| 16:20.39 | *** join/#brlcad ibot (~ibot@rikers.org) | |
| 16:20.39 | *** topic/#brlcad is BRL-CAD Open Source Solid Modeling || http://brlcad.org || logs: http://ibot.rikers.org/%23brlcad/ || BRL-CAD release 7.22.0 is forthcoming || fixing all our Coverity Scan Initiative defects | |
| 16:21.23 | ``Erik | your compiler isn't liking the __extension__ bit, which version of gcc is that? |
| 16:21.59 | jordisayol | gcc version 4.5.1 |
| 16:22.19 | ``Erik | and your libc headers are the most recent? |
| 16:22.45 | jordisayol | I think so |
| 16:22.58 | jordisayol | I've system up to date |
| 16:25.24 | jordisayol | glibc 2.11.3-12.21.1 |
| 16:27.29 | starseeker | ``Erik: awesome |
| 16:27.46 | starseeker | hope they produce it |
| 16:28.47 | ``Erik | lacks much in the way of linux machines, is going to try compiling with a new gcc on fbsd to see if its' a change in gcc O.o |
| 16:29.50 | ``Erik | but this is looking like an opensuse issue to me, your gcc doesn't like your system headers :/ any program that uses FD_SET() would be throwing this issue (might be some flags we use triggering it) |
| 16:32.10 | jordisayol | ok. I've compiled all previous rpm packages on this system so, is new the use of this function on brlcad? |
| 16:33.11 | jordisayol | if not, then is a new bug on opensuse glibc, isn't it? |
| 16:33.49 | ``Erik | hm, it's an old use of that macro in BRL-CAD, only a couple changes in that file in the last few months |
| 16:34.11 | ``Erik | one on dec 15, the other on nov 29 |
| 16:34.29 | ``Erik | built fine using gcc 4.7 on fbsd 8.2, ia32 |
| 16:35.05 | jordisayol | no no, it build fine on same opensuse 64-bit!!! |
| 16:35.49 | ``Erik | yeah, it's an assembly macro, so it's platform specific... |
| 16:36.13 | ``Erik | it's coming from /usr/include/asm-*/posix_types.h |
| 16:36.35 | ``Erik | so on 64, you get asm-x86_64, on 32 you get asm-i386... different files |
| 16:37.09 | jordisayol | can i do something to override it? |
| 16:37.29 | ``Erik | not to my knowledge, but I'm not a linux guy and have never logged into an opensuse box |
| 16:37.52 | jordisayol | or better, when will be released the next brlcad version? |
| 16:39.03 | ``Erik | sorry, dude, I'm out of ideas, maybe the opensuse guys can help? the __extension__ on line 53 of your paste is the bit causing the issue, and that's from /usr/include/asm-i386/posix_types.h *shrug* |
| 16:39.56 | ``Erik | next BRL-CAD release should be fairly soon, but I don't think it'll fix stuff or ya :/ |
| 16:40.10 | ``Erik | s/ or/ for/ |
| 16:41.03 | jordisayol | I ask you just to know if there is enough time to opensuse crew correct this |
| 16:41.19 | starseeker | how much time would they need? |
| 16:41.29 | jordisayol | no idea |
| 16:41.35 | starseeker | we won't be putting a release out in the next few days - there's too much to do |
| 16:42.07 | jordisayol | I'm not an opensuse user. I just use it in a virtualbox machine to build brlcad rpm packages |
| 16:42.39 | starseeker | my advice would be to contact them and see what they say |
| 16:43.06 | *** join/#brlcad packrat (~packrator@c-98-209-146-133.hsd1.mi.comcast.net) | |
| 16:43.56 | jordisayol | ok, i'll try, many thanks! |
| 16:44.22 | ``Erik | they may need to see your compile line, it might be something funky like -msse3 or something causing it to trigger |
| 17:08.32 | CIA-48 | BRL-CAD: 03brlcad * r49068 10/brlcad/trunk/src/ (13 files in 3 dirs): (log message trimmed) |
| 17:08.32 | CIA-48 | BRL-CAD: Can't take it any longer. Doggy bag all the gcc vomit about SCL's qualified type errors. Reduces the build log by about 16MB, 78k warning lines, approx 765 unique instances. |
| 17:08.32 | CIA-48 | BRL-CAD: Quell them all with a teeny little one-liner bit of scripting based on the build log output: |
| 17:08.32 | CIA-48 | BRL-CAD: for match in `grep qualifiers build.log | sort | uniq | cut -d: -f1,2` ; do |
| 17:08.33 | CIA-48 | BRL-CAD: export file="`echo $match | cut -d: -f1`" ; export line="`echo $match | cut -d: |
| 17:08.33 | CIA-48 | BRL-CAD: -f2`" ; export pline="`expr $line - 1`" ; if test "x`sed -n ${pline}p |
| 17:08.34 | CIA-48 | BRL-CAD: $file |grep const`" = "x" ; then sed -n ${line}p $file ; sed |
| 17:08.36 | *** join/#brlcad merzo (~merzo@248-93-200-46.pool.ukrtel.net) | |
| 17:10.33 | CIA-48 | BRL-CAD: 03brlcad * r49069 10/brlcad/trunk/src/other/step/src/clstepcore/ExpDict.h: |
| 17:10.33 | CIA-48 | BRL-CAD: more warning quellage. shouldn't pass string constants to char* parameters as |
| 17:10.33 | CIA-48 | BRL-CAD: they should be considered immutable/const. make SelectTypeDescriptor() |
| 17:10.33 | CIA-48 | BRL-CAD: constructor take a const char* instead, fortunately TypeDescriptor's constructor |
| 17:10.33 | CIA-48 | BRL-CAD: is correct. |
| 17:12.12 | CIA-48 | BRL-CAD: 03bob1961 * r49070 10/brlcad/trunk/src/mged/polyif.c: Variable declarations need to be at the top of the code block. |
| 17:22.41 | CIA-48 | BRL-CAD: 03brlcad * r49071 10/brlcad/trunk/src/other/step/src/clstepcore/ (8 files): |
| 17:22.41 | CIA-48 | BRL-CAD: might make merging a bit more difficult or (more likely) repeats work already |
| 17:22.41 | CIA-48 | BRL-CAD: done in the github tree, but helps ease our sanity in the meantime. eliminate a |
| 17:22.41 | CIA-48 | BRL-CAD: slew of gcc warnings about string literals being passed as char* parameters when |
| 17:22.41 | CIA-48 | BRL-CAD: they should be treated immutable. mark them all const. |
| 17:23.02 | brlcad | that takes care of the majority of the step vomit |
| 17:23.20 | brlcad | the remaining warnings look like they're other valid step-g issues |
| 17:24.58 | brlcad | now the warning log has but 520 or so lines in it (instead of 78k lines) |
| 17:26.06 | brlcad | jordisayol: what does man FD_SET list for headers? |
| 17:27.21 | jordisayol | brlcad: sorry, I don't understand you |
| 17:27.45 | brlcad | if you run "man FD_SET" it should give a usage man page |
| 17:27.59 | brlcad | it should list one or more #include headers near the top |
| 17:28.00 | brlcad | like |
| 17:28.01 | brlcad | #include <sys/select.h> |
| 17:28.03 | jordisayol | nop, no man pacge for it |
| 17:28.11 | brlcad | hm, okay, what about man select |
| 17:28.38 | jordisayol | not too |
| 17:29.01 | brlcad | no man page for select? |
| 17:29.25 | brlcad | sounds like you're missing manual pages |
| 17:29.40 | jordisayol | just a moment, i just use opensuse to build brlcad |
| 17:29.51 | jordisayol | maybe it's not installed |
| 17:30.17 | brlcad | no matter, found it online http://manpages.unixforum.co.uk/man-pages/linux/opensuse-10.2/0p/select.h-man-page.html |
| 17:31.29 | jordisayol | now i got man FD_SET |
| 17:31.56 | brlcad | grep HAVE_SYS_SELECT_H include/* |
| 17:32.03 | brlcad | from your build directory |
| 17:32.56 | jordisayol | include/brlcad_config.h:#define HAVE_SYS_SELECT_H 1 |
| 17:32.56 | jordisayol | include/brlcad_config.h.in:#cmakedefine HAVE_SYS_SELECT_H 1 |
| 17:32.56 | jordisayol | include/bselect.h:#ifdef HAVE_SYS_SELECT_H |
| 17:33.07 | brlcad | hm, okay that's good |
| 17:34.41 | brlcad | jordisayol: try this .. edit include/bselect.h and add this line before the #ifdef HAVE_SYS_SELECT_H line |
| 17:35.06 | brlcad | #include <sys/types.h> |
| 17:35.13 | brlcad | #include <signal.h> |
| 17:35.17 | brlcad | #include <time.h> |
| 17:35.26 | brlcad | those three lines |
| 17:35.40 | brlcad | see if that makes any difference at all |
| 17:35.51 | jordisayol | ok |
| 17:38.17 | *** join/#brlcad Stattrav (~Stattrav@61.12.114.82) | |
| 17:38.17 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) | |
| 17:38.37 | jordisayol | brlcad: it appears that pass the error point |
| 17:38.58 | jordisayol | still compiling... |
| 17:39.56 | jordisayol | it will take a wile (compiling in a virtualbox machine with just one cpu) |
| 17:40.13 | brlcad | wow |
| 17:40.26 | brlcad | yeah, that's a busted system header then |
| 17:40.43 | brlcad | remove the signal.h and time.h lines and retry |
| 17:40.56 | jordisayol | ok |
| 17:41.24 | jordisayol | ohhh, sorry, the error still there :-/ |
| 17:41.25 | brlcad | then assuming that works, remove the sys/types.h line just to verify that's what makes it work |
| 17:41.31 | brlcad | hm, okay |
| 17:46.03 | brlcad | ``Erik: nifty -E -o trick, handy thx |
| 17:50.18 | brlcad | jordisayol: you could try a newer version of gcc, one of the 4.6 line |
| 17:51.51 | jordisayol | I'll try, I'm not comfortable on opensuse... |
| 17:55.30 | *** join/#brlcad juan_man (~quassel@186.136.169.88) | |
| 17:55.58 | *** join/#brlcad juan_man (~quassel@unaffiliated/juanman) | |
| 17:56.52 | jordisayol | brlcad: anyway, on opensuse 11.4 64-bit with gcc 4.5.1 , properly compiles |
| 17:57.20 | CIA-48 | BRL-CAD: 03starseeker * r49072 10/brlcad/trunk/src/other/libpng/CMakeLists.txt: If we're conditionalizing the non-WIN32 files on generation, need to also conditionalize the related install rules. |
| 18:00.17 | CIA-48 | BRL-CAD: 03starseeker * r49073 10/brlcad/trunk/src/other/libpng/CMakeLists.txt: ws |
| 18:20.29 | CIA-48 | BRL-CAD: 03n_reed * r49074 10/brlcad/trunk/src/libbu/test_sscanf.c: Using macros to condense numeric conversion tests, now in their own routine. Moving on to more exotic conversions. |
| 18:22.39 | jordisayol | brlcad: I've upgraded to gcc 4.6.2 (from opensuse 12.1) and it appears to work |
| 18:22.46 | jordisayol | still compiling |
| 18:32.39 | jordisayol | ok, successfully compiled |
| 18:43.32 | CIA-48 | BRL-CAD: 03starseeker * r49075 10/brlcad/trunk/src/nirt/command.c: Send the default coordinate printouts to stdout - really should be handled like 'normal' nirt reporting... |
| 18:44.54 | brlcad | jordisayol: awesome |
| 18:45.17 | brlcad | did you see the forum message about the opensuse rpm not working? |
| 18:45.45 | jordisayol | nop :-( |
| 18:45.57 | brlcad | https://sourceforge.net/projects/brlcad/forums/forum/362510/topic/4972540 |
| 18:46.11 | brlcad | thought that might be why you were looking into opensuse :) |
| 18:49.09 | CIA-48 | BRL-CAD: 03starseeker * r49076 10/brlcad/trunk/src/nirt/command.c: another stdout fprintf change... |
| 18:49.17 | jordisayol | no no :-) just test to create deb/rpm packages randomly |
| 18:57.59 | brlcad | any idea what that error might be? |
| 18:58.25 | jordisayol | not really |
| 18:59.05 | brlcad | I know ogl_interface is a symbol in our libfb library, only compiled if you enable opengl |
| 18:59.27 | brlcad | is the rpm he's using supposed to have opengl on or off? |
| 18:59.46 | jordisayol | on |
| 19:00.15 | brlcad | so then the fact that there is a reference to ogl_interface somewhere is correct |
| 19:00.25 | brlcad | why it can't find it becomes the issue |
| 19:00.47 | brlcad | maybe he has another library on his system named libfb |
| 19:00.49 | jordisayol | so probably is a problem with his opengl. do you remember that some other person got problems on opensuse too? |
| 19:01.27 | brlcad | nope |
| 19:02.17 | jordisayol | well, this guy got problems running mged/archer |
| 19:02.46 | jordisayol | and finally he wasn't able to run glxgears on his system |
| 19:03.05 | brlcad | right, so that one sounds like an opengl issue |
| 19:03.16 | brlcad | not finding ogl_interface isn't, though |
| 19:04.28 | jordisayol | missing dependencies on rpm? |
| 19:05.02 | brlcad | you don't specify object files in the rpm, so not likely |
| 19:05.19 | brlcad | that's just a specific symbol inside libfb that is referenced by libdm and mged |
| 19:06.47 | brlcad | what I'm betting is happening is that it's loading mged, begins loading libs, loads libfb but finds *another* system libfb instead of ours, loads libdm, see's the ogl_interface symbol, gets through loading all the of rest of the libs and has no way to resolve it |
| 19:08.05 | jordisayol | aha |
| 19:08.39 | brlcad | maybe even another compile of BRL-CAD |
| 19:09.20 | jordisayol | aha |
| 19:11.54 | *** join/#brlcad Stattrav (~Stattrav@61.12.114.82) | |
| 19:11.54 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) | |
| 19:17.12 | jordisayol | brlcad: if there is a /usr/lib/libdm.so and a /usr/brlcad/lib/libdm.so, which one is used by mged? |
| 19:21.41 | ``Erik | 'ldd' will tell you. "ldd /path/to/mged" |
| 19:27.31 | jordisayol | I've installed libdm0 (installs /liblibdm.so.0.0.4) and ldd /usr/brlcad/mged returns: |
| 19:27.31 | jordisayol | ... |
| 19:27.31 | jordisayol | libdm.so.19 => /usr/brlcad/lib/libdm.so.19 (0x00007f38e9ac9000) |
| 19:27.31 | jordisayol | ... |
| 19:29.45 | jordisayol | s:/liblibdm.so.0.0.4:/lib/libdm.so.0.0.4: |
| 19:35.55 | brlcad | binaries are usually compiled with linker information pointing to where the library is expected to be |
| 19:36.44 | brlcad | so if you don't install where it was compiled, LD_LIBRARY_PATH becomes necessary or modifications to ld.config |
| 19:37.08 | brlcad | if you do install where it was compiled for, you don't generally need anything |
| 19:38.24 | jordisayol | rpm package always installs libs at /usr/brlcad/lib |
| 19:39.02 | brlcad | so then it's back to him having a libfb just somewhere else on his system and LD_LIBRARY_PATH is pointing to it |
| 19:42.49 | jordisayol | I've to go |
| 19:42.53 | ``Erik | รด/cl |
| 19:45.34 | brlcad | cya jordisayol |
| 19:45.37 | brlcad | thanks! |
| 19:47.25 | *** join/#brlcad pawleeq (~pawleeq@core1.humlnet.cz) | |
| 19:47.30 | pawleeq | hello |
| 19:47.55 | pawleeq | here is another czech article about brlcad: http://www.abclinuxu.cz/clanky/brl-cad-pokrocile-techniky-modelovani |
| 19:52.45 | brlcad | pawleeq: that is awesome! |
| 19:52.58 | brlcad | and you even figured out how to work that terrible pattern tool gui :D |
| 19:53.24 | pawleeq | yeah and it made me to think about few improvements ... |
| 19:53.33 | pawleeq | i will write a hot report soon a let you know |
| 19:53.41 | brlcad | the "clone" command is a lot better, at least easier to use |
| 19:53.42 | pawleeq | * a > and |
| 19:53.51 | brlcad | though it doesn't do the spherical/cylindrical patterns |
| 19:53.58 | brlcad | cool |
| 19:54.19 | brlcad | it'd be great to have those tutorials in docbook format :) |
| 19:54.38 | pawleeq | one of the readers asked specially about cylindrical patterns so I did it |
| 19:54.40 | brlcad | then we could get some english translations going and get them up into our docs |
| 19:55.36 | pawleeq | I can translate it, i even think about real book, but these articles are fortunately keeping my feet on the ground :) |
| 20:18.33 | brlcad | pawleeq: do you know docbook/xml? or html? |
| 20:19.36 | pawleeq | I know html |
| 20:20.49 | brlcad | docbook/xml is very very similar -- if you're interested, I could pull together a template example |
| 20:21.09 | brlcad | from docbook, we can autogenerate html, pdf, rtf, etc |
| 20:21.52 | brlcad | when was your first article published? |
| 20:22.53 | pawleeq | November 11th |
| 20:23.32 | pawleeq | ok send me a template and I will fill it with english translation of those articles (or just the articles you want) |
| 20:24.07 | brlcad | both english and czech would be best :) |
| 20:24.14 | pawleeq | ok |
| 20:24.18 | brlcad | we can track translations |
| 20:24.47 | brlcad | that way, we might get others willing to help translate your articles to other languages as well |
| 20:25.44 | pawleeq | ok :) |
| 20:30.47 | CIA-48 | BRL-CAD: 03brlcad * r49077 10/brlcad/trunk/AUTHORS: credit pavel jansa (pawleeq) with his numerous brl-cad tutorial articles (in czech): http://www.abclinuxu.cz/autori/pavel-jansa |
| 20:32.47 | CIA-48 | BRL-CAD: 03brlcad * r49078 10/brlcad/trunk/src/libbn/noise.c: |
| 20:32.47 | CIA-48 | BRL-CAD: clean up the magic number handling to be consistent with our uint32_t magic |
| 20:32.47 | CIA-48 | BRL-CAD: type. expand the size of the hash table from shorts to that as well and do some |
| 20:32.47 | CIA-48 | BRL-CAD: minor refactoring so that the table size (4096) is defined in just one place and |
| 20:32.47 | CIA-48 | BRL-CAD: everything keys off of it. |
| 20:37.06 | CIA-48 | BRL-CAD: 03brlcad * r49079 10/brlcad/trunk/src/libbu/badmagic.c: |
| 20:37.06 | CIA-48 | BRL-CAD: not clear what is exactly being checked but the magic number alignment checking |
| 20:37.06 | CIA-48 | BRL-CAD: looks to be incorrect now that we're using uint32_t's everywhere. getting |
| 20:37.06 | CIA-48 | BRL-CAD: misalignment on big-endian 64-bit system in shaders regression, fixed by |
| 20:37.06 | CIA-48 | BRL-CAD: checking our magic type. |
| 20:42.04 | *** join/#brlcad Stattrav (~Stattrav@61.12.114.82) | |
| 20:42.05 | *** join/#brlcad Stattrav (~Stattrav@unaffiliated/stattrav) | |
| 20:52.17 | CIA-48 | BRL-CAD: 03n_reed * r49080 10/brlcad/trunk/src/libbu/sscanf.c: Forgot that we need to search for closing brace of character classes. Was causing truncated conversion specification. |
| 21:01.04 | CIA-48 | BRL-CAD: 03starseeker * r49081 10/brlcad/trunk/src/nirt/sfiles/csv-gap.nrt: Stray commas in csv-gap formatting were causing error when specifying -f csv-gap in a nirt run. |
| 21:48.10 | CIA-48 | BRL-CAD: 03n_reed * r49082 10/brlcad/trunk/src/libbu/sscanf.c: Doh! Accidently overwrote conversion type var before switching on it. |
| 21:52.38 | CIA-48 | BRL-CAD: 03starseeker * r49083 10/brlcad/trunk/ (NEWS src/nirt/if.c): Attribute printing was haywire in nirt, printing ALL attributes and ignoring the list. Now printing only specified attributes. |
| 21:56.44 | n_reed | hah, accidentally used accidently |
| 22:10.34 | CIA-48 | BRL-CAD: 03starseeker * r49084 10/brlcad/trunk/src/nirt/command.c: err -> stdout for nirt default units reporting. |
| 22:14.37 | CIA-48 | BRL-CAD: 03n_reed * r49085 10/brlcad/trunk/src/libbu/test_sscanf.c: Add some string-type tests now that character-class conversion is debugged. |
| 22:16.01 | CIA-48 | BRL-CAD: 03starseeker * r49086 10/brlcad/trunk/ (NEWS src/nirt/nirt.c src/nirt/nirt.h): Add -h option to nirt to suppress header, couple other tweaks. |
| 22:21.31 | CIA-48 | BRL-CAD: 03starseeker * r49087 10/brlcad/trunk/ (TODO regress/CMakeLists.txt regress/nirt.sh): Add regression testing for nirt based on the nirt DocBook manual. |
| 22:34.57 | CIA-48 | BRL-CAD: 03brlcad * r49088 10/brlcad/trunk/doc/docbook/articles/en/ (CMakeLists.txt TEMPLATE.xml): add an initial template that demonstrates some of the basic docbook/xml elements for adding new articles. |
| 22:57.01 | *** join/#brlcad pawleeq (~pawleeq@core1.humlnet.cz) | |
| 23:18.25 | brlcad | pawleeq: http://brlcad.svn.sourceforge.net/viewvc/brlcad/brlcad/trunk/doc/docbook/articles/en/TEMPLATE.xml?view=log |
| 23:18.42 | brlcad | there's a template you can use, it'll generate something like this: http://brlcad.org/~sean/tmp/TEMPLATE.html |
| 23:19.14 | pawleeq | ok thank you |
| 23:19.34 | brlcad | it's in svn, so you can copy that file, update the doc/docbook/articles/en/CMakeLists.txt file with your article(s) |
| 23:19.49 | brlcad | and when you run "make", it'll generate the html |
| 23:20.14 | brlcad | don't worry about how the html looks exactly, it's more that it's structurally all there described and easy enough to read |
| 23:20.23 | brlcad | if that wasn' |
| 23:20.36 | brlcad | if that wasn't clear, folks in here can easily help ;) |
| 23:24.40 | pawleeq | ok I will ask if I get unsure about something, and there always is google :) I am going to bed, so good bye everybody |
| 23:55.32 | *** join/#brlcad starseeker (~starseeke@BZ.BZFLAG.BZ) | |
| 23:57.03 | *** join/#brlcad piksi (piksi@pi-xi.net) | |
| 23:58.41 | *** join/#brlcad ChanServ (ChanServ@services.) | |
| 23:58.41 | *** mode/#brlcad [+o ChanServ] by pratchett.freenode.net | |