| 03:15.51 | *** join/#brlcad louipc (~louipc@archlinux/fellow/louipc) | |
| 06:12.20 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) | |
| 06:36.51 | *** join/#brlcad louipc (~louipc@archlinux/fellow/louipc) | |
| 08:16.53 | *** join/#brlcad merzo (~merzo@193.254.217.44) | |
| 10:52.10 | *** join/#brlcad Stattrav (~Stattrav@61.12.114.82) | |
| 10:52.10 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) | |
| 12:04.50 | CIA-48 | BRL-CAD: 03Rabotykrovelnye 07http://brlcad.org * r3273 10/wiki/Index.php: New page: = ???????????????????? ???????????? = ?? ?????????? ?????? ???????????? ???? ?????????? ?????????????????? ????????????????, ???? ?????? ?????? ???????????????????????????? '''????????????... |
| 12:13.54 | CIA-48 | BRL-CAD: 03Rabotykrovelnye 07http://brlcad.org * r3274 10/wiki/User:Rabotykrovelnye: New page: = ???????????????????? ???????????? = ?? ?????????? ?????? ???????????????? ???? ?????????? ?????????????????? ????????????????, ???? ?????? ?????? ?????????????????????????? '''??????????... |
| 13:15.17 | CIA-48 | BRL-CAD: 03Sean 07http://brlcad.org * r0 10/wiki/Special:Log/delete: deleted "[[Index.php]]": spam |
| 13:15.36 | CIA-48 | BRL-CAD: 03Sean 07http://brlcad.org * r0 10/wiki/Special:Log/block: blocked [[User:Rabotykrovelnye]] with an expiry time of infinite (account creation disabled, e-mail blocked): Inserting nonsense/gibberish into pages |
| 14:33.42 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) | |
| 14:37.38 | CIA-48 | BRL-CAD: 03Sean 07http://brlcad.org * r0 10/wiki/Special:Log/delete: deleted "[[User:Rabotykrovelnye]]": content was spam |
| 15:04.22 | brlcad | starseeker: curious (because I don't remember) but why is the standards conformance set to gnu99? commit comment didn't say anything useful as to why it got set to that |
| 15:08.31 | brlcad | explains some compilation failure differences I've been seeing |
| 15:12.36 | brlcad | never mind, I traced down some commentary |
| 15:20.33 | starseeker | brlcad: didn't you decide to add that flag? (memory spotty, but I doubt I would have added it on my own...) |
| 16:07.36 | brlcad | it's not the flag itself, it's the value |
| 16:08.13 | brlcad | that's one of those regression flags where I'd really like to set it to four different things, make sure all keep working |
| 17:06.28 | CIA-48 | BRL-CAD: 03brlcad * r49185 10/brlcad/trunk/src/other/step/include/express/basic.h: use __inline__ instead of inline for c89 compilation |
| 17:07.11 | CIA-48 | BRL-CAD: 03brlcad * r49186 10/brlcad/trunk/src/other/step/src/express/ (expr.c expscan.l): use static_inline instead of 'static inline' so that we're consistent with other uses and expand correctly to the right symbols for this compilation |
| 17:08.23 | CIA-48 | BRL-CAD: 03brlcad * r49187 10/brlcad/trunk/src/libicv/Makefile.am: use PNG_CPPFLAGS instead of hard-coding them so that we don't get a conflict if png is disabled |
| 17:08.39 | CIA-48 | BRL-CAD: 03brlcad * r49188 10/brlcad/trunk/src/libpkg/example/server.c: c89-style comments |
| 17:08.55 | CIA-48 | BRL-CAD: 03brlcad * r49189 10/brlcad/trunk/src/libbu/Makefile.am: realpath missing |
| 17:09.46 | CIA-48 | BRL-CAD: 03brlcad * r49190 10/brlcad/trunk/src/ (conv/obj-g.c shapes/coil.c): break up the long string literal for c89 compliance |
| 17:10.16 | CIA-48 | BRL-CAD: 03brlcad * r49191 10/brlcad/trunk/src/libged/analyze.c: no // comments in C files for portability |
| 17:26.21 | CIA-48 | BRL-CAD: 03brlcad * r49192 10/brlcad/trunk/src/other/step/src/express/fedex.c: include getopt.h directly in case we're compiling in strict c89 mode |
| 17:26.57 | CIA-48 | BRL-CAD: 03brlcad * r49193 10/brlcad/trunk/src/other/step/src/fedex_plus/ (classes.c classes_misc.c fedex_idl.c): more static_inline propagation so we support strict c89 mode |
| 17:33.05 | CIA-48 | BRL-CAD: 03brlcad * r49194 10/brlcad/trunk/misc/CMake/Distclean.cmake: |
| 17:33.06 | CIA-48 | BRL-CAD: add a distclean target so we can wipe out common cmake turds and other built |
| 17:33.06 | CIA-48 | BRL-CAD: system entities without wiping out an entire build tree. derived from distclean |
| 17:33.06 | CIA-48 | BRL-CAD: code by Jan Woetzel with some minor tweaks for other compiler product dirs and |
| 17:33.06 | CIA-48 | BRL-CAD: more clear wording. |
| 17:35.55 | *** join/#brlcad Elrohir (~kvirc@p579F02E6.dip.t-dialin.net) | |
| 18:03.48 | *** join/#brlcad Yoshi47 (~jan@64.235.102.210) | |
| 18:12.20 | CIA-48 | BRL-CAD: 03n_reed * r49195 10/brlcad/trunk/src/libbu/ (sscanf.c test_sscanf.c): make maximum field width mandatory for %s and %[...] conversions |
| 18:24.10 | CIA-48 | BRL-CAD: 03brlcad * r49196 10/brlcad/trunk/misc/CMake/LEMON_Util.cmake: delete out lemon output files during make clean |
| 19:29.47 | CIA-48 | BRL-CAD: 03brlcad * r49197 10/brlcad/trunk/misc/CMake/LEMON_Util.cmake: also try to clean up the lemoninput file we copied. i may be doing something wrong here, though.. because it's not evident that this section of code is getting called. |
| 19:30.36 | brlcad | starseeker: would you sanity check that for me? I tried putting in message and echo statements in there, but they never seemed to get printed |
| 19:31.16 | brlcad | moreover expparse.y isn't being deleted on make clean still (src/other/step/src/express copied input file) |
| 20:13.15 | CIA-48 | BRL-CAD: 03brlcad * r49198 10/brlcad/trunk/CMakeLists.txt: |
| 20:13.15 | CIA-48 | BRL-CAD: use the CheckTypeSize module's CHECK_TYPE_SIZE() macro for explicitly evaluating |
| 20:13.15 | CIA-48 | BRL-CAD: the size of a pointer instead of assuming cmake will figure it out for us. this |
| 20:13.15 | CIA-48 | BRL-CAD: takes care of the problems assuming 32-bit on a 64-bit platform when the size |
| 20:13.15 | CIA-48 | BRL-CAD: can be determined. unclear how cmake (repeatedly) got into states where the |
| 20:13.16 | CIA-48 | BRL-CAD: CMAKE_SIZEOF_VOID_P ends up unset, but since it's a built-in test, likely just |
| 20:13.17 | CIA-48 | BRL-CAD: some bug internal to cmake. |
| 20:14.36 | starseeker | brlcad: checking... |
| 20:19.58 | starseeker | brlcad: oh, I think I know what's going on - when we add src/other directories, their copies of *.cmake files get called before our misc copies |
| 20:20.01 | starseeker | one sec... |
| 20:21.24 | CIA-48 | BRL-CAD: 03starseeker * r49199 10/brlcad/trunk/ (misc/CMake/LEMON_Util.cmake src/libgcv/wfobj/CMakeLists.txt): Make case labeling consistent. |
| 20:21.44 | starseeker | That's why I have to update all the src/other copies of FindX11.cmake - one of those gets triggered (usually) before our misc copy |
| 20:22.31 | CIA-48 | BRL-CAD: 03starseeker * r49200 10/brlcad/trunk/src/libgcv/wfobj/CMakeLists.txt: Gah - remove debug message |
| 20:22.58 | brlcad | is there a way to append the top-level search path to the lists of paths searched, after the cmake sub-paths are added? |
| 20:23.07 | brlcad | s/append/prepend/ |
| 20:24.32 | CIA-48 | BRL-CAD: 03starseeker * r49201 10/brlcad/trunk/src/other/ (re2c/CMake/LEMON_Util.cmake step/CMake/LEMON_Util.cmake): Sync src/other copies of LEMON_Util.cmake |
| 20:25.14 | starseeker | there are two things at work - the src/other builds are doing their own thing, and CMake remembers if it has defined macros previously and doesn't re-define them |
| 20:25.47 | starseeker | when a src/other .cmake macro defining file is loaded, those macro definitions are global - that's why you weren't seeing any changes |
| 20:26.57 | starseeker | We can try forcing the src/other builds to look in our misc/CMake directory, but that's not totally reliable - they can still override that path in their own logic |
| 20:28.06 | starseeker | Would be nice if we could get CMake upstream to accept some of those .cmake files - then we could require a version that has them and not include them... |
| 20:33.50 | starseeker | hmm... wait... am I right about what it's doing??? |
| 20:33.54 | starseeker | introspects |
| 20:35.55 | CIA-48 | BRL-CAD: 03brlcad * r49202 10/brlcad/trunk/src/libged/combmem.c: gcc correctly detects an array overrun read where a vect_t was being passed as a hvect_t with the H component getting used. looks like at least a couple mistakes were made, fixed by just making them hvect_t's. |
| 20:45.18 | CIA-48 | BRL-CAD: 03brlcad * r49203 10/brlcad/trunk/misc/CMake/CompilerFlags.cmake: toggle whether we're aiming for c89 or c99 based on the build configuration, just so we get more portability coverage. as long as we're compliant, might as well stay that way for as long as we can bear. |
| 20:53.39 | starseeker | it's gnarly... I'm loading macro definition, but not using find_package to do so... how should I do that... |
| 21:02.51 | CIA-48 | BRL-CAD: 03starseeker * r49204 10/brlcad/trunk/src/other/ (27 files in 27 dirs): Clean up some ws in src/other |
| 21:40.28 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) | |
| 21:45.17 | brlcad | n_reed: library code shouldn't be calling bu_exit(), bu_bomb() if it's not recoverable or kick back an error code to the caller |
| 21:47.49 | brlcad | we have just a few instances of misuse the have propagated, but that's something I've been meaning to add a regression test for |
| 21:48.39 | brlcad | moreover to r49195, I'd expect 0 to be a valid width .. it'll read zero chars, maybe init with -1 and check if < 0 |
| 21:49.29 | brlcad | or just default to 0-width, and proceed |
| 22:00.36 | CIA-48 | BRL-CAD: 03starseeker * r49205 10/brlcad/trunk/ (16 files in 15 dirs): Try to be better about using misc/CMake .cmake files by default. Tweak the listing of additional files to clean from lemon output. |
| 22:02.01 | CIA-48 | BRL-CAD: 03starseeker * r49206 10/brlcad/trunk/misc/CMake/CompilerFlags.cmake: Hmm... make test match comment - was checking for gnu90 on debug, gnu99 on release - check for both if doing release. How come these checks are release only? |
| 22:03.09 | CIA-48 | BRL-CAD: 03starseeker * r49207 10/brlcad/trunk/ (3 files in 3 dirs): remove debug message |
| 22:03.45 | starseeker | brlcad: give that a whirl and see if it achieves what you were looking for behavior wise |
| 22:08.51 | CIA-48 | BRL-CAD: 03starseeker * r49208 10/brlcad/trunk/misc/CMake/CompilerFlags.cmake: Oh, I see now. gnu99 when doing a Release build, but gnu90 when developing (normally done in Debug mode) |
| 22:16.38 | brlcad | yeah, the comment was unclear but you can't do both |
| 22:17.02 | starseeker | no biggie - I'm just being slow today ;-) |
| 22:17.39 | brlcad | been trying to get non-gnu mode.. but that's really a bear |
| 22:18.07 | brlcad | we compile gnu1x, though, that was nice to see |
| 22:18.15 | starseeker | cool! |
| 22:19.15 | starseeker | brlcad: is the compiler flags setup tolerable? |
| 22:21.42 | brlcad | what do you mean? |
| 22:21.46 | ``Erik | compile fail on osX, C++/c99 style comments in /System/Library/Frameworks/OpenGL.framework/Headers/gl.h via fb.h |
| 22:22.14 | brlcad | right, there was a separate fix for that in autotools |
| 22:22.55 | starseeker | brlcad: the way the CMake detection and assignment of compilation flags is set up - does it seem workable? |
| 22:23.28 | brlcad | starseeker: it's working well enough thusfar, save for one change I had to make |
| 22:23.36 | ``Erik | brlcad: did you do something to bz to drop cpu consumption? I'm used to load >1, which isn't the case lately? ( http://crit.brlcad.org/perfmon/ ) |
| 22:24.11 | starseeker | ``Erik: which component is failing to build? |
| 22:24.26 | ``Erik | starseeker: libfb, but anything that uses fb.h will fail |
| 22:26.29 | ``Erik | starseeker: http://paste.lisp.org/display/127469 (on my work desktop) |
| 22:26.36 | brlcad | yay, neat stats |
| 22:26.39 | ``Erik | puts on pants and heads home |
| 22:26.57 | brlcad | ``Erik: yes, the bzflag services were migrated to a different host |
| 22:34.41 | starseeker | the autotools build was adding the C99 flag specifically for that library... |
| 22:34.46 | starseeker | um |
| 23:25.09 | CIA-48 | BRL-CAD: 03brlcad * r49209 10/brlcad/trunk/src/mged/Makefile.am: hideline.c is no more |
| 23:34.45 | CIA-48 | BRL-CAD: 03brlcad * r49210 10/brlcad/trunk/misc/CMake/CompilerFlags.cmake: |
| 23:34.45 | CIA-48 | BRL-CAD: make the CHECK_C_FLAG macro a little more robust. the autonaming feature is |
| 23:34.45 | CIA-48 | BRL-CAD: nice, but it needs to account for all possible characters that aren't valid |
| 23:34.45 | CIA-48 | BRL-CAD: variable names. (e.g., --std=c++98) also, the build type comparison is wrong |
| 23:34.45 | CIA-48 | BRL-CAD: as mixed case for this variable. |
| 23:44.52 | ``Erik | <-- waits for other services to be migrated to crit O:-) |
| 23:50.37 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) | |