IRC log for #brlcad on 20161019

00:24.12 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
00:41.27 Notify 03BRL-CAD:starseeker * 69125 brlcad/trunk/src/libbu/temp.c: Add a refinement from Sean to avoid the temp file routine from generating identical names when called within a one second window. This was (one of) the issues we were encountering on Windows with the bn_vlist tests.
00:47.44 *** join/#brlcad smigtmuzrafltdhi (~armin@dslb-088-066-131-085.088.066.pools.vodafone-ip.de)
00:50.57 Notify 03BRL-CAD:starseeker * 69126 brlcad/trunk/src/libbu/temp.c: Untested as yet, but attempt to disable a too aggressive check by MSVC that prevents close operations from working on already closed files. The cleanup function in libbu is *designed* to ensure that temp files are closed, and as a consequence is very likely to be doing redundant close operations if the users already cleaned up. Approach identifed by Sean.
01:06.24 Notify 03BRL-CAD:starseeker * 69127 brlcad/trunk/src/libbu/temp.c: Make a few changes, guided by Microsoft's _set_invalid_parameter_handler example in their documentation. This suppresses the assertion successfully.
01:07.25 Notify 03BRL-CAD:starseeker * 69128 brlcad/trunk/src/libbn/tests/bn_list.c: Set errno
01:11.57 Notify 03BRL-CAD:starseeker * 69129 brlcad/trunk/src/libbn/tests/bn_list.c: include the header
04:25.22 Notify 03BRL-CAD:brlcad * 69130 brlcad/trunk/src/libbu/tests/bu_booleanize.c: remove the debug printing
04:35.57 Notify 03BRL-CAD:brlcad * 69131 brlcad/trunk/src/libbu/tests/bu_bitv.c: setting a nul char isn't right, almost certainly masking a real bug that was caught. bu_vls_putc() and other vls printing functions already manually nul-terminate and bu_vls_cstr() is defined to guarantee only returning a nul-terminated string. so any vls string lacking 0-termination would be a bug.
06:54.47 *** join/#brlcad teepee (~teepee@unaffiliated/teepee)
07:07.26 *** join/#brlcad Caterpillar (~caterpill@unaffiliated/caterpillar)
07:39.50 *** join/#brlcad ``Erik_ (~erik@pool-100-16-14-17.bltmmd.fios.verizon.net)
07:57.28 *** join/#brlcad amarjeet (~amarjeet@202.164.53.117)
09:48.26 *** join/#brlcad teepee] (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
10:22.56 *** join/#brlcad KimK (~Kim__@2600:8803:7a85:6d00:f44f:f08c:c43b:6224)
10:39.55 *** join/#brlcad LordOfBikes (~armin@dslb-088-066-131-085.088.066.pools.vodafone-ip.de)
11:45.42 *** join/#brlcad yorik (~yorik@2804:431:f720:5fc3:290:f5ff:fedc:3bb2)
13:10.21 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
14:09.21 Notify 03BRL-CAD:starseeker * 69132 brlcad/trunk/src/libbu/vls.c: Stepping through the bitv test, this is the point in MSVC debugging where the display of the string starts to look strange. If we extend the copy length one past the vls_len in bu_vls_prepend, we pick up the terminating null and the test completes successfully.
14:27.00 Notify 03BRL-CAD:starseeker * 69133 brlcad/trunk/src/libbu/tests/bu_progname.c: If we've got a Windows style path with a drive letter + colon prefix (i.e. C:) the check for directory separator in the first char isn't going to work. Also check for a colon in second char + directory separator in the third char.
14:57.18 Notify 03BRL-CAD:starseeker * 69134 brlcad/trunk/src/libbu/tests/bu_progname.c: Cute. When running the test program straight up, the file path for Case 7 uses back slashes. When run by ctest, it gets forward slashes instead. Check for both cases.
15:14.39 *** join/#brlcad teepee] (bc5c2133@gateway/web/freenode/ip.188.92.33.51)
16:03.04 Notify 03BRL-CAD:starseeker * 69135 brlcad/trunk/src/libbu/tests/bu_vls_vprintf.c: Switch a few of the tests that use formatting strings vls supporst but MSVC doesn't over to a string based correctness test.
16:26.11 Notify 03BRL-CAD:starseeker * 69136 brlcad/trunk/src/libbu/vls_vprintf.c: In addition to z and t, MSVC also doesn't support 'j'
16:30.51 *** join/#brlcad amarjeet (~amarjeet@169.149.163.135)
16:33.03 starseeker finally
18:18.55 *** join/#brlcad yorik (~yorik@2804:431:f720:5fc3:290:f5ff:fedc:3bb2)
18:55.14 Notify 03BRL-CAD:starseeker * 69137 brlcad/trunk/misc/CMake/BRLCAD_Targets.cmake: See if we can categorize things a bit better for executables.
18:55.47 *** join/#brlcad teepee_ (~teepee@unaffiliated/teepee)
19:06.53 Notify 03BRL-CAD:starseeker * 69138 (brlcad/branches/tcltk86/CHANGES brlcad/branches/tcltk86/CMakeLists.txt and 416 others): Sync with trunk thru r69137
19:08.48 Notify 03BRL-CAD:starseeker * 69139 (brlcad/branches/qtged/CHANGES brlcad/branches/qtged/CMakeLists.txt and 416 others): Sync with trunk thru r69137
19:34.02 Notify 03BRL-CAD:starseeker * 69140 (brlcad/trunk/src/libbu/tests/CMakeLists.txt brlcad/trunk/src/libbu/tests/tests_bitv.cmake brlcad/trunk/src/libbu/tests/tests_vls.cmake): Rename for less verbosity
19:47.14 Notify 03BRL-CAD:starseeker * 69141 brlcad/trunk/CMakeLists.txt: benchmark isn't really a unit test - it's more of an integration test.
19:50.29 Notify 03BRL-CAD:starseeker * 69142 brlcad/trunk/src/libbn/tests/bntester.c: At some point need to come back and digest tester into the more standard setup - good intro project for a beginning student.
19:53.05 Notify 03BRL-CAD:starseeker * 69143 brlcad/trunk/regress/repository.sh: fix dirname.c sed expression in repository.h
19:58.57 Notify 03BRL-CAD:starseeker * 69144 brlcad/trunk/src/libbn/tests/CMakeLists.txt: rename libbn tests
20:00.12 Notify 03BRL-CAD:starseeker * 69145 brlcad/trunk/src/libbg/tests/CMakeLists.txt: flag bg test executables as such
20:12.47 Notify 03BRL-CAD:starseeker * 69146 brlcad/trunk/src/librt/tests/CMakeLists.txt: Do some renaming with the librt tests
20:29.23 Notify 03BRL-CAD:starseeker * 69147 brlcad/trunk/src/conv/gcv/CMakeLists.txt: some conversion capabilities are exposed only through gcv - enable install.
20:30.23 Notify 03BRL-CAD:starseeker * 69148 brlcad/trunk/src/libicv/tests/CMakeLists.txt: rename icv test files
21:10.25 *** join/#brlcad kintel (~kintel@unaffiliated/kintel)
21:10.39 *** part/#brlcad kintel (~kintel@unaffiliated/kintel)
22:48.34 *** join/#brlcad teepee (~teepee@unaffiliated/teepee)

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