| 00:49.06 | abhi2011 | ok so I am trying to add a new .cpp file to src/libged where all the c++ code to access bullet physics will go |
| 00:49.40 | abhi2011 | the cpp file has a header which I include in the mged command implementation file |
| 00:50.43 | abhi2011 | So my c++ test file is simphysics.cpp with some test code |
| 00:51.45 | abhi2011 | http://bin.cakephp.org/view/1739302378 |
| 00:53.23 | abhi2011 | I have a header for it and this is included in simulate.c which has the libged implementation of the simulate command : http://bin.cakephp.org/view/1706385166 |
| 00:54.24 | abhi2011 | the cpp file has the definition of a single function single_step_sim() which uses the cout object to print out some test message |
| 00:55.19 | abhi2011 | I get a linking error when I compile which says single_step_sim() is an undefined reference |
| 00:56.39 | abhi2011 | However since I have included the header (which declares single_step_sim() ) in simulate.c the compiler should be able to see the declaration |
| 00:57.14 | abhi2011 | Perhaps it has to be exported using GED_EXPORT, but its not called from outside the library so it shouldnt be required to export it |
| 01:06.51 | abhi2011 | this is the linking error : http://bin.cakephp.org/view/1997617995 |
| 05:15.44 | *** join/#brlcad Stattrav (~Stattrav@203.196.190.162) |
| 05:15.44 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) |
| 06:04.43 | *** join/#brlcad emagdalenag (~emagdalen@109.Red-88-4-184.dynamicIP.rima-tde.net) |
| 10:23.01 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) |
| 11:32.54 | CIA-62 | BRL-CAD: 03Kunigami 07http://brlcad.org * r3056 10/wiki/User:Kunigami/GSoc2011/Reports: /* Week 11 (August 1st to August 8th) */ log for past week |
| 12:30.41 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) |
| 12:36.39 | abhi2011 | has anyone tried to add c++ files to existing code, that is mixing c with c++ |
| 12:38.07 | brlcad | abhi2011: adding c++ to c files would make them c++ files |
| 12:40.40 | brlcad | you'll want to write a bridge, i.e., one or more C functions that wrap all of your C++ logic |
| 12:40.42 | abhi2011 | ok I am trying to integrate bullet c++ code into brlcad |
| 12:40.48 | brlcad | I know |
| 12:40.49 | abhi2011 | right |
| 12:41.08 | brlcad | those bridging functions will need to be marked extern "C" since they're in c++ files |
| 12:41.16 | brlcad | so that they're not name-mangled |
| 12:42.52 | abhi2011 | yes I have written a separate function in a cpp file simphysics.cpp that has all the c++ code, I have declared this function in a header simphysics.h |
| 12:42.54 | *** join/#brlcad emagdalenag (~emagdalen@179.Red-88-4-185.dynamicIP.rima-tde.net) |
| 12:43.10 | abhi2011 | i include this header in simulate.c which is the libged simulate command implementation |
| 12:43.41 | abhi2011 | in the header where i declare the c++ function, I can declare it as extern |
| 12:43.55 | brlcad | it's not just extern |
| 12:43.58 | brlcad | it's extern "C" |
| 12:44.05 | abhi2011 | ah ok |
| 12:44.20 | brlcad | search for that in existing include/*.h headers |
| 12:44.21 | abhi2011 | i get it |
| 12:44.27 | abhi2011 | right |
| 12:44.39 | brlcad | stop saying right :) |
| 12:45.05 | abhi2011 | :P, yes |
| 12:51.46 | brlcad | heh, http://www.sbir.gov/sbirsearch/detail/93829 |
| 12:52.06 | brlcad | apparently ARA got half a million back in the mid-90's to develop that .. interesting, never saw it |
| 12:52.35 | brlcad | (assuming they got phase 2 funding, maybe only phase 1) |
| 13:40.55 | abhi2011 | interestng |
| 14:19.10 | *** join/#brlcad emagdalenag (~emagdalen@41.Red-88-21-190.staticIP.rima-tde.net) |
| 14:52.47 | *** join/#brlcad abhi2011 (~chatzilla@wlan-145-94-187-155.wlan.tudelft.nl) |
| 15:25.29 | CIA-62 | BRL-CAD: 03brlcad * r45808 10/brlcad/trunk/src/libwdb/skt.c: there already exists a copy function for deep copying, so use it. |
| 15:27.33 | CIA-62 | BRL-CAD: 03brlcad * r45809 10/brlcad/trunk/src/libwdb/ (CMakeLists.txt Makefile.am sketch.c skt.c): rename from skt.c to sketch.c because it's easier to find the file if the filename matches the primitive's canonical short name. |
| 15:35.56 | *** join/#brlcad kunigami (~kunigami@201.53.206.27) |
| 15:49.05 | *** join/#brlcad fosburg (~steve@or-67-238-17-118.dhcp.embarqhsd.net) |
| 15:49.59 | fosburg | anyone here |
| 16:04.54 | *** join/#brlcad kunigami (~kunigami@201.53.206.27) |
| 16:07.57 | *** join/#brlcad kunigami (~kunigami@201.53.206.27) |
| 16:11.46 | *** join/#brlcad fosburg (~steve@or-67-238-17-118.dhcp.embarqhsd.net) |
| 16:16.06 | kunigami | rt is running out of single letter options :) shouldn't we allow more for more mnemonic options? |
| 16:16.46 | kunigami | I'd like to add a framebuffer mode. Bothb,B, f and F are already taken, but -fb would be better anyway |
| 16:23.47 | brlcad | kunigami: we don't (yet) have proper infrastructure in place for --long-options |
| 16:25.02 | brlcad | with getopt, -fb is equivalent to "-f -b", so at best you'd need something like --fb |
| 16:25.12 | brlcad | that said, what do you mean by framebuffer mode? |
| 16:27.20 | CIA-62 | BRL-CAD: 03brlcad * r45810 10/brlcad/trunk/include/wdb.h: meant to commit this with r45808, made the passed parameter const now that it's properly copied before export. |
| 16:28.36 | CIA-62 | BRL-CAD: 03brlcad * r45811 10/brlcad/trunk/include/raytrace.h: add a new type, rt_curve, for use by sketch and extrude definitions. |
| 16:38.40 | abhi2011 | does anyone else get warnings about the version infomation being missing from libz : /usr/bin/cmake: /usr/brlcad/dev-7.20.3/lib/libz.so.1: no version information available (required by /usr/bin/cmake) |
| 16:38.58 | abhi2011 | I get this during cmake and during make install as well |
| 16:39.38 | abhi2011 | maybe the libz version info needs to be patched |
| 17:03.11 | *** join/#brlcad fosburg (~steve@or-67-238-17-118.dhcp.embarqhsd.net) |
| 17:05.32 | kunigami | brlcad: something like we discussed on email: the frequency on which the framebuffer is updated and how rays are shot. this would form two options. I want to add an option for the "how rays are shot". fb was just an example for a long letter example. Which name would better describe this parameter? tr? |
| 17:08.05 | brlcad | that rabbit hole runs a little deep :) |
| 17:09.25 | brlcad | my point wasn't specific to -fb, -abc == -a -b -c to getopt (which is highly desireable in some contexts) |
| 17:09.42 | brlcad | the feature missing is long-opts support, which would let you specify longer name options in addition to one-letter optiosn |
| 17:10.15 | brlcad | half of rt's short options would be better as long-only options for some form of advanced control |
| 17:11.38 | brlcad | but like I said, the infrastructure for long options isn't implemented, so better to just pick one of the few remaining one-letter options for now unless you want to tackle implementing long options properly (new libbu api) |
| 17:12.12 | kunigami | hehe ok. Will choose one of the remaining :) |
| 17:12.19 | brlcad | :) |
| 17:13.01 | brlcad | looks like -m -y -z -L -Y -Z plus some punctuation are all that remain |
| 17:15.39 | brlcad | given -l is lighting model, -L for firing pattern might work well |
| 17:18.36 | brlcad | another option might be to overload the -Q option or -B option |
| 17:18.51 | brlcad | looks like -b might be fair game too.. |
| 17:22.08 | brlcad | definitely fair game, it's the same as -j |
| 17:22.54 | brlcad | so I suggest -L, -Q, or -b |
| 17:23.40 | brlcad | the latter two require a little refactoring, but seem most appropriate |
| 17:25.43 | brlcad | kunigami: suggest writing up the proposed usage change like what bhinesley did for edit, either as a comment or wiki dev page would be sufficient |
| 17:26.05 | brlcad | to make sure the usage is clean and impact minimal before you run down a path |
| 17:27.14 | brlcad | rt is a critical path, so should make sure there's extra thought put into how the new options will get used |
| 17:33.31 | CIA-62 | BRL-CAD: 03brlcad * r45812 10/brlcad/trunk/src/libwdb/ (CMakeLists.txt Makefile.am extr.c extrude.c): rename extr.c to extrude.c so that the filename matches the primitiv'es canonical short name. makes it easier to identify what this file contains. |
| 17:35.33 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) |
| 17:44.47 | kunigami | brlcad: comment right on the code? like rt/main.c or rt/opt.c? |
| 17:45.40 | brlcad | anywhere, just some place we can collectively see it |
| 17:45.52 | kunigami | ok! |
| 17:58.49 | *** join/#brlcad fosburg (~steve@or-67-238-17-118.dhcp.embarqhsd.net) |
| 18:12.14 | brlcad | excellent |
| 18:12.35 | brlcad | couple usage examples would be helpful too in addition to more formalized usage syntax |
| 18:23.00 | CIA-62 | BRL-CAD: 03brlcad * r45813 10/brlcad/trunk/src/tclscripts/menu_override.tcl: |
| 18:23.00 | CIA-62 | BRL-CAD: example of why it's useful to see the tclsh warnings.. remove a handful of |
| 18:23.00 | CIA-62 | BRL-CAD: ::tk:: functions that curiously are living in our sources, yet they don't appear |
| 18:23.00 | CIA-62 | BRL-CAD: to have any overridden behavior. menus in mged tested fine without, so let the |
| 18:23.01 | CIA-62 | BRL-CAD: tk guys manage their own code. remove our fork. |
| 18:25.58 | ``Erik | starseeker: tclUnixInit.c:367 or so, #include <sys/utsname.h> for uname(3) on line 471 or so |
| 18:26.22 | CIA-62 | BRL-CAD: 03brlcad * r45814 10/brlcad/trunk/misc/Doxyfile: output to 'doxygen_output' in whatever our current directory is, no point putting it into misc |
| 18:28.06 | CIA-62 | BRL-CAD: 03Kunigami 07http://brlcad.org * r3057 10/wiki/User:Kunigami/GSoc2011/RT_Parameters_Proposal: draft to present the new options to be added RT application. Waiting for review before implementing |
| 18:28.28 | CIA-62 | BRL-CAD: 03Kunigami 07http://brlcad.org * r3058 10/wiki/User:Kunigami/GSoc2011/RT_Parameters_Proposal: /* Examples */ correct formatting |
| 18:29.28 | *** join/#brlcad tofu (~sean@BZ.BZFLAG.BZ) |
| 18:29.38 | kunigami | may I already add osl and oiio code to the svn? |
| 18:30.03 | tofu | how big are they? |
| 18:30.48 | *** mode/#brlcad [+o brlcad] by ChanServ |
| 18:31.01 | kunigami | osl is 5MB |
| 18:31.24 | kunigami | oiio is 9.4MB |
| 18:31.50 | kunigami | or we'll let them be optional external code? |
| 18:31.56 | brlcad | do they have any dependencies of their own? |
| 18:32.09 | brlcad | it depends just how complex they are |
| 18:32.20 | kunigami | yes, a quite long one by the way |
| 18:33.35 | brlcad | do you have the build documented somewhere? |
| 18:33.48 | kunigami | http://brlcad.org/wiki/User:Kunigami/GSoc2011/OSL_Tutorial |
| 18:33.59 | brlcad | excellent, thought you had |
| 18:34.58 | starseeker | as i recall, OSL has quite the nasty dependency list |
| 18:35.02 | kunigami | but I didn't mentioned all the dependencies. Most of them are aviable like in mac ports or ubuntu synaptic |
| 18:35.08 | starseeker | llvm |
| 18:35.10 | kunigami | starseeker: yup |
| 18:36.24 | brlcad | ah, right, llvm .. that's certainly a pickle |
| 18:37.14 | brlcad | pretty much guarantees it'll be disabled on release configurations unless we do some heavy lifting to make it easier |
| 18:37.16 | starseeker | boost, imath, openimageio |
| 18:38.25 | starseeker | (imath looks like it may be part of openexr) |
| 18:38.45 | kunigami | yes, I had to install openexr |
| 18:38.53 | brlcad | kunigami: how about checking all of the deps into a separate module |
| 18:39.24 | starseeker | not surpring, openexr really has taken off in the movie industry and that's sony's target use case for this... |
| 18:39.31 | brlcad | we can hook those in later as an svn external if we want, or extra download step |
| 18:39.41 | starseeker | votes for that appraoch |
| 18:39.41 | brlcad | but that way, the N deps are all in one place |
| 18:39.43 | starseeker | approach even |
| 18:40.51 | brlcad | then their compilation can be wrapped up with even a simple Makefile that will compile/install all in the proper order |
| 18:41.01 | brlcad | <PROTECTED> |
| 18:41.17 | brlcad | kunigami: do you know how to create a new module? |
| 18:41.18 | starseeker | osl itself uses cmake |
| 18:41.29 | kunigami | kunigami: nope |
| 18:41.34 | kunigami | ops |
| 18:41.37 | kunigami | brlcad: nope |
| 18:41.38 | brlcad | sure, but it's just one of about a half-dozen or so |
| 18:42.08 | starseeker | openimagio has one, it looks like |
| 18:42.13 | starseeker | pretty sure llvm does |
| 18:42.39 | kunigami | you mean cmake module? like FindXYZ.cmake? |
| 18:43.00 | starseeker | no, CMakeLists.txt file |
| 18:43.08 | brlcad | kunigami: basically, you'll checkout https://brlcad.svn.sourceforge.net/svnroot/brlcad (perhaps with max-depth set to 0 so it doesn't check out the whole repo) and add an 'osl' dir with three subdirs, 'trunk', 'branches', 'tags' |
| 18:43.56 | brlcad | then add the various OSL packages as subdirs under /svnroot/brlcad/osl/trunk/. |
| 18:44.30 | brlcad | one dir for each dep, openexr, llmbase, boost, etc |
| 18:44.52 | brlcad | can skip any external deps that are in /svnroot/brlcad/brlcad/trunk/src/other (like libpng, libz, etc) |
| 18:45.43 | starseeker | hmm... openexr doesn't have one |
| 18:45.52 | starseeker | wonder if they'd accept one |
| 18:46.16 | brlcad | unless we plan on fully managing them, wouldn't bother |
| 18:46.55 | brlcad | only reason to add them is to have them in revision control so they can be conveniently downloaded if we want to push out binaries with osl enabled |
| 18:47.57 | brlcad | exr-pix would be cool, that'd be a reason :) |
| 18:49.10 | starseeker | isn't quite sure how exr would integrate with our tools - we might have to make rt and friends output straight to exr if we wanted to take full advantage of it... |
| 19:03.39 | CIA-62 | BRL-CAD: 03kunigami * r45815 10/osl/ (. branches/ tags/ trunk/): setting up initial structure for osl code and dependences |
| 19:06.27 | *** join/#brlcad fosburg (~steve@or-67-238-17-118.dhcp.embarqhsd.net) |
| 19:12.31 | CIA-62 | BRL-CAD: 03starseeker * r45816 10/brlcad/trunk/src/other/tcl/CMakeLists.txt: Add one missing file to Windows source list - this new setup, for the first time, allows for a successful run of wish.exe |
| 19:22.44 | *** join/#brlcad emagdalena (~emagdalen@216.Red-88-4-184.dynamicIP.rima-tde.net) |
| 19:38.33 | CIA-62 | BRL-CAD: 03brlcad * r45817 10/brlcad/trunk/misc/ (CMakeLists.txt Makefile.am): the Doxygen paths are already relative, so no need to auto-generate the file with full-paths shoved in. write out to the doc dir. |
| 19:56.07 | CIA-62 | BRL-CAD: 03Kunigami 07http://brlcad.org * r3059 10/wiki/User:Kunigami: /* Links */ added link to my RT parameters proposal |
| 20:01.04 | *** join/#brlcad dli (~dli@dsl-67-204-17-156.acanac.net) |
| 20:03.49 | CIA-62 | BRL-CAD: 03brlcad * r45818 10/brlcad/trunk/ (16 files in 10 dirs): |
| 20:03.50 | CIA-62 | BRL-CAD: fix a FIXME, rename the struct curve definition that was embedded into |
| 20:03.50 | CIA-62 | BRL-CAD: rt_sketch_internal. this creates a new rt_curve structure and moves the |
| 20:03.50 | CIA-62 | BRL-CAD: associated segment structures into rtgeom. unfortunately requires nmg.h for the |
| 20:03.50 | CIA-62 | BRL-CAD: knot_vector and cascades a set of name changes, but improvement all around. |
| 20:17.54 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) |
| 20:32.24 | CIA-62 | BRL-CAD: 03kunigami * r45819 10/osl/trunk/osl/: add just the empty dir for mime-type testing |
| 20:35.00 | CIA-62 | BRL-CAD: 03kunigami * r45820 10/osl/trunk/osl/CHANGES: add just the empty dir for mime-type testing |
| 21:54.45 | *** join/#brlcad ibot (~ibot@rikers.org) |
| 21:54.46 | *** 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! |
| 22:05.20 | CIA-62 | BRL-CAD: 03brlcad * r45827 10/brlcad/trunk/ (include/raytrace.h src/librt/primitives/sketch/sketch.c): segment magic numbers are presently unsigned long, update pointers accordingly |
| 22:15.48 | CIA-62 | BRL-CAD: 03brlcad * r45828 10/brlcad/trunk/src/proc-db/sketch.c: now that mk_sketch() releases memory, we can simplify by creating the sketch and all segments on the stack. |
| 22:19.13 | CIA-62 | BRL-CAD: 03kunigami * r45829 10/osl/trunk/oiio/ (399 files in 82 dirs): adding missing files to oiio |
| 22:21.18 | CIA-62 | BRL-CAD: 03brlcad * r45830 10/brlcad/trunk/src/proc-db/sketch.c: come to think of it, we don't need ANY stinking dynamic memory allocation. huzzah. |
| 22:25.58 | CIA-62 | BRL-CAD: 03brlcad * r45831 10/brlcad/trunk/src/proc-db/sketch.c: reduce some more and document the vars |
| 23:38.46 | CIA-62 | BRL-CAD: 03kunigami * r45832 10/osl/trunk/ilmbase/ (245 files in 69 dirs): adding latest stable version of ilmbase, 1.0.1 |
| 23:46.50 | CIA-62 | BRL-CAD: 03kunigami * r45833 10/osl/trunk/ilmbase/config/Makefile.in: added missing file to ilmbase |
| 23:51.29 | CIA-62 | BRL-CAD: 03kunigami * r45834 10/osl/trunk/ilmbase/ (7 files in 7 dirs): added missing file to ilmbase (cant understand why some files are not being added) |