IRC log for #brlcad on 20140820

00:01.05 *** join/#brlcad mihaineacsu (~mihaineac@92.85.29.79)
00:12.21 Notify 03BRL-CAD:starseeker * 62269 (brlcad/branches/dm-work/src/libdm/dm-qt.cpp brlcad/branches/dm-work/src/mged/dm-qt.c): Move qt MGED logic to libdm
00:31.30 Notify 03BRL-CAD:starseeker * 62270 brlcad/trunk/src/libbu/parallel.c: Fix build
00:52.50 *** join/#brlcad emagdalena (~emagdalen@57.159.165.83.dynamic.mundo-r.com)
01:15.42 *** join/#brlcad Zhao_Anqing (~clouddrif@183.157.160.31)
01:51.58 Notify 03BRL-CAD:brlcad * 62271 brlcad/trunk/include/bu/parallel.h: update and expand the documentation for the venerable bu_parallel() function with details on setting ncpu=0 and support for recursive invocation. formally deprecate bu_is_parallel().
01:55.38 Notify 03BRL-CAD:brlcad * 62272 brlcad/trunk/src/libbu/parallel.c: separate debug from logic
01:59.31 *** join/#brlcad FreezingCold (~FreezingC@135.0.41.14)
02:00.01 Notify 03BRL-CAD:brlcad * 62273 brlcad/trunk/src/libbu/parallel.c: basic arg sanity, make sure the callback function is not NULL
02:08.11 *** join/#brlcad Zhao_Anqing (clouddrift@222.205.3.53)
03:00.36 Notify 03BRL-CAD:brlcad * 62274 brlcad/trunk/src/libbu/parallel.c: oof, don't mix decls and code. have to replicate the arg check in order to come after the var decls for both the parallel and non-parallel ifdef sections.
03:24.25 Notify 03BRL-CAD:brlcad * 62275 brlcad/trunk/src/libbu/tests/CMakeLists.txt: add an initial simple set of unit tests for validating expected bu_parallel behavior.
03:57.26 Notify 03BRL-CAD:brlcad * 62276 brlcad/trunk/src/libbu/tests/bu_parallel.c: implement a couple more tests with considerably more collision potential, make the counter per-cpu in order to avoid using semaphores.
05:09.37 *** join/#brlcad gagan (~gagan@220.227.40.234)
05:11.56 Notify 03BRL-CAD:zhaoanqing * 62277 (brlcad/branches/nmgreorg/include/nurb.h brlcad/branches/nmgreorg/src/libnmg/fuse.c and 30 others): change functions' name in libnurb, make them start with nurb_
06:45.39 Notify 03BRL-CAD:brlcad * 62278 brlcad/trunk/src/libbu/tests/bu_parallel.c: narrowing in on a semaphore issue, looks to be related to thread tracking in ThreadLocal and the semaphores it uses.
06:59.25 *** join/#brlcad ries (~ries@D979EA84.cm-3-2d.dynamic.ziggo.nl)
07:51.27 *** join/#brlcad d_rossberg (~rossberg@66-118-151-70.static.sagonet.net)
08:19.19 *** join/#brlcad infobot (ibot@rikers.org)
08:19.19 *** topic/#brlcad is BRL-CAD || http://brlcad.org || logs: http://ibot.rikers.org/%23brlcad/ || GCI winners: Jacob Burroughs and Peter Amidon! || GSoC 2014 selections are announced! Thank you to all we got to work with. Remember that SOCIS is coming up right around the corner and you don't need a summer of code to get involved with open source.
08:26.56 *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
09:50.48 *** join/#brlcad albertcoder (~albertcod@202.164.53.117)
10:04.53 *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
10:10.49 *** join/#brlcad albertcoder (~albertcod@202.164.53.117)
10:32.31 *** join/#brlcad ries (~ries@D979EA84.cm-3-2d.dynamic.ziggo.nl)
11:59.55 *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
12:19.56 *** join/#brlcad FreezingCold (~FreezingC@135.0.41.14)
12:44.32 *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
13:15.48 *** join/#brlcad ries (~ries@D979EA84.cm-3-2d.dynamic.ziggo.nl)
13:16.50 Notify 03BRL-CAD:starseeker * 62279 (brlcad/trunk/src/conv/step/STEPWrapper.cpp brlcad/trunk/src/conv/step/STEPWrapper.h brlcad/trunk/src/conv/step/step-g/step-g.cpp): Write out log file, instead of to std::out. Tweak formatting
13:32.02 *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
13:38.05 *** join/#brlcad clock (~clock@77-58-143-135.dclient.hispeed.ch)
13:38.23 *** join/#brlcad Zhao_Anqing (~clouddrif@183.157.160.24)
13:39.24 Notify 03BRL-CAD Wiki:203.110.247.221 * 7718 /wiki/User:Ankeshanand/GSoC14/logs: /* Week 13 */
13:41.02 Notify 03BRL-CAD:brlcad * 62280 brlcad/trunk/src/libbu/semaphore.c: significantly rework the semaphore api to no longer require explicit initialization. we use static data to avoid library initialization statefulness and a common source of application dev frustration (parallel apps forgetting a bu_semaphore_init() resulted in obscure crashes and random behavior). also, avoiding heap-allocated data makes memory management
13:41.04 Notify and cleanup more simple. current semaphore limit is arbitrarily set at 1024 (two orders higher than our current known uses).
13:41.40 Notify 03BRL-CAD:indianlarry * 62281 brlcad/trunk/src/conv/step/step-g/OpenNurbsInterfaces.cpp: Added code to reparameterize face UV using surface 3d width and height. Finding occasions when the face UV is parametrrized narrow to the point of causing problems with the 2d pullback. Also includes some debugging hacks that should be CPP'ed out.
13:44.22 Notify 03BRL-CAD:brlcad * 62282 (brlcad/trunk/include/bu/parallel.h brlcad/trunk/include/raytrace.h): no longer need to call bu_semaphore_init()
13:46.38 *** join/#brlcad Izakey (~Isaac@195.24.220.16)
13:46.40 Notify 03BRL-CAD:brlcad * 62283 brlcad/trunk/CHANGES: no longer need to init semaphores
13:53.15 *** join/#brlcad albertcoder (~albertcod@101.215.140.199)
13:53.43 Notify 03BRL-CAD:brlcad * 62284 (brlcad/trunk/src/adrt/librender/camera.c brlcad/trunk/src/gtools/beset/fitness.c and 14 others): no longer need to manually initialize semaphores. they are auto-initialized on first access and expanded as used.
13:56.25 Notify 03BRL-CAD:brlcad * 62285 brlcad/trunk/src/libbu/semaphore.c: reinit is no longer needed, minimally impacting to remove (keeping explicit init for now)
14:06.58 *** join/#brlcad albertcoder (~albertcod@101.215.140.199)
14:07.26 *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
14:10.44 *** join/#brlcad Zhao_Anqing (~clouddrif@183.157.160.24)
14:15.57 *** join/#brlcad albertcoder (~albertcod@101.215.140.199)
14:18.27 Notify 03BRL-CAD:brlcad * 62286 brlcad/trunk/src/libbu/semaphore.c: muuuuuu
14:26.27 Notify 03BRL-CAD:starseeker * 62287 brlcad/trunk/src/libged/exists.c: Working on clearing a valgrind report - this isn't it, but go ahead and initiazlie anyway.
14:37.57 *** join/#brlcad albertcoder (~albertcod@101.215.140.199)
14:42.35 Notify 03BRL-CAD:ejno * 62288 brlcad/trunk/src/conv/3dm/conv3dm-g.cpp: append newline to output
15:08.54 Notify 03BRL-CAD:brlcad * 62289 brlcad/trunk/src/libbu/semaphore.c: more muuu!
15:18.23 Notify 03BRL-CAD:brlcad * 62290 brlcad/trunk/src/libbu/tests/bu_semaphore.c: turns out that windows CRITICAL_SECTIONS are actually thread-aware such that you can multiply request/acquire semaphores from the same thread without blocking. this means our test assertion here is incorrect, so removing it.
15:24.45 Notify 03BRL-CAD:brlcad * 62291 brlcad/trunk/BUGS: wgl is not lingering
15:26.40 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
15:37.10 Notify 03BRL-CAD:zhaoanqing * 62292 (brlcad/branches/nmgreorg/include/bu/magic.h brlcad/branches/nmgreorg/include/nurb.h and 39 others): change macros name from RT_NURB_XXX to NURB_XXX.
15:41.10 Notify 03BRL-CAD:carlmoore * 62293 brlcad/trunk/doc/docbook/system/man1/en/dbcp.xml: use <command> construct with 'dd', and provide proper line breaks in the EXAMPLES
15:46.31 *** join/#brlcad Izakey (~Isaac@195.24.220.134)
15:50.00 Notify 03BRL-CAD:zhaoanqing * 62294 brlcad/branches/nmgreorg/src/libnmg/rt_segs.c: change a comment. small update.
15:50.43 Notify 03BRL-CAD:carlmoore * 62295 brlcad/trunk/doc/docbook/system/man1/en/dbupgrade.xml: add comma, hyphens (replacing some blanks); reverse period and right parenthesis
15:54.30 *** join/#brlcad ries (~ries@188.207.75.185)
15:56.43 Notify 03BRL-CAD:zhaoanqing * 62296 (brlcad/branches/nmgreorg/src/libnmg/mk.c brlcad/branches/nmgreorg/src/libnmg/plot.c and 4 others): make sure there is no *.c files named rt_nurb_XXX or macros named RT_NURB_XXX in libnurb. make sure there is no files named rt_nmg_XXX or macros named RT_NMG_XXX in libnmg.
16:08.24 Notify 03BRL-CAD:carlmoore * 62297 brlcad/trunk/doc/docbook/system/man1/en/dxf-g.xml: a line-break specifier does not seem to work where I am, so I added 2 semicolons; other touch-up of dxf-g.xml
16:20.28 *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
16:31.33 *** join/#brlcad clock (~clock@77-58-143-135.dclient.hispeed.ch)
16:45.52 *** join/#brlcad ries (~ries@188.207.75.185)
17:00.45 ``Erik brlcad: sorry I missed your call. ya guys doing the offsite out of the qri office? when does it end? O.o
17:03.18 Notify 03BRL-CAD:zhaoanqing * 62298 (brlcad/branches/nmgreorg/include/nurb.h brlcad/branches/nmgreorg/src/conv/dxf/dxf-g.c and 12 others): change the calling of functions in libnurb after rename, make program can be compiled on Linux
17:04.31 *** join/#brlcad clock (~clock@77-58-143-135.dclient.hispeed.ch)
17:14.44 kanzure ON_BrepTrim &trim = brep->NewSingularTrim(brep->m_V[i], ....
17:14.58 kanzure a few lines earlier, i was set to brep->m_V.Count()
17:19.25 *** join/#brlcad albertcoder (~albertcod@101.215.172.12)
17:41.41 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
17:57.33 Notify 03BRL-CAD:ejno * 62299 brlcad/trunk/src/conv/3dm/conv3dm-g.cpp: fix windows build - neglected to specify a comparison function for std::set::const_iterator
18:02.36 *** join/#brlcad albertcoder (~albertcod@101.215.172.12)
18:15.37 *** join/#brlcad ries (~ries@D979EA84.cm-3-2d.dynamic.ziggo.nl)
18:20.05 *** join/#brlcad albertcoder (~albertcod@202.164.45.204)
18:21.43 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
18:27.29 kanzure nevermind
18:28.06 Notify 03BRL-CAD:carlmoore * 62300 (brlcad/trunk/doc/docbook/system/man1/en/enf-g.xml brlcad/trunk/doc/docbook/system/man1/en/euclid-g.xml and 3 others): touch-up; use <command> when mentioning command in the DESCRIPTION; do not underscore parentheses, punctuation, or (in EXAMPLE) the command name
18:40.36 *** part/#brlcad Izakey (~Isaac@195.24.220.134)
19:15.57 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
19:17.51 Notify 03BRL-CAD:brlcad * 62301 brlcad/trunk/src/libbu/parallel.c: since we dont' actually ensure anything, we're just printing logging messages, keeping track of the threading start/stop counts is unhelpful. we may need a variant to track recursive IDs, but don't need them at the moment (and they're stateful).
19:42.02 Notify 03BRL-CAD:n_reed * 62302 brlcad/trunk/src/libbrep/opennurbs_ext.cpp: fix valgrind uninitialized value warning; should initialize bbox from surface in case there aren't any loops to set it from
19:49.02 *** join/#brlcad albertcoder (~albertcod@101.214.25.171)
19:59.44 Notify 03BRL-CAD:carlmoore * 62303 (brlcad/trunk/include/bu/parallel.h brlcad/trunk/src/conv/step/STEPWrapper.cpp): remove a trailing blank and fix spellings
20:00.47 Notify 03BRL-CAD:brlcad * 62304 brlcad/trunk/src/libbu/tests/bu_parallel.c: add support for testing recursive callbacks into bu_parallel(). it works!
20:02.44 Notify 03BRL-CAD:brlcad * 62305 brlcad/trunk/src/libbu/parallel.c: begin working on a book-keeping construct for tracking recursive invocation and being aware of our parent context.
20:11.09 *** join/#brlcad ries (~ries@D979EA84.cm-3-2d.dynamic.ziggo.nl)
20:12.05 *** join/#brlcad vladbogo (~vlad@86.121.102.20)
20:14.17 Notify 03BRL-CAD:starseeker * 62306 brlcad/trunk/src/conv/step/step-g/Path.cpp: Propagate LOAD_ERROR up to Edge_Loop - still not getting up to BRep...
20:32.55 Notify 03BRL-CAD:carlmoore * 62307 (brlcad/trunk/doc/docbook/system/man1/en/fb-orle.xml brlcad/trunk/doc/docbook/system/man1/en/fb-pix.xml and 2 others): make citerefentry references to brlcad(1)
20:55.56 Notify 03BRL-CAD:starseeker * 62308 (brlcad/trunk/src/conv/step/step-g/ConnectedFaceSet.cpp brlcad/trunk/src/conv/step/step-g/Face.cpp and 4 others): Propagate errors (and error status) back up to the Brep level
21:40.02 Notify 03BRL-CAD:carlmoore * 62309 (brlcad/trunk/doc/docbook/system/man1/en/fbcbars.xml brlcad/trunk/doc/docbook/system/man1/en/fbcmap.xml and 2 others): improve brlcad(1) references and list of INSPIRATIONS; do not underscore parentheses
21:40.42 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
21:45.57 Notify 03BRL-CAD:carlmoore * 62310 brlcad/trunk/doc/docbook/system/man1/en/pictx.xml: fix references to ir-X
21:48.30 Notify 03BRL-CAD:carlmoore * 62311 brlcad/trunk/doc/docbook/system/man1/en/ir-X.xml: fix a reference to showtherm
22:07.17 Notify 03BRL-CAD:n_reed * 62312 brlcad/trunk/src/libbrep/opennurbs_ext.cpp: initialize bbox from surface in case there aren't any loops
22:58.07 *** join/#brlcad vladbogo_ (~vlad@86.121.102.20)
23:13.36 Notify 03BRL-CAD:n_reed * 62313 brlcad/trunk/src/libbrep/opennurbs_ext.cpp: whoops, to get minimal bbox, only use surface bbox if we fail to calculate one from the loops
23:31.23 Notify 03BRL-CAD:brlcad * 62314 brlcad/trunk/src/libbu/tests/bu_parallel.c: clean up our expectations. our recursive children should have unique ids, so our expected results are larger.
23:34.29 Notify 03BRL-CAD:brlcad * 62315 brlcad/trunk/src/libbu/parallel.c: lots of progres towards throttled child thread creation. this keeps track of all threads being invoked, giving each a unique id. this change also aims to clean up the nthreadc/nthreade creation/completion tracking along with making the windows impl typewise consistent with the others (code reduction).
23:38.05 Notify 03BRL-CAD:n_reed * 62316 brlcad/trunk/include/brep.h: initialize m_start/m_end from bbox when given it so the GetBBox member returns the correct result
23:40.11 Notify 03BRL-CAD:brlcad * 62317 brlcad/trunk/src/libbu/parallel.c: make sure we release our thread id for windows and sun/unix platforms too
23:49.22 *** join/#brlcad nanicoar_ (~n@87-92-148-237.bb.dnainternet.fi)

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