| 00:19.13 | Twingy | it just built for me *shrug* |
| 00:19.36 | Twingy | Done. |
| 00:19.36 | Twingy | BRL-CAD Release 7.3.0, Build 20050617 |
| 00:19.36 | Twingy | Elapsed compilation time: 19 minutes, 18 seconds |
| 00:19.54 | ``Erik | build for me before I left the office |
| 00:19.57 | ``Erik | built, even |
| 00:28.43 | archivist | i seem to have a borked download and the animal is refusing to do another cvs checkout (my winbox is quite happy though) |
| 00:29.49 | Twingy | that might explain things |
| 00:33.54 | archivist | hehe must type ALL the cvs cmd |
| 00:37.08 | archivist | now getting aclocal: configure.ac: 1409: macro `AM_PATH_SDL` not found in library |
| 00:37.17 | archivist | aclocal failed |
| 00:49.32 | Twingy | hrm |
| 00:49.34 | Twingy | what OS? |
| 00:49.36 | CIA-3 | BRL-CAD: 03twingy * 10brlcad/src/adrt/libtie/ (define.h struct.h tie.c tie.h): Added licensing info to libtie files. |
| 00:50.53 | archivist | debian sarge the noo one |
| 00:51.11 | Twingy | it's wanting SDL for ADRT the new realtime raytracer |
| 00:51.37 | Twingy | I suppose I can have it skip ADRT if configure cannot find SDL on the system.. |
| 00:51.47 | Twingy | I think SDL comes standard with redhat now |
| 00:52.10 | Twingy | archivist, are you familiar with SDL? |
| 00:52.14 | archivist | or tell user to fetch it |
| 00:52.18 | Twingy | yah |
| 00:52.20 | archivist | no |
| 00:52.30 | Twingy | libsdl.org |
| 00:52.49 | Twingy | lemme poke the configure.ac |
| 00:54.57 | Twingy | do you have sdl-config on your machine? |
| 00:55.35 | archivist | dunno |
| 00:55.40 | Twingy | fine out :) |
| 00:55.45 | Twingy | type "sdl-config" |
| 00:55.47 | Twingy | *find |
| 00:57.22 | archivist | nope |
| 00:57.25 | Twingy | k |
| 00:57.38 | Twingy | for now just grab sdl 1.2 and install it |
| 00:57.53 | Twingy | you can obtain it via aptget I think |
| 01:01.45 | Twingy | k? |
| 01:02.28 | ``Erik | configure.ac isn't the place to put that, it needs the m4, period... the autogen.sh machine needs sdl.m4, period |
| 01:02.43 | ``Erik | did I repeat myself enough? :) |
| 01:03.20 | archivist | command not found for aptget on this box, they want one to use the new synaptic thing (which in this case doesnt know about the web on this box) |
| 01:03.31 | Twingy | so ... you want to include it in the distribution or enforce it must be installed on all users machines? |
| 01:04.08 | ``Erik | when the dist gets build, the m4 data is in the thing... dist tarballs are ok and don't need sdl installed on the box... only the machine running auto* has the requirement |
| 01:04.40 | Twingy | so in archivist's case we'll just enforce that he needs libsdl installed |
| 01:04.46 | ``Erik | indeed |
| 01:04.57 | ``Erik | (apt-get, not aptget) |
| 01:05.04 | Twingy | archivist, talk to erik, I only know slackware and redhat |
| 01:06.38 | ``Erik | (apt-get install libsdl-dev) |
| 01:08.51 | archivist | its coming |
| 01:10.02 | CIA-3 | BRL-CAD: 03twingy * 10brlcad/src/adrt/libtienet/ (8 files): Added licensing info to libtienet files. |
| 01:21.04 | archivist | going forrunning make again i get the following configure.ac:110 required file misc/config.guess not found also misc/config.sub and misc/ltmain.sh |
| 01:21.19 | Twingy | erm |
| 01:21.22 | Twingy | do an autoreconf -vis |
| 01:21.25 | Twingy | err |
| 01:21.27 | Twingy | ./autogen.sh |
| 01:21.37 | archivist | ok |
| 01:23.11 | archivist | i need a faster box |
| 01:23.48 | Twingy | hehe |
| 01:24.05 | Twingy | got got a couple dual cpu dual core 2ghz opterons with 8GB coming in in 2 weeks |
| 01:24.27 | Twingy | one of those should compile brlcad in under 4 minutes |
| 01:24.49 | archivist | mine is a dual 300meg pentium2 |
| 01:24.55 | Twingy | ouch |
| 01:25.00 | Twingy | you're doing make -j3 right? |
| 01:25.23 | archivist | no |
| 01:25.30 | Twingy | what are you doing? |
| 01:25.56 | archivist | just make |
| 01:26.35 | Twingy | punk |
| 01:26.42 | Twingy | all I had was quesadillas |
| 01:26.55 | brlcad | big thing of "premium" sake too.. |
| 01:26.56 | brlcad | yum |
| 01:27.24 | Twingy | I fixed samba today |
| 01:30.03 | brlcad | archivist: it sounds like you probably got a partial set of cvs commits at some update |
| 01:30.14 | Twingy | o.O |
| 01:30.21 | Twingy | *boom* |
| 01:30.34 | CIA-3 | BRL-CAD: 03twingy * 10brlcad/src/adrt/libutil/ (10 files): Added licensing info to libutil files. |
| 01:30.58 | brlcad | Twingy: you do know that there is a script that will add the header for you? |
| 01:31.07 | brlcad | sh/header.sh |
| 01:31.08 | Twingy | I figured there was |
| 01:31.25 | Twingy | does it operate on files with existing header info? |
| 01:31.36 | brlcad | yes, should |
| 01:31.54 | Twingy | cause I had no standard when creating the header stuff |
| 01:32.02 | Twingy | so some of them have cvs stuff, some don't etc |
| 01:32.10 | brlcad | read the header.sh header for examples how to run it |
| 01:32.56 | brlcad | basically sh/header.sh LGPL some_file.c |
| 01:33.13 | brlcad | couple that with find and you can do batches at onces |
| 01:33.34 | brlcad | there's also footer.sh |
| 01:34.11 | Twingy | hrm |
| 01:34.16 | Twingy | it doesn't do author and purpose does it? |
| 01:34.25 | brlcad | no, just the license header block |
| 01:34.29 | Twingy | ah |
| 01:34.38 | Twingy | well I'm just doing a cut and paste and editing descriptions for each one... |
| 01:34.45 | Twingy | so... |
| 01:35.01 | brlcad | heh, running the script once would still be faster :) |
| 01:35.21 | brlcad | it does the file T I T L E . C expansion thing too |
| 01:35.32 | ``Erik | (also, using an editor that could do multiple buffers with shared clipboards, etc... but... :) |
| 01:35.59 | Twingy | *shrug* I don't see the big advantage |
| 01:36.27 | brlcad | the big advantage? you've spent how many minutes cut and pasting now? |
| 01:36.43 | Twingy | mostly editing description stuff |
| 01:36.46 | brlcad | I did all 5000 or whatever existing files in less than a minute |
| 01:36.51 | Twingy | yes |
| 01:36.55 | Twingy | but did you enter descriptions for all of them? |
| 01:37.04 | Twingy | "This is librt it contains functions that do blah blah blah" |
| 01:37.32 | brlcad | already existed, and it prepends or merges with an existing header |
| 01:37.43 | Twingy | yah, see with my stuff it's all in my head |
| 01:37.53 | Twingy | 'sides I'm already 20% done |
| 01:38.00 | Twingy | I'll be done in an hour for the entire project |
| 01:38.05 | archivist | ./configure elapsed time 4m 25 |
| 01:38.16 | brlcad | slow machine :) |
| 01:38.18 | Twingy | archivist, make sure you do make -j3 when you compile |
| 01:38.28 | brlcad | he has an smp machinE? |
| 01:38.39 | Twingy | yes |
| 01:38.51 | Twingy | otherwise I wouldn't recommend it o.O |
| 01:38.59 | brlcad | never know with you |
| 01:39.02 | Twingy | though make -j2 would have a small benefit |
| 01:39.16 | Twingy | thanks for that moral booster |
| 01:39.16 | archivist | are but i dont think debian enabled smp |
| 01:39.33 | brlcad | archivist: that doesn't matter |
| 01:39.35 | ``Erik | um, cat /proc/cpuinfo |
| 01:39.41 | ``Erik | if two cpu's are listed, the kernel is using two cpu's |
| 01:40.41 | brlcad | cat: /proc/cpuinfo: This ain't no damn linux box |
| 01:40.46 | archivist | only 1 |
| 01:41.12 | archivist | dozy debian install |
| 01:41.37 | Twingy | brlcad, I think you had too much Sake |
| 01:41.44 | Twingy | we're refering to archivist's machine |
| 01:41.55 | brlcad | one can never have enough |
| 01:42.08 | Twingy | uh huh, and who was the one that passed out that night at siggraph? :) |
| 01:42.16 | Twingy | wasn't me ;) |
| 01:42.17 | brlcad | it was like 2am |
| 01:42.24 | Twingy | and your point? :) |
| 01:43.05 | archivist | its 2:42 am here and no sake |
| 01:45.13 | brlcad | archivist: sorry to hear that :) |
| 01:45.58 | archivist | so running plain make as box is brain dead |
| 01:46.12 | Twingy | as box? |
| 01:46.30 | archivist | only 1 cpu used |
| 01:46.39 | Twingy | sux 2 b u |
| 01:47.16 | brlcad | others should have gone to this conference |
| 01:47.38 | Twingy | if you're interested in solid modeling *shrug* |
| 01:50.21 | brlcad | lots of great papers on mesh generation, mesh healing, segmentation for efficient FEM purposes, lots lots more |
| 01:51.06 | brlcad | the most exciting to me was a knapsack packing algorithm using simulated annealing |
| 01:52.09 | brlcad | I think I can extend that to arbitrary shapes pretty easily using our ray-based volume computation |
| 01:52.41 | Twingy | the FEM stuff might have interested me |
| 01:53.09 | Twingy | ingo will be at siggraph |
| 01:53.18 | brlcad | multilevel tetra meshes are pretty cool |
| 01:54.06 | brlcad | the lead of the visible human project was there |
| 01:54.30 | brlcad | clammoring that there needs to be more open source collaboration |
| 01:54.53 | brlcad | there toolkit was/is apparently open source now too |
| 01:56.32 | Twingy | hrm, I think it's time for some video games |
| 01:57.00 | CIA-3 | BRL-CAD: 03twingy * 10brlcad/src/adrt/libcommon/ (13 files): Added licensing info to libcommon files. |
| 01:58.03 | archivist | i think it stime this box finished compiling |
| 01:58.11 | Twingy | heh |
| 01:58.20 | Twingy | might wanna recompile your kernel tonight |
| 01:58.26 | Twingy | cut your brl-cad build time in half |
| 01:58.45 | Twingy | or spend $188 on a 1.5ghz walmart pc |
| 01:59.39 | archivist | i do PIC stuff as well |
| 02:00.22 | Twingy | I've got a 16F88 in my rocket control system |
| 02:01.21 | archivist | i use 18f627 for a clock we manufacture and other control thingies |
| 02:01.32 | Twingy | ah |
| 02:01.49 | Twingy | you run it at 32,768 hz then? |
| 02:02.01 | Twingy | or 131,072 hz |
| 02:02.46 | archivist | no its mains controled to force the user to keep it switched on and to time (its in a cigarette display) |
| 02:03.06 | archivist | so use the rc clock at 4mhz |
| 02:03.16 | Twingy | oh, those aren't very accurate :-\ |
| 02:03.21 | brlcad | egads! sdl can't be required to build |
| 02:03.28 | Twingy | why not? |
| 02:03.34 | brlcad | are you kidding? |
| 02:03.55 | brlcad | sdl is fine |
| 02:04.02 | brlcad | as a requirement to build is not |
| 02:04.04 | Twingy | well, most people will only be sucking down a dist |
| 02:04.13 | Twingy | not the cvs archive |
| 02:04.19 | Twingy | so they'll be fine |
| 02:04.19 | brlcad | that's specifically one thing we do NOT do |
| 02:04.27 | Twingy | hrm? |
| 02:04.28 | brlcad | i.e. make people go get things to build |
| 02:04.47 | Twingy | like I said, if they get a make dist of brl-cad |
| 02:04.50 | Twingy | it'll have sdl.m4 in there |
| 02:04.59 | Twingy | i.e the released version |
| 02:05.13 | brlcad | the sdl check aborts configure |
| 02:05.16 | Twingy | then they compile and it'll just not use sdl instead of fail |
| 02:05.18 | Twingy | yes |
| 02:05.22 | Twingy | cause you're using the cvs version |
| 02:05.32 | brlcad | eh? |
| 02:05.36 | brlcad | they still ahve to run configure |
| 02:05.42 | Twingy | a make dist should include the m4 in the tarball |
| 02:05.47 | Twingy | then when it look for AM_PATH_SDL |
| 02:05.49 | Twingy | it'll have the m4 there |
| 02:05.57 | Twingy | and just say "oh my, you don't have sdl, oh well" |
| 02:06.13 | brlcad | no, it says "you don't have sdl, abort" |
| 02:06.20 | Twingy | you did a make dist? |
| 02:06.26 | brlcad | AM_PATH_SDL(1.2.0,,AC_MSG_ERROR([Cannot find SDL.])) |
| 02:06.41 | Twingy | because you're compiling from the cvs repo |
| 02:06.50 | Twingy | not from a make dist tarball |
| 02:06.56 | Twingy | so you're assumed to be a developer |
| 02:06.58 | Twingy | not a user |
| 02:07.02 | Twingy | users will be using a make dist |
| 02:07.05 | Twingy | right? |
| 02:07.09 | brlcad | no |
| 02:07.17 | ``Erik | heh |
| 02:07.17 | brlcad | well, yes they will |
| 02:07.20 | ``Erik | twiggly |
| 02:07.29 | brlcad | but that's not how AC_MSG_ERROR works |
| 02:07.32 | ``Erik | that should set a 'I have sdl" variable in the true and send an AC_MSG_WARNING() |
| 02:07.37 | Twingy | jey |
| 02:07.41 | Twingy | op me on #siggraph plz |
| 02:07.45 | Twingy | and riisbot too |
| 02:08.00 | ``Erik | no |
| 02:08.05 | brlcad | heh |
| 02:08.37 | Twingy | I dunno |
| 02:08.41 | Twingy | you two can fight it out |
| 02:09.02 | ``Erik | I don't think there's anything to fight about |
| 02:09.18 | brlcad | it's just incomplete checks |
| 02:09.51 | ``Erik | the need to have sdl installed to autogen is expected and I think acceptable, but we do need to fix up the configure.ac so sdl isn't required on the host during ./configure&&make |
| 02:10.14 | brlcad | the stuff needed for sdl in configure is slightly ugly |
| 02:10.54 | brlcad | like i said, sdl in itself is fine .. it just shouldn't halt the build if it's not there or there and not a good version |
| 02:11.17 | Twingy | sounds good to me |
| 02:11.24 | brlcad | checking for sdl-config won't be sufficient |
| 02:12.21 | brlcad | there are actually distros of sdl that don't include it |
| 02:12.34 | brlcad | probably a bug on their part, but a reality nonetheless |
| 02:12.42 | ``Erik | hrm, that's what the m4 that comes with sdl does |
| 02:12.50 | brlcad | yeah, i know |
| 02:12.54 | brlcad | it's a freaking pita |
| 02:13.08 | brlcad | had to deal with it in bz too |
| 02:13.25 | ``Erik | heh, pkgconfig? :} |
| 02:13.41 | Twingy | then just copy&paste from bz |
| 02:13.46 | brlcad | no, just more checks for the libraries/facilities like other libs |
| 02:13.53 | ``Erik | there's so much that sdl-config fu's... :-/ like the cflags/ldflags |
| 02:14.16 | Twingy | I would just not support machines without sdl-config |
| 02:14.20 | brlcad | bz's "fix" is not suitable, their configure.in is rather different |
| 02:14.22 | ``Erik | sdl-config on winderz gives some really... interesting... results |
| 02:16.02 | brlcad | yeah, though most of the sdl-config cflag/ldflag stuff usually boils down to an -L/usr/whatever/lib -lSDL -I/usr/whatever/inclue |
| 02:16.04 | Twingy | hrm |
| 02:16.08 | Twingy | wish I had a 3d scanner here |
| 02:16.48 | ``Erik | and some -D stuff |
| 02:17.15 | brlcad | heh _THREAD_SAFE? :P |
| 02:17.47 | ``Erik | reentrant |
| 02:17.58 | ``Erik | thread_safe on mac |
| 02:23.06 | brlcad | wonder what actually does on mac |
| 02:23.28 | brlcad | there's not _THREAD_SAFE declarations in the headers and no libc_r |
| 02:23.50 | brlcad | it's a freebsd'ism afaik |
| 02:24.07 | brlcad | or maybe glibc'ism |
| 02:25.13 | brlcad | gah Preparing build ... aclocal: configure.ac: 1413: macro `AM_PATH_SDL' not found in library |
| 02:26.40 | archivist | hehe |
| 02:26.49 | archivist | i know that error |
| 02:27.01 | ``Erik | <PROTECTED> |
| 02:27.20 | brlcad | reentrant is the same as thread_safe |
| 02:27.52 | ``Erik | in meaning, not spelling :) |
| 02:28.03 | brlcad | true true |
| 02:28.57 | CIA-3 | BRL-CAD: 03brlcad * 10brlcad/m4/ (sdl.m4 Makefile.am): add sdl.m4 to unbreak my build |
| 02:29.11 | Twingy | http://graphics.stanford.edu/data/3Dscanrep/armadillo.jpg |
| 02:32.12 | Twingy | hrm |
| 02:32.23 | Twingy | ply-g doesn't seem to like this .ply |
| 02:32.57 | brlcad | hmm? i converted several of their ply's just a couple weeks ago |
| 02:33.13 | Twingy | -rw-r--r-- 1 justin justin 6919168 Jun 17 22:34 Armadillo.ply |
| 02:33.21 | Twingy | [justin@borg:~]$ ply-g Armadillo.ply armadillo.g |
| 02:33.29 | Twingy | ignoring face with 0 vertices |
| 02:33.32 | Twingy | times a billion |
| 02:34.03 | Twingy | straight off stanford's site |
| 02:34.30 | brlcad | *shrug* |
| 02:34.39 | Twingy | I'll see if I can make it smarter |
| 02:34.54 | brlcad | there's a -v verbose and -d debug flag |
| 02:36.04 | Twingy | there's color indicies in this file |
| 02:36.17 | brlcad | ahh |
| 02:36.25 | Twingy | so what's the deal with sdl? |
| 02:36.42 | brlcad | what do you mean? |
| 02:36.52 | Twingy | we figure out what we want to do behavior wise? |
| 02:37.20 | brlcad | ideally, it should test for sdl, and if available it uses it of course |
| 02:37.53 | brlcad | if not available, it falls back to either not compiling whatever needed it, and/or providing an alternate display means in whatever used it |
| 02:38.42 | brlcad | the autogen.sh is fixed, just had to add the m4 to our m4/ |
| 02:40.58 | archivist | it just made iges.g how far through the compile am i |
| 02:41.23 | brlcad | about 2/3rds |
| 02:41.49 | brlcad | maybe more |
| 02:41.49 | Twingy | archivist, you're already out of date |
| 02:41.52 | archivist | dam slow box :(( |
| 02:41.52 | Twingy | time to start over |
| 02:42.54 | archivist | hmm 3:42 am |
| 02:44.10 | brlcad | Twingy: you should add everything that was added to the NEWS file, maybe with a short paragraph or two writeup for the next release |
| 02:45.00 | brlcad | a single line for each user-visible (non-dev) change since the previous release |
| 02:45.25 | Twingy | for each change? |
| 02:45.42 | Twingy | hrm |
| 02:45.47 | Twingy | mged core dumps on me in freebsd |
| 02:46.29 | Twingy | mged and mged -c both core dump :( |
| 02:46.31 | brlcad | pretty much, the previous release notes are good examples to follow |
| 02:46.41 | Twingy | each change like, every cvs add? |
| 02:46.57 | Twingy | ...we added adrt. |
| 02:47.07 | Twingy | that's 1 change |
| 02:47.12 | brlcad | heck no, not every cvs add |
| 02:47.19 | brlcad | user-visible changes |
| 02:47.19 | Twingy | k, 1 change then |
| 02:47.37 | brlcad | which would be the addition of a handful of commands |
| 02:47.42 | Twingy | any clue why mged core dumps? |
| 02:48.00 | brlcad | core dumps on startup? on file open? |
| 02:48.04 | brlcad | try console mode |
| 02:48.04 | Twingy | on startup |
| 02:48.11 | Twingy | Twingy mged and mged -c both core dump :( |
| 02:49.20 | brlcad | try mged -f |
| 02:49.30 | Twingy | [justin@borg:~/src/brlcad/src/adrt]$ mged -f |
| 02:49.30 | Twingy | Segmentation fault (core dumped) |
| 02:49.42 | brlcad | gdb --args mged -f |
| 02:49.52 | Twingy | I compiled optimized |
| 02:49.59 | brlcad | that's fine |
| 02:50.04 | brlcad | still has debug symbols |
| 02:50.20 | brlcad | you would have had to --disable-debug |
| 02:50.27 | Twingy | (gdb) run |
| 02:50.28 | Twingy | Starting program: /usr/local/bin/mged -f |
| 02:50.28 | Twingy | Program received signal SIGSEGV, Segmentation fault. |
| 02:50.28 | Twingy | [Switching to LWP 100318] |
| 02:50.28 | Twingy | 0x285347e0 in pthread_testcancel () from /usr/lib/libpthread.so.1 |
| 02:50.28 | Twingy | (gdb) where |
| 02:50.30 | Twingy | #0 0x285347e0 in pthread_testcancel () from /usr/lib/libpthread.so.1 |
| 02:50.32 | Twingy | #1 0x2851ff2b in open () from /usr/lib/libpthread.so.1 |
| 02:50.34 | Twingy | #2 0x2861f1e1 in glXChannelRectSyncSGIX () from /usr/X11R6/lib/libGL.so.1 |
| 02:50.36 | Twingy | #3 0x28642e00 in ?? () from /usr/X11R6/lib/libGL.so.1 |
| 02:50.38 | Twingy | #4 0x00000002 in ?? () |
| 02:50.39 | brlcad | eww |
| 02:50.54 | brlcad | that's familar |
| 02:51.24 | brlcad | should only happen if it's linking in the wrong pthread library |
| 02:51.35 | brlcad | did you add cflags/ldflags? |
| 02:51.39 | Twingy | nope |
| 02:51.44 | brlcad | hmm.. sdl-config probably did |
| 02:52.09 | Twingy | need an account on a fbsd box? |
| 02:52.41 | brlcad | nope, I got a couple |
| 02:52.42 | Twingy | oh |
| 02:52.52 | Twingy | heh, I forgot you have access to ones at work from there |
| 02:52.54 | brlcad | did you get an account on that power5? |
| 02:53.02 | Twingy | nope, no desire to |
| 02:53.03 | brlcad | ftp.brlcad.org is fbsd |
| 02:53.26 | brlcad | I got one on both machines, and supposedly 4 tickets to that movie |
| 02:53.35 | archivist | oo just bombed here its in tclscripts lt-btclsh libbn.so.19 cant open shared library |
| 02:53.44 | Twingy | that sux |
| 02:54.15 | brlcad | archivist: when did you last build ok? |
| 02:54.23 | archivist | never |
| 02:54.27 | brlcad | heh |
| 02:54.36 | brlcad | which OS is this? |
| 02:54.45 | brlcad | debian? |
| 02:54.51 | archivist | debian sarge the nooo one |
| 02:55.06 | brlcad | "nooo" is an understatement ;) |
| 02:55.39 | brlcad | ls -la /usr/lib/libbn/.libs/*.so* |
| 02:55.48 | brlcad | er, brlcad/src/libbn |
| 02:57.59 | archivist | not there |
| 02:58.15 | brlcad | ls -la brlcad/src/libbn/.libs/* |
| 02:58.22 | brlcad | what is there? |
| 02:58.47 | brlcad | sounds like you have a build that's out of sync still |
| 02:59.12 | brlcad | like you've done builds before and didn't clean up properly or have a clock out of sync or something |
| 02:59.45 | archivist | hehe clock WAS out of sync earlier |
| 03:01.52 | archivist | rm .o for a quick fix in libbn ? |
| 03:04.45 | brlcad | cd src/libbn |
| 03:04.45 | brlcad | make clean |
| 03:04.45 | brlcad | make |
| 03:06.23 | archivist | running again |
| 03:06.47 | archivist | 4 am and i wanna go home |
| 03:07.06 | brlcad | if this is the first time you've built, how did you get the other builds? |
| 03:07.17 | brlcad | used the posted binaries? |
| 03:07.42 | archivist | who said i had other builds |
| 03:08.18 | brlcad | hm, I thought I remembered you saying you'd run mged |
| 03:08.30 | archivist | no |
| 03:10.41 | brlcad | ok |
| 03:14.17 | archivist | still same error |
| 03:15.49 | *** join/#brlcad JBdiGriz (~JBdiGriz@adsl-67-120-233-89.dsl.sndg02.pacbell.net) | |
| 03:16.35 | brlcad | howdy JBdiGriz ! |
| 03:16.50 | JBdiGriz | howdy, I was just looking for you |
| 03:29.30 | *** join/#brlcad archivist2 (~djc@host217-35-103-47.in-addr.btopenworld.com) | |
| 03:30.56 | brlcad | archivist: how'd it go? |
| 03:33.02 | archivist | it still cant find libbn.so* |
| 03:33.26 | archivist | no erros from the mak in libbn |
| 03:34.28 | archivist | just setting up the other box so i can copy paste |
| 03:35.47 | brlcad | but what does it actually make in the libbn directory |
| 03:36.10 | brlcad | ls -la src/libbn/.libs/libbn* |
| 03:40.03 | archivist2 | -rw-r--r-- 1 root root 1977188 2005-06-18 04:38 .libs/libbn.a |
| 03:40.03 | archivist2 | lrwxrwxrwx 1 root root 11 2005-06-18 04:38 .libs/libbn.la -> ../libbn.la |
| 03:40.03 | archivist2 | -rw-r--r-- 1 root root 936 2005-06-18 04:38 .libs/libbn.lai |
| 03:40.03 | archivist2 | lrwxrwxrwx 1 root root 15 2005-06-18 04:38 .libs/libbn.so -> libbn.so.19.0.1 |
| 03:40.04 | archivist2 | lrwxrwxrwx 1 root root 15 2005-06-18 04:38 .libs/libbn.so.19 -> libbn.so.19.0.1 |
| 03:40.06 | archivist2 | -rwxr-xr-x 1 root root 1858391 2005-06-18 04:38 .libs/libbn.so.19.0.1 |
| 03:40.37 | archivist | looks ok |
| 03:41.00 | brlcad | yeah, okay, so it's there |
| 03:41.07 | brlcad | cd src/bwish |
| 03:41.08 | brlcad | make clena |
| 03:41.13 | brlcad | er, clean |
| 03:41.16 | brlcad | and make |
| 03:41.21 | brlcad | then ./btclsh |
| 03:43.16 | archivist2 | portacabin:/home/djc/brlcad/src/bwish# ./btclsh |
| 03:43.16 | archivist2 | /home/djc/brlcad/src/bwish/.libs/lt-btclsh: error while loading shared libraries: libbn.so.19: cannot open shared object file: No such file or directory |
| 03:43.48 | brlcad | hrm, libtool bug |
| 03:43.55 | brlcad | ls -la .libs |
| 03:44.56 | archivist2 | drwxr-xr-x 2 root root 4096 2005-06-18 04:43 . |
| 03:44.56 | archivist2 | drwxr-xr-x 4 djc djc 4096 2005-06-18 04:42 .. |
| 03:44.56 | archivist2 | -rwxr-xr-x 1 root root 285558 2005-06-18 04:42 btclsh |
| 03:44.56 | archivist2 | -rwxr-xr-x 1 root root 428812 2005-06-18 04:42 bwish |
| 03:44.56 | archivist2 | -rwxr-xr-x 1 root root 285686 2005-06-18 04:43 lt-btclsh |
| 03:45.42 | brlcad | try running: .libs/btclsh |
| 03:46.29 | archivist2 | .libs/btclsh |
| 03:46.45 | archivist | 4:45 am and its light outhehe wrong window |
| 03:46.56 | brlcad | does it give you a % prompt? |
| 03:47.27 | archivist | no an aerror |
| 03:47.51 | archivist2 | portacabin:/home/djc/brlcad/src/bwish# .libs/btclsh |
| 03:47.51 | archivist2 | .libs/btclsh: error while loading shared libraries: librt.so.19: cannot open shared object file: No such file or directory |
| 03:48.12 | brlcad | ahh |
| 03:48.24 | brlcad | did librt build? |
| 03:48.50 | archivist | dunno where to look? |
| 03:48.59 | brlcad | ls -la src/librt/.libs/librt* |
| 03:51.32 | archivist2 | -rw-r--r-- 1 root root 15082652 2005-06-18 03:14 librt/.libs/librt.a |
| 03:51.32 | archivist2 | lrwxrwxrwx 1 root root 11 2005-06-18 03:14 librt/.libs/librt.la -> ../librt.la |
| 03:51.32 | archivist2 | -rw-r--r-- 1 root root 961 2005-06-18 03:14 librt/.libs/librt.lai |
| 03:51.32 | archivist2 | lrwxrwxrwx 1 root root 15 2005-06-18 03:14 librt/.libs/librt.so -> librt.so.19.0.1 |
| 03:51.32 | archivist2 | lrwxrwxrwx 1 root root 15 2005-06-18 03:14 librt/.libs/librt.so.19 -> librt.so.19.0.1 |
| 03:51.33 | archivist2 | -rwxr-xr-x 1 root root 13841082 2005-06-18 03:14 librt/.libs/librt.so.19.0. |
| 03:52.04 | brlcad | i hope that last line was really 19.0.1 ? |
| 03:52.26 | archivist2 | yup was my error |
| 03:53.04 | brlcad | hmmm |
| 03:53.37 | brlcad | if you want, I'll debug on it from here if you want to provide up a temp account |
| 03:54.16 | brlcad | something is wrong with libtool |
| 03:54.55 | archivist | router wont let you in i could reconfigure monday some time |
| 03:55.06 | brlcad | libtoolize --version ? |
| 03:55.34 | archivist | 1.5.6 |
| 03:56.10 | brlcad | that 'should' be okay |
| 03:56.19 | brlcad | built on it just fine before |
| 03:56.47 | brlcad | oh, hrm |
| 03:56.53 | brlcad | do you have an LD_LIBRARY_PATH set? |
| 03:57.22 | brlcad | set|grep -i ld_lib |
| 03:58.26 | brlcad | unset it if it's set |
| 03:58.43 | archivist | hehe havnt set the kb properly yet cant do | |
| 03:59.00 | brlcad | just set, and look for it :) |
| 03:59.56 | archivist | heh fount it under~ its not set |
| 04:01.31 | archivist2 | set |
| 04:01.38 | brlcad | hm? |
| 04:01.40 | brlcad | it's set? |
| 04:01.56 | archivist2 | no |
| 04:02.05 | brlcad | okay |
| 04:03.55 | brlcad | export LD_LIBRARY_PATH=/home/djc/brlcad/src/librt/.libs:/home/djc/brlcad/src/libbu/.libs:/home/djc/brlcad/src/libbn/.libs |
| 04:04.27 | brlcad | src/bwish/btclsh |
| 04:04.44 | brlcad | src/bwish/.libs/btclsh |
| 04:09.28 | archivist2 | <PROTECTED> |
| 04:09.29 | archivist2 | portacabin:/home/djc/brlcad/src# bwish/btclsh |
| 04:09.29 | archivist2 | /home/djc/brlcad/src/bwish/.libs/lt-btclsh: error while loading shared libraries: libtcl8.4.so.19: cannot open shared object file: No such file or directory |
| 04:09.38 | brlcad | cool, progress |
| 04:10.20 | brlcad | export LD_LIBRARY_PATH=/home/djc/brlcad/src/librt/.libs:/home/djc/brlcad/src/libbu/.libs:/home/djc/brlcad/src/libbn/.libs:/home/djc/brlcad/src/other/libtcl/.libs |
| 04:12.36 | archivist | ooo btclsh> prompt |
| 04:14.51 | archivist | so i run make in brlcad and..... |
| 04:15.21 | archivist2 | make[2]: Entering directory `/home/djc/brlcad/src/tclscripts' |
| 04:15.21 | archivist2 | ../../src/bwish/btclsh ../../src/tclscripts/ampi.tcl ../../src/tclscripts |
| 04:15.21 | archivist2 | /home/djc/brlcad/src/bwish/.libs/lt-btclsh: error while loading shared libraries: libbn.so.19: cannot open shared obje |
| 04:16.33 | brlcad | expected |
| 04:16.41 | brlcad | libtool still hijacks it |
| 04:17.13 | brlcad | cp src/bwish/.libs/btclsh src/bwish/.libs/lt-btclsh |
| 04:17.19 | brlcad | make |
| 04:19.14 | archivist | permmission denied |
| 04:19.35 | brlcad | eh? |
| 04:19.49 | brlcad | you can't keep switching between root and djc :) |
| 04:19.56 | brlcad | pick one and stick with it :) |
| 04:20.22 | brlcad | given you have root files now, you have to be root |
| 04:20.39 | brlcad | you generally should not compile as root |
| 04:21.03 | archivist | i did a chown and chgrp -R to set it all to djc and it seems to have gone wrong |
| 04:21.45 | brlcad | you'd have to chown as root |
| 04:21.47 | brlcad | to djc |
| 04:22.53 | archivist | i did |
| 04:22.57 | archivist2 | cp: cannot create regular file `bwish/.libs/lt-btclsh': Text file busy |
| 04:23.29 | brlcad | did you quit that btclsh prompt? |
| 04:23.41 | archivist2 | no just doing it |
| 04:24.04 | archivist2 | done |
| 04:24.42 | archivist2 | time for a make |
| 04:24.47 | archivist2 | ? |
| 04:24.54 | brlcad | yeah |
| 04:26.28 | archivist2 | ../../src/bwish/btclsh ../../src/tclscripts/ampi.tcl ../../src/tclscripts |
| 04:26.28 | archivist2 | /home/djc/brlcad/src/bwish/.libs/lt-btclsh: error while loading shared libraries: librt.so.19: cannot open shared obje |
| 04:29.53 | brlcad | cp src/bwish/.libs/btclsh src/bwish/btclsh |
| 04:31.22 | brlcad | need to make that ampi script not occur until install time to avoid this.. |
| 04:33.22 | archivist2 | i need a better chat client to copy from |
| 04:33.39 | Twingy | hrm |
| 04:36.20 | archivist2 | ../../src/bwish/btclsh: error while loading shared libraries: librt.so.19: cannot open shared object file: |
| 04:36.39 | archivist2 | are we chasing a never ending nasty |
| 04:37.21 | brlcad | no, but it'd probably be easier if I dealt with it directly |
| 04:39.50 | archivist2 | 5:40 can i leave it till monday as I need a b it of sleeeeeeeeep |
| 04:40.03 | brlcad | ok |
| 04:40.07 | brlcad | thanks for your patience |
| 04:40.17 | brlcad | sorry it's not just working for you |
| 04:40.18 | archivist2 | thanks for your patience |
| 04:40.30 | Twingy | hrm |
| 04:40.33 | archivist2 | hehe |
| 04:40.37 | Twingy | 5:40 is an acceptable time to leavae |
| 04:41.20 | archivist2 | im supposed to go out tomorow |
| 04:41.47 | brlcad | archivist2: monday is fine :) |
| 04:43.38 | Twingy | bu_malloc is in librt right? |
| 04:43.41 | Twingy | well |
| 04:43.43 | Twingy | libbu |
| 04:43.46 | brlcad | bu |
| 04:43.50 | Twingy | but does that come with librt.a ? |
| 04:43.59 | brlcad | should |
| 04:44.06 | Twingy | hrm |
| 04:44.10 | Twingy | I gots some linking problems |
| 04:44.21 | brlcad | wouldn't rely on it in a .a |
| 04:44.26 | brlcad | should link all three |
| 04:44.30 | Twingy | yah |
| 04:44.36 | Twingy | I think I will |
| 04:44.45 | Twingy | db_walk_tree |
| 04:44.50 | Twingy | that's librt thought right? |
| 04:44.55 | brlcad | yeah |
| 04:45.11 | Twingy | hrm |
| 04:45.23 | Twingy | ../../../../src/librt/.libs/librt.a |
| 04:45.28 | Twingy | ../../../../src/adrt/libcommon/.libs/libcommon.a(g.o)(.text+0x1175):/usr/home/justin/src/brlcad/src/adrt/libcommon/g.c:497: undefined reference to `db_walk_tree' |
| 04:45.42 | brlcad | libbu is on librt's LIBADD line |
| 04:45.48 | Twingy | ah |
| 04:45.57 | Twingy | but I don't understand why it's not picking up db_walk_tree |
| 04:46.01 | Twingy | from librt.a |
| 04:48.35 | brlcad | what's your link line? |
| 04:49.36 | Twingy | one sec phone |
| 04:49.42 | brlcad | <PROTECTED> |
| 04:50.39 | brlcad | it's a libtool library, you can't just make it through flags |
| 04:50.49 | brlcad | have to set it as a libadd |
| 04:51.36 | Twingy | I did have in libadd |
| 04:51.43 | Twingy | k |
| 04:51.50 | brlcad | not the one in cvs |
| 04:52.03 | Twingy | night |
| 04:57.50 | brlcad | autogen runs autoreconf |
| 04:58.01 | Twingy | yah |
| 04:58.09 | Twingy | just on an order of magnitude more stuff :-\ |
| 04:58.10 | brlcad | it takes longer because the configure script is 20x longer |
| 04:58.59 | brlcad | you shouldn't have to reconf if everything is set up |
| 04:59.12 | Twingy | I'm tweaking Makefile.am's |
| 04:59.14 | brlcad | you edit the am and it should rebuild itself |
| 04:59.26 | Twingy | to et bu_* and db_walk_tree resolved |
| 04:59.34 | Twingy | hrm |
| 04:59.39 | Twingy | doesn't seem to be the case for me |
| 04:59.41 | brlcad | that's what I mean -- editing .am's should rebuild the makefile |
| 04:59.47 | brlcad | make Makefile |
| 04:59.49 | Twingy | yes |
| 04:59.49 | brlcad | make |
| 04:59.55 | Twingy | it used to do that before I merged into brlcad |
| 05:00.27 | brlcad | it's no different |
| 05:00.31 | Twingy | LDADD = ${top_builddir}/src/librt/.libs/librt.a |
| 05:00.36 | Twingy | in Makefile.am for libcommon |
| 05:00.47 | Twingy | and I still get lcad/src/adrt/libcommon/g.c:497: undefined reference to `db_walk_tree' |
| 05:00.47 | Twingy | *** Error code 1 |
| 05:00.47 | brlcad | it should be attached to the build target |
| 05:01.09 | brlcad | libcommon_la_LIBADD |
| 05:01.14 | Twingy | ah |
| 05:01.15 | brlcad | it's not an LDADD |
| 05:02.17 | brlcad | same goes for the CFLAGS if you want to be correct |
| 05:02.22 | Twingy | gcc -D__STRICT_ANSI__ -I.. -I../../../../src/adrt/libtie -I../../../../src/adrt/libtienet -I../../../../src/adrt/libtexture -I../../../../src/adrt/libcommon -I../../../../src/adrt/librender -I../../../../src/adrt/libutil -I../../../../src/adrt/rise -pipe -fno-strict-aliasing -fno-common -fPIC -O3 -ffast-math -fstrength-reduce -fexpensive-optimizations -fomit-frame-pointer -finline-functions -g -pipe -fno-strict-aliasing -fno-commo |
| 05:02.22 | Twingy | n -fPIC -O3 -ffast-math -fstrength-reduce -fexpensive-optimizations -fomit-frame-pointer -finline-functions -g -o rise_master dispatcher.o main.o master.o post.o -L/usr/X11R6/lib -L/usr/local/lib -lpthread ../../../../src/adrt/libcommon/.libs/libcommon.a ../../../../src/adrt/librender/.libs/librender.a ../../../../src/adrt/libutil/.libs/libutil.a -lm ../../../../src/adrt/libtexture/.libs/libtexture.a ../../../../src/adrt/libtiene |
| 05:02.23 | Twingy | t/.libs/libtienet.a ../../../../src/adrt/libtie/.libs/libtie.a -lz |
| 05:02.30 | brlcad | otherwise there are issues overriding |
| 05:03.12 | brlcad | you have to make the makefile firest |
| 05:03.24 | brlcad | otherwise it'll build the target before the makefile |
| 05:03.50 | Twingy | and I do that by typing? |
| 05:03.54 | brlcad | now you'll have to delete the target too |
| 05:04.08 | Twingy | make clean in libcommon? |
| 05:04.14 | brlcad | sure |
| 05:04.25 | brlcad | then make |
| 05:04.44 | Twingy | ../../../../src/adrt/libcommon/.libs/libcommon.a(g.o)(.text+0x1175):/home/justin/src/brlcad/src/adrt/libcommon/g.c:497: undefined reference to `db_walk_tree' |
| 05:05.09 | brlcad | is it on the link line? |
| 05:05.28 | Twingy | ../../../../src/adrt/libcommon/.libs/libcommon.a ../../../../src/adrt/librender/.libs/librender.a ../../../../src/adrt/libutil/.libs/libutil.a -lm ../../../../src/adrt/libtexture/.libs/libtexture.a ../../../../src/adrt/libtienet/.libs/libtienet.a ../../../../src/adrt/libtie/.libs/libtie.a -lz |
| 05:05.44 | Twingy | I thought by doing ldadd stuff it's now part of libcommon |
| 05:05.59 | Twingy | as one big honkin archive |
| 05:06.01 | brlcad | it's not ldadd(!) |
| 05:06.06 | Twingy | I didn't do ldadd |
| 05:06.14 | Twingy | lib_LTLIBRARIES = libcommon.la |
| 05:06.14 | Twingy | libcommon_la_SOURCES = canim.c cdb.c env.c pack.c unpack.c g.c |
| 05:06.14 | Twingy | libcommon_la_LIBADD = ${top_builddir}/src/librt/.libs/librt.a |
| 05:06.23 | Twingy | k? |
| 05:06.28 | Twingy | that's the stuff yo |
| 05:06.48 | brlcad | heh |
| 05:06.59 | brlcad | that's probably why |
| 05:07.04 | Twingy | hrm? |
| 05:07.35 | brlcad | not the right libadd for libtool |
| 05:07.42 | Twingy | k |
| 05:07.53 | Twingy | what's it want then? |
| 05:07.54 | brlcad | you've never suppoed to link/think/look at anything in .libs |
| 05:08.06 | Twingy | I do it for performance reasons |
| 05:08.13 | brlcad | do what? |
| 05:08.18 | Twingy | include the .a's |
| 05:08.25 | brlcad | that's not what I said |
| 05:08.35 | brlcad | you link static which has nothing to do with the .libs dir |
| 05:08.42 | brlcad | there's a -static libtool flag |
| 05:08.50 | Twingy | *shrug* I'm doing .libs stuff all over adrt |
| 05:08.59 | Twingy | and it works fine... cept with librt apparently |
| 05:09.00 | brlcad | yeah, and that's completely not portable |
| 05:09.06 | Twingy | not portable? |
| 05:09.12 | Twingy | it works on a bunch of OS's |
| 05:09.15 | brlcad | not according to libtool in the least |
| 05:09.21 | Twingy | *shrug* |
| 05:09.35 | brlcad | you didn't read the libtool manual did you |
| 05:09.38 | Twingy | newp |
| 05:09.43 | Twingy | I let erik do all that auto-fu |
| 05:10.03 | brlcad | use ${LIBRT} |
| 05:10.46 | brlcad | and add -static as an libcommon_la_LDFLAGS |
| 05:12.12 | brlcad | working on "a bunch of OS's" doesn't make it right, just happenstance |
| 05:12.15 | Twingy | meh, still not working, I'll figure something out, you should get some sleep |
| 05:12.27 | brlcad | thinks only really get different on systems we don't use much |
| 05:12.47 | Twingy | <-- only plans on supporting systems we use alot |
| 05:13.10 | Twingy | nobody will be using a realtime raytracer on a vax |
| 05:13.13 | brlcad | well, brl-cad supports more systems than you :) |
| 05:13.15 | Twingy | or a cray |
| 05:13.17 | Twingy | yah |
| 05:13.17 | Twingy | well |
| 05:13.23 | Twingy | lee had the bright idea of shoving it into brlcad repo |
| 05:13.37 | Twingy | *shrug* |
| 05:14.07 | brlcad | meh, it'll do more for adrt than it'd ever do on it's own |
| 05:14.24 | Twingy | in what sense? |
| 05:14.39 | brlcad | exposure/recognition |
| 05:14.44 | brlcad | downloads |
| 05:14.54 | Twingy | *shrug* I think a separate tarball would've been more effective |
| 05:15.07 | Twingy | quicker to d/l and compile if anything |
| 05:15.19 | Twingy | 12 minutes for cvs checkout and 12 minutes to build versions 1 minute to download and build |
| 05:15.27 | brlcad | perhaps, but the numbers would have been far fewer I bet |
| 05:15.28 | Twingy | s/versions/versus |
| 05:15.40 | brlcad | brl-cad's getting 2-4 thousand downloads a month right now |
| 05:15.46 | brlcad | with little publicity |
| 05:16.05 | Twingy | yah, but I dunno how many of those people are interested in a realtime raytracer... |
| 05:16.32 | Twingy | we could put the linux operating system in there too, but that doesn't tell me how many people are using linux |
| 05:17.11 | brlcad | sure, but you are getting installed on way more systems so exposure is intrinisically higher than it would have been |
| 05:17.22 | brlcad | and it will be in the documentation, release notes, etc |
| 05:17.30 | brlcad | so people will learn about it, try it |
| 05:17.42 | Twingy | yep |
| 05:23.24 | brlcad | the only real downside in being included in brl-cad is that adrt will have to build on more systems and more flexibly (e.g. sdl as an option eventually) |
| 05:24.27 | brlcad | the build time could be cut down with something like what was done for muves --enable-only-adrt-build for example |
| 05:28.54 | brlcad | is the path tracer in there as well? |
| 05:29.36 | Twingy | yes |
| 05:30.10 | brlcad | people will probably like that more than anything I bet |
| 05:30.33 | brlcad | suckers for pretty pictures, no matter what the cputime |
| 05:31.13 | Twingy | k fixed the rebuilding of Makefile |
| 05:31.19 | Twingy | it apparently wanted automake-1.9 |
| 05:31.22 | Twingy | and I had automake19 |
| 05:31.30 | Twingy | and I always ln -s automake19 to automake |
| 05:31.33 | Twingy | so now I |
| 05:31.37 | Twingy | ln -s automake19 automake-1.9 |
| 05:31.41 | Twingy | problem #1 solved |
| 05:31.45 | Twingy | problem #2 |
| 05:31.52 | brlcad | ahh, yeah cept there are like 3 automake tools |
| 05:32.00 | Twingy | fuck em |
| 05:32.24 | brlcad | it won't work right if they're not in sync |
| 05:32.37 | Twingy | if I see it complain |
| 05:32.38 | Twingy | I'll fix it. |
| 05:32.41 | brlcad | i mean it "might" work right .. but that would just be luck |
| 05:32.55 | brlcad | i've had that same problem bite me in the ass a couple times on fbsd |
| 05:32.59 | Twingy | right now I need to figure out why these archives are being to god damn bitchy |
| 05:34.49 | brlcad | how in the world do you get away with using lpthread on freebsd in rise's observer |
| 05:35.03 | brlcad | that really shouldn't work |
| 05:35.07 | Twingy | sure it should |
| 05:35.12 | Twingy | there's like 3 threading models in fbsd |
| 05:35.14 | Twingy | lpthread is one of em |
| 05:35.40 | brlcad | -pthread is the switch, however |
| 05:36.05 | Twingy | *shrug* |
| 05:36.09 | Twingy | not a problem right now |
| 05:36.30 | brlcad | do you actually use pthreads? |
| 05:36.37 | Twingy | of course |
| 05:36.54 | Twingy | grep for pthread |
| 05:36.58 | Twingy | you'll get tons of code |
| 05:38.23 | brlcad | I'm just surprised it actually doesn't crash |
| 05:38.36 | brlcad | pthread is pretty much required by the libc |
| 05:38.42 | brlcad | lpthread conflicts with it |
| 05:38.53 | brlcad | you should get a crash like you saw in mged |
| 05:39.39 | brlcad | wicked.. maybe that's why mged crashes on you.. you've swapped out your -pthread lib maybe |
| 05:40.02 | Twingy | ? |
| 05:40.15 | Twingy | I'm running a standard fbsd distro |
| 05:40.21 | Twingy | nothing fancy schmancy |
| 05:40.26 | Twingy | same config as branch server and shiva |
| 05:40.39 | brlcad | then yeah, it should crash on you |
| 05:40.40 | Twingy | so if it breaks here, good chance it breaks on shiva |
| 05:40.57 | Twingy | what should crash |
| 05:41.01 | Twingy | adrt stuff or mged? |
| 05:41.05 | brlcad | adrt |
| 05:41.35 | brlcad | mged works fine on branch server, shiva, etc as of a week ago |
| 05:41.53 | brlcad | regardless, the -pthread vs -lpthread thing is a fairly well known issue |
| 05:42.45 | Twingy | hrm |
| 05:42.51 | Twingy | this is surprisingly annoying me |
| 05:43.10 | Twingy | you feed it libbu.a |
| 05:43.16 | Twingy | and it's like, no effect |
| 05:43.32 | Twingy | it's working in the opposite way the rest of the libs in adrt are |
| 05:43.50 | Twingy | I say, each libblah.a bitch, and it's like "thank you, problem resolved" |
| 05:43.54 | Twingy | s/each/eat |
| 05:44.08 | Twingy | I feed it librt.a |
| 05:44.16 | Twingy | and it's like, "pardon? did you do something?" |
| 05:44.51 | brlcad | i told you, you're not supposed to do _anything_ inside .libs |
| 05:45.05 | brlcad | so it really is your own fault, even if it did work before |
| 05:45.13 | Twingy | well |
| 05:45.15 | brlcad | use the librt.la on LIBADD |
| 05:45.23 | brlcad | and add -static to LDFLAGS |
| 05:45.58 | Twingy | to rise/master/Makefile.am or libcommon/Makefile.am ? |
| 05:46.02 | brlcad | that's just how libtool works -- there are platforms that don't perform symbol resolution until you actually install anyways |
| 05:46.25 | brlcad | libraries use LIBADD, execs use LDADD |
| 05:46.45 | brlcad | both would need static if you only make static libraries |
| 05:48.05 | brlcad | basically, everywhere you have .libs/libblah.a, it can be swapped with libblah.la and add -static to LDFLAGS |
| 05:48.42 | Twingy | libcommon_la_SOURCES = canim.c cdb.c env.c pack.c unpack.c g.c |
| 05:48.42 | Twingy | libcommon_la_LDFLAGS = -static |
| 05:48.42 | Twingy | libcommon_la_LIBADD = ${top_srcdir}/src/librt/librt.la |
| 05:49.05 | Twingy | *** Error code 1 |
| 05:49.16 | Twingy | src/adrt/libcommon/g.c:497: undefined reference to `db_walk_tree |
| 05:49.39 | brlcad | that looks right, cept librt is provided via ${LIBRT} |
| 05:49.57 | brlcad | dude, the error doesn't help -- look at the link line |
| 05:50.04 | Twingy | <PROTECTED> |
| 05:50.05 | Twingy | <PROTECTED> |
| 05:50.05 | Twingy | config.status: creating src/adrt/libcommon/Makefile |
| 05:50.05 | Twingy | [justin@borg:~/src/brlcad/src/adrt/libcommon]$ |
| 05:50.07 | brlcad | if it's not linking librt, somethings wrong |
| 05:50.19 | Twingy | still aint werkin |
| 05:50.32 | Twingy | librt is hosered |
| 05:50.46 | brlcad | heh riight |
| 05:51.04 | Twingy | I mean |
| 05:51.06 | brlcad | what was the command before that missing line |
| 05:51.09 | brlcad | make? |
| 05:51.12 | Twingy | I see db_walk_tree via the grace of "nm" |
| 05:51.32 | Twingy | make? |
| 05:51.42 | brlcad | 01:50 < Twingy> cd ../../.. && /usr/local/bin/bash /home/justin/src/brlcad/misc/missing --run automake-1.9 --gnu --ignore-deps |
| 05:51.42 | Twingy | what if I just commit what I got |
| 05:51.47 | Twingy | and you poke at it |
| 05:51.51 | brlcad | what was the command that did that? |
| 05:52.03 | Twingy | that was from make |
| 05:52.10 | brlcad | well then that's the problem |
| 05:52.17 | Twingy | hrm? |
| 05:52.24 | brlcad | I said that you have to make the makefile before making the target |
| 05:52.28 | brlcad | separate steps |
| 05:52.32 | Twingy | I did a make clean |
| 05:52.32 | Twingy | make |
| 05:52.48 | Twingy | or something |
| 05:52.50 | Twingy | I don't remember |
| 05:52.56 | Twingy | but I've done make clean |
| 05:52.57 | Twingy | and make |
| 05:53.01 | Twingy | several hundred thousand times now |
| 05:53.03 | Twingy | and same error |
| 05:53.07 | brlcad | so then what's the link line? |
| 05:53.16 | Twingy | gcc -D__STRICT_ANSI__ -I.. -I../../../../src/adrt/libtie -I../../../../src/adrt/libtienet -I../../../../src/adrt/libtexture -I../../../../src/adrt/libcommon -I../../../../src/adrt/librender -I../../../../src/adrt/libutil -I../../../../src/adrt/rise -pipe -fno-strict-aliasing -fno-common -fPIC -O3 -ffast-math -fstrength-reduce -fexpensive-optimizations -fomit-frame-pointer -finline-functions -g -pipe -fno-strict-aliasing -fno-commo |
| 05:53.16 | Twingy | n -fPIC -O3 -ffast-math -fstrength-reduce -fexpensive-optimizations -fomit-frame-pointer -finline-functions -g -o rise_master dispatcher.o main.o master.o post.o -L/usr/X11R6/lib -L/usr/local/lib -lpthread ../../../../src/adrt/libcommon/.libs/libcommon.a ../../../../src/adrt/librender/.libs/librender.a ../../../../src/adrt/libutil/.libs/libutil.a -lm ../../../../src/adrt/libtexture/.libs/libtexture.a ../../../../src/adrt/libtiene |
| 05:53.17 | Twingy | t/.libs/libtienet.a ../../../../src/adrt/libtie/.libs/libtie.a -lz |
| 05:53.37 | brlcad | I see no librt there |
| 05:53.43 | Twingy | me either |
| 05:53.51 | Twingy | I never added anything to rise master |
| 05:53.57 | Twingy | cause I thought libcommon was taking care of this issue |
| 05:54.05 | Twingy | <-- has always dealt with this crap in the binary |
| 05:54.07 | Twingy | not the lib |
| 05:54.16 | Twingy | I used to have librt fu in rise/master/Makefile.am |
| 05:54.18 | Twingy | when it was standalone |
| 05:54.23 | Twingy | never had to touch libcommon |
| 05:54.42 | brlcad | so what's changed? |
| 05:55.00 | brlcad | it build for me here |
| 05:55.23 | Twingy | LDADD = @BRLCAD_LDFLAGS@ -lpthread ${top_builddir}/libcommon/.libs/l......................... |
| 05:55.28 | Twingy | in adrt standalone |
| 05:55.44 | Twingy | <PROTECTED> |
| 05:55.45 | brlcad | add |
| 05:55.58 | brlcad | you're not getting that any more |
| 05:56.15 | Twingy | no kidding :) |
| 05:56.20 | Twingy | that's what I said standalone |
| 05:57.11 | brlcad | libcommon_la_*=? |
| 05:57.26 | Twingy | libcommon_la_SOURCES = canim.c cdb.c env.c pack.c unpack.c g.c |
| 05:57.27 | Twingy | libcommon_la_LDFLAGS = -static |
| 05:57.27 | Twingy | libcommon_la_LIBADD = ${LIBRT} |
| 05:57.56 | brlcad | grep LIBRT Makefile |
| 05:58.09 | Twingy | LIBRT = ${top_builddir}/src/librt/librt.la |
| 05:58.09 | Twingy | libcommon_la_LIBADD = ${LIBRT} |
| 05:59.16 | brlcad | still sounds like the missing didn't rebuild the Makefile right |
| 05:59.24 | brlcad | because it's not using the LIBADD |
| 05:59.32 | brlcad | which means you ahve to autogen.sh |
| 05:59.35 | Twingy | yep |
| 06:00.29 | Twingy | alliteration, wee |
| 06:00.32 | brlcad | i hope your aclocal and acheader links at least match your automake version |
| 06:00.53 | brlcad | er, s/autoheader// |
| 06:00.59 | brlcad | just aclocal at least |
| 06:01.24 | brlcad | that could explain the libtool mismatch |
| 06:01.27 | Twingy | autoheader is 259 |
| 06:01.40 | brlcad | aclocal, forget autoheader |
| 06:01.44 | Twingy | aclocal is 1.9 |
| 06:01.57 | brlcad | and automake was 1.9 too? |
| 06:02.12 | Twingy | lrwxr-xr-x 1 root wheel 25 Jun 17 19:26 /usr/local/bin/automake -> /usr/local/bin/automake19 |
| 06:03.36 | Twingy | ...post.o -L/usr/X11R6/lib -L/usr/local/lib -lpthread ../../../../src/adrt/libcommon/.libs/libcommon.a ../../../../src/adrt/librender/.libs/librender.a ../../../../src/adrt/libutil/.libs/libutil.a -lm ../../../../src/adrt/libtexture/.libs/libtexture.a ../../../../src/adrt/libtienet/.libs/libtienet.a ../../../../src/adrt/libtie/.libs/libtie.a -lz |
| 06:03.44 | brlcad | there is a cache option to configure that you can set to make it always use cached values -- cuts the time by a third or better |
| 06:03.57 | Twingy | ../../../../src/adrt/libcommon/.libs/libcommon.a(g.o)(.text+0x1175):/usr/home/justin/src/brlcad/src/adrt/libcommon/g.c:497: undefined reference to `db_walk_tree' |
| 06:04.07 | Twingy | would you like me to just commit this? |
| 06:04.57 | brlcad | you sure librt isn't before your .o's or something? |
| 06:05.41 | Twingy | yep |
| 06:05.57 | Twingy | should rise/master/Makefile.am |
| 06:06.02 | Twingy | be picking up librt from libcommon? |
| 06:06.39 | brlcad | no |
| 06:06.46 | Twingy | okay |
| 06:06.52 | brlcad | oh, that was a link for rise |
| 06:06.55 | Twingy | then why would we be seeing librt on the link line |
| 06:07.33 | Twingy | libcommon compiles fine |
| 06:07.35 | Twingy | rise is bitching |
| 06:07.43 | Twingy | that it doesn't have db_walk_tree |
| 06:07.53 | Twingy | via libcommon |
| 06:08.02 | brlcad | it's because of the link against the .a instead of the .la |
| 06:08.11 | Twingy | for libcommon? |
| 06:08.15 | Twingy | I'll try la |
| 06:08.18 | brlcad | for rise |
| 06:08.23 | brlcad | everywhere really |
| 06:08.39 | Twingy | so in rise Makefile.am |
| 06:08.41 | Twingy | what do I need |
| 06:08.53 | Twingy | same thing |
| 06:08.54 | Twingy | -static |
| 06:08.58 | Twingy | and LDADD |
| 06:09.09 | brlcad | to make all the .libs/lib* be lib*.la and use -static |
| 06:09.10 | Twingy | rise_master_ldflags = -static |
| 06:09.44 | brlcad | rise_master_LDFLAGS = -static |
| 06:09.49 | Twingy | yah yah |
| 06:09.56 | Twingy | I'm just tired |
| 06:10.46 | Twingy | testing |
| 06:11.15 | brlcad | libtool will probably be pissy if the libs are still .a |
| 06:11.41 | Twingy | everything in rise Makefile is .la now |
| 06:11.56 | brlcad | all the rise makefile's you mean? |
| 06:12.02 | Twingy | just master to see if it works |
| 06:12.08 | brlcad | s/rise/adrt/ |
| 06:12.10 | Twingy | if master works I wiggle my nose and everything else works |
| 06:12.42 | brlcad | that's what I mean -- libtool requires everything to become libtool libraries to work correctly |
| 06:12.59 | Twingy | I have .la's in all those directories. |
| 06:13.07 | Twingy | in all of the libraries. |
| 06:13.13 | Twingy | k? |
| 06:13.17 | brlcad | la's the properly reference the .la's they depend on? |
| 06:13.26 | brlcad | or that reference .a's |
| 06:13.43 | Twingy | none of my libs rely on other libs |
| 06:13.46 | Twingy | err |
| 06:13.50 | Twingy | do add lib crap |
| 06:13.56 | Twingy | just check out libutil |
| 06:14.00 | Twingy | it's simple |
| 06:14.29 | brlcad | libcommon at least relies on one :) |
| 06:14.43 | Twingy | "Makefile", line 595: Need an operator |
| 06:14.43 | Twingy | make: fatal errors encountered -- cannot continue |
| 06:14.43 | Twingy | *** Error code 1 |
| 06:14.43 | Twingy | Stop in /usr/home/justin/src/brlcad/src/adrt/rise. |
| 06:14.43 | Twingy | *** Error code 1 |
| 06:14.49 | Twingy | for rise/Master/Makefile |
| 06:14.59 | brlcad | sounds like you typod |
| 06:15.20 | Twingy | ah |
| 06:15.27 | Twingy | part of irc buffer got put in there |
| 06:17.16 | brlcad | wopr rox |
| 06:17.36 | Twingy | if you're interested in developing remotely |
| 06:17.52 | Twingy | I have like graphical output and stuff |
| 06:17.58 | Twingy | and need to use keyboard mouse stuff to test |
| 06:20.18 | Twingy | hrm |
| 06:20.23 | Twingy | how about I commit this and you look at it |
| 06:20.27 | Twingy | output is... different |
| 06:21.38 | brlcad | mine compiles right now, so I dunno what you're referring to |
| 06:21.58 | brlcad | so commit if you like, maybe i'll look at it on the train |
| 06:21.59 | Twingy | yours compiles cause you don't have the version where I have g.c actually being used by pack.c |
| 06:22.12 | Twingy | cause of a HAVE_BRLCAD define |
| 06:22.15 | Twingy | that existed in the standlone |
| 06:22.18 | Twingy | that now isn't defined |
| 06:22.20 | Twingy | in the new system |
| 06:22.22 | brlcad | ah |
| 06:22.22 | Twingy | which I ended up removing |
| 06:22.25 | Twingy | so I could load a .g file |
| 06:22.27 | Twingy | :) |
| 06:22.29 | brlcad | so make HAVE_BRLCAD=1 |
| 06:22.37 | Twingy | then you'd get the same error :) |
| 06:22.47 | Twingy | cause that tells it to use g.c |
| 06:22.55 | Twingy | so my removing #if HAVE_BRLCAD |
| 06:22.56 | Twingy | in 2 places |
| 06:23.00 | Twingy | so it's _IN_ brlcad |
| 06:23.04 | Twingy | that would be retarded |
| 06:23.14 | Twingy | it's now just compiling g.c period. |
| 06:23.18 | Twingy | hence |
| 06:23.20 | Twingy | your version works |
| 06:23.22 | Twingy | and mine does not |
| 06:23.24 | Twingy | :) |
| 06:23.43 | brlcad | your engrish confusith me |
| 06:23.49 | Twingy | or lack there of |
| 06:23.58 | Twingy | cause I hate explaining shit in fine detail |
| 06:24.06 | brlcad | that's okay |
| 06:24.13 | brlcad | i hate understanding shit in fine detail |
| 06:24.59 | Twingy | autoconf depresses me |
| 06:25.20 | brlcad | it's really quite simple when you stop this hacking crap you keep doing to it |
| 06:26.38 | brlcad | GBS sucks ass, but it's what everyone uses/expects and it does work when used correctly |
| 06:26.40 | Twingy | turn-key :) |
| 06:26.51 | Twingy | GBS? |
| 06:26.57 | brlcad | gnu build system |
| 06:27.03 | Twingy | beat me to it |
| 06:27.05 | brlcad | autoconf+automake+libtool |
| 06:27.31 | Twingy | makes users happier |
| 06:27.34 | Twingy | makes developers bitter |
| 06:27.45 | brlcad | it's also more cross platform |
| 06:27.55 | Twingy | hrm |
| 06:27.57 | Twingy | I need a room |
| 06:28.01 | Twingy | where I can take sheets of glass |
| 06:28.04 | Twingy | and throw them on the ground |
| 06:28.07 | brlcad | scons would probably be the only alternate worth considering |
| 06:28.11 | Twingy | that'd make me happier |
| 06:28.22 | brlcad | a fireplace |
| 06:28.26 | brlcad | wine glasses |
| 06:28.29 | Twingy | hehe |
| 06:28.31 | Twingy | no fireplace |
| 06:28.41 | brlcad | arc-weld one |
| 06:28.46 | Twingy | maybe I'll just make some rocket fuel and blow shit up :) |
| 06:29.06 | Twingy | autoconf has turned me to violence |
| 06:29.52 | brlcad | i think i have a way to fix the tesselator now |
| 06:30.14 | brlcad | and even improve it's performance and guarantee convergence |
| 06:32.09 | Twingy | s/pump/pimp |
| 06:40.24 | *** join/#brlcad ibot_ (~ibot@codepoet.org) | |
| 06:41.11 | CIA-3 | BRL-CAD: 03twingy * 10brlcad/src/adrt/ (23 files in 7 dirs): fixing linking problems to get adrt to properly link to brlcad. |
| 07:09.08 | *** join/#brlcad ibot_ (ibot@166.70.99.142) | |
| 12:27.26 | *** join/#brlcad clock- (clock@twin.jikos.cz) | |
| 12:39.22 | brlcad | clock-: i don't know if you ever saw my response the other day, but the weight e+276 value happens when it ends up using a -1 density |
| 12:40.06 | brlcad | all densities are currently initialized to -1 so when it cannot find the material you set, that can happen |
| 12:40.15 | brlcad | bad behavior, of course |
| 13:09.31 | clock- | brlcad: thanks |
| 15:53.59 | *** join/#brlcad Twingy (~justin@pcp0011647505pcs.aberdn01.md.comcast.net) | |
| 21:25.03 | *** join/#brlcad clock- (clock@twin.jikos.cz) | |
| 21:31.00 | *** join/#brlcad clock-_ (clock@twin.jikos.cz) | |
| 21:59.17 | CIA-3 | BRL-CAD: 03twingy * 10brlcad/src/adrt/ (45 files in 6 dirs): Added licensing info to libtexture files. |
| 22:18.12 | CIA-3 | BRL-CAD: 03twingy * 10brlcad/src/adrt/igvt/ (master/master.c observer/observer.c): put Observer F12 to shutdown master back in. |
| 22:26.09 | CIA-3 | BRL-CAD: 03twingy * 10brlcad/src/adrt/igvt/master/master.c: Shotline visualization is working again. |
