| 00:00.35 | brlcad | need to check at least linux with and without system tcl, and mac (without of course) |
| 00:01.37 | brlcad | fwiw, make test and make distcheck aren't sufficient, have to kick off the GUI |
| 00:01.46 | starseeker | nods |
| 00:06.57 | brlcad | I had distchecks passing with all options on/off for 64-bit linux, testing mac now |
| 00:31.44 | CIA-53 | BRL-CAD: 03starseeker * r42869 10/brlcad/branches/cmake/TODO.cmake: Add a few more TODO items for CMake |
| 00:35.08 | CIA-53 | BRL-CAD: 03starseeker * r42870 10/brlcad/trunk/src/mged/mged.c: dbip_filename -> dbi_filename |
| 00:50.59 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) |
| 00:53.33 | starseeker | brlcad: I think I may have messed up. I am using BRLCAD_DATA as a relative path from BRLCAD_ROOT, but I'm seeing in trunk that both BRLCAD_DATA and BRLCAD_ROOT are full paths |
| 01:09.54 | *** join/#brlcad DX^ (~DX@c-71-59-50-121.hsd1.ga.comcast.net) |
| 01:10.01 | DX^ | I can't find the configure flag to disable compiling with X11 |
| 01:10.40 | starseeker | --disable-x11 ? |
| 01:14.14 | DX^ | I tried --disable-x |
| 01:14.19 | DX^ | but that didn't work, I'll try that.. |
| 01:15.55 | DX^ | still says its enabled |
| 01:16.40 | starseeker | --without-x11 ? |
| 01:17.34 | DX^ | trying |
| 01:17.58 | starseeker | brlcad: my apologies - I think I made my own trouble with the path logic, after all |
| 01:22.19 | DX^ | that worked, awesome |
| 01:22.38 | CIA-53 | BRL-CAD: 03starseeker * r42871 10/brlcad/trunk/src/libbu/brlcad_path.c: Dur. BRLCAD_DATA is supposed to be a full path, but I had a relative path in the CMake logic... hilarity ensued. Put it back the way it should be and make CMake do it right. |
| 01:29.59 | CIA-53 | BRL-CAD: 03starseeker * r42872 10/brlcad/trunk/configure.ac: Add note that BRLCAD_DATA must be a full path |
| 01:39.41 | CIA-53 | BRL-CAD: 03starseeker * r42873 10/brlcad/branches/cmake/ (CMakeLists.txt configure.ac src/libbu/brlcad_path.c): Do BRLCAD_DATA the right way in CMake. |
| 01:49.51 | CIA-53 | BRL-CAD: 03starseeker * r42874 10/brlcad/trunk/ (7 files in 4 dirs): Apply 42757 to trunk - without this, getting failures on gentoo with warnings enabled. |
| 01:55.54 | CIA-53 | BRL-CAD: 03starseeker * r42875 10/brlcad/trunk/configure.ac: Whoops - manuals/archer is still gone |
| 02:12.14 | CIA-53 | BRL-CAD: 03starseeker * r42876 10/brlcad/trunk/doc/docbook/Makefile.am: catalog.xml is no more |
| 02:35.33 | brlcad | starseeker: hm, yeah they should be full paths but it's an interesting thought |
| 02:35.51 | brlcad | relative path "should" work, but I can see how an assumption may be in there |
| 02:36.40 | starseeker | relative is already sort of there with the root/share/brlcad/version check - that's what you get with a relative BRLCAD_DATA |
| 02:36.52 | brlcad | BRLCAD_ROOT and BRLCAD_DATA are environment variables so they could conceivably be relative if we allow it |
| 02:37.10 | brlcad | bu_brlcad_root() and bu_brlcad_data() probably have assumptions otherwise |
| 02:37.27 | brlcad | ah, yeah, that would work |
| 02:37.36 | brlcad | though full-path root and relative data should work too |
| 02:37.58 | brlcad | or the limitation should be documented |
| 02:38.04 | starseeker | that's what I was doing with CMake (albeit accidentally) |
| 02:38.12 | starseeker | mentioned it in configure.ac |
| 02:38.22 | brlcad | yeah, i saw |
| 02:38.39 | brlcad | I mean to users though -- the environment variable isn't set by configure |
| 02:38.46 | brlcad | it sets the "hard wired" default |
| 02:38.48 | starseeker | wouldn't be too hard to make it accept a relative path - I did it initially before I realized |
| 02:38.59 | starseeker | ah, point |
| 02:39.52 | starseeker | Windows might be a problem if they stick a drive letter at the beginning of a path though |
| 02:40.33 | starseeker | brlcad: distcheck passes here |
| 02:41.00 | starseeker | although come to think of it, it did work on Windows |
| 02:48.18 | brlcad | cool |
| 02:48.22 | brlcad | mged runs? |
| 02:48.33 | brlcad | what platform/settings? |
| 02:49.02 | starseeker | gentoo x86_64 with --enable-all |
| 02:49.11 | starseeker | mged seems to run |
| 02:49.16 | starseeker | will build again with system tcl/tk |
| 02:49.59 | brlcad | okay, testing mac here |
| 02:50.08 | brlcad | 10.6 |
| 02:56.38 | brlcad | everyone ready for GSoC? |
| 02:57.09 | starseeker | ah, that's on again? |
| 02:57.10 | starseeker | sweet |
| 03:06.21 | starseeker | brlcad: just a thought - with the cmake build, all the final build products are right there in lib or bin |
| 03:08.21 | DX^ | IGES/STEP converters are segfaulting on me |
| 03:08.23 | DX^ | with the newest code |
| 03:08.51 | starseeker | on files that worked previously? |
| 03:09.02 | DX^ | no, but the STEP file I'm trying is just a box |
| 03:09.06 | DX^ | don't think it can get much simpler |
| 03:09.14 | starseeker | what's the error? |
| 03:09.35 | DX^ | "Segmentation fault" |
| 03:09.42 | starseeker | ah, that's it? |
| 03:09.56 | DX^ | yep |
| 03:09.59 | starseeker | DX^: can you post the file on a bug report at sourceforge? |
| 03:10.15 | starseeker | brlcad: local tcl/tk seems to work too |
| 03:10.32 | DX^ | starseeker: yeah |
| 03:10.47 | starseeker | cool - we'll need to hook up a debugger to see what happened |
| 03:32.13 | brlcad | starseeker: that's good to hear -- what version of tcl? |
| 03:32.16 | CIA-53 | BRL-CAD: 03brlcad * r42877 10/brlcad/trunk/misc/Makefile.am: include debian/source/include-binaries in dist |
| 03:32.27 | brlcad | suspects the crashes were 8.4 specific |
| 03:33.24 | starseeker | 8.5.9 |
| 03:37.28 | brlcad | starseeker: that's a good point regarding cmake build -- noprod might be obsolete via that alone since you could run targetted rm's and remake |
| 03:37.58 | brlcad | have to put it through the paces with use to see what's most efficient |
| 03:39.41 | CIA-53 | BRL-CAD: 03starseeker * r42878 10/brlcad/branches/cmake/src/other/togl/CMakeLists.txt: Stray character |
| 03:39.53 | CIA-53 | BRL-CAD: 03tbrowder2 * r42879 10/brlcad/trunk/HACKING: added info on using in-tree executables without installing\nadded words about developer responsibilities |
| 03:48.10 | CIA-53 | BRL-CAD: 03starseeker * r42880 10/brlcad/branches/cmake/src/other/togl/demo/gears.tcl: Just package require Togl for the gears demo - don't need to manually load anything if we're set up correctly. |
| 03:53.17 | CIA-53 | BRL-CAD: 03tbrowder2 * r42881 10/brlcad/trunk/INSTALL: added information about where file check sums are to be found |
| 04:14.11 | *** join/#brlcad yukonbob (~bch@S010600235a187d92.ok.shawcable.net) |
| 05:08.29 | CIA-53 | BRL-CAD: 03starseeker * r42882 10/brlcad/branches/cmake/bench/CMakeLists.txt: benchmark needs pixcmp |
| 05:36.33 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) |
| 06:33.42 | *** join/#brlcad Stattrav (~Stattrav@122.172.33.253) |
| 06:33.42 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) |
| 07:26.28 | *** join/#brlcad epileg (~epileg@unaffiliated/epileg) |
| 07:35.18 | *** join/#brlcad yukonbob (~bch@S010600235a187d92.ok.shawcable.net) |
| 09:59.48 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) |
| 12:49.28 | *** join/#brlcad juanman (~quassel@unaffiliated/juanman) |
| 12:52.28 | epileg | brlcad: what's better? build deb packages with or without pdf files? |
| 12:52.28 | epileg | with (â 85 MB.), without (â 55 MB.) |
| 13:58.38 | starseeker | epileg: for now I'd say without |
| 13:58.55 | epileg | ok, thanks! |
| 13:58.55 | starseeker | we have some work to do before the pdf output from most of our stuff is fit for consumption |
| 14:21.00 | CIA-53 | BRL-CAD: 03erikgreenwald * r42883 10/brlcad/branches/bottie/ (1260 files in 90 dirs): MFC 42842 |
| 14:37.09 | *** topic/#brlcad by brlcad -> BRL-CAD Open Source Solid Modeling || http://brlcad.org || http://sf.net/projects/brlcad || #brlcad logs: http://ibot.rikers.org/%23brlcad/ || BRL-CAD release prep for 7.18.2 under way (20110202) |
| 14:37.26 | brlcad | starseeker: did "make test" pass for you? |
| 14:37.56 | brlcad | I'm getting all sorts of failures here, but still have to look into why |
| 14:42.33 | starseeker | hmm - it's failing to create solids.rt, shaders.rt... |
| 14:44.42 | *** join/#brlcad Stattrav (~Stattrav@117.192.140.121) |
| 14:44.42 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) |
| 14:46.21 | CIA-53 | BRL-CAD: 03brlcad * r42884 10/brlcad/trunk/src/halftone/main.c: enable-optimized linux strictness failure, check fwrite() return value |
| 14:47.15 | starseeker | brlcad: opendb isn't creating solids.g |
| 14:48.02 | starseeker | when it runs opendb solids.g y the command completes, but a database is not opened |
| 14:48.28 | brlcad | okay, that should be easy to fix |
| 14:48.37 | brlcad | maybe even my doing with the -f flag |
| 14:48.48 | starseeker | that was my first guess |
| 14:49.00 | brlcad | an argc check off by one or something |
| 14:49.33 | brlcad | cool, optimized and unoptimized distchecks now all passing on linux |
| 14:56.54 | starseeker | brlcad: I'm guessing it's in 42564 |
| 14:57.21 | brlcad | hah, looks like the logic is just reversed |
| 14:57.34 | brlcad | opendb files.g n creates a database :) |
| 14:57.42 | starseeker | lol |
| 14:58.11 | starseeker | ok, so probably not that commit then |
| 15:01.21 | ``Erik | well, damn, I just spent the last 2 days trying to get things together and working to move bottie to trunk O.o |
| 15:01.38 | starseeker | why is that a problem? |
| 15:02.08 | ``Erik | if we're in the middle of release cycle, it might not be the best time to do it |
| 15:02.08 | brlcad | it can be the highlight feature for 7.18.4 or 7.20 even |
| 15:02.33 | brlcad | get more attention that way anyways |
| 15:02.46 | *** join/#brlcad mafm (~mafm@72.Red-83-45-72.dynamicIP.rima-tde.net) |
| 15:02.51 | ``Erik | apparently some weirdness happened, or svn sucks for this kinda merge, royal pain :/ |
| 15:03.28 | brlcad | ``Erik: it sucks until that last to-do item is dealt with, which I'm planning as soon as release is posted |
| 15:03.44 | brlcad | our repo is too old to support merge tracking |
| 15:03.52 | ``Erik | ah, --reintegrate? |
| 15:03.54 | brlcad | that's what has been causing some of the merge headache |
| 15:03.56 | ``Erik | fsfs 2 vs 3 or something |
| 15:04.07 | brlcad | 1.5+ backend repo |
| 15:04.12 | brlcad | we're a 1.4 repo or something iirc |
| 15:04.24 | starseeker | ``Erik: were you going to try and track down the big missing pieces in bottie raytracing prior to merge? |
| 15:04.49 | starseeker | or just stick with the insane high threshold for now |
| 15:05.23 | ``Erik | starseeker: was going to have an obnoxiously high # and work it after merge, I seem to spend way more time merging and resolving conflicts to try to track trunk than getting to actually work on it |
| 15:06.06 | ``Erik | wonders if there's a darcs/svn bridge O.o |
| 15:09.56 | starseeker | ``Erik: then you should be pretty safe to merge, yes? |
| 15:11.14 | ``Erik | probably *shrug* |
| 15:16.13 | brlcad | I wouldn't call that merge-safe in the least .. heh |
| 15:16.20 | brlcad | it's a huge body of change to code and a new feature |
| 15:18.38 | brlcad | release safe things are like minor bug fixes, updating documentation, tweaking a bu_log, removing dead code, adding comments, ... |
| 15:21.13 | brlcad | that really is great feature for the next release where it can be spoken to and highlighted, maybe erik can write up a mini article about it all even |
| 15:22.00 | CIA-53 | BRL-CAD: 03erikgreenwald * r42885 10/brlcad/trunk/src/halftone/main.c: more missed size_t changes |
| 15:24.38 | CIA-53 | BRL-CAD: 03erikgreenwald * r42886 10/brlcad/trunk/src/libged/ (vdraw.c wdb_vdraw.c): cast to long unsigned int to match wdb_vdraw.clu |
| 15:25.05 | ``Erik | ugh, tons of those in src/conv |
| 15:26.01 | brlcad | want me to disable strict or you going to hit em up? |
| 15:26.30 | brlcad | I don't have a plat that warns on that one opt or non-opt |
| 15:26.37 | ``Erik | I can hit them, bit nervous about casting read pointers like that blindly |
| 15:26.41 | ``Erik | crit should do it |
| 15:27.14 | ``Erik | FreeBSD 8.2-PRERELEASE #3: Thu Dec 2 13:59:53 EST 2010 |
| 15:28.08 | ``Erik | using system tcl, tk, tnt, debug off (that seems to be a big one), optimized on |
| 15:28.50 | CIA-53 | BRL-CAD: 03brlcad * r42887 10/brlcad/trunk/TODO: two release issues being worked on now, opendb failure and strict failures on fbsd |
| 15:28.57 | brlcad | ah, debug off .. haven't tries that in a while |
| 15:29.52 | CIA-53 | BRL-CAD: 03erikgreenwald * r42888 10/brlcad/trunk/src/libtclcad/tkImgFmtPIX.c: more size_t fixing |
| 15:40.57 | CIA-53 | BRL-CAD: 03erikgreenwald * r42889 10/brlcad/trunk/src/conv/ (6 files in 3 dirs): casts/changes to cope with size_t vs vprintf |
| 15:46.12 | ``Erik | that seems to be all the type warnings, there is still the tcl85 vs tcl8.5 link error when building tcl with BRL-CAD, though :/ |
| 15:49.35 | CIA-53 | BRL-CAD: 03brlcad * r42890 10/brlcad/trunk/src/libbu/booleanize.c: |
| 15:49.36 | CIA-53 | BRL-CAD: bu_booleanize() FAIL. several problems: 1) \!str == 0 means we return false for |
| 15:49.36 | CIA-53 | BRL-CAD: all non-null strings, 2) strtol() will return 0 if there are no numbers (boo, |
| 15:49.36 | CIA-53 | BRL-CAD: hiss), and 3) it wasn't taking whitespace into consideration. stash the input |
| 15:49.36 | CIA-53 | BRL-CAD: into a vls, trim whitespace, then perform our tests. |
| 15:49.49 | starseeker | brlcad: heh, you just beat me to it |
| 15:49.53 | starseeker | was writting my commit message |
| 15:50.55 | CIA-53 | BRL-CAD: 03brlcad * r42891 10/brlcad/trunk/src/libbu/booleanize.c: remove debug statements |
| 15:51.35 | brlcad | starseeker: ah, heh -- what was your change? |
| 15:51.41 | brlcad | there's another problem in mged.c |
| 15:52.09 | starseeker | was doing str == NULL instead of !str |
| 15:52.24 | starseeker | !str was returning 0 here even for a valid string |
| 15:52.39 | starseeker | good call on the whitespace |
| 15:53.34 | CIA-53 | BRL-CAD: 03brlcad * r42892 10/brlcad/trunk/src/mged/mged.c: logic is reversed. if NOT true, i.e. the user responded 'n', then report that no database is open. |
| 15:53.41 | brlcad | !str is 0 for all valid strings ;) |
| 15:53.49 | brlcad | !str is only true if str is null :) |
| 15:54.03 | starseeker | ah, gotcha |
| 15:54.38 | starseeker | one possible improvement I can still make... |
| 15:54.43 | brlcad | it should have been "str == NULL" or "!str" .. but "!str == NULL" (i.e., "!str == 0") means that all valid strings will return negative |
| 15:54.51 | starseeker | hehe |
| 15:54.53 | CIA-53 | BRL-CAD: 03starseeker * r42893 10/brlcad/trunk/src/libbu/booleanize.c: Can still use strtol, but need to make sure endptr is '\0' (i.e., that a valid number was actually processed). |
| 15:55.53 | starseeker | brlcad: if I'm reading the strtol documentation right, that will work |
| 15:56.23 | brlcad | hm |
| 15:56.25 | brlcad | "If there were no digits at all, however, strtol() stores the original value of str in *endptr. |
| 15:56.39 | starseeker | right |
| 15:56.48 | brlcad | okay, I think I see it |
| 15:58.42 | CIA-53 | BRL-CAD: 03starseeker * r42894 10/brlcad/trunk/src/libbu/booleanize.c: whoops - val is an int |
| 15:59.07 | brlcad | what do you think about bu_true() instead of bu_booleanize() ? |
| 15:59.22 | brlcad | for readability, if (bu_true(some_string)) ... |
| 15:59.43 | starseeker | it's a bit shorter... does true imply anything that booleanize doesn't? (or vice versa?) |
| 16:00.15 | starseeker | have expects bu_true to always return true, but that's just a silly reflex |
| 16:00.21 | starseeker | s/have/half |
| 16:00.26 | brlcad | the name bu_booleanize() doesn't imply a truth return value, so it's funky to read and get the logic right |
| 16:00.46 | brlcad | the mged.c change for example |
| 16:01.09 | starseeker | I'm cool with bu_true |
| 16:01.13 | brlcad | bu_str_true() |
| 16:01.22 | starseeker | ah, yeah that's better |
| 16:01.25 | starseeker | little more specific |
| 16:04.08 | CIA-53 | BRL-CAD: 03starseeker * r42895 10/brlcad/trunk/src/libbu/booleanize.c: Let's stay consistent - long, not int |
| 16:04.18 | brlcad | namingwise, bu_str_to_rgb() is the only other bu_str_ along with the C wrappers bu_strdup, bu_strcpy, etc |
| 16:05.02 | starseeker | well, either way |
| 16:05.12 | starseeker | doesn't have a real strong opinion |
| 16:05.46 | brlcad | bu_str_to_bool() is no better than bu_booleanize() |
| 16:06.13 | brlcad | heh, bu_truthify() |
| 16:06.46 | brlcad | bu_test_whether_string_is_yes() |
| 16:07.00 | *** join/#brlcad Zaebos (~irc@217.91.127.94) |
| 16:07.10 | ``Erik | why do I feel like the description needs the phrase "and truthishly" in it |
| 16:07.27 | CIA-53 | BRL-CAD: 03starseeker * r42896 10/brlcad/trunk/src/adrt/libtie/tie.c: tie.h includes common.h, so don't need to include it twice - fixes regress common.h inclusion order error. |
| 16:09.16 | starseeker | brlcad: we still have bio.h in cmd.h - I guess the thing to do is take it out and see what fails on Windows? |
| 16:11.25 | brlcad | sure, though the comment says why it's there |
| 16:11.29 | brlcad | timeval from windows.h |
| 16:11.31 | starseeker | was included in 41077... if it does need to be there (and since some guy called brlcad committed it I'm guessing we really do) we should probably special case exempt that in the regression test |
| 16:12.12 | brlcad | so either cmd.h needs the same windows wrapper logic like bio.h has or bio.h has to be included before all instances of cmd.h |
| 16:12.35 | starseeker | or we tweak the regression test... |
| 16:12.49 | brlcad | just because I committed it does mean it's right.. looks wrong to me :) |
| 16:13.18 | brlcad | having to put exceptions usually implies something is wrong |
| 16:14.42 | starseeker | mutter... mged has its own cmd.h file, clutters up grep |
| 16:14.49 | starseeker | tunes... |
| 16:16.12 | starseeker | wow, cmd.h is popular |
| 16:17.00 | brlcad | to be a self-contained header including what it needs, cmd.h should probably just have the windows.h logic |
| 16:17.11 | starseeker | is doing that now... |
| 16:19.02 | CIA-53 | BRL-CAD: 03starseeker * r42897 10/brlcad/trunk/include/cmd.h: grab the relevant part of bio.h's windows.h inclusion logic and put it in cmd.h, rather than including bio.h (should fix regression test failure. |
| 16:21.51 | CIA-53 | BRL-CAD: 03brlcad * r42898 10/brlcad/trunk/ (include/bu.h src/libbu/booleanize.c src/mged/mged.c): |
| 16:21.51 | CIA-53 | BRL-CAD: rename bu_booleanize() to bu_str_true() which is not only two chars shorter but |
| 16:21.51 | CIA-53 | BRL-CAD: reads much better within if-statements. not as cool a name, but hopefully less |
| 16:21.51 | CIA-53 | BRL-CAD: prone to logic errors. also add the counterpart bu_str_false() to have a |
| 16:21.52 | CIA-53 | BRL-CAD: matching set and for readability. |
| 16:24.13 | CIA-53 | BRL-CAD: 03starseeker * r42899 10/brlcad/branches/cmake/ (22 files in 14 dirs): Update cmake branch to trunk r42898 - need to do some manual checking, probably missed a few things earlier when working in both trees at once. |
| 16:42.20 | CIA-53 | BRL-CAD: 03starseeker * r42900 10/brlcad/trunk/regress/repository.sh: have repository.sh look in TOPSRC for the INSTALL file too, not just configure.ac. |
| 16:43.55 | CIA-53 | BRL-CAD: 03starseeker * r42901 10/brlcad/trunk/INSTALL: blt is no more |
| 16:44.38 | CIA-53 | BRL-CAD: 03bob1961 * r42902 10/brlcad/trunk/src/tclscripts/archer/ArcherCore.tcl: Modified ArcherCore::updateRtControl to call rtcntrl's updateControlPanel instead of configuring -size. |
| 16:45.31 | CIA-53 | BRL-CAD: 03starseeker * r42903 10/brlcad/trunk/ (INSTALL configure.ac): It's just tkhtml now |
| 16:47.03 | CIA-53 | BRL-CAD: 03bob1961 * r42904 10/brlcad/trunk/src/tclscripts/archer/Archer.tcl: Added code to remember Archer's window size. |
| 16:48.00 | CIA-53 | BRL-CAD: 03starseeker * r42905 10/brlcad/trunk/INSTALL: tkimg -> tkpng |
| 16:49.41 | brlcad | ``Erik: ah, I see what you're saying about the casts |
| 16:49.52 | CIA-53 | BRL-CAD: 03starseeker * r42906 10/brlcad/trunk/configure.ac: remove togl logic from configure.ac altogether - togl CMake logic is succeeding, and it is unlikely we will go back to do autotools version. |
| 16:49.53 | brlcad | yeah, casting a scanf probably isn't safe |
| 16:50.20 | brlcad | probably need a temp var that can be set to the size_t after the scanf |
| 16:52.00 | brlcad | printing casts should be fine, scanning casts probably not |
| 16:52.08 | brlcad | reviewing now |
| 16:52.30 | CIA-53 | BRL-CAD: 03starseeker * r42907 10/brlcad/trunk/configure.ac: add alias for tktable |
| 16:52.48 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) |
| 16:54.13 | CIA-53 | BRL-CAD: 03starseeker * r42908 10/brlcad/trunk/ (INSTALL configure.ac): add tkpng-build alias |
| 17:01.13 | brlcad | that alias already exists |
| 17:01.15 | brlcad | "tktable-build" |
| 17:01.23 | brlcad | see the comment before |
| 17:01.39 | brlcad | (it's not an alias, it's the actual) |
| 17:01.52 | brlcad | same for tkpng |
| 17:01.57 | starseeker | oh, ok - sorry |
| 17:03.26 | CIA-53 | BRL-CAD: 03starseeker * r42909 10/brlcad/trunk/configure.ac: remove unneeded alias lines |
| 17:10.11 | *** join/#brlcad mafm_ (~mafm@72.Red-83-45-72.dynamicIP.rima-tde.net) |
| 17:17.42 | CIA-53 | BRL-CAD: 03starseeker * r42910 10/brlcad/trunk/configure.ac: I imagine there are more of these issues, but it looks like rtgl should be a 'with', not an 'enable' in the build logic. |
| 17:20.05 | CIA-53 | BRL-CAD: 03starseeker * r42911 10/brlcad/trunk/INSTALL: remainder of INSTALL documentation options - repository-regress should now succeed. |
| 17:22.29 | CIA-53 | BRL-CAD: 03starseeker * r42912 10/brlcad/branches/cmake/ (5 files in 3 dirs): MFC 42911 |
| 17:35.50 | CIA-53 | BRL-CAD: 03starseeker * r42913 10/brlcad/branches/cmake/regress/CMakeLists.txt: Ah right, DEPENDS can't tell if the target's done anything if it's set up this way - just have regress run 'em all. Regression test fully succeeded. |
| 17:35.52 | starseeker | woo-hoo! |
| 17:35.56 | starseeker | heads in |
| 17:41.09 | dloman | Ice clearing up? |
| 17:42.31 | brlcad | starseeker: awesome, thanks for fixing those! |
| 17:46.13 | brlcad | --enable-* turns compilation of things on/off |
| 17:46.26 | brlcad | s/things/code/ |
| 17:46.53 | brlcad | --with-* turns usage of system components on/off |
| 17:48.02 | brlcad | with/without X11, OpenGL, Mac Frameworks, particular threading types, etc |
| 17:50.09 | brlcad | enable/disable compiling things in our code like building src/other, compilation modes, subcompiles, etc |
| 17:50.46 | brlcad | another way to think of it is --enable/--disable => internal and --with/--without => external |
| 17:53.31 | CIA-53 | BRL-CAD: 03brlcad * r42914 10/brlcad/trunk/configure.ac: rtgl was okay as an --enable flag. enable/disable for inward code selection, with/without for outward system selection. |
| 17:53.50 | ``Erik | mmm, pröst coma |
| 17:54.57 | brlcad | mmm |
| 17:55.26 | brlcad | I need to stop coding and get back to studying soon |
| 18:04.43 | CIA-53 | BRL-CAD: 03erikgreenwald * r42915 10/brlcad/trunk/bench/pixcmp.c: break help message into two strings to bring it under 509 characters (ISO C90) |
| 18:10.39 | CIA-53 | BRL-CAD: 03brlcad * r42916 10/brlcad/trunk/INSTALL: document rtgl and expand the explanation of configuration options to detail what is different between the enable and with options. reword a little for clarity too. |
| 18:11.05 | CIA-53 | BRL-CAD: 03brlcad * r42917 10/brlcad/trunk/configure.ac: comment on rtgl option |
| 18:30.29 | CIA-53 | BRL-CAD: 03brlcad * r42918 10/brlcad/trunk/src/libged/vdraw.c: scanning into a coerced pointer probably isn't a good idea, so revert back down to unsigned long for the scan and upconvert the comparisons accordingly |
| 18:33.07 | CIA-53 | BRL-CAD: 03brlcad * r42919 10/brlcad/trunk/src/libged/wdb_vdraw.c: this duplication of effort really sucks. fix the same size_t uind issues that were in vdraw.c |
| 18:34.50 | brlcad | that should do it, beginning another round of testing |
| 18:36.16 | CIA-53 | BRL-CAD: 03tbrowder2 * r42920 10/brlcad/trunk/HACKING: stick with rough English translation--rest is covered in later parts |
| 18:39.15 | brlcad | mm, crit is full |
| 18:40.05 | ``Erik | in /home ? there's some issue with rsync duping things or something |
| 18:40.21 | brlcad | intentional |
| 18:40.27 | brlcad | not duping, just not deleting |
| 18:40.34 | brlcad | i'm cleaning up some space |
| 18:41.23 | brlcad | easy enough to clear out the big offenders and let them re-rsync later .. checking usage now |
| 18:51.45 | dloman | anything i did? |
| 18:53.17 | *** join/#brlcad Ralith (~ralith@d142-058-094-230.wireless.sfu.ca) |
| 18:53.18 | ``Erik | nah |
| 18:53.56 | ``Erik | stuff like me doing BRL-CAD builds, people collating/cleaning/deleting irc logs on bz and having dups show up on crit, etc |
| 18:54.45 | ``Erik | the way it's set up, if I go on bz and decide to compress every text file in a directory, crit will have both the compressed and uncompressed ones sitting next to eachother |
| 18:57.11 | CIA-53 | BRL-CAD: 03brlcad * r42921 10/brlcad/trunk/HACKING: |
| 18:57.11 | CIA-53 | BRL-CAD: split the difference. emphasize build breaking and maintainability concern in |
| 18:57.11 | CIA-53 | BRL-CAD: the second point as developer mantra. also, add a brief hint to the first point |
| 18:57.11 | CIA-53 | BRL-CAD: that it's speaking to people's character and the community, not just the code. |
| 18:58.27 | epileg | is there someone working on the creation of rpm package for fedora? |
| 19:02.52 | brlcad | there, lots of space |
| 19:03.11 | brlcad | ``Erik: big offender was a massive core dump that got syncd :) |
| 19:03.58 | brlcad | there's 40GB |
| 19:04.01 | ``Erik | ahhh |
| 19:04.21 | ``Erik | that'd do it |
| 19:05.50 | ``Erik | epileg: I think starseeker tried to set up a fedora image to try to get the spec.in up to date, it's not exactly a priority for us unless someone wants to step up and take care of it |
| 19:06.33 | brlcad | epileg: go for it ;) |
| 19:06.57 | brlcad | sh/make_rpm.sh :D |
| 19:12.59 | epileg | Ok, I'll try, but in rpm systems the situation is different than in debian like ones. They share the same packaging system (rpm), but the package nomenclature is different between distributions. |
| 19:14.25 | epileg | so, the fedora rpm package will be only for fedora, and not installable in opensuse |
| 19:14.37 | ``Erik | starseeker: dm-rtgl is chock full of GLX stuff, allowing it to try to build on winderz won't work |
| 19:16.57 | CIA-53 | BRL-CAD: 03Sean 07http://brlcad.org * r2454 10/wiki/Community_Publication_Portal: update date, still in draft |
| 19:17.50 | brlcad | epileg: ah, interesting -- didn't know that |
| 19:18.32 | brlcad | there can be a --fedora or --generic flag to the script for making the other type |
| 19:19.27 | brlcad | right now, there is no rpm generation whatsoever, so it almost goes without saying that anything you do will be an improvement ;) |
| 19:23.09 | ``Erik | well, brlcad.spec.in and just run rpm -someflag on it, iirc, but it needs write access to like /usr/src/redhat/SOURCES/ and stuff, iirc |
| 19:25.33 | brlcad | so the script is probably a one-liner or maybe few-liner with error-checking on environment assumptions |
| 19:25.49 | brlcad | that's still one line I don't want to memorize :) |
| 19:26.43 | ``Erik | yeh, I built it into a make target a long time ago, um, dunno if it still exists in BRL-CAD, but I can dig it up in other old projects (if it's still even pertinent... I think the spec format changed, that's why starseeker was trying to work on it) |
| 19:34.58 | brlcad | *shrug* |
| 19:35.36 | brlcad | I don't think there's anything in brl-cad any more -- if something doesn't get maintained and can't be verified that it works, I generally yank it so it doesn't slow down the next guy |
| 19:48.59 | ``Erik | misc/brlcad.spec.in is there and it's still in the AC_OUTPUT *shrug* |
| 19:50.33 | CIA-53 | BRL-CAD: 03brlcad * r42922 10/brlcad/trunk/HACKING: set jordi up as the .deb maintainer |
| 19:50.38 | brlcad | but no rpm line hooking it in |
| 19:50.45 | brlcad | just there for those that know what a spec file is |
| 19:51.09 | ``Erik | line 159 of /Makefile.am |
| 19:51.16 | ``Erik | :D |
| 19:51.48 | brlcad | ah, okay very good then |
| 19:51.57 | brlcad | probably should move that up into the make_rpm.sh script |
| 19:52.17 | brlcad | hm, distcheck is not happy on fbsd |
| 19:52.50 | brlcad | apparently its doing a "make distdir" on something in src/other that doesn't have that rule |
| 19:57.25 | starseeker | ``Erik: gaaaahhh. OK. |
| 19:57.33 | starseeker | rtgl is busted anyway |
| 20:03.41 | CIA-53 | BRL-CAD: 03starseeker * r42923 10/brlcad/branches/cmake/CMakeLists.txt: Don't try rtgl unless we've got X11 |
| 20:06.16 | starseeker | grrrrr.... now unistd.h and glxext.h are arguing on my Mac |
| 20:07.10 | starseeker | may be time to upgrade this baby |
| 20:10.43 | brlcad | ``Erik: soo, guess what |
| 20:10.56 | brlcad | I'm upgrading crit to better hardware :) |
| 20:11.16 | brlcad | they're going to attach the current hard drive via USB to the new hardware |
| 20:11.21 | brlcad | should more than double the performance |
| 20:13.12 | ``Erik | hah, cool |
| 20:13.30 | ``Erik | wonders how many hw iterations the 'new server' will go through before migration O:-) |
| 20:14.09 | ``Erik | might try doing "gmake distcheck" instead of "make distcheck"? |
| 20:14.28 | ``Erik | there're still a couple flakey rough spots between bsdmake and auto* |
| 20:18.18 | brlcad | I'm hoping this motivates me since it's in context now |
| 20:18.30 | brlcad | plus hella faster and bigger |
| 20:18.49 | brlcad | less bandwidth, but even .bz will only consume about 50% on average |
| 20:19.22 | brlcad | 1TB disk, dual P4 3.0, 2GB, 100Mbs |
| 20:19.45 | brlcad | 500GB transfer cap |
| 20:20.24 | ``Erik | nice, enough cpu that the load may drop below the number of cores O.O |
| 20:21.17 | ``Erik | and enough memory that it won't be in swapland like bz |
| 20:23.43 | *** join/#brlcad Ralith (~ralith@d142-058-076-096.wireless.sfu.ca) |
| 20:29.04 | *** join/#brlcad epileg (~epileg@unaffiliated/epileg) |
| 20:37.01 | *** join/#brlcad Ralith_ (~ralith@d142-058-076-096.wireless.sfu.ca) |
| 20:46.00 | CIA-53 | BRL-CAD: 03starseeker * r42924 10/brlcad/trunk/ (NEWS src/nirt/command.c): Ow. nirt units command was reporting all unit changes as invalid (even 'm') - this seems to fix it, but someone else should verify. |
| 20:46.41 | CIA-53 | BRL-CAD: 03starseeker * r42925 10/brlcad/trunk/NEWS: # -> * |
| 20:57.54 | *** join/#brlcad kanzure (~kanzure@131.252.130.248) |
| 21:08.34 | CIA-53 | BRL-CAD: 03bob1961 * r42926 10/brlcad/trunk/src/ (libged/rect.c libtclcad/ged_obj.c): Call dm_draw_rect() only if grs_draw and grs_line width are not zero. |
| 21:22.36 | brlcad | server transfer under way |
| 21:25.41 | epileg | who's the maintainer of "misc/brlcad.spec.in"? |
| 21:26.19 | brlcad | epileg: "svn log misc/brlcad.spec.in" will say who all has edited the file |
| 21:26.43 | epileg | thanks |
| 21:27.09 | brlcad | starseeker: remember that the last comment is the one that will show up in the configuraiton review report |
| 21:29.05 | brlcad | looks like distcheck and release tests are passing on mac now with only one test needing investigating |
| 21:29.24 | brlcad | fastgen test is reporting "realpath: No such file or directory" several times |
| 21:29.36 | brlcad | followed by a Tcl_Init failure |
| 21:30.13 | starseeker | brlcad: oh, sorry |
| 21:30.32 | brlcad | http://pastebin.mozilla.org/1015503 |
| 21:31.06 | CIA-53 | BRL-CAD: 03starseeker * r42927 10/brlcad/branches/cmake/src/other/step/CMake/SCL_Utils.cmake: Make the SCL install commands more generic |
| 21:31.26 | brlcad | ``Erik: 8.1 is no longer pre-release, stable okay? |
| 21:31.48 | ``Erik | stable is 8.2RC3 right n ow |
| 21:32.01 | ``Erik | once 8.2 comes out, I'll upgrade the system |
| 21:32.12 | ``Erik | so any old 8.x, right? um |
| 21:32.22 | brlcad | crit is presently 8.1-prerelease |
| 21:32.42 | ``Erik | it's running 8.1-prerelease, it has 8.2rc2 installed iirc, I just haven't been rebooting it |
| 21:32.44 | brlcad | you want 8.2RC3 or 8.1 stable? |
| 21:32.54 | brlcad | or doesn't matter |
| 21:32.59 | ``Erik | doesn't matter |
| 21:33.02 | brlcad | k |
| 21:33.28 | ``Erik | we'll upgrade when 8.2 goes stable and lock on that branch for security updates until 8.3 comes out, right? |
| 21:33.39 | starseeker | would be happy never to have to deal with path logic again... |
| 21:33.45 | starseeker | brlcad: that's on a 10.6 mac? |
| 21:33.51 | ``Erik | http://www.freebsd.org/releases/8.2R/schedule.html |
| 21:38.32 | CIA-53 | BRL-CAD: 03starseeker * r42928 10/brlcad/trunk/NEWS: nirt units command was reporting all units as invalid - fixed |
| 21:41.36 | brlcad | starseeker: yes |
| 21:41.58 | starseeker | nevermind, I see it here too. |
| 21:42.02 | starseeker | g_diff is failing |
| 21:42.56 | starseeker | that may be the tclcad_tcl_library() removal |
| 21:44.23 | starseeker | no, wait... |
| 21:44.31 | starseeker | maybe not g_diff... |
| 21:44.47 | brlcad | g_diff probably has its own tcl interpreter |
| 21:45.00 | brlcad | so if a change was made to bwish and mged, might need something there too |
| 21:45.02 | starseeker | it does, but running it in isolation succeeds |
| 21:47.22 | starseeker | ah, there we go |
| 21:47.36 | starseeker | running it from toplevel dir succeeds, not from regress |
| 21:47.40 | starseeker | digs... |
| 21:55.00 | starseeker | that's why... |
| 21:55.05 | brlcad | realpath blather found/fixed |
| 21:55.10 | CIA-53 | BRL-CAD: 03brlcad * r42929 10/brlcad/trunk/src/libbu/brlcad_path.c: there's no need to perror() on realpath failure |
| 21:56.59 | CIA-53 | BRL-CAD: 03starseeker * r42930 10/brlcad/trunk/src/gtools/g_diff.c: By the time g_diff was doing Tcl_FindExecutable, argv[0] referred to the first file supplied and not the executable name. Instead, use the 'invoked_as' string |
| 21:57.00 | starseeker | brlcad: that should do it. |
| 21:57.37 | starseeker | has to go take cats to vet... |
| 21:57.50 | CIA-53 | BRL-CAD: 03brlcad * r42931 10/brlcad/trunk/configure.ac: comment tweak |
| 21:57.55 | brlcad | awesome, thanks! |
| 21:58.13 | brlcad | enjoy your delicious dinner ;) |
| 22:06.22 | brlcad | well now that's really odd |
| 22:07.07 | brlcad | running "../src/gtools/g_diff fastgen_unix.g fastgen_dos.g" stand-alone works, but inside the script with the same pwd it blathers a tcl failure |
| 22:09.46 | brlcad | er... someone changed the fastgen regression script |
| 22:10.37 | brlcad | supposed to create a unix line-ending fastgen conversion, create a dos line-ending fast conversion, and compare the two |
| 22:10.51 | brlcad | it's creating the dos version by simply copying the unix version! |
| 22:11.50 | epileg | @MAJOR_VERSION@, @MINOR_VERSION@ and @PATCH_VERSION@ variables do not change when "misc/brlcad.spec" is created |
| 22:14.10 | brlcad | ahh, my bad .. was reading the script wrong .. turning my internal alarm off |
| 22:15.27 | brlcad | epileg: those don't sound like the right variables |
| 22:15.32 | ``Erik | hm, not being AC_SUBST'd anymore |
| 22:15.33 | brlcad | probably out of date |
| 22:15.50 | epileg | aha |
| 22:15.52 | ``Erik | probably switch the version to @BRLCAD_VERSION@ and the release to '1' or something, or take release out |
| 22:16.46 | epileg | yes. where can i find the variable list? |
| 22:18.06 | ``Erik | configure.ac |
| 22:18.12 | epileg | ok, thanks! |
| 22:18.14 | ``Erik | grep AC_SUBST configure.ac |
| 22:31.30 | CIA-53 | BRL-CAD: 03Sean 07http://brlcad.org * r2455 10/wiki/Community_Publication_Portal: less is more. be briefer. |
| 22:33.05 | CIA-53 | BRL-CAD: 03Sean 07http://brlcad.org * r2456 10/wiki/Community_Publication_Portal: move note down to publication section |
| 22:35.19 | CIA-53 | BRL-CAD: 03Sean 07http://brlcad.org * r2457 10/wiki/Community_Publication_Portal: reduce intro further |
| 23:34.44 | starseeker | growls... worst traffice I think I've ever run into on 22 |
| 23:34.50 | starseeker | missed the appointment |
| 23:35.04 | *** join/#brlcad Ralith (~ralith@d142-058-094-230.wireless.sfu.ca) |
| 23:36.04 | starseeker | brlcad: it's still failing? that should have fixed it... |
| 23:47.33 | CIA-53 | BRL-CAD: 03starseeker * r42932 10/brlcad/branches/cmake/ (15 files in 9 dirs): MFC r42931 |
| 23:50.26 | CIA-53 | BRL-CAD: 03starseeker * r42933 10/brlcad/branches/cmake/ (. src/other/step/ src/other/tkhtml/): mergeinfo property is annoying, not using it - remove from cmake branch |