IRC log for #brlcad on 20100802

00:28.43 starseeker brlcad: that's probably that autogenerated file
00:28.45 starseeker grr
00:29.01 starseeker I thought the rest of the build depended on that rule being done...
03:22.14 CIA-43 BRL-CAD: 03brlcad * r39984 10/brlcad/trunk/src/rt/ (view.c viewedge.c): sort the option usage statement so that more common options are at the top, related options are grouped together, and the wording is made slightly less cryptic.
03:22.15 brlcad starseeker: they are both autogenerated
03:26.33 brlcad yeah, probably is just some detail in the dependencies
03:26.59 brlcad but since they're autogenerated, they'll need some sort of locking step before all of the things that depend on it try to use it
03:27.08 brlcad otherwise it's a race condition
03:32.15 brlcad one example: A depends on C, B depends on C, two processes evaluate A and B, process 1 checks A, process 2 checks B simultaneously, C doesn't exist for process1 and begins C rule, C exists (partially) for process2 so tries to compile against it and fails
03:32.41 brlcad so need a stamp or semaphore file or similar
04:41.58 starseeker brlcad: yech. Too bad the build systems aren't smart enough to recognize that case and handle it
04:56.14 ``Erik hm, wonder if that was what was causing issue for me when I started adjusting the makefile
04:56.39 ``Erik there's a sports bar down here called, uh, 'continental', it's very... old bowling alley
04:59.17 starseeker wonders if CMake is actually smart enough...
05:05.01 starseeker hmm: linkinghub.elsevier.com/retrieve/pii/S0010448508001851
05:11.29 starseeker O.o http://handle.dtic.mil/100.2/ADP012016
05:12.26 brlcad that could be very handy for proper bounding box estimating
05:12.31 brlcad and space partitioning
05:13.25 brlcad as well as for draping support
05:15.43 starseeker ah ha http://public.kitware.com/Wiki/CMake_FAQ#How_can_I_generate_a_source_file_during_the_build.3F
05:19.50 CIA-43 BRL-CAD: 03brlcad * r39985 10/brlcad/trunk/TODO: red passes a quick round of manual testing. looks like all the previous problems are working well now.
05:21.52 starseeker holy smoke, intersection curves for deforming surfaces: http://www.cs.utah.edu/~xchen/papers/ssi-cad.pdf
05:22.08 starseeker makes a note to check that reference list for surface/surface papers
05:25.00 brlcad looks like a money shot there
05:31.02 CIA-43 BRL-CAD: 03brlcad * r39986 10/brlcad/trunk/TODO: nirt seems to be working again
05:48.13 CIA-43 BRL-CAD: 03brlcad * r39987 10/brlcad/trunk/src/mged/tedit.c:
05:48.13 CIA-43 BRL-CAD: help mac systems out when running from Terminal where X11 paths are not included
05:48.13 CIA-43 BRL-CAD: (better than getting back null and crashing). also don't assume anything about
05:48.13 CIA-43 BRL-CAD: emacs for non-mac/non-windows systems, since there isn't an apparent way to test
05:48.14 CIA-43 BRL-CAD: it. at worst, it just creates an extra window this way instead of hanging mged.
05:51.28 CIA-43 BRL-CAD: 03brlcad * r39988 10/brlcad/trunk/src/libged/editit.c: (log message trimmed)
05:51.29 CIA-43 BRL-CAD: fix a bug on Mac where TextEdit blathers warnings about being unable to register
05:51.29 CIA-43 BRL-CAD: itself with system services. this is due to running multiple instances of
05:51.29 CIA-43 BRL-CAD: TextEdit and while the warning is benign, we quell it (via closing stdin AND
05:51.29 CIA-43 BRL-CAD: stderr for the forked child) as it's not useful to the user. add a few more
05:51.29 CIA-43 BRL-CAD: sanity checks in along with keeping track of our (null) options with the ugly
05:51.30 CIA-43 BRL-CAD: argv-in-an-argv hack. make the Invoking statement more generally diagnostic and
05:58.42 CIA-43 BRL-CAD: 03brlcad * r39989 10/brlcad/trunk/src/libged/combmem.c: fix whitespace bug introduced in r39975 where there's badness in a macro after a line continuator. was breaking windows build
06:12.21 CIA-43 BRL-CAD: 03brlcad * r39990 10/brlcad/trunk/misc/win32-msvc8/libged/libged.vcproj: add select.c to the windows build for libged
06:50.36 *** join/#brlcad merzo (~merzo@smartbussiness.mobicom.net.ua)
07:20.59 *** join/#brlcad hyarion (c05ben@peppar.cs.umu.se)
09:25.16 CIA-43 BRL-CAD: 03d_rossberg * r39991 10/brlcad/trunk/misc/win32-msvc/CMakeLists.txt: at least the brlcad.dll can be build this way in x64 mode
10:10.51 d-lo Mernin all
10:52.08 Ralith mern
10:54.37 *** join/#brlcad Yoshi47 (~jan@64.235.102.210)
10:57.52 louipc monin
12:15.07 CIA-43 BRL-CAD: 03Linomone 07http://brlcad.org * r2256 10/wiki/Documentation:
13:54.46 d-lo brlcad: what flag was it that helped me get past my system tk? --enable-all ?
14:16.39 brlcad yep
14:34.29 CIA-43 BRL-CAD: 03brlcad * r39992 10/brlcad/trunk/misc/nsis/updateVersion.tcl: make sure it's the root, look for include dir
14:37.34 *** join/#brlcad Zaebos (~irc@pd95b7f5e.dip0.t-ipconnect.de)
15:05.27 CIA-43 BRL-CAD: 03brlcad * r39993 10/brlcad/trunk/misc/win32-msvc8/asc2g/asc2g.vcproj:
15:05.27 CIA-43 BRL-CAD: fix the windows build. for whatever reason, it's insisting that the windows
15:05.27 CIA-43 BRL-CAD: executable extension be included on asc2g and tclsh now. add some simple
15:05.27 CIA-43 BRL-CAD: scripting logic to detect whether NSIS is available before trying to run it,
15:05.27 CIA-43 BRL-CAD: only warning if it's missing. clean up echo quotes too.
15:05.29 brlcad and with that, we seem to have a full build tested on mac and windows now, last up is linux
15:05.33 brlcad <PROTECTED>
15:52.28 starseeker kicks off a linux distcheck
16:16.39 starseeker distcheck succeeds here (redhat)
16:25.22 *** join/#brlcad kanzure (~kanzure@131.252.130.248)
17:09.00 brlcad starseeker: awesome!
18:13.02 *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1)
18:48.49 *** join/#brlcad roberthl (~robert@2001:ba8:1f1:f03d::2)
18:48.49 *** join/#brlcad roberthl (~robert@mediawiki/RobertL)
22:43.55 *** join/#brlcad Nohla (~Nohla@201.255.255.197)

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