| 00:01.15 | Ralith | okay |
| 03:33.38 | *** join/#brlcad epileg (~epileg@unaffiliated/epileg) |
| 04:56.44 | *** join/#brlcad Stattrav (~Stattrav@122.172.16.143) |
| 04:56.44 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) |
| 07:55.10 | brlcad | Ralith: was waiting since you were gone and not away -- he didn't write up any of his plans or work from anything written up |
| 07:55.38 | brlcad | if he made any progress, it wasn't made publicly available that I'm aware of |
| 07:56.10 | Ralith | brlcad: just got ahold of him after a month of trying, actually. Will be resuming his work. |
| 07:56.17 | Ralith | he sent me his full git repo |
| 07:56.58 | brlcad | excellent |
| 07:57.07 | brlcad | I have a copy of some python code he wrote |
| 07:57.20 | brlcad | don't know if it's the same as what was in his repo |
| 07:57.35 | Ralith | nope, this is C |
| 07:57.38 | brlcad | but at a glance, it was mainly a way to do in python some of the things you can do in mged |
| 07:57.45 | brlcad | k |
| 07:57.57 | Ralith | this is a partly complete slicer based on the raytracer |
| 07:58.06 | Ralith | not sure how partly; I suspect mostly. |
| 07:58.08 | brlcad | that's good |
| 07:58.20 | Ralith | outputs a sequence of line segments and circular arcs |
| 07:58.24 | brlcad | that's one of two respectable approaches, the easier of the two |
| 07:58.33 | Ralith | what's the other, out of curiosity? |
| 07:59.09 | brlcad | we'll, you're either using ray-tracing and sampling the space to determine an approximation for toolpaths |
| 07:59.57 | brlcad | or you're projecting a contour image to 2D (either sampled or in parametric form) and using that for toolpaths |
| 08:00.46 | Ralith | contour image being something along the lines of the sketch-from-plane-region-intersection thing I was asking about earlier? |
| 08:00.54 | Ralith | projected contour image* |
| 08:01.13 | brlcad | basically |
| 08:01.26 | brlcad | that'd be one way to attempt it |
| 08:01.54 | Ralith | that was what I was planning on doing had I not been able to recover this code. |
| 08:02.04 | brlcad | three or four come to mind but they all involve projecting an object onto a viewing plane, not a simple thing to do |
| 08:02.05 | Ralith | looks like that will be unnecessary, though |
| 08:02.28 | brlcad | the sampled approach is MUCH much easier and doable today |
| 08:02.35 | Ralith | nod |
| 08:02.37 | Ralith | good to know |
| 08:02.45 | brlcad | and only limited by the resolution you're willing to sample |
| 08:03.05 | Ralith | and I have a concrete upper limit to that: the precision of the target machine. |
| 08:03.19 | Ralith | typically worse than 0.1mm. |
| 08:03.32 | brlcad | you can probably even sample below some specified machining tolerance and it'd never matter |
| 08:03.37 | brlcad | yeah |
| 08:03.48 | Ralith | and a typical build volume of 20cm means quite reasonable sample counts, I believe. |
| 08:04.22 | brlcad | yeah, that'd just be a 2k x 2k sample grid |
| 08:04.34 | Ralith | trickier on ultraprecise CNC lathes and such, but my first priority is reprap support, which means relatively low-res additive. |
| 08:04.42 | brlcad | that'd take just a second or two to calculate them all |
| 08:04.47 | Ralith | sweet |
| 08:05.08 | Ralith | if the line/arc approximation math jonored's already finished isn't too heavy, this'll easily outstrip everything else people are using |
| 08:05.29 | Ralith | all of which are improvised STL processors |
| 08:05.54 | brlcad | crank that up to a 200cm piece or 20cm at a 0.01mm tolerance, and you jack up the grid to 20k x 20k |
| 08:06.11 | brlcad | which would be a couple minutes (10x order of mag. increase) |
| 08:06.55 | Ralith | jonored also mentioned that he was put off by some bug wherein the second derivative of toroids was mis-reported? |
| 08:06.59 | Ralith | know anything about that? |
| 08:07.14 | brlcad | nope |
| 08:07.41 | Ralith | okay, will see about hunting it down if/when I encounter it myself |
| 08:07.41 | brlcad | first I've heard of toroids having a problem |
| 08:08.03 | Ralith | it sounded like a minor one: |
| 08:08.16 | Ralith | 19:56:53 <jonored> Yep. Either puts them in the wrong slots, or gives a direction of the larger curvature at right angles to what it should, don't recall which. |
| 08:08.44 | brlcad | there is one thing that comes to mind, but it's not second-derivative related. it's failure to converge on a hit/miss if you perfectly graze a torus under certain math conditions |
| 08:08.52 | brlcad | but then it just counts it as a miss |
| 08:09.04 | Ralith | seems like a pretty minor issue, at least for my purposes |
| 08:09.16 | brlcad | ah, curvature is a different routine |
| 08:09.46 | Ralith | he referred to it as having the "curvature flipped" |
| 08:09.51 | brlcad | not used by much, so that would be a little less surprising -- but also trivial to investigate -- one tiny func |
| 08:09.58 | Ralith | okay then |
| 08:10.10 | Ralith | will mention it if I find anything |
| 08:10.28 | Ralith | hopefully with a patch in tow, though I'm still a bit worried about my ability to keep up with the math |
| 08:10.35 | brlcad | src/librt/primitives/tor/tor.c:831 |
| 08:10.39 | brlcad | rt_tor_curve() |
| 08:10.45 | Ralith | thanks :) |
| 08:10.52 | brlcad | if there's a curvature bug, that's where it'd be |
| 08:11.45 | brlcad | it'd be easy to get a sign wrong |
| 08:12.32 | Ralith | looks like this might well be pretty straightforward. |
| 08:14.29 | brlcad | "rt -l4 -s1024 file.g object" will render 'object' in 'file.g' to a window 1024x1024 with a curvature visualization lighting model, so you can see them in an image |
| 08:14.52 | Ralith | ooh, fun! |
| 08:17.49 | Ralith | hm, behaves oddly. |
| 08:17.54 | Ralith | perhaps I should install a stable version |
| 08:18.50 | Ralith | frame buffers are hanging around and not responding to close requests, and display sharply skewed output despite it appearing normal mid-render |
| 08:19.17 | brlcad | are you (or your window manager) resizing the window? |
| 08:19.24 | Ralith | window manager is. |
| 08:19.42 | brlcad | that'd be a problem |
| 08:19.58 | Ralith | tiling wm; tends to be more forceful about that than much code expects |
| 08:20.05 | Ralith | not a very well behaved tiling wm, though. |
| 08:20.20 | brlcad | fb windows aren't resizable unfortunately |
| 08:20.23 | Ralith | will switch to something more compliant and with float support in the near future |
| 08:20.28 | brlcad | sucks, but nobody has fixed |
| 08:20.31 | brlcad | try fbserv instead |
| 08:20.39 | brlcad | fbserv 0 /dev/X & |
| 08:20.49 | brlcad | rt -F0 -s1024 file.g object |
| 08:21.35 | Ralith | that works nicely! |
| 08:21.36 | Ralith | thanks :) |
| 08:22.47 | brlcad | manpage shows other options if the need arises |
| 08:22.50 | Ralith | nod |
| 09:43.04 | *** join/#brlcad mafm (~mafm@5.Red-83-45-73.dynamicIP.rima-tde.net) |
| 10:38.21 | CIA-29 | BRL-CAD: 03tbrowder2 * r42406 10/brlcad/trunk/src/libged/attr.c: modify and rename attr sort comp function for extern use |
| 10:40.01 | CIA-29 | BRL-CAD: 03tbrowder2 * r42407 10/brlcad/trunk/src/libged/ged_private.h: add extern for attr comp function; add flags for extension of mged tree command args |
| 10:42.17 | CIA-29 | BRL-CAD: 03tbrowder2 * r42408 10/brlcad/trunk/src/libged/ls.c: cast to int to quell compiler warnings about signed/unsigned comparison |
| 10:44.21 | *** join/#brlcad epileg (~epileg@unaffiliated/epileg) |
| 11:13.47 | *** join/#brlcad mafm (~mafm@5.Red-83-45-73.dynamicIP.rima-tde.net) |
| 11:14.02 | CIA-29 | BRL-CAD: 03tbrowder2 * r42409 10/brlcad/trunk/HACKING: list extensions for Perl code |
| 11:44.18 | DaveLo | Mernin all |
| 12:32.38 | CIA-29 | BRL-CAD: 03Dloman 07http://brlcad.org * r2422 10/wiki/GeometryServiceNetworkProtocol: |
| 13:07.39 | *** join/#brlcad d_rossberg (~rossberg@BZ.BZFLAG.BZ) |
| 13:37.10 | DaveLo | opens a new can of Photoshop-fu. Will need lots today. |
| 13:38.50 | CIA-29 | BRL-CAD: 03Dloman 07http://brlcad.org * r0 10/wiki/Special:Log/upload: uploaded "[[Image:GSNetProtocol.png]]" |
| 13:55.05 | CIA-29 | BRL-CAD: 03Dloman 07http://brlcad.org * r2424 10/wiki/GeometryServiceNetworkProtocol: Added in specifics about GSNet Header stufff. WIP still. |
| 13:57.11 | CIA-29 | BRL-CAD: 03Dloman 07http://brlcad.org * r0 10/wiki/Special:Log/upload: uploaded "[[Image:GSNetMsgBreakdown.png]]" |
| 13:59.45 | CIA-29 | BRL-CAD: 03Dloman 07http://brlcad.org * r2426 10/wiki/GeometryServiceNetworkProtocol: New image name, down size a bit. |
| 14:03.33 | starseeker | is probably going to end up volunteering to maintain a few Find*.cmake modules |
| 14:09.06 | DaveLo | that'll be fun! |
| 14:10.03 | starseeker | apparently back in 07 they switched to a system where people volunteer to maintain specific modules, on the (quite reasonable) theory that Kitware didn't have the resources to properly test them all |
| 14:10.59 | *** join/#brlcad cjdevlin (~devlin@d118-75-252-178.try.wideopenwest.com) |
| 14:11.04 | starseeker | submitted patches for Tcl, OpenGL and X11 anyway on the theory that Kitware probably wouldn't want to trust them to a 3rd party maintainer, give how core they were |
| 14:11.07 | starseeker | silly me |
| 14:11.17 | *** join/#brlcad kanzure (~kanzure@131.252.130.248) |
| 14:11.20 | *** join/#brlcad 15SABD4WY (~CIA@208.69.182.149) |
| 14:12.26 | starseeker | in one sense though it's not too big a deal, since I'll either be maintaining our local copy of those files as a "mini-fork" or doing it for CMake proper |
| 14:13.28 | *** join/#brlcad CIA-38 (~CIA@208.69.182.149) |
| 14:13.32 | *** join/#brlcad packrat (~packrator@99-67-225-40.lightspeed.livnmi.sbcglobal.net) |
| 14:13.48 | starseeker | X11 and OpenGL should be pretty mature anyway, but the FindTCL module is a radical departure |
| 14:14.42 | DaveLo | Hows the roads down there? |
| 14:15.45 | CIA-38 | BRL-CAD: 03Dloman 07http://brlcad.org * r0 10/wiki/Special:Log/upload: uploaded "[[Image:GSNet Symbol.png]]" |
| 14:15.53 | *** join/#brlcad R0b0t1 (~Enigma@64-136-219-55.dyn.everestkc.net) |
| 14:15.53 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) |
| 14:17.59 | *** join/#brlcad ``Erik_ (Here@c-69-140-109-104.hsd1.md.comcast.net) |
| 14:18.33 | *** join/#brlcad epileg1 (~epileg@188.119.210.222) |
| 14:20.36 | *** join/#brlcad mafm (~mafm@5.Red-83-45-73.dynamicIP.rima-tde.net) |
| 14:21.53 | starseeker | grits his teeth and does a reboot into Windows for another test round... |
| 14:22.49 | CIA-38 | BRL-CAD: 03Dloman 07http://brlcad.org * r2428 10/wiki/GeometryServiceNetworkProtocol: Add in GSnet graphic and align. |
| 14:23.13 | DaveLo | starseeker: Hows the roads down there? made the trek in to work yet? |
| 14:24.11 | *** join/#brlcad Elrohir (~kvirc@p5B14977D.dip.t-dialin.net) |
| 14:27.59 | starseeker | no, not yet - they plowed but there is a lot of slush in our area |
| 14:28.13 | starseeker | just got done scraping the driveway |
| 14:29.16 | CIA-38 | BRL-CAD: 03Dloman 07http://brlcad.org * r2429 10/wiki/GSNet_String: Stub in GSNet String page. |
| 14:31.33 | starseeker | will head in once he sees what happens on Windows... |
| 14:33.09 | starseeker | oooo - lot of updates to pull |
| 14:37.44 | *** join/#brlcad Elrohir (~kvirc@p5B14977D.dip.t-dialin.net) |
| 14:59.43 | DaveLo | lotsa ice up here as well. |
| 15:00.00 | DaveLo | just called in to leave a telework notice. Heh, no one answered! =D |
| 15:06.09 | *** join/#brlcad epileg (~epileg@unaffiliated/epileg) |
| 15:10.43 | ``Erik_ | snow line said post was closed until 9 |
| 15:13.05 | CIA-38 | BRL-CAD: 03erikgreenwald * r42410 10/brlcad/trunk/src/librt/Makefile.am: move nmg_junk to EXTRA_DIST |
| 15:28.03 | CIA-38 | BRL-CAD: 03starseeker * r42411 10/brlcad/branches/cmake/src/CMakeLists.txt: Put the CMAKE_HEADERS variable in the src add_definitions (Windows seems to like this better, needs more testing.) |
| 15:36.01 | DaveLo | ``Erik: righto, but I called at 10 and still no answer :) |
| 15:39.33 | CIA-38 | BRL-CAD: 03erikgreenwald * r42412 10/brlcad/trunk/src/libged/ged.c: type fix |
| 15:44.38 | *** join/#brlcad Stattrav (~Stattrav@117.202.19.46) |
| 15:44.38 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) |
| 16:22.12 | CIA-38 | BRL-CAD: 03starseeker * r42413 10/brlcad/branches/cmake/CMakeLists.txt: |
| 16:22.12 | CIA-38 | BRL-CAD: Successfully generate an NSIS installer, although it doesn't work yet due to |
| 16:22.12 | CIA-38 | BRL-CAD: src/other libraries being installed to the wrong dir - need to correct the CMake |
| 16:22.12 | CIA-38 | BRL-CAD: files in question or add additional logic in src/other/CMakeLists.txt, whichever |
| 16:22.12 | CIA-38 | BRL-CAD: makes sense. |
| 18:03.41 | DaveLo | whoooo doggy. There's some serious ice out there! Im still chipping away at getting clear. |
| 18:03.49 | CIA-38 | BRL-CAD: 03starseeker * r42414 10/brlcad/branches/cmake/ (11 files in 11 dirs): Make the CMake library target logic a bit more generic. A lot of these files should probably be worked into being proper stand-alone CMakeLists.txt files. |
| 18:03.56 | DaveLo | looks like two layers of ice then snow. Blech |
| 18:04.12 | starseeker | nods - I'll be heading in soon, but yeah - figured the more melted the better |
| 18:04.21 | starseeker | at least it's above freezing, or moving would be out of the question |
| 18:04.37 | DaveLo | Temps up here are cycling above then below here. |
| 18:04.45 | starseeker | ah, we're a few degrees above |
| 18:04.59 | DaveLo | half the stuff that melted and ran off the jeep refroze on the ground :/ |
| 18:05.18 | DaveLo | We're getting more this evening, right? i havent checked the forecast in a bit. |
| 18:05.27 | starseeker | still had to shovel - if it had re-frozen tonight, would have had 3/4 inch or better of ice all over the driveway |
| 18:05.42 | starseeker | dunno - haven't seen any indications of more, but perhaps that's changed |
| 18:06.09 | starseeker | makes one file stab at Windows before heading in |
| 18:07.02 | DaveLo | checks forecast |
| 18:07.56 | DaveLo | heh, yeah. Starting around 10pm, more ice and snow... at least up here. |
| 18:08.40 | DaveLo | looks like mostly rain down your way, with temps never dipping below freezing. |
| 18:08.53 | DaveLo | have a nice day at work! muwahahaha =P |
| 18:09.35 | CIA-38 | BRL-CAD: 03Dloman 07http://brlcad.org * r2430 10/wiki/GeometryServiceNetworkProtocol: Change links to start eliminating the iBME references. |
| 18:10.01 | CIA-38 | BRL-CAD: 03Dloman 07http://brlcad.org * r0 10/wiki/Special:Log/delete: deleted "[[IBME NETWORKPROTO STRING]]": iBME references are antiquated and being updated. |
| 18:20.41 | *** part/#brlcad cjdevlin (~devlin@d118-75-252-178.try.wideopenwest.com) |
| 18:20.53 | *** join/#brlcad cjdevlin (~devlin@d118-75-252-178.try.wideopenwest.com) |
| 18:49.54 | *** join/#brlcad alex_joni (~alex_joni@emc/board-of-directors/alexjoni) |
| 19:24.53 | CIA-38 | BRL-CAD: 03Dloman 07http://brlcad.org * r2431 10/wiki/GeometryServiceNetworkProtocol: Spacing |
| 19:30.10 | CIA-38 | BRL-CAD: 03Dloman 07http://brlcad.org * r2432 10/wiki/GeometryServiceNetworkProtocol: Specifics about the libPKG header implementation |
| 19:48.20 | brlcad | starseeker: note that the current nsis installer is "wrong" in several aspects as to where files are placed |
| 19:49.22 | brlcad | you don't need to accommodate the existing mistakes -- keep the install tree like it is supposed to be so it's the same on all platforms |
| 19:49.25 | brlcad | the nsis installer can be fixed to accomodate |
| 19:52.00 | CIA-38 | BRL-CAD: 03Dloman 07http://brlcad.org * r2433 10/wiki/GeometryServiceNetworkProtocol: Use CSS to make the table purty. |
| 19:55.41 | CIA-38 | BRL-CAD: 03brlcad * r42415 10/brlcad/trunk/src/libged/ (attr.c ged.c ged_private.h): |
| 19:55.41 | CIA-38 | BRL-CAD: private functions used across commands should be in the _ged_ namespace, |
| 19:55.41 | CIA-38 | BRL-CAD: otherwise, they don't need to be declared in ged_private.h and can remain HIDDEN |
| 19:55.41 | CIA-38 | BRL-CAD: _cmd_ prefixed. rename _attr_cmpstringp() to _ged_cmpattr() to reflect that it |
| 19:55.41 | CIA-38 | BRL-CAD: specifically compares bu_attribute_value_pair objects. |
| 20:12.27 | CIA-38 | BRL-CAD: 03brlcad * r42416 10/brlcad/trunk/src/libged/ (attr.c ged.c ged_private.h): ws consistency update, elim forward decls |
| 20:15.04 | starseeker | nods |
| 20:15.30 | starseeker | trying to figure out why mged isn't finding anything on Windows... |
| 20:15.47 | starseeker | needs to generalize the pkgIndex.tcl files a bit, doing that first |
| 20:19.42 | CIA-38 | BRL-CAD: 03brlcad * r42417 10/brlcad/trunk/src/libged/ (ged_private.h ls.c): convert up to size_t instead of down to int. |
| 20:56.07 | CIA-38 | BRL-CAD: 03brlcad * r42418 10/brlcad/trunk/src/librt/parse.c: quell warnings, size_t upconversions |
| 20:58.54 | CIA-38 | BRL-CAD: 03brlcad * r42419 10/brlcad/trunk/src/librt/Makefile.am: |
| 20:58.55 | CIA-38 | BRL-CAD: compile parse.c and nmg_junk.c so we make sure that they keep compiling as the |
| 20:58.55 | CIA-38 | BRL-CAD: library evolves. they don't need to be installed, but they should still be |
| 20:58.55 | CIA-38 | BRL-CAD: maintained until they're integrated or removed. (alas, they both have useful |
| 20:58.55 | CIA-38 | BRL-CAD: routines that should be integrated) |
| 20:59.38 | brlcad | starseeker: be wary of #ifdef _WIN32 and if {$platform == "windows"} statements in C/Tcl code that Bob added to override search paths |
| 21:00.33 | brlcad | some of them are sprinkled in unsuspecting places, have to trace the logic when something fails to load that should be loading to make sure it's not a hack job screwing things up |
| 21:00.54 | starseeker | nods - will do |
| 21:22.42 | CIA-38 | BRL-CAD: 03starseeker * r42420 10/brlcad/branches/cmake/src/other/ (6 files in 6 dirs): Fix/generalize pkgIndex.tcl logic, make incrTcl libs follow convention. |
| 21:40.08 | CIA-38 | BRL-CAD: 03brlcad * r42421 10/brlcad/trunk/src/librt/primitives/metaball/metaball.c: |
| 21:40.09 | CIA-38 | BRL-CAD: not entirely clear what the compiler is complaining about here with the two |
| 21:40.09 | CIA-38 | BRL-CAD: (const point_t *) parameters to rt_metaball_find_intersection(). the problem |
| 21:40.09 | CIA-38 | BRL-CAD: has something to do with point_t being an array typedef. instead of getting the |
| 21:40.09 | CIA-38 | BRL-CAD: address during the call, get the point_t address beforehand with an explicit |
| 21:40.09 | CIA-38 | BRL-CAD: cast to keep things quiet. (gcc 4.2.1) |
| 21:49.15 | ``Erik | hm, I disabled nmg_junk.c last week due to 'defined but not used' warnings stopping the compile |
| 22:08.28 | brlcad | everything in librt_xxx.la is not used, so maybe it'll be quiet -- seems to work here with 4.2.1 |
| 22:08.46 | brlcad | otherwise, I can add some extern decls |
| 22:08.54 | brlcad | that'll definitely make it shut up |
| 22:15.54 | CIA-38 | BRL-CAD: 03starseeker * r42422 10/brlcad/branches/cmake/src/libbu/brlcad_path.c: |
| 22:15.54 | CIA-38 | BRL-CAD: Interesting... when launching bwish from the root install dir (e.g. ./bin/bwish) |
| 22:15.54 | CIA-38 | BRL-CAD: the return of bu_brlcad_data was not a full path (or even strictly speaking a |
| 22:15.54 | CIA-38 | BRL-CAD: relative path unless you add an implicit '.') and archer didn't start. This is |
| 22:15.54 | CIA-38 | BRL-CAD: a bit of a corner case, but still should work - try to make it more robust. |
| 22:29.09 | CIA-38 | BRL-CAD: 03starseeker * r42423 10/brlcad/branches/cmake/src/archer/plugins/Utility/botUtilityP.tcl: shouldn't need brlcadDataPath variable here. |
| 22:38.48 | CIA-38 | BRL-CAD: 03starseeker * r42424 10/brlcad/branches/cmake/src/libbu/brlcad_path.c: Whoops, try to do snprintf right |
| 22:47.42 | starseeker | mutter... bwish starts, but it's auto_path is rather underpopulated |
| 22:50.13 | starseeker | and bu_brlcad_data doesn't know what's up on Windows (at least from build-dir, which I suppose isn't too surprising) |
| 23:06.08 | *** join/#brlcad epileg (~epileg@unaffiliated/epileg) |
| 23:46.11 | brlcad | starseeker: that could be related to path separator, or trying to mix one style of separator with another |