| 00:20.51 | starseeker | will have to test with newer versions on mac to see if it makes a difference |
| 00:23.45 | CIA-77 | BRL-CAD: 03starseeker * r43592 10/brlcad/branches/cmake/ (5 files in 4 dirs): MFC r43591 |
| 00:28.44 | CIA-77 | BRL-CAD: 03starseeker * r43593 10/brlcad/trunk/src/libged/search.c: gentoo wants search_results initialized |
| 00:33.39 | brlcad | that's just fantastic.. mged crash |
| 00:34.03 | starseeker | winces - what's the issue? |
| 00:34.14 | brlcad | don't know yet, just crashed opening an nmg |
| 00:34.19 | starseeker | ah |
| 00:34.30 | starseeker | fwiw - search seems to be working on gentoo |
| 00:55.57 | brlcad | excellent |
| 01:07.46 | *** join/#brlcad Ralith (~ralith@S010600221561996a.vc.shawcable.net) |
| 02:49.35 | *** join/#brlcad ibot (~ibot@rikers.org) |
| 02:49.35 | *** topic/#brlcad is BRL-CAD Open Source Solid Modeling || http://brlcad.org || http://sf.net/projects/brlcad || #brlcad logs: http://ibot.rikers.org/%23brlcad/ || BRL-CAD release prep for 7.18.2 under way (20110202) |
| 02:51.00 | starseeker | nods - I'll have to look at the current tests and see what he got running |
| 02:51.17 | starseeker | he breaks things out into analytical and numerical categories in the wiki page |
| 02:51.57 | starseeker | I'm assuming numerical would involve either floating point or some form of quantizing |
| 02:55.44 | starseeker | huh - looks numerical... |
| 03:04.01 | starseeker | wait... gecode has a flatzinc support setup, which does have some sort of Float support... |
| 03:51.28 | CIA-77 | BRL-CAD: 03brlcad * r43594 10/brlcad/trunk/NEWS: release write-up on the integration of TIE with LIBRT and the new shp-g shapefile importer. |
| 04:10.03 | CIA-77 | BRL-CAD: 03brlcad * r43595 10/brlcad/trunk/BUGS: encountered an infinite loop bug trying to smooth a bot. might be the same traversal bug encountered earlier, but needs testing / investigating. |
| 04:42.01 | brlcad | ``Erik_: harrumph.. I can't seem to get a render through to the tie code |
| 04:42.35 | brlcad | created a volume bot, smoothed it, rh orientation, mintie set to 1 .. still kicks over to old way |
| 04:42.41 | brlcad | any suggestions? |
| 04:42.51 | *** join/#brlcad yukonbob (~bch@S0106002129e399fc.ok.shawcable.net) |
| 04:42.53 | yukonbob | oh hai |
| 04:42.56 | brlcad | howdy |
| 04:43.31 | yukonbob | my friend. |
| 04:43.35 | yukonbob | how're things? |
| 04:44.31 | brlcad | crazy busy, trying to push a release out and hitting speed bumps left and right |
| 04:44.33 | brlcad | bbl |
| 04:46.03 | yukonbob | working through a bump w/ .2 current release + tk... |
| 04:49.20 | ``Erik_ | does it prep for tie? I may've borked some logic... |
| 04:49.50 | yukonbob | ``Erik_: that for me? |
| 04:50.14 | ``Erik_ | no, unless you're trying to do bot/tie raytracing |
| 04:50.16 | brlcad | ``Erik_: prep is where I've added some debug and it's not getting there |
| 04:50.48 | yukonbob | stands back from the prep ties. |
| 04:51.05 | ``Erik_ | hm, bot.c:208 is untested logic... I can look tomorrie |
| 04:51.44 | CIA-77 | BRL-CAD: 03brlcad * r43596 10/brlcad/trunk/src/librt/primitives/bot/bot.c: commit the debug code so it's clear what's going on with the tie integration |
| 04:53.07 | ``Erik_ | should be in really early tomorrow, so I'll be able to crank up the tunes, chill the office and be somewhat productive |
| 05:38.10 | yukonbob | brlcad: can I squeak in a patch for configure? |
| 05:38.48 | yukonbob | will see if still has commit bit... |
| 05:54.24 | CIA-77 | BRL-CAD: 03bharder * r43597 10/brlcad/trunk/configure.ac: Direct use of interp->result is deprecated; Use Tcl_GetStringResult(). |
| 06:05.04 | yukonbob | I'll look for more later... is not an immediate problem, but will be in future. |
| 06:32.15 | *** join/#brlcad yukonbob (~bch@S010600235a187d92.ok.shawcable.net) |
| 06:47.18 | brlcad | ``Erik: nod, i'll be in probably late morning after a package arrives, but will poke at it some more with the debugger |
| 06:47.41 | brlcad | addressing a couple other bugs at the moment too |
| 06:49.36 | *** join/#brlcad Stattrav (~Stattrav@122.172.12.71) |
| 06:49.36 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) |
| 06:52.21 | CIA-77 | BRL-CAD: 03brlcad * r43598 10/brlcad/trunk/TODO: |
| 06:52.21 | CIA-77 | BRL-CAD: bio.h header was de-netted, bin.h added. found a few other show-stoppers, |
| 06:52.21 | CIA-77 | BRL-CAD: though. nmg is busted due to a host/net bug. have to test tie optimization for |
| 06:52.21 | CIA-77 | BRL-CAD: release notes (as it's the major highlight). search implements '.' so it needs |
| 06:52.21 | CIA-77 | BRL-CAD: a quick test. and finally, verify red/ted work since tom is reporting a failure |
| 06:52.22 | CIA-77 | BRL-CAD: in .2 with red. |
| 07:15.15 | brlcad | it looks like the tclvar hook into rt_bot_mintie isn't sticking if set from mged |
| 07:16.59 | brlcad | heh, got it to go in and it crashed |
| 07:17.32 | brlcad | (during prep) |
| 07:20.10 | brlcad | aha, possibly completely unrelated! |
| 07:33.36 | brlcad | ooor, maybe it is related.. |
| 07:36.53 | yukonbob | getting build failure on ./src/libbu/bitv.c, w/ "array subscript has type 'char'" at lines 251, 255... |
| 07:37.03 | yukonbob | also hitting hay.... ciao. |
| 07:44.16 | CIA-77 | BRL-CAD: 03brlcad * r43599 10/brlcad/trunk/src/librt/primitives/bot/btg.c: not sure if this is right, but it takes care of rt bombing out with a bad matrix during do_ae() due to zero-sized bounding box |
| 07:53.02 | CIA-77 | BRL-CAD: 03brlcad * r43600 10/brlcad/trunk/src/rt/do.c: add a sanity check in case code ends up computing an empty bounding box, so we don't bomb out during bn_mat_inv(). make sure the viewsize is at least not zero. |
| 07:53.59 | brlcad | that takes care of the crash and bomb, but still no picture, reports invalid segments |
| 08:14.35 | *** join/#brlcad epileg (~epileg@unaffiliated/epileg) |
| 10:24.35 | *** join/#brlcad packrat (~packrator@99-67-225-40.lightspeed.livnmi.sbcglobal.net) |
| 11:28.17 | CIA-77 | BRL-CAD: 03starseeker * r43601 10/brlcad/branches/cmake/CMakeLists.txt: New policy in 2.8.4 - we prefer the old behavior for now |
| 11:31.39 | CIA-77 | BRL-CAD: 03starseeker * r43602 10/brlcad/branches/cmake/src/other/libutahrle/ (4 files in 2 dirs): Can't lean on the BRL-CAD checks for subdirs now - utah needs M_LIBRARY |
| 11:33.40 | CIA-77 | BRL-CAD: 03starseeker * r43603 10/brlcad/branches/cmake/ (9 files in 4 dirs): MFC r43602 |
| 11:36.33 | dloman | Merning all |
| 11:37.04 | CIA-77 | BRL-CAD: 03starseeker * r43604 10/brlcad/trunk/src/rt/do.c: Comment needs terminating |
| 11:43.35 | CIA-77 | BRL-CAD: 03starseeker * r43605 10/brlcad/trunk/src/librt/primitives/bot/tie_kdtree.c: gentoo seems to want cmax initialized - hopefully max is ok... |
| 11:45.59 | CIA-77 | BRL-CAD: 03starseeker * r43606 10/brlcad/trunk/src/librt/primitives/dsp/dsp.c: initalize points in dsp |
| 11:48.07 | *** join/#brlcad b0ef (~b0ef@157.26.202.84.customer.cdi.no) |
| 11:54.28 | starseeker | O.o |
| 11:54.34 | starseeker | src/librt/primitives/extrude/extrude.c: In function ?classify_sketch_loops?: |
| 11:54.34 | starseeker | src/librt/primitives/extrude/extrude.c:394: error: ?center2d? may be used uninitialized in this function |
| 11:54.37 | starseeker | src/librt/primitives/extrude/extrude.c:1509: note: ?center2d? was declared here |
| 11:54.39 | starseeker | src/librt/primitives/extrude/extrude.c:394: error: ?rb? may be used uninitialized in this function |
| 11:54.42 | starseeker | src/librt/primitives/extrude/extrude.c:1507: note: ?rb? was declared here |
| 11:54.43 | dloman | ``Erik: Hey man, is there still a tiny proxy running somewhere on seans box? |
| 11:54.45 | starseeker | src/librt/primitives/extrude/extrude.c:394: error: ?ra? may be used uninitialized in this function |
| 11:54.48 | starseeker | src/librt/primitives/extrude/extrude.c:1507: note: ?ra? was declared here |
| 11:54.49 | starseeker | that's got me puzzled |
| 11:55.04 | dloman | that is pretty wierd? |
| 11:55.11 | dloman | oops :/ |
| 11:56.01 | starseeker | unless I'm missing something, line 394 doesn't have center2d in it |
| 12:02.32 | dloman | starseeker: in fn isect_2D_loop_ray(..) on line 1545, isect_line_earc(..) is called and center2d is passed in. |
| 12:03.25 | dloman | isect_line_earc(..) in turn calls isect_line2(..), again passing in center2d |
| 12:03.39 | starseeker | but at that point... why is it reporting uninitalized? |
| 12:03.45 | dloman | and line 394 is in isect_line2_ellipse() |
| 12:04.22 | starseeker | I tried a V2SET on center2d right after it's declared, and it changes nothing |
| 12:05.31 | dloman | ....so that means what? center2d is initialized? |
| 12:06.18 | starseeker | it means I don't know what to do to convince the compiler it is initialized |
| 12:07.57 | starseeker | even better, it's only on the static build |
| 12:08.05 | starseeker | they dynamic lib succeeds |
| 12:08.44 | dloman | nice. |
| 12:09.01 | dloman | well, outside of tracing things visually, im of no use :) |
| 12:09.50 | starseeker | the only difference I see in the extruce.o compile lines is the presence of -Dlibrt_EXPORTS in the dynamic line |
| 12:11.01 | starseeker | no, that's not it... |
| 12:11.03 | starseeker | what am I missing |
| 12:14.24 | ``Erik | <PROTECTED> |
| 12:14.29 | starseeker | oh -fPIC |
| 12:16.41 | starseeker | I don't believe it |
| 12:16.45 | starseeker | -fPIC is the difference |
| 12:16.51 | starseeker | why would that matter??? |
| 12:17.52 | starseeker | gcc version 4.4.5 |
| 12:23.14 | ``Erik | -fPIC instructs the linker to generate position independent code |
| 12:23.27 | ``Erik | relative jmp's instead of absolute, etc |
| 12:23.50 | ``Erik | gcc -E and gcc -S might be handy if you wanna figure it |
| 12:25.28 | CIA-77 | BRL-CAD: 03starseeker * r43607 10/brlcad/trunk/src/librt/primitives/nmg/nmg_misc.c: More initializations (why doesn't VSETALL get old_pl[3]??) |
| 12:26.45 | CIA-77 | BRL-CAD: 03starseeker * r43608 10/brlcad/trunk/src/librt/primitives/nmg/nmg_plot.c: another initialization |
| 12:27.32 | CIA-77 | BRL-CAD: 03erikgreenwald * r43609 10/brlcad/trunk/src/librt/primitives/bot/tie_kdtree.c: force cmin/cmax using VSETALL. This should be unnecessary, but some compiler/OS combinations aren't smart enough to catch "int a; if(something) a=0; else a=1;" and assume a is unset. |
| 12:28.15 | CIA-77 | BRL-CAD: 03starseeker * r43610 10/brlcad/trunk/src/librt/primitives/nmg/nmg_misc.c: whoops, plane not point |
| 12:28.28 | ``Erik | the amin vs min in btg.c was important, they're different numbers :/ (actual geometric min vs fpu friendly padding min, iirc) |
| 12:33.41 | CIA-77 | BRL-CAD: 03starseeker * r43611 10/brlcad/trunk/src/librt/search.c: initialize new to NULL |
| 12:44.32 | CIA-77 | BRL-CAD: 03starseeker * r43612 10/brlcad/branches/cmake/CMakeLists.txt: Hmm - how did this not trigger before? make sure HAVE_STRINGS_H actually makes it into brlcad_config.h |
| 12:54.01 | CIA-77 | BRL-CAD: 03starseeker * r43613 10/brlcad/branches/cmake/src/librt/CMakeLists.txt: I can't believe it, but without -fPIC in the static build possibly uninitialized warnings appear with gcc 4.4.5 on gentoo. Only on extrude.c |
| 12:58.02 | CIA-77 | BRL-CAD: 03starseeker * r43614 10/brlcad/trunk/src/liboptical/sh_toyota.c: another initialization |
| 13:02.59 | ``Erik | hm, musta lost that in merge *sigh* |
| 13:05.18 | CIA-77 | BRL-CAD: 03erikgreenwald * r43615 10/brlcad/trunk/ (include/tie.h src/librt/primitives/bot/tie_kdtree.c): remove obsolete kdtree cache stuff |
| 13:09.38 | CIA-77 | BRL-CAD: 03starseeker * r43616 10/brlcad/branches/cmake/src/conv/intaval/CMakeLists.txt: Use BRLCAD_ADD_EXEC for tgf-g |
| 13:10.26 | brlcad | I'l make sure you have access to the new server sometime this week |
| 13:11.02 | brlcad | after release is posted |
| 13:12.04 | CIA-77 | BRL-CAD: 03starseeker * r43617 10/brlcad/trunk/src/conv/nastran-g.c: initializations |
| 13:13.43 | CIA-77 | BRL-CAD: 03erikgreenwald * r43618 10/brlcad/trunk/ (3 files in 2 dirs): revive the actual min/max (amin/amax) |
| 13:14.12 | brlcad | starseeker: zombie is back to life: https://sourceforge.net/tracker/?func=detail&atid=640802&aid=3197208&group_id=105292 |
| 13:18.13 | CIA-77 | BRL-CAD: 03starseeker * r43619 10/brlcad/branches/cmake/src/conv/iges/CMakeLists.txt: Use BRLCAD_ADDEXEC for iges |
| 13:21.06 | CIA-77 | BRL-CAD: 03starseeker * r43620 10/brlcad/trunk/src/conv/iges/ (conv_drawings.c trimsurf.c): initializations for iges |
| 13:21.18 | starseeker | brlcad: oh god |
| 13:21.22 | starseeker | ok, I'll have a look |
| 13:26.51 | CIA-77 | BRL-CAD: 03starseeker * r43621 10/brlcad/trunk/src/librtserver/rtserverTest.c: Quell a few warnings - rts_clean doesn't seem to be defined anywhere |
| 13:31.56 | CIA-77 | BRL-CAD: 03starseeker * r43622 10/brlcad/trunk/src/anim/anim_hardtrack.c: initializations for anim |
| 13:32.09 | CIA-77 | BRL-CAD: 03erikgreenwald * r43623 10/brlcad/trunk/src/librt/primitives/bot/btg.c: woops, point_t, not TIE_3 |
| 13:36.36 | CIA-77 | BRL-CAD: 03starseeker * r43624 10/brlcad/trunk/src/burst/grid.c: initialization for burst |
| 13:37.56 | CIA-77 | BRL-CAD: 03erikgreenwald * r43625 10/brlcad/trunk/include/raytrace.h: set RT_DEFAULT_MINTIE to 0 (disabled) |
| 13:41.46 | CIA-77 | BRL-CAD: 03starseeker * r43626 10/brlcad/trunk/src/fbed/fbed.c: initializations for fbed |
| 13:42.25 | brlcad | wishes one of his compilation environments would have picked up those warnings |
| 13:42.54 | CIA-77 | BRL-CAD: 03starseeker * r43627 10/brlcad/trunk/src/gtools/remapid.c: initialization for remapid |
| 13:46.12 | CIA-77 | BRL-CAD: 03starseeker * r43628 10/brlcad/trunk/src/shapes/picket_fence.c: initialize matrix |
| 13:48.45 | CIA-77 | BRL-CAD: 03starseeker * r43629 10/brlcad/trunk/src/util/pixborder.c: initialize for pixborder |
| 13:49.14 | starseeker | finally. there we go |
| 13:50.18 | brlcad | ~starseeker++ |
| 13:50.23 | brlcad | thanks |
| 13:50.53 | starseeker | dingnabbit. red problem confirmed |
| 13:55.11 | CIA-77 | BRL-CAD: 03starseeker * r43630 10/brlcad/branches/cmake/CMakeLists.txt: Don't care about warnings on the time delta utilties |
| 13:57.15 | starseeker | brlcad: interesting thing about those warnings - they popped up only when I did a release build - optimizations flags, etc |
| 13:57.31 | starseeker | (well, some of them) |
| 13:59.35 | ``Erik | um, lots of funky things go on depending on debug.. like 'HIDDEN' is "" on debug and "static" in release, etc |
| 14:03.25 | starseeker | actually, cancel that - red problem is NOT confirmed, with one possible exception |
| 14:03.36 | starseeker | you can't rename the region during a red command |
| 14:03.43 | starseeker | other than that, I have success using it here |
| 14:05.43 | starseeker | IIRC, I decided not to allow changing the object name during a red - it complicated things a bit |
| 14:07.08 | brlcad | starseeker: heh, I just saw yesterday's commit -- didn't realize that you just deleted all of those variables :) |
| 14:07.53 | ``Erik | scrum ~1300 |
| 14:07.59 | ``Erik | er, ~1330 |
| 14:08.07 | brlcad | if it can't be changed, it should probably not be in the file ... or it should allow it ;) |
| 14:08.27 | brlcad | otherwise people are going to try |
| 14:08.51 | brlcad | shouldn't be too hard to allow, just keep a copy before and after |
| 14:08.59 | starseeker | ugh |
| 14:09.00 | brlcad | apply edits to old, and if name changed, rename |
| 14:10.26 | brlcad | interesting approach to search "." .. I think I like it |
| 14:18.54 | starseeker | brlcad: adding renaming support will take a little time - is that something you want prior to release? |
| 14:22.37 | starseeker | brlcad: if you decide search looks good we can nuke the TODO item for that particular tweak and update NEWS - I was waiting until I was sure it was "done" |
| 14:25.07 | starseeker | saddles up |
| 14:40.55 | brlcad | starseeker: if it works, then that's stable for release -- but would you follow up with tom to see if that's what he was running into? |
| 14:41.45 | brlcad | the search todo is my action to look at since I had that particular request |
| 15:37.06 | *** join/#brlcad PrezKennedy (MK@whitecalf.net) |
| 15:38.17 | CIA-77 | BRL-CAD: 03tbrowder2 * r43631 10/brlcad/trunk/src/tab/script.l: eliminate unused function warning |
| 15:55.28 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) |
| 16:33.41 | starseeker | brlcad: I thought I emailed him for more details, but I see it didn't pop up... |
| 16:34.29 | starseeker | huh - sent over 2 hours ago |
| 16:43.04 | CIA-77 | BRL-CAD: 03starseeker * r43632 10/brlcad/branches/cmake/ (21 files in 17 dirs): MFC -r43631 |
| 16:46.58 | CIA-77 | BRL-CAD: 03starseeker * r43633 10/brlcad/branches/cmake/CMakeLists.txt: Lovely - check cmake version before we set CMP0017 |
| 16:55.18 | starseeker | heh: http://spacewar.oversigma.com/ |
| 16:55.47 | starseeker | also http://spacewar.oversigma.com/html5/ |
| 16:56.30 | starseeker | ``Erik: whadya think, original BRL-CAD on a PDP1 emulator in html5? :-P |
| 17:00.45 | ``Erik | I don't think it ever ran on a pdp1... mebbe an 11/70, though it might need a vax11/780 by the time it did anything interesting |
| 17:01.09 | ``Erik | um, jove appeared in the repo in '83, then mid 85 is when really basic raytracing started happening |
| 17:02.00 | ``Erik | the date flag to svn up makes for some amusing code spelunking if ya find yourself with a bit of free time :D |
| 17:03.48 | starseeker | heh |
| 17:08.12 | CIA-77 | BRL-CAD: 03brlcad * r43634 10/brlcad/trunk/src/librt/primitives/nmg/nmg_plot.c: use VINIT_ZERO macro symbol instead of {0.0, 0.0, 0.0} for point_t/vect_t. less error-prone, less typing, and supports converion to fixed point arithmetic down the road. |
| 17:08.39 | starseeker | brlcad: oops, sorry - there'll be a few of those |
| 17:09.13 | CIA-77 | BRL-CAD: 03brlcad * r43635 10/brlcad/trunk/src/librt/primitives/nmg/nmg_misc.c: same with HINIT_ZERO for 4D vector initialization. |
| 17:43.32 | CIA-77 | BRL-CAD: 03erikgreenwald * r43636 10/brlcad/trunk/src/librt/primitives/bot/btg.c: rewrite prep to be correct and only make one tie_push call |
| 18:26.59 | CIA-77 | BRL-CAD: 03erikgreenwald * r43637 10/brlcad/trunk/src/librt/primitives/bot/tie_kdtree.c: reincorporate some of the lost changes |
| 21:23.41 | CIA-77 | BRL-CAD: 03starseeker * r43638 10/brlcad/branches/cmake/src/mged/points/CMakeLists.txt: Ugh. Go vanilla with the cflags in points dir until it's clear what the issue is |
| 21:57.55 | starseeker | hmm http://www.research.ibm.com/haifa/projects/verification/fpgen/papers/plsrange.pdf |
| 22:02.19 | starseeker | http://portal.acm.org/citation.cfm?id=1048200 |
| 22:03.19 | starseeker | maybe we can recruit this guy :-) http://johnsogg.blogspot.com/2011_01_01_archive.html |
| 22:16.12 | starseeker | http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.112.3073 |
| 22:18.41 | starseeker | http://code.google.com/p/sketchsolve/ |
| 22:28.59 | starseeker | hah, cool - not related directly to geometric constrants but possibly useful for other things: http://adaptagrams.sourceforge.net/ |
| 22:38.04 | starseeker | reflects that the constraint system is going to require some paper diving... |
| 22:42.20 | starseeker | hmm http://www.cs.purdue.edu/homes/cmh/electrobook/intro.html |
| 23:16.50 | CIA-77 | BRL-CAD: 03starseeker * r43639 10/brlcad/branches/cmake/src/conv/step/CMakeLists.txt: Use BRLCAD_ADDEXEC for step-g |
| 23:17.47 | CIA-77 | BRL-CAD: 03starseeker * r43640 10/brlcad/branches/cmake/src/libpc/CMakeLists.txt: Yes boost, we know, be quiet already |
| 23:21.03 | CIA-77 | BRL-CAD: 03starseeker * r43641 10/brlcad/branches/cmake/ (5 files in 3 dirs): |
| 23:21.03 | CIA-77 | BRL-CAD: Rework handling of compile flags - no longer setting strict flags at the |
| 23:21.03 | CIA-77 | BRL-CAD: individual target levels, since BRL-CAD is going fully strict. There are a few |
| 23:21.03 | CIA-77 | BRL-CAD: local overrides for issues not yet resolved, but the default will now be that |
| 23:21.03 | CIA-77 | BRL-CAD: anything new automatically gets the struct flags with all warnings unless |
| 23:21.03 | CIA-77 | BRL-CAD: options are turned off. |
| 23:25.29 | CIA-77 | BRL-CAD: 03starseeker * r43642 10/brlcad/branches/cmake/misc/CMake/test_srcs/time.c.in: Add the zero for day, not just month. |
| 23:59.25 | *** join/#brlcad yukonbob_ (~bch@20-144.wireless.kamloops.net) |