| 02:52.29 | *** join/#brlcad ibot (~ibot@rikers.org) |
| 02:52.29 | *** topic/#brlcad is BRL-CAD Open Source Solid Modeling || http://brlcad.org || http://sf.net/projects/brlcad || #brlcad logs: http://ibot.rikers.org/%23brlcad/ || BRL-CAD release 7.20.2 is posted (20110701) || BRL-CAD is participating in the ESA Summer of Code in Space! |
| 03:49.08 | *** join/#brlcad Technicus (~Technicus@DSLPool-net208-2.wctc.net) |
| 07:37.45 | *** join/#brlcad alex_joni (~alex_joni@emc/board-of-directors/alexjoni) |
| 08:11.18 | *** join/#brlcad packrat (~packrator@c-98-209-146-133.hsd1.mi.comcast.net) |
| 09:14.28 | *** join/#brlcad packrat (~packrator@c-98-209-146-133.hsd1.mi.comcast.net) |
| 13:01.00 | *** join/#brlcad abhi2011 (~chatzilla@117.200.90.204) |
| 13:58.00 | *** join/#brlcad abhi2011 (~chatzilla@117.200.88.113) |
| 14:30.48 | *** join/#brlcad abhi2011 (~chatzilla@117.200.88.113) |
| 14:44.31 | brlcad | starseeker: I know, and that was one of the points being made .. they're never "worth it" in the long run :) |
| 14:45.23 | brlcad | at least not when maintaining excessive portability (backwards and forewards) is the goal |
| 14:59.38 | brlcad | and it's worth saying that "it's all good" given past misinterpretations of ranting -- just a lot to say on the topic ;) |
| 15:12.49 | starseeker | do I undersdand you correctly that you want to change the conditionalized mechanism currently being used in the .h files? |
| 15:13.01 | starseeker | s/undersdand/understand |
| 15:13.07 | starseeker | kick brain into gear |
| 15:13.55 | starseeker | or just change how the CMake logic triggers it? |
| 15:19.16 | starseeker | is certainly in favor of excessive portability :-) - just not sure how that blasted import/export trick Windows needs can be made to play nicely |
| 15:21.12 | starseeker | both the shared library and executable targets currently need BRLCAD_DLL with MSVC, but if I understand correctly the static libraries *shouldn't* have it |
| 15:23.43 | CIA-109 | BRL-CAD: 03brlcad * r47582 10/brlcad/trunk/src/libbu/vls.c: |
| 15:23.43 | CIA-109 | BRL-CAD: it was an interesting idea, but not a great one. did a quick test to see how |
| 15:23.43 | CIA-109 | BRL-CAD: much time might be gained if we skipped the initial vls allocation. looked to |
| 15:23.43 | CIA-109 | BRL-CAD: be about 25% for bu_vls_printf() which is marginally interesting at best. |
| 15:23.43 | CIA-109 | BRL-CAD: probably not worth the complexity and long-term maintenance (error-prone), at |
| 15:23.44 | CIA-109 | BRL-CAD: least for now. |
| 15:23.57 | starseeker | which rules out any global setting of it, unless... perhaps we want to have toplevel BRLCAD_SHARED_COMPILE_FLAGS and BRLCAD_STATIC_COMPILE_FLAGS variables? |
| 15:32.18 | brlcad | starseeker: the .h files still are conditionalized, changing how cmake triggers |
| 15:41.35 | brlcad | probably don't need different flag variables |
| 15:43.25 | brlcad | if you conditionally set flags and add them (as vars), then those variables you add them to are implicitly conditionalized too |
| 15:45.15 | brlcad | I think the problem stems from the logic in bu.h presently only providing BU_EXPORT_DLL with no corresponding BU_IMPORT_DLL |
| 15:45.41 | brlcad | not export does not mean import .. e.g., when compiling static |
| 15:46.46 | brlcad | so that could simplify to a three-way if/elseif/else toggling on just those two variables -- then cmake has to set either BU_EXPORT_DLL or BU_IMPORT_DLL or neither |
| 15:48.19 | brlcad | with that, BRLCAD_DLL can go away and a cmake test is needed to determine whether __declspec(dllimport) works .. if it does, then variables get triggered |
| 15:51.45 | brlcad | if (dllimport_works) then LIBBU_CPPFLAGS+="-DBU_EXPORT_DLL" ; LIBBU_STATIC_CPPFLAGS+="..nada.." ; bu-using non-static apps CPPFLAGS+="-DBU_IMPORT_DLL" |
| 15:52.48 | brlcad | may need a layer of variables in there to avoid duplicating information all over the place but that's the gist in pseudocode |
| 16:40.41 | CIA-109 | BRL-CAD: 03brlcad * r47583 10/brlcad/trunk/NEWS: |
| 16:40.41 | CIA-109 | BRL-CAD: butler added an initial stab and providing ambient occlusion to rt. this is |
| 16:40.41 | CIA-109 | BRL-CAD: presently disabled by default and enabled with the ambSamples and ambRadius rt |
| 16:40.41 | CIA-109 | BRL-CAD: variables. more work is needed on controlling the sample pattern and noise. |
| 16:42.23 | *** join/#brlcad n_reed (~molto_cre@BZ.BZFLAG.BZ) |
| 17:43.04 | CIA-109 | BRL-CAD: 03n_reed * r47584 10/brlcad/trunk/doc/bison_to_lemon.txt: more on assigning types to symbols |
| 17:45.49 | CIA-109 | BRL-CAD: 03n_reed * r47585 10/brlcad/trunk/src/other/perplex/ (scanner.re template.c): don't allocate new token string without freeing existing string |
| 18:04.28 | CIA-109 | BRL-CAD: 03n_reed * r47586 10/brlcad/trunk/src/other/perplex/ (parser.y scanner.re): fix separator pattern; properly close output scanner |
| 18:26.59 | *** join/#brlcad jordisayol (~jordisayo@unaffiliated/jordisayol) |
| 19:30.09 | *** join/#brlcad Yoshi47 (~jan@64.235.102.210) |
| 19:55.47 | *** join/#brlcad Forth (~Forth@92.242.118.253) |
| 19:57.53 | *** part/#brlcad Forth (~Forth@92.242.118.253) |
| 20:13.07 | CIA-109 | BRL-CAD: 03n_reed * r47587 10/brlcad/trunk/src/other/perplex/ (Makefile.local perplex.h scanner.re template.c): fixed start condition initialization; removed requirement for EOF rule in input |
| 20:27.47 | brlcad | starseeker: n_reed: e-mail sent, assistance requested |
| 20:28.04 | *** join/#brlcad merzo (~merzo@19-255-132-95.pool.ukrtel.net) |
| 21:25.33 | *** join/#brlcad ``Erik (Here@c-69-140-109-104.hsd1.md.comcast.net) |
| 22:46.58 | *** join/#brlcad velociostrich (~nicholas@c-24-0-153-224.hsd1.pa.comcast.net) |
| 22:47.18 | *** part/#brlcad velociostrich (~nicholas@c-24-0-153-224.hsd1.pa.comcast.net) |
| 22:53.09 | *** part/#brlcad n_reed (~molto_cre@BZ.BZFLAG.BZ) |