| 00:25.22 | *** join/#brlcad bqkqxkphkutjtpji (~armin@dslb-088-064-046-111.088.064.pools.vodafone-ip.de) |
| 00:46.42 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) |
| 02:54.32 | Notify | 03BRL-CAD:brlcad * 68910 brlcad/trunk/src/conv/dbupgrade.c: add some basic input/output protection to make sure we don't clobber an existing output |
| 03:04.14 | *** join/#brlcad teepee (~teepee@unaffiliated/teepee) |
| 03:17.35 | brlcad | starseeker: what's the status of freetype in the build? I'm seeing a build failure on Mac now where it was working before (failing to find -lfreetype, missing -L/usr/X11/lib on linker flags) |
| 05:55.44 | *** join/#brlcad KimK (~Kim__@2600:8803:7a85:6d00:20e1:82bf:a13c:a385) |
| 08:40.38 | *** join/#brlcad merzo (~merzo@92.60.189.225) |
| 09:33.54 | *** join/#brlcad Caterpillar3 (~caterpill@unaffiliated/caterpillar) |
| 09:53.42 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) |
| 10:41.09 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) |
| 11:39.33 | starseeker | blinks - our freetype is turned off |
| 11:40.03 | starseeker | if Tk is trying to pull it in, it should be finding the system tk |
| 11:40.08 | starseeker | er, system freetype |
| 11:43.21 | starseeker | yeah... openscenegraph uses freetype, but that's still off - Tk is probably what's trying to bring it in |
| 11:43.58 | starseeker | so it sounds like FREETYPE_INCLUDE_DIRS on the Mac isn't pulling what's needed - it is finding the non-X11 version of freetype somehow? |
| 11:45.33 | starseeker | scowls... really need to get OSX building with native Aqua to dodge this whole stupid mess of X vs non-X libs... |
| 11:59.27 | *** join/#brlcad teepee (~teepee@unaffiliated/teepee) |
| 12:32.44 | *** join/#brlcad Caterpillar (~caterpill@unaffiliated/caterpillar) |
| 12:39.36 | brlcad | starseeker: yes, tk was trying to link it in first (followed by anything that linked tk) |
| 12:40.38 | brlcad | mac only has a /usr/X11/lib/libfreetype.6.dylib, there isn't another system freetype |
| 12:41.12 | brlcad | the system libs have a completely different font rendering system, not freetype-based |
| 13:03.49 | Notify | 03BRL-CAD:brlcad * 68911 brlcad/trunk/src/liboptical/sh_light.c: fix infinite loop bug in rt when using non-infinite shadow-casting lights. multiple issues going on, first being to not loop indefinitely when calling rt_shootray to find light samples if we don't acquire the sample (for whatever reason). since light samples were being randomly created using BN_RANDOM which has a *very* small cycle, use that limit as |
| 13:03.51 | Notify | termination criteria here. probably want a better way (and better randomness).the second problem was introduced in 28642 when I was apparently chasing a null derefernce, I'd added an if(&array) check. as always true, it was logically harmless until in 65111 jon noticed the always true logic and presumed the intent was if(array) which made the logic always false thus causing light sources to never get initialized and |
| 13:03.53 | Notify | samples to spin infinitely.... |
| 13:08.35 | Notify | 03BRL-CAD:brlcad * 68912 brlcad/trunk/NEWS: fixed an infinite loop bug in rt that was occurring on non-infinite light sources that cast shadows. the logic for initializing sample points (so we cast shadows in the shape of the light) was faulty causing it to never terminate. fixed logic and added additional protection in 68911. |
| 13:14.23 | Notify | 03BRL-CAD:brlcad * 68913 brlcad/trunk/src/liboptical/sh_light.c: fix a regression introduced way back in 2000 (r19476) that breaks old models with directional light sources. this was long before we had deprecation policy, but a detrimental regression nonetheless when the d/dir keyword was changed to t/target without adding any logic to upgrade the old shader strings. restore the ability to read those geometry files |
| 13:14.25 | Notify | and have their lights work by adding d/dir back to the list of recognized keywords. |
| 13:14.27 | Notify | ... |
| 13:19.59 | Notify | 03BRL-CAD:brlcad * 68914 brlcad/trunk/NEWS: directional lights created prior to rel 6.0 used the 'd {x y z}' notation to specify their direction. this was changed to t/target, which broke all light sources and required the modeler to redefine the light. this undoes the regression, restoring the ability to read the old .g files without error by putting recognition for d/dir keywords back. |
| 13:27.44 | Notify | 03BRL-CAD:brlcad * 68915 brlcad/trunk/NEWS: consolidate all of the rtwizard window sizing changes into one entry: 1) this addresses user request from Josh Baker to have rtwizard name its window something other than the generic 'tk' to make it easier to find when minimized. 2) If the user sets the hierarchy view width, we want that width to stay constant between windows. Also uniformly initialize to width 250 rather |
| 13:27.46 | Notify | than on the lenght of the file name. 3) Have the default rtwizard image size key off of the mgedObj size. Can be overridden by the FbPage. |
| 13:27.48 | Notify | ... |
| 13:29.28 | Notify | 03BRL-CAD:brlcad * 68916 brlcad/trunk/NEWS: slight tweak for readability. original note: Improve keep error reporting when file can not be created. |
| 13:30.39 | *** join/#brlcad yorik (~yorik@2804:431:f721:45a4:290:f5ff:fedc:3bb2) |
| 13:31.59 | Notify | 03BRL-CAD:brlcad * 68917 brlcad/trunk/NEWS: g_diff man page is replaced by gdiff man page. consolidate into one line item. |
| 13:36.03 | Notify | 03BRL-CAD:brlcad * 68918 brlcad/trunk/NEWS: jon added support for specifying colors using 0 to 255 or 0.0 to 1.0 ranges to all of the rt/rtedge/etc tools |
| 13:38.14 | Notify | 03BRL-CAD:brlcad * 68919 brlcad/trunk/NEWS: brlman and man(n) now have manual pages. (reworded for clarity) |
| 13:42.50 | Notify | 03BRL-CAD:brlcad * 68920 brlcad/trunk/NEWS: consolidate man page removals, removed erase_all, export_body, vrmgr, dall. also avoid using ' - ' in description as it breaks automatic parsing of names. |
| 13:46.58 | starseeker | brlcad: I'm not quite sure offhand what the fix is for the OSX freetype issue... looking at the build logic I don't see anything obviously wrong |
| 13:47.17 | starseeker | brlcad: just to be sure, you were starting from a clean build dir? |
| 13:49.02 | Notify | 03BRL-CAD:brlcad * 68921 brlcad/trunk/NEWS: gdiff now uses a 100mm default tolerance and a default grid spacing based on the bounding box. (consolidated two gdiff default changes into one entry) |
| 13:49.25 | brlcad | starseeker: not clean, it was a rebuild after svn up - I'll check clean |
| 13:52.13 | Notify | 03BRL-CAD:brlcad * 68922 brlcad/trunk/NEWS: group the gdiff options together, man page movement is arguably not visible. |
| 13:55.04 | Notify | 03BRL-CAD:brlcad * 68923 brlcad/trunk/TODO: multiple bugs found, invokes regression test requirement. |
| 13:56.39 | Notify | 03BRL-CAD:brlcad * 68924 brlcad/trunk/src/conv/dbupgrade.c: need to upgrade d/dir shader strings if we're going to remove them |
| 14:44.20 | *** join/#brlcad Caterpillar (~caterpill@unaffiliated/caterpillar) |
| 14:54.37 | Notify | 03BRL-CAD:starseeker * 68925 brlcad/trunk/src/conv/dbupgrade.c: unused variable |
| 17:23.36 | *** join/#brlcad Caterpillar (~caterpill@unaffiliated/caterpillar) |
| 18:34.54 | Notify | 03BRL-CAD:starseeker * 68926 brlcad/trunk/src/libbu/booleanize.c: Restore original tests for on/off for the nN and yY cases to make unit testing work again. Need to check if the new behavior was/is preferred - if so, need to update tests. |
| 20:54.25 | Notify | 03BRL-CAD:starseeker * 68927 brlcad/trunk/regress/repository.sh: in-src-dir regress-repository needs to ignore mime.c |
| 21:15.35 | Notify | 03BRL-CAD Wiki:Luigipacetti * 0 /wiki/User:Luigipacetti: |
| 22:00.44 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) |
| 22:01.54 | Notify | 03BRL-CAD:starseeker * 68928 brlcad/trunk/src/conv/k-g.cpp: Fix while test and apply mk_bot fixes (thanks to Vasco Alexandre da Silva Costa for catching these) |
| 22:13.03 | Notify | 03BRL-CAD:starseeker * 68929 brlcad/trunk/src/libbu/color.c: initialize separator char (build problem reported by Vasco Alexandre da Silva Costa) |
| 22:19.21 | Notify | 03BRL-CAD:vasco_costa * 68930 brlcad/trunk/src/librt/primitives/bot/bot.c: Fix compilation error due to uninitialized variable. |