| 00:52.24 | CIA-13 | BRL-CAD: 03brlcad * 10brlcad/src/ (19 files in 11 dirs): more libbu memory management and header cleanup |
| 01:40.26 | CIA-13 | BRL-CAD: 03brlcad * 10brlcad/src/conv/ (5 files in 3 dirs): update converters to the API changes, avoiding the deprecated calls |
| 03:49.16 | *** join/#brlcad ChanServ (ChanServ@services.) |
| 03:49.16 | *** mode/#brlcad [+o ChanServ] by irc.freenode.net |
| 07:44.47 | *** join/#brlcad DTRemenak (n=DTRemena@DHCP-170-143.caltech.edu) |
| 08:38.10 | *** join/#brlcad PKMOBILE (n=Apathy@c-68-33-243-45.hsd1.md.comcast.net) |
| 08:39.59 | CIA-13 | BRL-CAD: 03brlcad * 10brlcad/src/ (9 files in 6 dirs): update to api changes from rt_* to bu_* for certain routines. |
| 11:30.45 | *** join/#brlcad phcoder (n=phcoder@c-68-55-177-228.hsd1.md.comcast.net) |
| 20:36.59 | *** join/#brlcad egsavage (n=egsavage@user-0c8h828.cable.mindspring.com) |
| 20:37.33 | egsavage | Hello everyone |
| 20:38.55 | egsavage | First, does anyone know of a binary distribution for ubuntu? Failing that, I'm trying to get past some 'configure' issues on my ubuntu box. Issue is the message: configure: error: C++ preprocessor "gcc -E" fails sanity check |
| 20:59.58 | brlcad | hello |
| 21:00.49 | brlcad | egsavage: if you install g++, that error should go away, alternatively you should be able to run something like: ./configure CXXCPP=cpp |
| 21:02.43 | egsavage | hmm, i have g++ installed already |
| 21:03.03 | egsavage | odd i shouldn't be fine without that other configure option |
| 21:03.03 | brlcad | hmm indeed |
| 21:03.10 | brlcad | it should |
| 21:03.19 | brlcad | can you post the config.log file? |
| 21:03.28 | brlcad | or send it to me |
| 21:03.45 | egsavage | post where? |
| 21:04.40 | egsavage | shall i send you the file via irc or some website? |
| 21:05.11 | brlcad | whatever is convenient |
| 21:05.21 | egsavage | let me try it again though - i just installed the kernel source tree package - i saw some reference somewhere saying that could fix the issue |
| 21:05.45 | brlcad | did it work with the CXXCPP? |
| 21:05.51 | egsavage | same failure |
| 21:05.55 | brlcad | k |
| 21:06.05 | egsavage | i can try with that configure option if you want me to try |
| 21:06.08 | brlcad | at the end of the config.log file, it should have the compile test that failed |
| 21:06.22 | brlcad | ahh, yeah try with that option |
| 21:06.27 | brlcad | if you would |
| 21:07.02 | egsavage | now it says "cpp" fails sanity check |
| 21:08.19 | brlcad | which shouldn't happen :) |
| 21:08.32 | brlcad | need to see what the error is |
| 21:09.27 | brlcad | you can run: less config.log, hit shift-g which takes you to the end of the file, then hit 'b' to go back a page.. keep going back until you get to an error |
| 21:10.18 | egsavage | trying to dcc the file to you - may not work because of my firewall config though |
| 21:10.50 | brlcad | got it |
| 21:11.33 | brlcad | configure:13889: gcc -E -I/usr/X11R6/include -I/usr/local/include conftest.cc |
| 21:11.34 | brlcad | gcc: installation problem, cannot exec 'cc1plus': No such file or directory |
| 21:12.03 | egsavage | hmm |
| 21:12.16 | egsavage | i noticed in the configure log it said there is no g++ too - but there is |
| 21:12.28 | egsavage | i can invoke it on the command line (g++ that is) |
| 21:12.59 | egsavage | hmm, wait, i thought I could - now it says not found... gcc worked though |
| 21:13.37 | brlcad | still "shouldn't" need it |
| 21:13.51 | brlcad | but there's some tests external to brl-cad that expect it |
| 21:13.59 | brlcad | even though the code is C |
| 21:14.48 | egsavage | aha, got past it! ;-) |
| 21:15.10 | egsavage | for whatever reason, g++-3.4 was installed - i installed g++-4.0 with synaptic and its rolling through it now |
| 21:15.51 | egsavage | configure finished ! |
| 21:15.59 | egsavage | time for the build now |
| 21:16.58 | egsavage | make is cruising now |
| 21:28.15 | brlcad | is this a production install? :) |
| 21:28.20 | brlcad | if so.. you should add --enable-optimized |
| 21:28.55 | brlcad | otherwise raytrace performance will be about half what it should be |
| 21:30.14 | egsavage | doh! |
| 21:30.29 | brlcad | it's still fully functional |
| 21:30.30 | egsavage | i guess I should recompile? do a make clean and redo? |
| 21:30.33 | brlcad | and not a big deal |
| 21:30.42 | brlcad | but .. it will run considerably faster ;) |
| 21:30.50 | egsavage | might as well 'do it right' |
| 21:30.51 | brlcad | yeah, make clean |
| 21:31.08 | egsavage | 12 mins, 48 seconds for the make |
| 21:31.10 | egsavage | not bad |
| 21:31.36 | brlcad | not too shabby |
| 21:31.37 | egsavage | add those options to the command line or the config header file somewhere? |
| 21:31.54 | brlcad | ./configure --help |
| 21:32.11 | egsavage | k |
| 21:32.27 | brlcad | the INSTALL file also talks about the build in detail |
| 21:33.11 | egsavage | any other recommended options? |
| 21:33.18 | egsavage | i see there are plenty in the configure help |
| 21:35.41 | brlcad | nah, default is to auto-detect dependencies, so you don't need to install anything |
| 21:36.04 | egsavage | and it uses opengl by default? |
| 21:36.07 | brlcad | could turn off run-time debugging for even more performance, but if this is your first go at it, I wouldn't bother |
| 21:36.13 | brlcad | by default yet |
| 21:36.28 | brlcad | though it will look the same (wireframe) regardless of it being opengl |
| 21:36.33 | egsavage | k, hopefully i have all of the opengl libs needed |
| 21:36.59 | brlcad | if you don't it'll use the X11 interface, which looks/acts the same |
| 21:37.02 | egsavage | started the make again |
| 21:37.24 | brlcad | if this is a dual/quad processor machine, you can make in parallel |
| 21:37.35 | egsavage | i was thinking about that -j option on make |
| 21:37.47 | egsavage | but i'd like some CPU for other stuff i'm doing on the box too ;-) |
| 21:37.48 | brlcad | yeah, it'll work like a charm |
| 21:38.19 | egsavage | wow! |
| 21:38.38 | egsavage | maybe next go around if i rebuild i'll dare to crank it up |
| 21:38.47 | egsavage | 3.2ghz p4 laptop here |
| 21:39.12 | brlcad | you can ctrl-c safely, and make -j# -- it will pick up safely |
| 21:39.29 | egsavage | let me try 3 |
| 21:39.43 | egsavage | it's going |
| 21:40.31 | brlcad | it's been a while since an ubuntu compile, so there might be some minor build issue |
| 21:40.34 | brlcad | hopefully not |
| 21:40.38 | brlcad | what version are you compiling? |
| 21:40.45 | egsavage | it built last time, no errors |
| 21:40.52 | egsavage | 7.6.6 |
| 21:40.56 | brlcad | ah, right, excellent |
| 21:41.12 | egsavage | latest/greatest, although not the HEAD from CVS |
| 21:41.24 | egsavage | the tar ball from sourceforge |
| 21:43.53 | egsavage | error: |
| 21:43.55 | egsavage | gcc -DHAVE_CONFIG_H -I. -I. -I../../../include -I/usr/X11R6/include -I/usr/local/include -I../../../include -pipe -fno-strict-aliasing -fno-common -g -O3 -c jove_buf.c |
| 21:43.55 | egsavage | In file included from jove_buf.c:64: |
| 21:43.55 | egsavage | ./jove.h:256: warning: built-in function ‘exp’ declared as non-function |
| 21:43.55 | egsavage | ./jove.h:477: error: conflicting types for ‘malloc’ |
| 21:44.31 | egsavage | look familiar? |
| 21:48.55 | egsavage | been awhile since I've been hacking C (Java mostly now), but it's extern char * malloc - shouldn't it be void *? |
| 21:50.41 | brlcad | ahh yes |
| 21:50.53 | brlcad | odd that you didn't get that the first time |
| 21:51.04 | brlcad | yeah, it should be |
| 21:51.14 | brlcad | that code was written well before ansi C |
| 21:51.20 | brlcad | when it used to be char * |
| 21:51.41 | brlcad | you can --disable-jove if you want, or edit that header and remove the extern decl |
| 21:53.25 | egsavage | k, let me comment that malloc extern out |
| 21:54.01 | egsavage | moving forward past there now |
| 21:54.24 | egsavage | strange others haven't seen that issue |
| 21:54.33 | egsavage | that jove stuff has been there for sometime, right? |
| 21:55.53 | brlcad | it's been there for over 15 years ;) |
| 21:56.22 | brlcad | jove isn't always compiled though -- that's the relatively recent change |
| 21:56.30 | egsavage | never used jove myself, emacs, sure |
| 21:56.36 | egsavage | ahh, ok |
| 21:56.41 | brlcad | some compilers don't care also, though new build flags make conformance more strict |
| 21:56.56 | egsavage | am i a guinea pig? ;-) |
| 21:56.59 | brlcad | jove is "jonathan's own version of emacs" |
| 21:57.10 | egsavage | gnu or xemacs for me |
| 21:57.21 | brlcad | basically a streamlined version of emacs with the 'wrong' keybindings ;) |
| 21:57.58 | brlcad | back when the invocation of emacs took 30-60 seconds or more, jove was 'fast' |
| 21:58.20 | egsavage | for quick stuff i've used vim - not emacs, but quick |
| 21:58.26 | brlcad | no longer really an issue, but it's still shipped for historic reasons -- old users demand it ;) |
| 21:58.48 | brlcad | ever try to take someone's editor away from them? :) |
| 21:59.06 | egsavage | oh, i'm sure its not pleasant |
| 21:59.48 | brlcad | i'd be fine with it myself since it's emacs-ish enough, but the bindings are foreign |
| 22:00.03 | egsavage | show how well does brl-cad work with autocad drawings? good conversions? |
| 22:00.29 | brlcad | hmm, depends entirely on what you mean by drawings ;) |
| 22:00.45 | brlcad | for strict 2D drawings, it doesn't like them much because they're not solid geometry |
| 22:01.03 | brlcad | for 3D models in dxf, it's pretty good -- better than most converters |
| 22:01.16 | brlcad | more complaint than blender, for example |
| 22:01.21 | brlcad | compliant even |
| 22:01.31 | egsavage | i see dxf and dwg (?) files that I would like to view/edit |
| 22:02.02 | brlcad | dxf may be fine, dwg are often/usually 2d-only |
| 22:02.28 | brlcad | brl-cad mostly only cares about solid geometry |
| 22:02.30 | egsavage | i'm new to cad myself ... |
| 22:02.57 | brlcad | might as well say that patches (and new devs) are always welcome too ;) |
| 22:03.37 | egsavage | ok, i'll remember that ... can you run with that jove.h fix though? |
| 22:04.07 | egsavage | run, meaning, carry that fix through the CVS tree |
| 22:05.27 | egsavage | build is still going - optimized takes a bit more compilation time |
| 22:17.06 | *** join/#brlcad tegtmeye (n=tegtmeye@pool-70-17-225-27.balt.east.verizon.net) |
| 22:17.07 | brlcad | indeed |
| 22:18.20 | egsavage | install finished... |
| 22:19.32 | egsavage | now i guess i need to do the 'hard stuff' - read the docs and figure out how to use this! ;-) |
| 22:19.34 | CIA-13 | BRL-CAD: 03brlcad * 10brlcad/src/other/jove/jove.h: fix the malloc/realloc and exp decl conflicts, include stdlib.h |
| 22:19.51 | egsavage | thx brlcad for the jove fix |
| 22:20.20 | egsavage | some other externs for other *alloc(), but when in doubt and no error, leave it I guess |
| 22:20.45 | brlcad | the ealloc is jove's |
| 22:20.51 | egsavage | ok |
| 22:21.01 | CIA-13 | BRL-CAD: 03brlcad * 10brlcad/src/other/jove/ (11 files): s/exp/jove_exp/ to avoid conflicting with exp() in math.h |
| 22:21.36 | egsavage | ah, i did see some warnings about that fly by too |
| 22:22.51 | egsavage | i'll have to go read that tutorial ... any quick test to see if my basic compile worked? |
| 22:22.57 | brlcad | yeah |
| 22:23.05 | brlcad | make test |
| 22:23.08 | brlcad | and make benchmark |
| 22:23.10 | egsavage | k |
| 22:23.31 | egsavage | as regular joe, not root, right? |
| 22:23.37 | brlcad | the latter will test system performance |
| 22:23.38 | brlcad | yeah |
| 22:23.57 | brlcad | root would only be required to do a make install into a system directory |
| 22:24.06 | egsavage | right, did that |
| 22:24.08 | egsavage | test is going |
| 22:24.17 | egsavage | sucking up that CPU ;-) |
| 22:24.23 | egsavage | it passed |
| 22:24.38 | egsavage | let me do the benchmark |
| 22:24.59 | brlcad | if you did an install, you can also run the 'benchmark' tool to get the benchmark suite to run |
| 22:25.44 | egsavage | oh, too late - doing it via make now |
| 22:25.55 | egsavage | this sends or runs forever? |
| 22:25.58 | brlcad | that's fine, they're the same |
| 22:26.02 | egsavage | ends i mean |
| 22:26.07 | brlcad | it ends |
| 22:26.22 | brlcad | it iterates until stable numbers are computed |
| 22:26.33 | brlcad | should take about 10 minutes |
| 22:26.42 | egsavage | moss said RIGHT |
| 22:27.01 | brlcad | you could have cut that down considerably by running the tool directly and decreasing the iteration window ;) |
| 22:27.12 | brlcad | but you'll get slightly better results with the default window |
| 22:27.18 | brlcad | there are 6 tests |
| 22:27.42 | egsavage | i noticed there are lots of executables... are these all integrated in the UI too? |
| 22:27.43 | brlcad | it raytraces various models testing the performance and behavior of the raytracer |
| 22:28.16 | brlcad | no they're unfortunately not all integrated, but maybe 1/5 or 1/4 are integrated into mged |
| 22:28.46 | egsavage | is that the one i should try after the benchmark? |
| 22:28.55 | brlcad | the new modeler integrates most all of them, but that's active development that probably wont hit hands until late summer at best |
| 22:29.07 | brlcad | yeah, mged is a good starting point |
| 22:29.10 | egsavage | k |
| 22:29.17 | brlcad | the converters are all external to mged |
| 22:29.31 | brlcad | so you would probably interact with them early as well, dxf-g for example |
| 22:29.54 | egsavage | ok |
| 22:30.01 | egsavage | on bldg391 now |
| 22:30.55 | pra5ad | hmm i use ubuntu |
| 22:31.11 | pra5ad | (reading log now) |
| 22:31.18 | egsavage | got it going on ubuntu now |
| 22:31.23 | egsavage | not that painful |
| 22:31.38 | egsavage | just make sure you have the latest g++ installed so configure can proceed |
| 22:32.42 | brlcad | i had a similar issue on a debian box a year or so ago, but the problem was worked around by setting CXXCPP (which configure now tries automatically iirc) |
| 22:33.28 | brlcad | ah, yes.. here it is |
| 22:33.29 | brlcad | # libtool's configuration check has a bug that causes a /lib/cpp |
| 22:33.29 | brlcad | # sanity check failure if a C++ compiler is not installed. This makes |
| 22:33.30 | brlcad | # the sanity test pass regardless of whether there is a c++ compiler. |
| 22:34.19 | brlcad | might be a new test added in another part of the autotools m4 sourcse that also provokes the same problem now |
| 22:34.30 | egsavage | benchmark results (for the debug build): |
| 22:34.31 | egsavage | Benchmark results indicate an approximate VGR performance metric of 1610 |
| 22:34.31 | egsavage | Logarithmic VGR metric is 3.21 (natural logarithm is 7.38) |
| 22:34.43 | brlcad | not bad at all |
| 22:34.48 | brlcad | especially for a P4 |
| 22:34.56 | brlcad | single cpu? |
| 22:34.58 | egsavage | oh, cool, good to hear ;-) |
| 22:35.13 | egsavage | p4 HT, but I haven't booted the SMP kernel yet |
| 22:35.22 | egsavage | just built it - need to boot it |
| 22:35.31 | brlcad | HT doesn't buy much |
| 22:35.36 | egsavage | yeah |
| 22:36.23 | egsavage | started mged |
| 22:37.43 | brlcad | for what it's worth, a VGR of 1 is the performance of an old VAX 11/780 supercomputer (circa 1977) |
| 22:38.08 | egsavage | i saw that in the log... |
| 22:38.15 | brlcad | which is one of the oldest machines to run brl-cad |
| 22:38.17 | egsavage | what should i try for a quick test in mged? |
| 22:38.32 | brlcad | hmm, a quick test.. |
| 22:38.42 | brlcad | did you create a new database? |
| 22:38.45 | egsavage | something to load up |
| 22:38.50 | brlcad | ah, okay |
| 22:39.06 | brlcad | there are simple example geometry databases that get installed |
| 22:39.18 | brlcad | if you installed into /usr/brlcad, they are in share/brlcad/7.6.6/db |
| 22:39.34 | brlcad | open havoc.g |
| 22:39.55 | brlcad | then 'e havoc' in the command window |
| 22:40.09 | brlcad | maybe then: rt -F/dev/Xl |
| 22:40.48 | egsavage | a nice heli appeared ;-) |
| 22:41.14 | brlcad | russian attack helicopter, relatively simple model |
| 22:41.29 | brlcad | but more detailed than most of the example .g files |
| 22:41.46 | egsavage | nice solid view too |
| 22:42.08 | egsavage | can i zoom in that rt? |
| 22:42.31 | brlcad | right/left click to zoom in/out |
| 22:43.23 | egsavage | zoom works on the wireframe, not solid |
| 22:43.26 | brlcad | there are various shift-option-control clicks as well for various options |
| 22:43.45 | brlcad | ah, you mean the window that popped up on the rt |
| 22:43.52 | brlcad | that's a raster image, raytraced |
| 22:43.57 | egsavage | ok |
| 22:44.44 | brlcad | there's means to zoom them in/out but not as a /dev/Xl window |
| 22:45.03 | egsavage | thanks for your help! seems to be working well... Time to read/learn brlcad in the tutorial... |
| 22:45.17 | egsavage | My wife just said dinner is ready so I'll be back later - thx |
| 22:45.22 | brlcad | no problem! |
| 22:45.28 | brlcad | if you have any questions, someone's usually here ;) |
| 22:45.34 | egsavage | thanks again! |
| 22:45.34 | brlcad | or will eventually answer |
| 22:59.50 | pra5ad | i had no probs compiling from head on breezy |
| 22:59.57 | pra5ad | *shrug* |