IRC log for #brlcad on 20100119

00:14.09CIA-11BRL-CAD: 03starseeker * r37289 10/brlcad/trunk/doc/docbook/lessons/es/ (10 files in 2 dirs): Commit Spanish translations of MGED lessons 3 and 6, courtesy of Jesica Giudice.
01:56.04brlcadhola Nohla
02:05.55*** join/#brlcad akafubu (n=akafubu@unaffiliated/akafubu)
04:18.54*** join/#brlcad ibot_ (i=ibot@rikers.org)
04:18.55*** topic/#brlcad is BRL-CAD Open Source Solid Modeling || http://brlcad.org || http://sf.net/projects/brlcad || #brlcad logs at http://ibot.rikers.org/%23brlcad/ || Happy Open Source Anniversary! (December 21st) || Release 7.16.4 in prep, should be posted 20100114
04:22.28starseekerdigs for info on how to make a custom automake primary (like _SOURCES, _DATA, etc.)
04:27.40starseekermight not need to go that deep to get the TEA_ADD_SOURCES logic out of configure.in, but I'm a little worried that it's there in the first place... why...
04:50.35*** join/#brlcad ibot (i=ibot@rikers.org)
04:50.35*** topic/#brlcad is BRL-CAD Open Source Solid Modeling || http://brlcad.org || http://sf.net/projects/brlcad || #brlcad logs at http://ibot.rikers.org/%23brlcad/ || Happy Open Source Anniversary! (December 21st) || Release 7.16.4 in prep, should be posted 20100114
04:51.12CIA-11BRL-CAD: 03starseeker * r37290 10/brlcad/trunk/src/other/ (8 files in 3 dirs): (log message trimmed)
04:51.12CIA-11BRL-CAD: tkhtml subconfigure test - it's unlikely this is correct, but somewhat
04:51.13CIA-11BRL-CAD: surprisingly it built and installed on a Gentoo Linux box, so since we're past
04:51.13CIA-11BRL-CAD: the 7.16.4 tag stick it in to give it a try on a few other systems - maybe it
04:51.13CIA-11BRL-CAD: can be polished up. Constitutes a minimal set of changes to try to get the
04:51.15CIA-11BRL-CAD: original configure.in and Makefile.in files to act like configure.ac and
04:51.17CIA-11BRL-CAD: Makefile.am files, so they don't look much like other configure.ac and
05:01.34brlcadheh, release generally isn't over until the include/conf is bumped to an odd, but close enough
05:01.48starseekerah, crud
05:01.48brlcadjust finished validating stable
05:01.52brlcadno worries
05:01.59brlcadstable gets tagged
05:02.08brlcadbump it
05:02.17starseekernods
05:03.15starseekerscrolls up looking for the commit that bumped all the version numbers to 16.4...
05:03.43starseekerah, there it is
05:11.37brlcadnews, readme, include/conf are the important ones
05:11.57brlcadthe windows .bat files are the additional annoyances until they can be tested
05:12.00starseekerprods CIA...
05:12.01CIA-11BRL-CAD: 03starseeker * r37291 10/brlcad/trunk/ (7 files in 6 dirs): Bump version numbers to 7.16.5 - will be tagging off of STABLE, so back to odd number on trunk.
05:12.05starseekerah :-)
05:12.24brlcadmisc isn't really necessary, but hadn't touched it in a while
05:12.31starseekerhehe
05:13.13starseekereyes the fans on his computer case... please don't die...
05:13.43brlcadREADME and NEWS aren't right, should be next expected release number
05:13.53starseekerah
05:14.05brlcadeverything else is good
05:14.48CIA-11BRL-CAD: 03starseeker * r37292 10/brlcad/trunk/ (NEWS README): whoops - set NEWS and README to the release to come.
05:19.53brlcadjesica's lessons should be news items too :)
05:20.01brlcadgets back to what he was doing
05:23.06starseekerNohla: there's an incentive to get committing to svn ;-)
05:27.46starseekerNohla: when we both have some time, I need to walk you through the svn commit process
06:31.51CIA-11BRL-CAD: 03brlcad * r37293 10/brlcad/tags/rel-7-16-4/: tagging release 7.16.4 (belated tag from 20100115)
06:50.48CIA-11BRL-CAD: 03brlcad * r37294 10/brlcad/trunk/HACKING: change to the new dir
07:43.24CIA-11BRL-CAD: 03d_rossberg * r37295 10/rt^3/tags/rel-7-16-4/: tag the C++ core interface with the corresponding BRL-CAD version 7.16.4
11:21.50*** join/#brlcad mafm_ (n=mafm@121.Red-81-32-105.dynamicIP.rima-tde.net)
13:23.15*** join/#brlcad R0b0t1 (n=Enigma@unaffiliated/r0b0t1)
14:18.16*** join/#brlcad _clock_ (n=_sushi_@80-218-244-105.dclient.hispeed.ch)
15:39.16CIA-11BRL-CAD: 03starseeker * r37296 10/brlcad/trunk/src/other/tkhtml3/ (4 files in 2 dirs): More tkhtml3 build tweaks - these work on OSX and Redhat, but clearly more work to do
16:06.20*** join/#brlcad Ralith (n=ralith@69.90.48.97)
16:08.20CIA-11BRL-CAD: 03brlcad * r37297 10/brlcad/trunk/NEWS: ah, runtime toggling of display managers WAS already announced in 7.16.2; remove.
16:22.09CIA-11BRL-CAD: 03starseeker * r37298 10/brlcad/trunk/src/other/tkhtml3/Makefile.am: Tweak comments to refer to Makefile.am.
16:41.49*** join/#brlcad ``Erik_ (n=erik@c-69-140-109-104.hsd1.md.comcast.net)
16:41.52*** join/#brlcad Computer (n=Computer@unaffiliated/computer) [NETSPLIT VICTIM]
16:41.52*** join/#brlcad d-lo (n=claymore@63.246.136.16) [NETSPLIT VICTIM]
16:42.06*** join/#brlcad Ralith (n=ralith@69.90.48.97)
16:46.34CIA-11BRL-CAD: 03brlcad * r37299 10/brlcad/trunk/HACKING: the rock linux package maintainer, stefan fiedler, is not responsive with no valid point-of-contact. remove from release notifications.
17:01.24starseekeraaand that hack job blows distcheck all to hell
17:02.02CIA-11BRL-CAD: 03irpguardian * r37300 10/brlcad/trunk/src/rt/view.c:
17:02.02CIA-11BRL-CAD: Added two more functions to view in relation to the heat-graph. The first,
17:02.02CIA-11BRL-CAD: timetable_input is a reworked timetable_init that focuses only on inputting
17:02.02CIA-11BRL-CAD: values into the timetable array. Timetable_init has been reworked to only
17:02.02CIA-11BRL-CAD: initialize the timetable array. The second, is timetable_process, which will
17:02.04CIA-11BRL-CAD: do the normalization and placing pixels into the file buffer.
17:11.58CIA-11BRL-CAD: 03starseeker * r37301 10/brlcad/trunk/src/other/tkhtml3/Makefile.am: Can't call out the headers in this fashion - they're in src or not there yet. Let autotools do the distclean rule so it doesn't complain, and add an EXTRA_DIST to grab src since so far src isn't yet a proper subdir.
17:17.18brlcadstarseeker: perhaps you can provide some tutelage to irpguardian .. he's getting very magic number lazy
17:17.40brlcadassuming you're within earshot of course, or anyone for that matter
17:18.00brlcadshouldn't be an afterthought, particularly for new code ..
17:18.03brlcadpart of coding complete
17:18.21starseekerhe's throwing in magic numbers or ignoring ours?
17:18.28brlcadthrowing in new ones
17:18.34starseekerah
17:18.40starseekerchecks his last commit
17:19.02brlcadmax dimensions on image sizes he can compute his lighting model heat value
17:19.11brlcad4096x4096
17:19.32brlcadshould just init to null and alloc what he needs if null
17:19.46brlcadanother is 1000000 as a max assumed compute time
17:19.53brlcadarbitrary pointless limit
17:19.57starseekeryeah, saw that one
17:20.03starseekerheads over...
17:20.04brlcadif he really wants to clamp, INFINITY
17:20.22brlcador MAX_DBL or whatever it is
17:21.05brlcadhe should also not have any more static vars than are absolutely necessary
17:21.13brlcadhe's made several things static that do not need to be static
17:25.50starseekerpointed him to the MAX* stuff, and has him looking into dynamic image sizes
17:27.35CIA-11BRL-CAD: 03starseeker * r37302 10/brlcad/trunk/src/other/tkhtml3/Makefile.am: Oh yeah, bring doc along for the ride too.
17:41.57brlcaddiscovers that all of openNURBS point classes use exact floating point comparisons, except 4d
17:42.23CIA-11BRL-CAD: 03irpguardian * r37303 10/brlcad/trunk/src/rt/view.c:
17:42.23CIA-11BRL-CAD: Timetable is now dynamically sized via malloc, instead of hard coded to have
17:42.23CIA-11BRL-CAD: size 4096. Also, max and min times are now properly initialized.
17:45.07brlcadheh, that probably won't work .. a_x and a_y are ray coordinates, not image dimensions
17:45.16brlcadshould just pass the size into init
17:46.19brlcadnot that there should be an explicit init call either, though ..
17:46.21brlcadlets him figure it out
17:54.35CIA-11BRL-CAD: 03starseeker * r37304 10/brlcad/trunk/src/other/tkhtml3/Makefile.am: Ah, right, gonna have to specify the .o files to clean up - take a first stab.
18:16.18CIA-11BRL-CAD: 03irpguardian * r37305 10/brlcad/trunk/src/rt/view.c:
18:16.18CIA-11BRL-CAD: timetable_init now takes the Frame width & height, as opposed to the current
18:16.18CIA-11BRL-CAD: pixel location.
18:19.06brlcadhehe
18:19.13brlcadthat comment lies
18:26.17CIA-11BRL-CAD: 03brlcad * r37306 10/brlcad/trunk/src/rt/view.c: ws, style, consistency cleanup and register keyword elimination
18:32.41CIA-11BRL-CAD: 03brlcad * r37307 10/brlcad/trunk/src/rt/view.c: some notes about FIXME items including one HACKING code convention failure (use of malloc()).
18:39.51*** join/#brlcad Ralith (n=ralith@69.90.48.97)
18:45.08CIA-11BRL-CAD: 03irpguardian * r37308 10/brlcad/trunk/src/rt/view.c:
18:45.08CIA-11BRL-CAD: Moved timeTable_input time funtions into timeTable_process
18:45.08CIA-11BRL-CAD: Made timeTable process not access a global variable inproperly.
18:56.07CIA-11BRL-CAD: 03irpguardian * r37309 10/brlcad/trunk/src/rt/view.c: Changed some variable settings in timetable_process
19:02.45brlcadstarseeker: did you have a pdf or screenshot of one of the spanish translations?
19:08.08CIA-11BRL-CAD: 03irpguardian * r37310 10/brlcad/trunk/src/rt/view.c:
19:08.08CIA-11BRL-CAD: Changed mallocs to bu_mallocs, and added timeTable_free() for freeing up
19:08.08CIA-11BRL-CAD: allocated space for timeTable, by using bu_free()
19:13.17CIA-11BRL-CAD: 03brlcad * r37311 10/brlcad/trunk/src/rt/view.c: calling malloc was the only reason it was blocked out. remove ifdef/svn diff! note.
19:17.22brlcadstarseeker: never mind, I found a copy
19:27.39*** join/#brlcad akafubu (n=akafubu@unaffiliated/akafubu)
19:27.51CIA-11BRL-CAD: 03erikgreenwald * r37312 10/brlcad/trunk/src/rt/view.c: Add missing parameter to bu_mallocs. Dereference last timeTable_input call arg to match function definition (hope that's right...)
19:30.57CIA-11BRL-CAD: 03brlcad * r37313 10/brlcad/trunk/src/librt/db5_alloc.c:
19:30.57CIA-11BRL-CAD: if the user hasn't built a directory yet before attempting to write out
19:30.57CIA-11BRL-CAD: geometry, build one for them instead of just failing (as dbi_eof will be -1).
19:30.57CIA-11BRL-CAD: this has been observed on simple snippets that create an object and try to write
19:30.58CIA-11BRL-CAD: it out.
19:31.56brlcadguess I shoulda compiled
19:37.09``Eriktesting is for wimps :D
19:50.59*** join/#brlcad Ralith (n=ralith@69.90.48.97)
20:17.04starseekerheh, cool - that tkhtml3 stuff just made it all the way through distcheck
20:43.17*** join/#brlcad louipc (n=louipc@archlinux/trusteduser/louipc)
20:58.49CIA-11BRL-CAD: 03starseeker * r37314 10/brlcad/branches/dmtogl/src/other/tkhtml3/ (Makefile.am Makefile.in configure.ac src/Makefile.am): Tantalizingly closer to a proper working autotools/TEA hybrid tkhtml build - this builds something and it doesn't list the files in configure.ac, but it's not yet got all the 'correct' names, versions, etc. plugged in.
20:59.16CIA-11BRL-CAD: 03starseeker * r37315 10/brlcad/branches/dmtogl/src/other/tkhtml3/configure.in: Won't need configure.in in this version any more...
21:14.55CIA-11BRL-CAD: 03erikgreenwald * r37316 10/brlcad/trunk/src/librt/primitives/nmg/nmg_tri_mc.c: re-source table. (hopefully not losing/damaging data this time)
21:28.06*** join/#brlcad talcite (n=matthew@dhcp-143-177.mcme-students.carleton.ca)
21:28.30*** join/#brlcad R0b0t1 (n=Enigma@unaffiliated/r0b0t1)
22:05.57*** join/#brlcad Ralith_ (n=ralith@69.90.48.97)
22:06.43CIA-11BRL-CAD: 03erikgreenwald * r37317 10/brlcad/trunk/src/librt/primitives/nmg/nmg_tri_mc.c: Print full triangle information. Move nmg_cmface into place, make some comments about structure/tasks, remove dead code.
22:12.43starseekerbrlcad: is there anything in recent commit history that could be causing bot raytracing to fail?
22:12.51starseekerspecifically unoriented bots?
22:16.43brlcadstarseeker: plenty of librt and libbn changes have that potential, do you know if previous release had issue?
22:17.02brlcadnothing specific comes to mind
22:17.14starseeker7.16.2 is ok
22:20.00starseekerso far seeing it only in mged, which I suppose includes ged_rtcheck in the list of things to check
22:23.31starseekerdingnabbit, why do the bots alway have to be so blasted fragile?
22:31.09``Erikcuz it's amusing watching ya wig out? :D
22:31.37starseekerwould be less wigged if he hadn't been making sudden progress on the TEA/autotools stuff when the call came through
22:32.23``Erikthat's a risk you decided to take on yourself when you went and started making progress... O.o *duck* D
22:32.26``Erik:D
22:32.31starseekerheh
22:33.11starseekerwell, I'll probably go back and find out I was doing it all wrong again - my track record with autotools puts the probability of such an event at 0.9+
22:34.51starseekerthinks college should replace their intro programming course with a "doing crap with common unix tools" course - teach bash, sed, awk, grep, etc. as a semester topic
22:36.13starseeker'course, I suppose that's too much power to be put in casual hands...
22:39.19*** join/#brlcad Nohla (i=a8e2b37b@gateway/web/freenode/x-hhaqyxbjvosouglr)
22:41.30brlcadstarseeker: what do you mean "only in mged"?  they only raytrace from if you call rt from within mged?
22:41.47brlcads/raytrace from/raytrace wrong/
22:43.36starseekerrtcheck in mged gives the bad magic failures
22:43.49starseekerfrom the command line it just complains about feeding binary data to the terminal
22:44.09starseekerand if I give it a -o file to dump that into, it seems to run
22:44.21brlcadrtcheck complains about binary data to terminal?
22:44.26brlcadredirect
22:44.32brlcadrtcheck's default output is plot data
22:44.36starseekeron my Mac
22:45.01brlcad.. rtcheck or rt ..
22:45.08starseekerinitial report was rtcheck
22:45.25starseekerapparently all bot raytracing is foobared right now though
22:45.37starseekerat least, unoriented
22:45.43brlcadbad magic could be a simple bad badmagic check that was added as part of quelling warnings
22:46.46starseekeris getting set up to try and do a binary search to hone in on it
22:47.12brlcadif rtcheck in mged gave a badmagic failure, there should be a bomb log
22:47.19brlcadthat will point directly at the check
22:48.54starseekerletsee... do_run in worker.c:714, called from do-frame in do.c:818
22:49.20starseekercalled from cmd_end, do.c:314
22:49.35starseekercalled from rt_do_cmd, cmd.c:159
22:58.32starseekerok, looks like it was failing before 37015...
23:00.26*** join/#brlcad jesica__ (n=jesica@168.226.179.123)
23:11.51*** join/#brlcad ``Erik (n=erik@c-69-140-109-104.hsd1.md.comcast.net) [NETSPLIT VICTIM]
23:12.27*** join/#brlcad CIA-82 (n=CIA@208.69.182.149)
23:20.22CIA-82BRL-CAD: 03starseeker * r37318 10/brlcad/branches/dmtogl/src/other/tkhtml3/src/Makefile.am: Playing around with tkhtml3 Makefile.am some more...
23:22.49starseekerbetween 36500 and 36466
23:27.37brlcadworker.c:714 is the wrong thread
23:27.49brlcadshould either be a separate section, or further down the report
23:28.13brlcadthat's the main thread that kicks off processes, you need to find the thread that crashed
23:33.11``Erikfind the bad magic macro line and break on it? O.o :D
23:43.12*** join/#brlcad Ralith (n=ralith@69.90.48.97)
23:54.35brlcadif you know the badmagic line, it should be obvious
23:55.03brlcadthe CKMAG functions don't know anything about types, so it's a runtime halt
23:55.53starseekerERROR: bad pointer x71899e0: s/b rt_bot_internal(x626f7472), was Unknown_Magic(x102), file ../../../brlcad/src/librt/primitives/bot/./g_bot_include.c, line 489
23:56.51brlcadyep, that'd be it
23:57.13brlcadnote type type of pointer passed and the macro being called
23:57.27starseekerah, wait - got a non-parallel backtrace
23:57.31starseekerrt_bot_unoriented_segs_double
23:57.34``Erikxglue hack, eck
23:57.42brlcadRT_BOT_CK_MAGIC() and "struct bot_specific *"
23:57.53brlcadRT_BOT_CK_MAGIC checks .. *drumroll*
23:57.56``Erik<-- wonders if he should redo the libtie stuff to use the xglue hack style opposed to his build hack style
23:58.03brlcadstruct rt_bot_internal *
23:58.31starseekerbut aren't we looking for how the bad magic got into the bot?
23:58.50brlcadstarseeker: not at all
23:58.57brlcadit's not a bad magic number
23:59.01brlcadit's a bad check
23:59.11brlcadthink of it this way..
23:59.40brlcadint main(int argc, char *argv[]) { RT_BOT_CK_MAGIC(argv); return 0; }
23:59.53brlcadthat will fail with a bad magic number

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.