| 00:00.21 | abhi2011 | though in my original program, I was detecting a region by passing over the region list in the rtip and checking for name matches |
| 00:03.43 | abhi2011 | so to detect the fact that the user has passed a region in a struct rt_db_internal , there should be a field in the struct rt_db_internal that identifies the contents as a region |
| 00:04.47 | abhi2011 | or the user would need to pass the model's directory so I can try name matches to detect |
| 00:20.10 | CIA-62 | BRL-CAD: 03kunigami * r45912 10/osl/trunk/osl/src/cmake/externalpackages.cmake: force cmake to ilmbase, openexr and llvm in the path os osl/trunk |
| 00:21.32 | CIA-62 | BRL-CAD: 03kunigami * r45913 10/osl/trunk/osl/src/testshade/CMakeLists.txt: disable testshade_dso app because it uses an oiio version that is not stable (its not used by brlcad) |
| 00:52.05 | CIA-62 | BRL-CAD: 03bhinesley * r45914 10/brlcad/trunk/src/libged/edit.c: |
| 00:52.05 | CIA-62 | BRL-CAD: With the improved edit_*_get_arg_head()'s, individual subcommand init functions |
| 00:52.05 | CIA-62 | BRL-CAD: are no longer necessary; a generic edit_cmd_init() can handle it all. Reduces |
| 00:52.05 | CIA-62 | BRL-CAD: the number of functions needed for each subcommand to 4; nice. Also in this |
| 00:52.05 | CIA-62 | BRL-CAD: commit: noticed a variable for a return value being initialized to a literal |
| 00:52.06 | CIA-62 | BRL-CAD: value, rather than GED_OK. |
| 01:53.16 | CIA-62 | BRL-CAD: 03bhinesley * r45915 10/brlcad/trunk/src/libged/edit.c: regrouping/sorting functions a bit |
| 02:05.14 | starseeker | bhinesley: um. that's probably the html viewer not finding its stuff |
| 02:05.25 | starseeker | looks at what mged is up too... |
| 02:06.11 | bhinesley | starseeker: yeah, it's the one that I wrote :-P |
| 02:06.31 | starseeker | can you see where it's failing? |
| 02:06.37 | bhinesley | yeah |
| 02:07.05 | bhinesley | man_browser.tcl:124 |
| 02:07.48 | starseeker | what's in $path? |
| 02:07.56 | bhinesley | I'm checking |
| 02:08.01 | starseeker | will see if he can duplicate the failure, one sec... |
| 02:08.44 | starseeker | regardless, we probably want to add -nocomplain to that glob - that's not a reason for archer to fail to start |
| 02:08.55 | bhinesley | ah there it is; its using the default for the class itself on :68 |
| 02:09.17 | bhinesley | starseeker: true |
| 02:11.26 | bhinesley | seems the other help browser needs it as well |
| 02:14.33 | bhinesley | hm, or not. |
| 02:14.47 | bhinesley | I'm still getting an error, even with -nocomplain |
| 02:17.00 | CIA-62 | BRL-CAD: 03starseeker * r45916 10/brlcad/trunk/src/tclscripts/mged/help.tcl: I think we want to get helplib.tcl from the tclscripts dir? |
| 02:18.15 | starseeker | bhinesley: I'm guessing share/brlcad/7.20.3/html doesn't have anything in it? |
| 02:18.59 | bhinesley | there are folders and files |
| 02:19.05 | starseeker | O.o |
| 02:19.06 | starseeker | weird |
| 02:19.22 | starseeker | what does puts $path print out? |
| 02:21.55 | bhinesley | well it's looking for ./share/brlcad/7.20.3/html/mann/en/ |
| 02:22.09 | starseeker | which does exist? |
| 02:22.20 | bhinesley | mann doesn't |
| 02:22.34 | starseeker | urm. that's the problem then |
| 02:32.04 | CIA-62 | BRL-CAD: 03starseeker * r45917 10/brlcad/trunk/src/tclscripts/man_browser.tcl: Don't refuse to start archer just because the html docs aren't around. |
| 02:32.22 | bhinesley | was just about to do that ;) |
| 02:33.28 | starseeker | one remaining wrinkle - when I nuked all of html and then re-ran make, the docs rebuilt but toc.html wasn't present - it's copied during the configure process, not the make process |
| 02:33.49 | bhinesley | meh... your way looks better anyways |
| 02:33.56 | starseeker | probably should have the archer help viewer check for that file as well as one of the generated files |
| 02:35.51 | starseeker | wonders reworking the "copy during configure" cases into build rules... wonder if that's practical/worth it... |
| 02:40.52 | CIA-62 | BRL-CAD: 03starseeker * r45918 10/brlcad/trunk/src/tclscripts/archer/Archer.tcl: |
| 02:40.52 | CIA-62 | BRL-CAD: Check for toc.html as well, since it comes from the configure process and not |
| 02:40.52 | CIA-62 | BRL-CAD: the make process in the build directory. May want to consider adding a copy |
| 02:40.52 | CIA-62 | BRL-CAD: build rule for things like this so make puts everything back where it belongs, |
| 02:40.52 | CIA-62 | BRL-CAD: but not a huge deal since re-running cmake takes care of it. |
| 02:44.15 | CIA-62 | BRL-CAD: 03starseeker * r45919 10/brlcad/trunk/NEWS: A bug in Archer's help browser code resulted in Archer failing to start if it was unable to find the html files used for the help system - it now starts even if those files are not present. |
| 02:44.28 | starseeker | ok, we should be good to go again |
| 02:44.40 | starseeker | abhi2011: does that fix it for you? |
| 02:45.15 | starseeker | bhinesley: still begs the question of why your mann directory wasn't present |
| 02:45.57 | bhinesley | I reconfigured and I'm rebuilding... perhaps it is now. Do you periodically reconfigure? I tend to just 'svn update' and make |
| 02:47.15 | bhinesley | it should trigger a reconfigure if it needs one, right? |
| 02:54.06 | bhinesley | starseeker: no dice |
| 02:59.11 | bhinesley | ok iiii'm an idiot. BRLCAD-BUILD_EXTRADOCS=OFF |
| 03:01.10 | bhinesley | that's what I get for using ccmake |
| 03:15.46 | starseeker | bhinesley: yeah, normally cmake is pretty good about reconfiguring/rebuilding when it needs to |
| 03:16.08 | starseeker | earlier case where brlcad had to recommend a clean rebuild was a bit surprising |
| 03:16.56 | starseeker | heh - yeah, if you tell it not to build the html docs it shouldn't :-P |
| 03:17.10 | bhinesley | imaging that! :) |
| 03:17.14 | bhinesley | *imagine |
| 03:17.31 | starseeker | did that fix everything? |
| 03:17.37 | bhinesley | yes |
| 03:17.50 | bhinesley | thank you |
| 03:18.18 | starseeker | np |
| 03:18.30 | bhinesley | looks like there was no cmake issue? |
| 03:19.13 | starseeker | I didn't see one - looked like just a few tcl file tweaks |
| 03:20.52 | bhinesley | oh I see it now, help.tcl |
| 03:22.14 | bhinesley | well, at least it exposed some problems |
| 03:22.52 | starseeker | sure - good stuff :-) |
| 03:23.33 | CIA-62 | BRL-CAD: 03bhinesley * r45920 10/brlcad/trunk/src/tclscripts/man_browser.tcl: set -nocomplain on glob, just in case; we don't want to fail to start due to some missing files |
| 04:35.47 | brlcad | starseeker: the earlier clean rebuild case wasn't cmake's fault -- just coincidentally hitting a magic number failure after the magic numbers were converter to uint32_t, but he was passing a bad pointer (so the magic failure was the right thing to do) |
| 04:37.54 | brlcad | abhi2011: they can pass in any object, maybe even non-geometry |
| 04:38.13 | brlcad | note that the name, rt_bound_dbfullpath(), is no longer right |
| 04:43.55 | brlcad | finishes organizing the logo submissions |
| 06:43.01 | CIA-62 | BRL-CAD: 03bhinesley * r45921 10/brlcad/trunk/src/libged/edit.c: |
| 06:43.01 | CIA-62 | BRL-CAD: Now that there are functions to convert path+objects+offsets to coords, |
| 06:43.01 | CIA-62 | BRL-CAD: arguments that had to be split into multiple structs (due to -x/-y/-z options) |
| 06:43.01 | CIA-62 | BRL-CAD: can be consolidated. It would be slightly more efficient to do this as the |
| 06:43.01 | CIA-62 | BRL-CAD: arguments are parsed, but IMO not worth muddying up ged_edit() over. |
| 06:58.41 | CIA-62 | BRL-CAD: 0399.125.86.110 07http://brlcad.org * r3069 10/wiki/User:Bhinesley: /* Log */ today |
| 07:22.18 | *** join/#brlcad d_rossberg (~rossberg@BZ.BZFLAG.BZ) | |
| 09:14.25 | CIA-62 | BRL-CAD: 03Abhi2011 07http://brlcad.org * r3070 10/wiki/User:Abhijit: /* Log */ |
| 09:42.04 | abhi2011 | brlcad: so I would need to find the list of primitives in the region represented by the rt_db_internal |
| 11:02.42 | abhi2011 | bhinesley: did the mged and archer launching errors get resolved after the latest checkout |
| 11:13.16 | CIA-62 | BRL-CAD: 03kunigami * r45922 10/osl/trunk/boost_1_46_1/: adding latest compatible boost version with osl. doing it by parts |
| 11:14.45 | CIA-62 | BRL-CAD: 03kunigami * r45923 10/osl/trunk/boost_1_46_1/boost/: adding latest compatible boost version with osl. part 2 |
| 11:38.34 | abhi2011 | ok, db_walk_tree() allows a nice set of functions to be provided for accepting and rejecting regions while building a boolean tree of the regions, I will try and use it for adding the primitives of the passed regions to the wdb |
| 11:52.11 | kunigami | hmm just saw that boost has ~ 500MB. should upload it to svn anyway? |
| 11:53.03 | kunigami | I didn't have to change anything on it from the original version. maybe add a line on the script to download it directly? |
| 12:12.37 | kunigami | llvm is pretty big too ~ 250MB |
| 12:36.21 | abhi2011 | kunigami: doesnt boost source already ship with brlcad in src/other/boost |
| 12:42.44 | abhi2011 | ok passing a leaf function to db_walk_tree() wont work because its never called even after a missing primitive is detected, was hoping to use the function to add the primitive |
| 13:31.06 | starseeker | abhi2011: mged and archer should launch now |
| 13:52.25 | abhi2011 | starseeker: yes its fine now, thanks! |
| 14:01.18 | *** join/#brlcad kunigami (~kunigami@201.53.206.27) | |
| 14:05.50 | brlcad | kunigami: that's a bit large to add, maybe see if you can identify the portion used? boost has a tool to identify the headers/deps needed so you don't have to download the kitchen sink |
| 14:06.05 | brlcad | llvm can be expected as a system install |
| 14:19.41 | CIA-62 | BRL-CAD: 03n_reed * r45924 10/brlcad/trunk/src/libgcv/wfobj/ (5 files): Using more readable names. Removed unused typedefs and some cryptic size checks of questionable utility. |
| 14:36.21 | *** join/#brlcad abhi2011 (~chatzilla@wlan-145-94-184-068.wlan.tudelft.nl) | |
| 15:12.02 | *** join/#brlcad Yoshi47 (~jan@64.235.102.210) | |
| 16:27.57 | *** join/#brlcad abhi2011 (~chatzilla@wlan-145-94-184-068.wlan.tudelft.nl) | |
| 17:20.55 | *** join/#brlcad dtidrow_desk (~dtidrow@c-68-84-167-135.hsd1.mi.comcast.net) | |
| 17:58.58 | *** join/#brlcad abhi2011 (~chatzilla@wlan-145-94-184-068.wlan.tudelft.nl) | |
| 18:05.35 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) | |
| 18:36.53 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) | |
| 18:47.58 | kunigami1 | brlcad: ok |
| 18:51.24 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) | |
| 19:12.44 | abhi2011 | brlcad: currently the librt function for finding the bb works on shapes, but if I pass regions then the db_walk_tree() detects that the shapes of the regions are absent and causes the bounding box to not be calculated |
| 19:13.44 | abhi2011 | so I am going to use the code in db_walk_tree() to add primitive leaf nodes when they are detected to be absent, which will involve significant amounts of duplicate code |
| 19:14.56 | abhi2011 | so I was wondering if there is an easier way to find and add the primitives to the rt_db_internal representation of a region |
| 19:17.02 | abhi2011 | db_walk_tree() allows call backs when a region is started and ended and for leaf nodes, but the leaf node callback is never called if a leaf (primitive) is found to be missing from the model tree |
| 19:40.03 | CIA-62 | BRL-CAD: 03starseeker * r45925 10/brlcad/trunk/src/other/incrTcl/itcl/ (4 files in 2 dirs): Take a stab at moving the itcl compilation logic over to the cleaned up logic being used for tcl itself |
| 19:43.56 | CIA-62 | BRL-CAD: 03bob1961 * r45926 10/brlcad/trunk/src/tclscripts/lib/Ged.tcl: |
| 19:43.56 | CIA-62 | BRL-CAD: Need to destroy the ray object whenever the ged object is destroyed or when |
| 19:43.56 | CIA-62 | BRL-CAD: opening a different database. This fix was prompted by database turds being left |
| 19:43.56 | CIA-62 | BRL-CAD: on Windows platforms whenever the ray object was used. That is, the ray object |
| 19:43.56 | CIA-62 | BRL-CAD: also has the database copy (i.e. the turd) open and so the code that removes the |
| 19:43.57 | CIA-62 | BRL-CAD: database copy fails. |
| 19:49.30 | brlcad | abhi2011: I don't think you can call db_walk_tree .. you don't know the name of the rt_db_internal that you're trying to calculate a bounding box for |
| 20:07.48 | brlcad | at least, you can't call it for that dbi -- you could call it for all of the comb's members to recursively get their bbs |
| 20:08.48 | brlcad | it might be easier to just implement ft_prep for combs |
| 20:10.22 | CIA-62 | BRL-CAD: 03starseeker * r45927 10/brlcad/trunk/src/tclscripts/archer/ArcherCore.tcl: Do as MGED does and default to Navy |
| 20:12.33 | brlcad | abhi2011: aha, I think I have a solution, but it depends on what your code looks like now |
| 20:12.59 | brlcad | since you DO have it working for primitive, just work on cleaning up the code and make your patch |
| 20:13.47 | brlcad | make it gracefully fail on combs for now, then once your patch is integrated, can work on bb of comb |
| 20:14.31 | abhi2011 | brlcad: hint to the solution :P |
| 20:15.06 | brlcad | if you have an rt_db_internal that is a comb, then you have a union tree pointer |
| 20:15.17 | abhi2011 | yes |
| 20:15.22 | brlcad | if you have a union tree pointer, then you can call the other/existing bbox routine, rt_bound_tree() |
| 20:15.53 | abhi2011 | ah yes thats the other function i used in my program before too |
| 20:17.05 | brlcad | you may still need to call rt_gettree/rt_gettrees to load/evaluate the tree, but that's easy |
| 20:17.05 | abhi2011 | rt_bound_tree(regp->reg_treetop, reg_min, reg_max) |
| 20:17.31 | brlcad | that'd only be for regions |
| 20:17.34 | brlcad | you have combs |
| 20:17.40 | brlcad | combp->tree |
| 20:17.40 | abhi2011 | ah yes |
| 20:17.52 | brlcad | see rt_comb_internal in raytrace.h |
| 20:18.15 | abhi2011 | ok |
| 20:18.26 | brlcad | basically, implementing much of what _ged_get_obj_bounds() does |
| 20:18.43 | brlcad | just instead of using db_full_path objects you're using rt_db_internal objects |
| 20:23.30 | abhi2011 | ok |
| 20:31.27 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) | |
| 20:55.31 | CIA-62 | BRL-CAD: 03bob1961 * r45928 10/brlcad/trunk/src/tclscripts/archer/ArcherCore.tcl: The backgroundColor routine should be calling ::cadwidgets::Ged::get_rgb_color instead of getRgbColor for consistency. |
| 20:57.16 | CIA-62 | BRL-CAD: 03bhinesley * r45929 10/brlcad/trunk/src/libged/edit.c: oops... edit_cmd initialization routine introduced r45921 tried to set pointer that was pointed to by uninitialized pointer to NULL |
| 21:18.00 | CIA-62 | BRL-CAD: 03starseeker * r45930 10/brlcad/trunk/src/other/ (7 files in 4 dirs): |
| 21:18.00 | CIA-62 | BRL-CAD: Make a stab at itk, probably the trickiest of these extensions. Need to do some |
| 21:18.00 | CIA-62 | BRL-CAD: more logic consolidation into tcl.cmake and the src/other CMakeLists.txt |
| 21:18.00 | CIA-62 | BRL-CAD: settings probably need some more study (there are a lot of possible cases) but |
| 21:18.00 | CIA-62 | BRL-CAD: getting there. Need to study STUBS usage in the standard Tcl/Tk build more and |
| 21:18.01 | CIA-62 | BRL-CAD: see if I need some conditionalization logic for those flags... |
| 22:10.31 | CIA-62 | BRL-CAD: 03starseeker * r45931 10/brlcad/trunk/src/libbu/progname.c: We need a static buffer here, otherwise our path disappears on us. |
| 22:11.33 | *** join/#brlcad abhi2011 (~chatzilla@ip170-79-211-87.adsl2.static.versatel.nl) | |
| 22:45.45 | CIA-62 | BRL-CAD: 03starseeker * r45932 10/brlcad/trunk/src/libbu/progname.c: avoid overwriting the full argv0 path - bu_getprogname was storing its basename in the save variable as argv0's full path information. |