | indianlarry: why do you think there's a bed in my office? :) |
| 02:45.09 | brlcad | good for an emergency early morning crash |
| 02:59.42 | werneck | is there any way to not hide lines in rtedge? |
| 03:26.20 | CIA-28 | BRL-CAD: 03ralith * r34714 10/rt^3/trunk/src/g3d/ (CMakeLists.txt OgreGLWidget.cxx OgreGLWidget.h): Untested Ogre Qt widget |
| 03:26.38 | Ralith | got cmake and qt's metastuff playing together nicely |
| 03:31.12 | *** join/#brlcad schwinn434 (n=schwinn4@cpe-75-81-202-25.we.res.rr.com) |
| 03:34.09 | CIA-28 | BRL-CAD: 03ralith * r34715 10/rt^3/trunk/src/g3d/ (OgreGLWidget.cxx OgreGLWidget.h): Narrowed include scope, included resource loading, fixed copyright/doc headers |
| 03:39.35 | Ralith | brlcad: any conventions on exceptions yet? |
| 03:52.22 | CIA-28 | BRL-CAD: 03ralith * r34716 10/rt^3/trunk/src/g3d/ (OgreGLWidget.cxx OgreGLWidget.h): Added useful destructor, scene setup; moved resource loading into its own function |
| 03:53.13 | Ralith | ahh, it's good to be making commits. |
| 04:25.21 | starseeker | posts this link to himself in case it comes in handy... http://tom.cs.byu.edu/~tom/papers/bezclip.pdf |
| 04:32.05 | starseeker | indianlarry: much as I hate to suggest this, what about the following approach: In the case where we detect that two intersections are a possibility, and current methods fail to identify two different intersection points, temporarily subdivide the leaf bounding square in uv space into 4 subsquares, take the center of each and iterate - if they all leave the boxes, then the other root (if it exists) must be in the remaining subbox with the existing root |
| 04:33.07 | starseeker | a bit expensive, but wouldn't it work? |
| 04:44.11 | *** join/#brlcad madant_ (n=d@117.196.134.174) |
| 07:11.28 | *** join/#brlcad _clock_ (n=_sushi_@84-72-91-14.dclient.hispeed.ch) |
| 07:37.13 | *** join/#brlcad IriX64 (n=IriX64@bas2-sudbury98-1128564968.dsl.bell.ca) |
| 08:12.56 | *** join/#brlcad IriX64 (n=IriX64@bas2-sudbury98-1128564968.dsl.bell.ca) |
| 08:57.36 | *** join/#brlcad mafm (n=mafm@126.Red-83-45-252.dynamicIP.rima-tde.net) |
| 09:04.14 | *** join/#brlcad ChanServ (ChanServ@services.) |
| 09:04.14 | *** mode/#brlcad [+o ChanServ] by irc.freenode.net |
| 09:10.12 | *** join/#brlcad ChanServ (ChanServ@services.) |
| 09:10.12 | *** mode/#brlcad [+o ChanServ] by irc.freenode.net |
| 09:54.54 | *** join/#brlcad geocalc (n=geocalc@lns-bzn-38-82-253-99-25.adsl.proxad.net) |
| 12:07.40 | *** join/#brlcad docelic (n=docelic@78.134.196.16) |
| 12:30.44 | starseeker | boots his brain back into gear |
| 13:02.17 | *** join/#brlcad madant (n=d@117.196.137.241) |
| 14:05.20 | brlcad | ``Erik: can metaballs have per point weights? |
| 14:05.40 | brlcad | or are they uniform across a metaballs set |
| 14:12.06 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-135.sbndin.btas.verizon.net) |
| 14:15.30 | brlcad | goes with the per-point theory |
| 14:24.43 | CIA-28 | BRL-CAD: 03brlcad * r34717 10/brlcad/trunk/src/proc-db/metaball.c: ws while peeking. huh, looks like this was never finished/implemented. |
| 14:44.27 | ``Erik | heh, they combine a per point weight with a per primitive value and let 'em duke it out |
| 14:46.43 | CIA-28 | BRL-CAD: 03davidloman * r34718 10/rt^3/trunk/ (cmake/ cmakemodules/): Refactor in prep for converting rt^3 module's build system to cmake. |
| 15:05.43 | d-lo | Ralith: Since you are making Qt an external dependancy (which is good imho), what about making orge an external one also? |
| 15:08.35 | starseeker | d-lo: Problem there is we might need to patch Ogre to get it to work, based on last year (IIRC, anyway) |
| 15:09.16 | d-lo | starseeker: as in a homegrown patch, or a patch from the Ogre devs? |
| 15:15.46 | starseeker | I THINK it was a homegrown patch, but it's been a while |
| 15:16.07 | starseeker | 'course, a year might have fixed things on the Ogre side too |
| 15:18.12 | d-lo | do you recall what issue the patch was concerning? |
| 15:18.50 | starseeker | nope - I wasn't paying as much attention at that point :-/ |
| 15:19.00 | d-lo | kk thanks anyways |
| 15:19.06 | starseeker | first thing to do is try the latest Ogre though |
| 15:19.20 | starseeker | doubts we've synced in a while |
| 15:20.19 | starseeker | oh, wait - it's coming back a bit - I think RBgui needed some feature that was either being removed or wasn't working in the "release" version of Ogre at that time |
| 15:20.38 | starseeker | so might be moot for Qt |
| 15:21.44 | d-lo | ..which would be good :) |
| 15:21.58 | d-lo | ogre isnt exactly small :) |
| 15:22.17 | starseeker | agreed |
| 15:22.18 | d-lo | one might say... its a bit of a beast. |
| 15:22.29 | d-lo | slaps his knee. |
| 15:25.26 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-135.sbndin.btas.verizon.net) |
| 15:33.47 | *** join/#brlcad fgfdg (n=fdgdfg@71.161.80.190) |
| 15:46.53 | *** part/#brlcad fgfdg (n=fdgdfg@71.161.80.190) |
| 17:17.20 | *** join/#brlcad BigATo1 (n=BigAToo@pool-96-230-124-25.sbndin.btas.verizon.net) |
| 17:25.29 | *** join/#brlcad mafm (n=mafm@199.Red-88-26-141.staticIP.rima-tde.net) |
| 17:45.17 | *** join/#brlcad samrose (n=samrose@adsl-68-73-194-95.dsl.sfldmi.ameritech.net) |
| 18:44.39 | *** join/#brlcad hippieindamakin8 (n=hippiein@202.3.77.38) |
| 19:24.03 | CIA-28 | BRL-CAD: 03indianlarry * r34719 10/brlcad/trunk/ (3 files in 3 dirs): |
| 19:24.03 | CIA-28 | BRL-CAD: fixed trim H/V tangent checks, now using curve estimate of trim versus |
| 19:24.03 | CIA-28 | BRL-CAD: linear, grew 3D bounding boxes to cover surface curvature extruding box |
| 19:24.03 | CIA-28 | BRL-CAD: need to fix correctly |
| 19:42.25 | Ralith | starseeker, d-lo: Ogre works fine as an external dep; shall I kill it from the repo? |
| 19:43.52 | starseeker | Ralith: sure, go for it |
| 19:46.46 | CIA-28 | BRL-CAD: 03ralith * r34720 10/rt^3/trunk/src/other/ogre/: Dropped Ogre from internal repository, as the official releases should work fine for us. |
| 19:52.04 | louipc | tcl/tk next :P |
| 19:52.17 | starseeker | yeah, that's... harder |
| 19:53.02 | CIA-28 | BRL-CAD: 03ralith * r34721 10/rt^3/trunk/ (13 files in 3 dirs): Moved cmake modules into root cmake module directory |
| 19:54.34 | Ralith | afks for another few hours |
| 19:58.02 | mafm | Ralith: starseeker: last year brlcad/Sear suggested me (under torture, even if he doesn't admit) to have all dependencies inside |
| 19:58.21 | mafm | similar to the rest of libraries in main brl-cad module |
| 19:59.00 | mafm | that is, I didn't put Ogre there because we needed to have a heavily patched version or anything |
| 20:00.56 | starseeker | mafm: really? huh. Well, we can always add them in later at need |
| 20:01.35 | mafm | yes well, just noting it |
| 20:37.04 | *** join/#brlcad IriX64 (n=IriX64@bas2-sudbury98-1128564968.dsl.bell.ca) |
| 22:19.03 | *** join/#brlcad madant (n=d@117.196.131.16) |
| 22:29.01 | Ralith | weird, ohloh doesn't seem to be updating commit stats |
| 22:47.00 | brlcad | they lag by a few days |
| 22:55.26 | *** join/#brlcad pacman87 (n=pacman87@pool-173-57-41-37.dllstx.fios.verizon.net) |
| 22:56.24 | *** join/#brlcad jdoliner (n=jdoliner@c-98-227-157-38.hsd1.il.comcast.net) |
| 23:44.14 | ``Erik | w00t, a/c is fixed O.o |