| 00:02.27 | starseeker | if it works, should we add it to allow a tweaked indent.sh to work "out of the box", as it were? |
| 00:07.13 | brlcad | you mean add astyle to our repo? nah |
| 00:09.36 | brlcad | it's enough to point an offender at the download with instructions that they need to install/run it |
| 00:10.21 | brlcad | it's a thought, but still not necessary until a complete style is defined |
| 00:10.56 | starseeker | alrightie |
| 00:11.02 | brlcad | which would still take at least a day I think to get right, test, verify |
| 00:11.43 | brlcad | next step is for someone(tm) to work on a matching style definition |
| 00:12.00 | brlcad | that would have been a good GCI task :) |
| 00:12.51 | starseeker | heh - next year :-) |
| 00:13.04 | brlcad | maybe :) |
| 00:13.35 | brlcad | it's still a highly dubious program from an efficiency perspective -- most I talked to said it's a wash at best |
| 00:13.56 | starseeker | hmm. Maybe just as well then |
| 00:14.01 | brlcad | I was interested more from a community building/awareness perspective, get 'em aware young |
| 00:17.06 | starseeker | sees the core chromium browser is BSD licensed... cool. Might have to try that. Been avoiding Chrome itself. |
| 00:18.18 | starseeker | Oo - lots of design docs |
| 00:18.41 | starseeker | yow - multi *process* architecture? |
| 00:19.00 | starseeker | backs away slowly... |
| 00:22.11 | brlcad | you do realize why it's designed that way though, right? |
| 00:22.23 | brlcad | flash crashing doesn't bring down your browser |
| 00:22.27 | starseeker | sure - so the browser is robust when individual pages fail |
| 00:22.35 | starseeker | er, yeah :-) |
| 00:22.59 | starseeker | not very useful for our own problem domain though |
| 00:22.59 | brlcad | if a plugin fails, no impact other than maybe a page load failing |
| 00:23.15 | brlcad | if we had lots of unreliable plugins, perhaps |
| 00:53.55 | CIA-59 | BRL-CAD: 03starseeker * r47672 10/brlcad/trunk/src/librt/CMakeLists.txt: Because librt is using so many different DLL_IMPORTS/EXPORTS flags, we're going to have to handle a few things manually. Any reason not to roll TIE_DLL and DB5_DLL into RT_DLL? |
| 01:01.45 | brlcad | no reason comes to mind, wondered why they were separate to begin with |
| 01:05.41 | CIA-59 | BRL-CAD: 03starseeker * r47673 10/brlcad/trunk/ (include/db5.h include/tie.h src/librt/CMakeLists.txt): DB5 and TIE export prefixes switched to RT |
| 02:09.46 | *** join/#brlcad louipc (~louipc@archlinux/fellow/louipc) |
| 02:15.08 | CIA-59 | BRL-CAD: 03starseeker * r47674 10/brlcad/trunk/src/ (8 files in 7 dirs): Make a stab at dealing with more MSVC build issues with new setup. Not tested yet. |
| 02:59.06 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) |
| 03:11.55 | starseeker | huh, cool - wonder if features could be snarfed from this for astyle: http://gcgreatcode.cvs.sourceforge.net/viewvc/gcgreatcode/GC/GC.txt?revision=1.4&view=markup |
| 03:22.42 | starseeker | BRL-CAD builds on Windows once more |
| 03:22.51 | starseeker | is outta here |
| 03:22.57 | louipc | bye |
| 03:27.17 | louipc | Anyone know the address/url to checkout/commit via svn+ssh? or is brlcad not set up for that? |
| 05:35.56 | *** join/#brlcad IriX64 (~kvirc@bas2-sudbury98-1177872364.dsl.bell.ca) |
| 05:36.07 | IriX64 | http://pastebin.com/6qAwbny1 |
| 05:36.16 | IriX64 | just checked out and happened |
| 05:36.22 | IriX64 | ciao |
| 06:19.11 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) |
| 13:28.37 | brlcad | ~cadsvn |
| 13:28.37 | ibot | To obtain BRL-CAD from Subversion: svn checkout https://brlcad.svn.sourceforge.net/svnroot/brlcad/brlcad/trunk brlcad |
| 13:32.32 | brlcad | louipc: it should work, though I admittedly haven't used that method in a very long time |
| 13:33.24 | brlcad | I believe svn+ssh requires uploading your ssh key ( https://sourceforge.net/apps/trac/sourceforge/wiki/SSH%20keys ) or it will prompt for password multiple times for a single checkout |
| 13:41.21 | brlcad | louipc: I take it back, looks like they changed it to be https-only some time ago (about a year ago) though they keys will still work and get used |
| 13:45.53 | brlcad | louipc: ahh, and their new beta website interface will support it in the upcoming months: https://sourceforge.net/p/forge/documentation/svn%20-%20Beta/ |
| 13:46.57 | brlcad | brl-cad has not yet migrated to the beta interface |
| 14:00.48 | ``Erik | TIE_DLL being seperate is legacy from when libtie linked librt |
| 14:27.47 | CIA-59 | BRL-CAD: 03d_rossberg * r47675 10/brlcad/trunk/src/ (librt/CMakeLists.txt libwdb/CMakeLists.txt): the brlcad.dll needs an openNURBS.dll: setting the compiler switches |
| 14:31.33 | CIA-59 | BRL-CAD: 03d_rossberg * r47676 10/rt^3/trunk/src/coreInterface/ConstDatabase.cpp: the result of a test for NULL should be honored |
| 14:36.56 | starseeker | hmm... IriX64's error raises the question of how hard we should look for jni.h... at least on my system, no special effort is required |
| 14:37.36 | starseeker | not sure if that should be regarded as a system configuration issue... |
| 14:39.21 | starseeker | oh, right, I've got that config thing working |
| 14:39.24 | starseeker | hmm... |
| 14:47.22 | starseeker | might have to treat it as a config issue, or there are other issues |
| 14:47.42 | starseeker | my system has both java itself and gcc supplying jni.h headers |
| 14:48.42 | starseeker | (incompatible ones, I might add) |
| 14:50.53 | starseeker | shakes head - need help for this one, don't know enough about Java/JNI |
| 15:07.07 | ``Erik | at the moment, the only jni using stuff is incredibly muves specific, so I'd say "not at all" for the time being... :D make 'em set the variable |
| 15:46.39 | CIA-59 | BRL-CAD: 03brlcad * r47677 10/brlcad/trunk/src/rt/viewedge.c: carl reported a typo |
| 15:54.25 | CIA-59 | BRL-CAD: 03brlcad * r47678 10/brlcad/trunk/src/tclscripts/rtwizard/examples/ (5 files in 5 dirs): various additional corrections from carl for rtwizard. |
| 16:18.00 | CIA-59 | BRL-CAD: 03brlcad * r47679 10/brlcad/trunk/src/adrt/isst_tcltk.c: TIE_EXPORT is dead |
| 16:42.40 | CIA-59 | BRL-CAD: 03bob1961 * r47680 10/brlcad/trunk/src/rt/do.c: Remove call to do_ae() from within cm_ae(). This causes a segmentation fault. Besides, do_ae() gets called later in main.c if necessary. |
| 17:22.06 | CIA-59 | BRL-CAD: 03n_reed * r47681 10/brlcad/trunk/src/other/perplex/ (9 files): borrowing re2c's option parser |
| 17:22.28 | CIA-59 | BRL-CAD: 03starseeker * r47682 10/brlcad/trunk/CMakeLists.txt: Do a little more to disable librtserver - may have to eventually do an actual compile test on this, but hopefully this will avoid some issues. |
| 17:50.14 | CIA-59 | BRL-CAD: 03bob1961 * r47683 10/brlcad/trunk/ (4 files in 4 dirs): Added a data_polygons subcommand for saving a polygon as a sketch. Also added one for importing polygon data from a sketch. |
| 18:07.07 | CIA-59 | BRL-CAD: 03starseeker * r47684 10/brlcad/trunk/src/other/libutahrle/CMakeLists.txt: Hmm, must have had a stale libutahrle lib file around. Tweak MSVC build flags. |
| 18:11.56 | CIA-59 | BRL-CAD: 03n_reed * r47685 10/brlcad/trunk/src/other/perplex/ (parser.y perplex.cpp perplex.h scanner_template.c): added option for specifying output file |
| 19:07.16 | *** join/#brlcad merzo (~merzo@29-12-132-95.pool.ukrtel.net) |
| 19:09.21 | *** join/#brlcad Forth (~Forth@92.242.118.253) |
| 19:11.58 | *** part/#brlcad Forth (~Forth@92.242.118.253) |
| 19:12.50 | CIA-59 | BRL-CAD: 03bob1961 * r47686 10/brlcad/trunk/src/rt/do.c: Revert the previous commit. |
| 19:22.12 | CIA-59 | BRL-CAD: 03bob1961 * r47687 10/brlcad/trunk/src/rt/do.c: Check rtip for NULL before trying to use withing do_ae(). |
| 19:33.44 | CIA-59 | BRL-CAD: 03erikgreenwald * r47688 10/brlcad/trunk/src/libtclcad/tclcad_obj.c: remove uninitialized (and basically unused) ret variable |
| 19:34.34 | CIA-59 | BRL-CAD: 03starseeker * r47689 10/brlcad/trunk/ (3 files in 3 dirs): Clean up the OSX framework detection |
| 19:43.38 | CIA-59 | BRL-CAD: 03erikgreenwald * r47690 10/brlcad/trunk/src/librt/CMakeLists.txt: Set OBJ_BREP on all statics, not just windows |
| 19:59.17 | CIA-59 | BRL-CAD: 03starseeker * r47691 10/brlcad/trunk/src/libdm/CMakeLists.txt: the libdm static build does need some defines - add those. |
| 20:04.14 | CIA-59 | BRL-CAD: 03starseeker * r47692 10/brlcad/trunk/src/librt/CMakeLists.txt: just BUILD_STATIC_LIBS |
| 20:05.34 | CIA-59 | BRL-CAD: 03n_reed * r47693 10/brlcad/trunk/src/other/perplex/ (parser.y perplex.cpp perplex.h scanner.re scanner_template.c): add ability to generate a header file |
| 20:17.05 | CIA-59 | BRL-CAD: 03erikgreenwald * r47694 10/brlcad/trunk/src/libged/CMakeLists.txt: Move FB includes to end to try to shuffle macports include dir out... |
| 20:32.12 | starseeker | ``Erik: brlcad said defining that broke some other platform |
| 20:32.27 | starseeker | apparently defining that can have Bad Side-effects... |
| 20:35.48 | ``Erik | hm, thought I had it safely wrapped *shrug* |
| 21:05.24 | CIA-59 | BRL-CAD: 03bob1961 * r47695 10/brlcad/trunk/src/tclscripts/lib/Ged.tcl: ae2dir doesn't take a view parameter |
| 21:08.27 | brlcad | ``Erik: you did have it wrapped, and it still caused a problem because it was a compiler variable |
| 21:09.33 | brlcad | that's why the _* vars aren't supposed to be used, compiler is allowed to claim imminent domain on them and screw up caller code in seeminly unreasonable ways |
| 21:11.31 | brlcad | if I remember the problem correctly, it would fail the build even if you #undef _VAR ; #define _VAR 1 ; it would abort saying you're screwing with an already defined _VAR and no #ifdef logic could get around it |
| 21:12.42 | ``Erik | hm *shrug* if fails on my laptop, too (very up to date), but turning off strict allows it to pass |
| 21:12.49 | brlcad | so need some other fix (like always putting X11 includes last) |
| 21:14.31 | brlcad | so great, you have a repeatable platform environment where you can find another solution :P |
| 21:14.35 | CIA-59 | BRL-CAD: 03starseeker * r47696 10/brlcad/trunk/misc/CMake/BRLCAD_Util.cmake: Just remove duplicates regardless |
| 21:14.55 | starseeker | votes for smacking the zlib developers with a wet trout |
| 21:16.40 | ``Erik | also have slews of warnings from variatic macros in an X header, :D all good fun here |
| 21:17.50 | brlcad | I just got those for the first time today from an autotools build, FD_SET() and FD_ISSET() |
| 21:18.21 | brlcad | at least seemingly related |
| 21:18.45 | brlcad | error: ISO C forbids braced-groups within expressions |
| 21:28.32 | CIA-59 | BRL-CAD: 03n_reed * r47697 10/brlcad/trunk/src/other/perplex/ (parser.y scanner.re scanner_template.c): address compiler warnings |
| 21:35.12 | brlcad | starseeker: false alarm on the strict failure discrepancy, it wasn't compiling optimized which triggers the failure |
| 21:50.18 | CIA-59 | BRL-CAD: 03bob1961 * r47698 10/brlcad/trunk/src/librt/primitives/extrude/extrude.c: When using "put" to create an extrude primitive it's possible to get a segmentation fault if the sketch name is not specified. This updates rt_extrude_adjust() to check if sketch_name is NULL. |
| 21:53.44 | CIA-59 | BRL-CAD: 03erikgreenwald * r47699 10/brlcad/trunk/misc/CMake/Fink_MacPorts.cmake: add include dirs to remove command |
| 21:58.41 | starseeker | phew :-) |
| 21:59.26 | starseeker | looks like we can work around the zlib thing by disabling macports |
| 22:23.16 | CIA-59 | BRL-CAD: 03brlcad * r47700 10/brlcad/trunk/src/libged/polyclip.cpp: curr_lsg won't be initialized if the list is empty. make sure we don't dereference. |
| 22:28.54 | CIA-59 | BRL-CAD: 03brlcad * r47701 10/brlcad/trunk/include/ (CMakeLists.txt Makefile.am bselect.h): |
| 22:28.56 | CIA-59 | BRL-CAD: add a new header for wrapping the logic and functionality necessary for |
| 22:28.56 | CIA-59 | BRL-CAD: protecting against a bug in gcc 4.6.2 where the FD_*() macros aren't marked as |
| 22:28.56 | CIA-59 | BRL-CAD: extensions causing gcc to emit a warning during -O3 compilation about ISO C |
| 22:28.56 | CIA-59 | BRL-CAD: forbidding braced-groups within expressions. this may need some more |
| 22:28.56 | CIA-59 | BRL-CAD: precautions to ensure we only perform this workaround when absolutely necessary, |
| 22:28.57 | CIA-59 | BRL-CAD: but it seems to work well for the platform (linux) exhibiting the error. |
| 22:32.59 | CIA-59 | BRL-CAD: 03brlcad * r47702 10/brlcad/trunk/src/ (7 files in 5 dirs): |
| 22:32.59 | CIA-59 | BRL-CAD: replace includes of sys/select with the new bselect.h header so we can avoid a |
| 22:33.00 | CIA-59 | BRL-CAD: gcc 4.6.2 'ISO C forbids braced-groups within expressions' -O3 -pedantic error. |
| 22:33.00 | CIA-59 | BRL-CAD: looks like the '-Werror=edantic' problem is already reported. |
| 22:36.28 | CIA-59 | BRL-CAD: 03brlcad * r47703 10/brlcad/trunk/src/conv/enf-g.c: initialize variables before use. atof may fail. |
| 22:42.40 | CIA-59 | BRL-CAD: 03brlcad * r47704 10/brlcad/trunk/src/ (6 files in 3 dirs): few more select() caller stragglers that didn't include sys/select.h directly. include bselect.h ftw. |
| 23:08.31 | CIA-59 | BRL-CAD: 03brlcad * r47705 10/brlcad/trunk/src/librt/primitives/extrude/extrude.c: |
| 23:08.31 | CIA-59 | BRL-CAD: reworking more variables since compilation is halting on bugs that gcc 4.6.2 |
| 23:08.31 | CIA-59 | BRL-CAD: detected where we're calling V3ARGS on point2d_t variables being passed around. |
| 23:08.31 | CIA-59 | BRL-CAD: make isect_line2_ellipse() take 2d entities and promote the rest to 3d for now. |