| 01:05.52 | *** join/#brlcad FLOSSrookie (~brian@107-200-34-111.lightspeed.tulsok.sbcglobal.net) |
| 01:32.44 | brlcad | pretty cool: http://brlcad.org/tmp/command_line_progress.png |
| 01:44.18 | brlcad | also cool: https://code.google.com/p/ccons/ |
| 01:49.48 | *** join/#brlcad FLOSSrookie (~brian@107-200-34-111.lightspeed.tulsok.sbcglobal.net) |
| 03:01.08 | Notify | 03BRL-CAD:starseeker * 58290 (brlcad/trunk/src/CMakeLists.txt brlcad/trunk/src/librt/CMakeLists.txt): Based on patch from Cristina Precup, break out libnmg from librt. |
| 03:10.08 | brlcad | nice starseeker |
| 03:10.32 | brlcad | the only thing I notice there is that nmg.c probably doesn't belong in libnmg |
| 03:10.43 | brlcad | we'll have to break that up, it's a cyclic dependency |
| 03:11.17 | brlcad | librt needs to use libnmg similar to opennurbs, and have libnmg NOT use anything in librt |
| 05:50.24 | *** join/#brlcad milamber (~devlin@d118-75-244-176.try.wideopenwest.com) |
| 06:10.13 | *** join/#brlcad d_rossberg (~rossberg@66-118-151-70.static.sagonet.net) |
| 08:13.05 | *** join/#brlcad kesha (~kesha@14.139.122.114) |
| 08:19.33 | Notify | 03BRL-CAD Wiki:KeshaSShah * 6241 /wiki/Google_Code_In/2013/TODO: |
| 08:31.09 | Notify | 03BRL-CAD Wiki:CatinaonxkvrhlhzDorpinghaus * 0 /wiki/User:CatinaonxkvrhlhzDorpinghaus: |
| 11:31.54 | *** join/#brlcad kesha (~kesha@14.139.122.114) |
| 12:16.27 | Notify | 03BRL-CAD:starseeker * 58291 brlcad/trunk/src/other/CMakeLists.txt: d_rossberg noticed that clearing the src/other flags on MSVC causes problems, since the MSVC defaults are different from the CMake defaults and we don't wipe-and-reset all flags on MSVC currently. |
| 12:17.06 | starseeker | d_rossberg: give that a go and see how it works |
| 12:17.47 | starseeker | d_rossberg: the "real" solution would involve fully managing the MSVC flags like we do the gcc/clang flags, and that's a much larger undertaking... |
| 12:26.33 | Notify | 03BRL-CAD:starseeker * 58292 brlcad/trunk/misc/CMake/Distcheck.cmake: Quote CMake path in distcheck cmds |
| 12:29.38 | Notify | 03BRL-CAD:starseeker * 58293 brlcad/trunk/src/libnmg/CMakeLists.txt: Fix quote location |
| 12:38.44 | d_rossberg | starseeker: i was just testing another work around (the "TCL way"), but i'll have a look at yours too |
| 12:49.23 | Notify | 03BRL-CAD:phoenixyjll * 58294 brlcad/trunk/src/libbrep/intersect.cpp: Planar optimizations for SSI. |
| 13:27.38 | Notify | 03BRL-CAD:bob1961 * 58295 brlcad/trunk/src/tclscripts/archer/ArcherCore.tcl: Fixed a bug that breaks ArcherCore::redrawWho. This appears to have been introduced in r56397. |
| 14:25.13 | Notify | 03BRL-CAD Wiki:Mandeepkaur * 6242 /wiki/Google_Code_In/Checklis: |
| 14:26.19 | Notify | 03BRL-CAD Wiki:Mandeepkaur * 6243 /wiki/Google_Code_In/Checklis: |
| 14:34.49 | starseeker | urm. Is it a bad thing that rt_bot_mintie is global? |
| 15:00.15 | d_rossberg | starseeker: is it a bad thing that libnmg has over 80 errors (unresolved symbols)? |
| 15:07.54 | *** join/#brlcad kesha (~kesha@14.139.122.114) |
| 15:12.07 | d_rossberg | the stepcode stuff seems to work now |
| 15:18.50 | Notify | 03BRL-CAD Wiki:Mandeepkaur * 6244 /wiki/Google_Code_In/Checklis: |
| 16:00.19 | d_rossberg | hmm, linux works, msvc has errors in libnmg, all build from scratch ... go on on monday |
| 16:30.11 | *** join/#brlcad kesha (~kesha@14.139.122.114) |
| 16:46.46 | *** join/#brlcad kesha_ (~kesha@14.139.122.114) |
| 16:59.08 | brlcad | finds this situation doubley unfortuante ... we probably should be using /MT, not /MD |
| 17:07.51 | *** join/#brlcad kesha_ (~kesha@14.139.122.114) |
| 17:10.48 | Notify | 03BRL-CAD:bob1961 * 58296 brlcad/trunk/src/libged/open.c: Updated ged_reopen to properly set the rt material head. This would have impacted Archer and its ilk. |
| 17:30.18 | brlcad | AHA! |
| 17:42.13 | Notify | 03BRL-CAD:brlcad * 58297 brlcad/trunk/CMakeLists.txt: these should be indented |
| 17:44.38 | Notify | 03BRL-CAD:brlcad * 58298 brlcad/trunk/CMakeLists.txt: ws cleanup |
| 17:46.01 | Notify | 03BRL-CAD:brlcad * 58299 brlcad/trunk/CMakeLists.txt: no space on linker opts, mention why it's there |
| 17:49.21 | Notify | 03BRL-CAD:brlcad * 58300 (brlcad/trunk/CMakeLists.txt brlcad/trunk/misc/CMake/CompilerFlags.cmake): separation of macro definition from macro use. move the compiler logic into the compiler section of the top-level CMakeLists.txt file so that we can unconditionally include the macro definitions like a template header. |
| 17:57.01 | Notify | 03BRL-CAD:brlcad * 58301 brlcad/trunk/misc/CMake/CompilerFlags.cmake: pull includes together, document the macros as api not as how they are implemented. |
| 17:57.55 | Notify | 03BRL-CAD:brlcad * 58302 brlcad/trunk/CMakeLists.txt: should no longer need inclusion of CompilerFlags.cmake to be conditional as it just defines a bunch of macros |
| 18:02.00 | Notify | 03BRL-CAD:brlcad * 58303 brlcad/trunk/misc/CMake/BRLCAD_CompilerFlags.cmake: this file is not (yet) included on Windows, simplify |
| 18:03.41 | Notify | 03BRL-CAD Wiki:37.130.224.21 * 6245 /wiki/Mailing_Lists: |
| 18:04.19 | Notify | 03BRL-CAD:brlcad * 58304 brlcad/trunk/CMakeLists.txt: looks like there are some assumptions about the ordering, so preserve the previous ordering by moving the BRLCAD_CompilerFlags.cmake include after the logic that was moved. |
| 18:08.20 | Notify | 03BRL-CAD:carlmoore * 58305 brlcad/trunk/src/proc-db/lens.c: initialize lens_1side_side in its declaration; usage will now include defaults and unit specification |
| 18:18.45 | Notify | 03BRL-CAD:starseeker * 58306 (brlcad/trunk/src/CMakeLists.txt brlcad/trunk/src/librt/CMakeLists.txt): revert r58290 - it's not going to be that simple. Looks like quite a lot of librt-isms have crept into the nmg code, will be a *lot* more work to properly extract it. |
| 18:19.39 | starseeker | might as well close that patch - there's a lot of work to do that right, some of which is probably going to involve logic re-thinking to avoid using librt features |
| 18:27.43 | Notify | 03BRL-CAD:brlcad * 58307 brlcad/trunk/src/other/CMakeLists.txt: needs testing, but should no longer need the MSVC conditionals. It should default to msvc's default linkage (/MT) and is synchronized with the top-level build settings. |
| 18:28.12 | brlcad | starseeker: poopie.. |
| 18:28.36 | brlcad | what about if you put just nmg.c back .. what rt symbols remain? |
| 18:29.02 | brlcad | that's one I think we could get a fair bit of GCI activity on |
| 18:30.44 | Notify | 03BRL-CAD Wiki:159.253.145.183 * 6246 /wiki/MGED_Commands: |
| 18:30.46 | starseeker | based on the MSVC build, there are a lot of them scattered across many files |
| 18:31.06 | brlcad | remember any of them? |
| 18:31.30 | starseeker | rt_nurb_s_eval |
| 18:31.40 | starseeker | rt_vlblock_free |
| 18:31.49 | starseeker | rt_plot_vlblock |
| 18:32.09 | starseeker | hmm. - got a wdb_fopen in there... |
| 18:32.10 | Notify | 03BRL-CAD Wiki:159.253.145.183 * 6247 /wiki/BRL-CAD_Commands: |
| 18:32.18 | brlcad | k |
| 18:32.23 | starseeker | rt_prep_parallel |
| 18:32.25 | starseeker | etc. |
| 18:32.34 | brlcad | rt_nurb_s_eval is almost certainly in the bspline code, which could move or get killed |
| 18:32.37 | starseeker | db_path_to_string |
| 18:32.39 | brlcad | the other two are interesting |
| 18:32.57 | starseeker | rt_init_resource |
| 18:33.05 | starseeker | rt_gettree |
| 18:33.12 | brlcad | I get it ;) |
| 18:33.15 | starseeker | hehe |
| 18:33.16 | brlcad | more than a few |
| 18:33.35 | starseeker | frankly, I'm wondering why the gcc build succeeded |
| 18:33.45 | Notify | 03BRL-CAD Wiki:159.253.145.183 * 6248 /wiki/BRL-CAD_Primitives: |
| 18:33.49 | brlcad | gcc by default doesn't care about unresolved symbols |
| 18:34.03 | starseeker | winces - ah |
| 18:34.12 | brlcad | there are flags we can add that would provoke the same error |
| 18:35.42 | starseeker | brlcad: wiping the CMake MSVC flags may be a big deal... |
| 18:36.00 | brlcad | yep, warrants another test |
| 18:36.22 | brlcad | no reason it shouldn't work though any more than it shouldn't for gcc |
| 18:36.49 | brlcad | would like to know exactly what if it doesn't (and if so, we should probably test for it instead of relying on cmake) |
| 18:37.09 | brlcad | if only just so we can compile with other compilers on windows |
| 18:37.32 | starseeker | yeah, that's exactly why I didn't do it previously - the "correct" answer requires an understanding of the MSVC flags I don't have time to acquire |
| 18:38.01 | Notify | 03BRL-CAD Wiki:159.253.145.183 * 6249 /wiki/FAQ: |
| 18:38.02 | brlcad | there's not really anything special about msvc |
| 18:38.13 | brlcad | they're just different names for the same concepts |
| 18:38.29 | brlcad | if we want MT or MD or MTd or whatever, we should be testing for those like anything else I think |
| 18:39.05 | brlcad | e.g., debug should probably be MTd and release should be MD |
| 18:39.18 | Notify | 03BRL-CAD Wiki:159.253.145.183 * 6250 /wiki/MgedFAQ: |
| 18:39.54 | starseeker | hooray, an even longer Windows configure step. If the concepts are similar, could we define a per-compiler conceptual map and load those? E.g. set(DEBUG_FLAG MTd) in the MSVC file and set(DEBUG_FLAG g) in the gcc file? |
| 18:40.39 | brlcad | if we have to sure, but I'm hoping we can avoid doing that altogether |
| 18:40.39 | starseeker | then just test ${DEBUG_FLAG} rather than all debug flags for all compilers? |
| 18:40.51 | brlcad | at least until we encounter something that requires it, which we haven't yet |
| 18:41.04 | brlcad | we just test for flags in a priority order |
| 18:42.09 | brlcad | we really need to get something different like suncc or borland's new compiler into the mix so we can break our gcc assumptions |
| 18:42.24 | starseeker | winces - the configure stage is already about 1 minute on Linux, which will be > build time in a few machine generations... On Windows, it's more like 15-20 minutes |
| 18:43.16 | brlcad | the problem is exemplified by this latest issue though |
| 18:43.49 | brlcad | I could have reconfigured several thousand builds in the time it took to figure out that there was an if (NOT MSVC) that caused all of this |
| 18:43.58 | starseeker | if I had defined a "wipe flags" macro from the get go and re-used it it wouldn't have been an issue, so that's my mistake |
| 18:44.41 | starseeker | not several thousand MSVC builds |
| 18:46.00 | brlcad | when we have active windows maintainers, they can work on optimizing the windows build properly by profiling and reducing time appropriately |
| 18:46.54 | brlcad | until then, and without a profile, I don't think it's an issue worth accommodating at the expense of build logic maintainability |
| 18:47.18 | brlcad | plus, I bet the build time could be reduced far more effectively in other ways |
| 18:47.48 | brlcad | like instead of avoiding tests, avoid file operations unnecessarily ... there's a lot we could be going on in memory |
| 18:49.38 | Notify | 03BRL-CAD:carlmoore * 58308 brlcad/trunk/src/proc-db/room.c: remove a blank line; remove degtorad variable, because DEG2RAD global variable provides that value |
| 18:51.18 | brlcad | by a quick, back-of-the-napkin calculation, probably could have built on MSVC about 100 times in the time it took to discuss and resolve (if it is resolved) ... |
| 18:51.32 | brlcad | we simply don't build that many times, that's several years worth |
| 18:58.07 | *** join/#brlcad kesha__ (~kesha@14.139.122.114) |
| 19:08.07 | kanzure | do you guys have a lot of windows users? |
| 19:57.10 | starseeker | kanzure: https://sourceforge.net/projects/brlcad/files/stats/os |
| 19:57.30 | kanzure | wtf |
| 19:57.47 | kanzure | alright |
| 19:57.59 | Notify | 03BRL-CAD Wiki:Sean * 6251 /wiki/MgedFAQ: Reverted edits by [[Special:Contributions/159.253.145.183|159.253.145.183]] ([[User talk:159.253.145.183|talk]]) to last revision by [[User:Tbrowder|Tbrowder]] |
| 19:58.01 | Notify | 03BRL-CAD Wiki:Sean * 6252 /wiki/FAQ: Reverted edits by [[Special:Contributions/159.253.145.183|159.253.145.183]] ([[User talk:159.253.145.183|talk]]) to last revision by [[User:Tbrowder|Tbrowder]] |
| 19:58.04 | Notify | 03BRL-CAD Wiki:Sean * 6253 /wiki/BRL-CAD_Primitives: Reverted edits by [[Special:Contributions/159.253.145.183|159.253.145.183]] ([[User talk:159.253.145.183|talk]]) to last revision by [[User:JoelDBenson|JoelDBenson]] |
| 19:58.06 | Notify | 03BRL-CAD Wiki:Sean * 6254 /wiki/BRL-CAD_Commands: Reverted edits by [[Special:Contributions/159.253.145.183|159.253.145.183]] ([[User talk:159.253.145.183|talk]]) to last revision by [[User:Mdupont|Mdupont]] |
| 19:58.10 | Notify | 03BRL-CAD Wiki:Sean * 6255 /wiki/MGED_Commands: Reverted edits by [[Special:Contributions/159.253.145.183|159.253.145.183]] ([[User talk:159.253.145.183|talk]]) to last revision by [[User:Phoenix|Phoenix]] |
| 19:58.12 | Notify | 03BRL-CAD Wiki:Sean * 6256 /wiki/Mailing_Lists: Reverted edits by [[Special:Contributions/37.130.224.21|37.130.224.21]] ([[User talk:37.130.224.21|talk]]) to last revision by [[User:Sean|Sean]] |
| 19:58.35 | Notify | 03BRL-CAD Wiki:Sean * 0 /wiki/User:159.253.145.183: Spamming links to external sites |
| 19:59.01 | Notify | 03BRL-CAD Wiki:Sean * 0 /wiki/User:37.130.224.21: Spamming links to external sites |
| 19:59.31 | Notify | 03BRL-CAD Wiki:Sean * 6257 /wiki/Google_Code_In: Reverted edits by [[Special:Contributions/117.207.157.190|117.207.157.190]] ([[User talk:117.207.157.190|talk]]) to last revision by [[User:71.198.169.135|71.198.169.135]] |
| 20:00.25 | starseeker | brlcad: should we set up the clang static analyzer stuff for GCI? Or are those likely to be a little too tricky for the target group? |
| 20:27.13 | Notify | 03BRL-CAD:starseeker * 58309 brlcad/trunk/src/libbn/obr.c: Add in the Melkman code needed for convex hulls... |
| 20:36.44 | *** join/#brlcad kesha__ (~kesha@14.139.122.114) |
| 21:34.24 | brlcad | starseeker: probably not a full-blown analysis, but perhaps categories of specific detections that have a well-documented fix |
| 21:35.11 | brlcad | not something generic like "fix 10 of these", but like "fix all the shadow warnings by renaming variables" |
| 21:54.27 | Notify | 03BRL-CAD:carlmoore * 58310 brlcad/trunk/src/proc-db/lens.c: shortened the code |
| 22:10.05 | *** join/#brlcad kesha__ (~kesha@14.139.122.114) |
| 22:54.07 | Notify | 03BRL-CAD Wiki:Arlo James Barnes * 0 /wiki/User:Arlo_James_Barnes: |
| 22:59.16 | Notify | 03BRL-CAD Wiki:Sean * 6258 /wiki/Community_Publication_Portal: fix the final review section |
| 23:00.22 | Notify | 03BRL-CAD Wiki:Sean * 6259 /wiki/Community_Publication_Portal: 7.24.0 was posted |
