| 00:09.53 | *** join/#brlcad jhujhiti (n=jhujhiti@64.252.100.253) | |
| 00:10.36 | jhujhiti | ../../src/libdm/.libs/libdm.so: undefined reference to `XFreeDeviceList' on debian testing. i cannot, for the life of me, figure out which package i'm missing |
| 00:14.51 | jhujhiti | with 7.10.4, by the way |
| 00:35.03 | ``Erik | um, it probably didn't link libXi.so ? |
| 00:35.11 | ``Erik | did you run autoreconf or autogen or anything? |
| 00:36.22 | jhujhiti | ah ha! libxi-dev wasn't installed. i'll try that |
| 00:36.39 | ``Erik | those'd just be headers... |
| 00:36.43 | jhujhiti | ``Erik: i didn't. i'm quite sure i just didn't install all of the -dev packages and the configure script didn't pick up on it |
| 00:38.43 | ``Erik | and now; conrad twitty. |
| 00:39.34 | jhujhiti | haha |
| 00:40.18 | ``Erik | anyways, that's a lib error, not a header error, so having libXi.so should be all you need... however; debian "fixes" libtool, which horribly breaks any nontrivial use of dependancy libs, and libdm makes a non-trivial use of the dependancy_libs variable in the libdm.la file |
| 00:40.46 | ``Erik | if you're working with stuff striaght out of the tarball and didn't touch any of the automake files, it SHOULD work |
| 00:41.01 | ``Erik | if you have libXi.so handy :) |
| 00:45.06 | jhujhiti | i have /usr/lib/libXi.so{,.6,.6.0.0} |
| 00:45.51 | jhujhiti | it's running its own libtool. |
| 00:46.01 | ``Erik | in theory, ldd /path/to/libdm.so SHOULD list one of those |
| 00:46.31 | jhujhiti | err, libdm isn't building? |
| 00:46.57 | jhujhiti | whoops. disregard that. |
| 00:47.55 | jhujhiti | it doesn't. |
| 00:48.05 | jhujhiti | it doesn't list *any* missing libraries. and no mention of libXi |
| 00:48.40 | ``Erik | weird |
| 00:48.59 | ``Erik | and in the build directory, what's the dependancies line in src/libdm/libdm.la look like? |
| 00:49.51 | jhujhiti | heh, probably too long to paste, eh? |
| 00:49.56 | jhujhiti | although there IS a newline in it |
| 00:50.06 | jhujhiti | dependency_libs=' -L/usr/local/lib /home/jhujhiti/build/brlcad-7.10.4/src/librt/librt.la -L/home/jhujhiti/build/brlcad-7.10.4/src/other/tcl/unix /home/jhujhiti/build/brlcad-7.10.4/src/libbn/libbn.la /home/jhujhiti/build/brlcad-7.10.4/src/other/libregex/libregex.la /home/jhujhiti/build/brlcad-7.10.4/src/libsysv/libsysv.la /home/jhujhiti/build/brlcad-7.10.4/src/other/openNURBS/libopenNURBS.la -lstdc++ /home/jhujhiti/build/brlcad-7.10.4/src/libfb/libfb.la |
| 00:50.45 | ``Erik | did it truncate? the last thing I saw was libfb.la |
| 00:51.00 | jhujhiti | then i guess the answer is yes. let me pastebin it |
| 00:51.07 | ``Erik | there *SHOULD* be an -lXi on it |
| 00:51.45 | ``Erik | mine looks like http://pastebin.bzflag.bz/d22fad3d6 (on a mac) |
| 00:52.10 | jhujhiti | http://qna.nu/p18vc/?raw |
| 00:52.29 | jhujhiti | there's no Xi in it |
| 00:53.23 | ``Erik | odd |
| 00:53.36 | jhujhiti | let me blow away the build directory and start from scratch.. |
| 00:54.18 | ``Erik | so if you grep ^X_LIBS Makefile |
| 00:54.21 | ``Erik | there's no Xi in that? |
| 00:54.26 | ``Erik | X_LIBS = -L/usr/X11R6/lib -lX11 -lX11 -lXext -lXi |
| 00:54.37 | jhujhiti | heh, too late =) |
| 00:54.48 | jhujhiti | is there a signature of the tarball somewhere? |
| 00:54.55 | jhujhiti | or an md5sum? sourceforge never makes it easy |
| 00:55.51 | ``Erik | I wonder if you'd need to do something like sudo ln -s /usr/lib/libXi.so.6 /usr/lib/libXi.so ? |
| 00:56.00 | ``Erik | http://www.freebsd.org/cgi/cvsweb.cgi/ports/cad/brlcad/distinfo?rev=1.15 has md5 and sha256 |
| 00:56.15 | jhujhiti | /usr/lib/libXi.so.6.0.0 |
| 00:56.15 | jhujhiti | /usr/lib/libXi.so |
| 00:56.16 | jhujhiti | /usr/lib/libXi.so.6 |
| 00:56.26 | jhujhiti | from find |
| 00:56.54 | ``Erik | um, does libXi.so actually resolve to the libXi.so.6.0.0 file? |
| 00:57.04 | jhujhiti | how would i check that? |
| 00:57.08 | ``Erik | could be a bad symlink, left over? *shrug* |
| 00:57.11 | ``Erik | um, ls -l it? |
| 00:57.12 | ``Erik | :) |
| 00:57.24 | ``Erik | or file, or if you're feeling brave, cat it... er... I mean, run file on it :D |
| 00:57.38 | jhujhiti | /usr/lib/libXi.so -> libXi.so.6.0.0 |
| 00:57.49 | ``Erik | weird |
| 00:58.04 | jhujhiti | configure just finished. which makefile should i grep on? |
| 00:58.21 | ``Erik | if 'grep ^X_LIBS Makefile' doesn't have -lXi in it, would you be able to post your config.log somewhere? |
| 00:58.25 | ``Erik | any of them shoudl work |
| 00:58.35 | *** join/#brlcad iraytrace (n=iraytrac@c-67-172-239-24.hsd1.ut.comcast.net) | |
| 00:58.38 | jhujhiti | X_LIBS = -lX11 -lX11 -lXext -lXi |
| 00:59.31 | jhujhiti | err, should the configure step be creating libdm.la? |
| 00:59.52 | ``Erik | no, that's during the compile |
| 01:00.05 | ``Erik | what about 'grep ^DM_LIBS Makefile' ? |
| 01:00.08 | jhujhiti | all right. well then.. is this thing make -j safe? |
| 01:00.18 | ``Erik | i do it all the time on 4 and 8 core machines, it SHOULD work fine |
| 01:00.36 | jhujhiti | i haven't been using it because i wasn't sure. but it takes a bit before it bails out. i'll try make again |
| 01:00.50 | jhujhiti | -lXi is at the very end of that grep. |
| 01:00.53 | ``Erik | ok |
| 01:00.55 | ``Erik | um |
| 01:00.56 | ``Erik | heh |
| 01:01.09 | ``Erik | you can, uh, go into src/libdm and do "make depends" |
| 01:01.21 | ``Erik | to bypass a fair bit of stuff ya don't need for libdm proper |
| 01:01.22 | ``Erik | :) |
| 01:01.41 | jhujhiti | let's see if a regular make bails first.. |
| 01:01.41 | ``Erik | (provided I did it right) |
| 01:01.45 | ``Erik | aight |
| 01:02.25 | jhujhiti | do you know off the top of your head what file i could look for to see if the build is past that point? |
| 01:03.06 | ``Erik | uhmmmmmm well, if src/mged/mged exists, then libdm is linking ok |
| 01:05.17 | jhujhiti | neat, -j2 is generating over 3.5 load usually it's just over 2 |
| 01:06.18 | jhujhiti | -rwxr-xr-x 1 jhujhiti 1000 9.0K 2007-12-02 20:06 src/mged/mged* |
| 01:06.29 | jhujhiti | *sigh* that was the third try. what the hell? |
| 01:07.31 | ``Erik | heh |
| 01:07.48 | ``Erik | like everything technological that doesn't work, it always works when someone else is looking :D |
| 01:07.58 | jhujhiti | i did upgrade from etch to testing between that second one, but i was getting the same error as with stable |
| 01:08.05 | jhujhiti | heh, indeed |
| 01:08.13 | jhujhiti | well, thanks for the help ( i guess ) |
| 01:08.24 | ``Erik | np, I guess O.o :D |
| 01:08.38 | jhujhiti | looks like all that heartache over moving away from stable was unnecessary. |
| 01:08.48 | jhujhiti | but i'm not installing debian again to go back. |
| 03:00.06 | *** join/#brlcad iraytrace (n=iraytrac@c-67-172-239-24.hsd1.co.comcast.net) | |
| 03:06.23 | *** join/#brlcad iraytrace (n=iraytrac@c-67-172-239-24.hsd1.co.comcast.net) | |
| 03:12.30 | *** join/#brlcad yukonbob (n=yukonbob@198.235.198.234) | |
| 05:53.28 | *** join/#brlcad dtidrow (n=dtidrow@c-69-255-182-248.hsd1.va.comcast.net) | |
| 06:32.35 | Axman6 | brlcad: yt? |
| 07:09.31 | *** join/#brlcad Z80-Boy (n=clock@77-56-93-5.dclient.hispeed.ch) | |
| 07:12.26 | *** join/#brlcad PrezKennedy (i=Matt@74.86.45.130) | |
| 07:23.25 | PrezKennedy | brlcad, have you seen http://wiki.vorratsdatenspeicherung.de/IRSeeK-en? |
| 07:30.00 | Axman6 | i'm hoping freenode find a way to ban them |
| 07:31.34 | PrezKennedy | i think they stopped em for now |
| 07:31.34 | Axman6 | also, anyone know a good tool for viewing or onverting .pix files on OS X? |
| 07:33.46 | PrezKennedy | this is what i posted on bram cohen's blog for his irseek "utility" |
| 07:34.03 | PrezKennedy | Hi, I'm here to log every website you operate. Since they're available publicly on the Internet, I'll be taking all of your content and making it searchable from my own site. Also, I'll be slapping up ads and trying to make a buck from it. If you don't like that you can opt-out of course. |
| 07:34.33 | *** join/#brlcad Defcon (n=def@74.17-246-81.adsl-static.isp.belgacom.be) | |
| 07:35.00 | PrezKennedy | *eran cohen... hmm getting too late to read |
| 07:38.49 | yukonbob | Axman6: you have BRLCAD installed? |
| 07:39.05 | yukonbob | if so, see pix-png(1) et al |
| 07:39.50 | Defcon | bw-png? |
| 07:48.57 | Axman6 | ah, thanks :) |
| 08:06.51 | Axman6 | and.. how do i raytrace to a .pix file? |
| 08:15.06 | Axman6 | ah got it |
| 08:59.40 | *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6) | |
| 09:14.41 | *** join/#brlcad d_rossberg (n=rossberg@bz.bzflag.bz) | |
| 09:34.40 | CIA-28 | BRL-CAD: 03d_rossberg * 10brlcad/src/libbu/brlcad_path.c: do not return a pointer to a temporary variable, made variable static |
| 09:37.34 | CIA-28 | BRL-CAD: 03d_rossberg * 10brlcad/src/librt/dg_obj.c: typing error in preceding commit |
| 09:39.22 | Axman6 | so what is it that makes brlcad's raytracer so fast anyway? |
| 09:42.03 | Defcon | some fine ass C coding prolly |
| 09:58.38 | *** join/#brlcad Z80-Boy (n=clock@zux221-122-143.adsl.green.ch) | |
| 10:07.03 | *** join/#brlcad elite01 (n=elite01@dslc-082-082-070-150.pools.arcor-ip.net) | |
| 11:50.36 | CIA-28 | BRL-CAD: 03brlcad * 10brlcad/src/conv/iges/g-iges.c: argv not ptr |
| 12:00.59 | brlcad | PrezKennedy: yes, I have been following some of that |
| 12:01.25 | brlcad | they were stopped, and then they stopped after the freenode ban |
| 12:06.41 | Axman6 | hey brlcad, had a chance to mess around with my mbp yet? |
| 12:47.33 | *** join/#brlcad b0ef (n=b0ef@062016141081.customer.alfanett.no) | |
| 14:49.00 | ``Erik | PrezKennedy: shoulda told him he was infringing on your copyright by reproducing your work without license or permission and told him that if he doesn't remove it all immediately, you'd be contacting a lawyer... :D |
| 14:50.20 | ``Erik | Axman6: BRL-CAD goes back to the days of pdp and vax, with computers measured in kilohertz, not gigahertz... it *HAD* to be tight back then, so even with the bloat through the ages, it's reasonably tight :) |
| 14:51.03 | Axman6 | heh, yeah :) |
| 14:52.08 | Axman6 | when i worked at the charity computer place here, we found some bits that had obviously been pulled off olllddd machines, little LED counter displays, showing the CPU's speed in MHz... with only two digits |
| 14:53.28 | ``Erik | heh, I had a few of those |
| 14:53.46 | ``Erik | often set to "HI" and "LO" |
| 14:54.08 | ``Erik | man, the slew of jumpers on the back of those things made configuring those a fun exercise O.o |
| 14:54.22 | Axman6 | we found some awesome stuff there like HDD's from seagate that were about 17x14x10cm, and 1GB |
| 14:54.34 | ``Erik | with I still had 'em :( now I'm buying straight line versions and having to breadboard them |
| 14:54.42 | Axman6 | wtf's a jumper? |
| 14:54.48 | Axman6 | i kis i kid, don't worry ;) |
| 14:54.54 | Axman6 | kid even |
| 14:54.55 | ``Erik | those're called "full height" |
| 14:55.05 | ``Erik | the winchesters you're used to seeing used to be called "half height" |
| 14:55.17 | Axman6 | heh |
| 14:55.28 | Axman6 | i used the 7 platters from one to make an artwork |
| 14:55.53 | ``Erik | m as in megs |
| 14:56.40 | ``Erik | <-- also remembers drooling over the $3000 20m hdd's :( but at the time, a 3.5" floppy was "almost hard drive like" compared to the 5.25"s |
| 14:56.56 | ``Erik | good old commodores |
| 14:59.50 | Defcon | lolz |
| 14:59.50 | Defcon | yeah |
| 14:59.53 | Defcon | good times |
| 15:03.10 | Z80-Boy | I started on a ZX Spectrum |
| 15:03.32 | PrezKennedy | ``Erik, that's SO cliche |
| 15:04.52 | Defcon | Z80-Boy : do u still need http://www.wazig.be/_dev/xor_text |
| 15:06.55 | Axman6 | our first comp was a Mac Plus. i still have it, it's almost as old as me :) |
| 15:08.07 | CIA-28 | BRL-CAD: 03brlcad * 10brlcad/NEWS: Daniel fixed a mged start-up initialization bug on Windows where a local var was being returned for the brlcad_data path. yuck. |
| 15:08.41 | Z80-Boy | Defcon: no |
| 15:09.05 | Defcon | ok |
| 15:09.58 | Axman6 | ah ha! brlcad is in my machine! |
| 15:10.13 | Axman6 | the fans just revved up :P |
| 15:10.37 | brlcad | :) |
| 15:12.00 | Axman6 | didn't know wtf was going on, saw perl running under your name |
| 15:14.32 | brlcad | you must be on the other side of the planet from me, aussie perhaps |
| 15:14.36 | brlcad | pretty big latency |
| 15:15.16 | brlcad | (not a problem) |
| 15:17.10 | Axman6 | yeah, australia indeed |
| 15:17.27 | brlcad | cool |
| 15:17.31 | Axman6 | and my connection sucks balls. 256k down, 64k up :( |
| 15:17.41 | brlcad | that explains why you keep going to sleep when I wake up |
| 15:17.42 | brlcad | :) |
| 15:17.49 | brlcad | that does suck |
| 15:20.30 | Axman6 | hopefully we'll be getting a lot faster these holidays |
| 15:21.48 | Axman6 | if you're compiling stuff on here, using -j 2 speeds it up a lot |
| 15:22.29 | Z80-Boy | Axman6: back in Prague I built myself an optical wireless Internet connection to my balcony - I couldn't complain :) Not many people have optics directly to their flat ;-) |
| 15:22.50 | Axman6 | heh |
| 15:23.26 | Axman6 | we can get optics here, i can see the cab;e outside my window. yet they defided our suburb didn't need it. wankers |
| 15:23.29 | Z80-Boy | Before that my Internet connection really sucked - it was 0 down and 0 up |
| 15:23.43 | Z80-Boy | Axman6: the same happened in our case with cable TV |
| 15:23.55 | Axman6 | lol |
| 15:23.58 | Z80-Boy | But I defined I want it and since I developed my own FSO... |
| 15:24.12 | Defcon | FileSystemObject? |
| 15:24.12 | Axman6 | how'd you do that anyway? |
| 15:24.23 | Z80-Boy | I just pointed my finger to my friend's house and defined the data will go there - and they did :) |
| 15:24.28 | Z80-Boy | Free Space Optics |
| 15:24.32 | Defcon | ohw :) |
| 15:24.45 | Z80-Boy | I went to my balcony and said "Let there be light" and there was |
| 15:24.47 | ``Erik | erm, but, brlcad, we're int he same time zone, and you tend to be going to sleep when I wake up... O.o :D |
| 15:25.19 | Defcon | lol |
| 15:25.41 | Z80-Boy | Axman6: we even had a coaxial cable that went stright through our house without a tap to another house over the street |
| 15:25.48 | Z80-Boy | The other house had catv, we didn't |
| 15:26.03 | Axman6 | cheers |
| 15:26.48 | Z80-Boy | Axman6: since then I think people should stop complaining and simple build their internet connection themselves if they don't like the commercial ones. |
| 15:27.10 | Axman6 | so what's the other end of yours connect to |
| 15:27.11 | Axman6 | ? |
| 15:27.15 | Z80-Boy | to a friend |
| 15:27.20 | Z80-Boy | and he connected to another friend |
| 15:27.27 | Z80-Boy | and that was connected with another friend with a cable |
| 15:27.31 | ``Erik | heh |
| 15:27.45 | Z80-Boy | and that one was connected through a professional grade guaranteed microwave link to fibre optics backbone |
| 15:28.06 | Z80-Boy | But since it was profi it was symmetric and they cared about us. |
| 15:28.20 | Z80-Boy | So I had fast both downloads and uploads |
| 15:28.37 | Z80-Boy | And it was cheap. And it was already in time when broadband connection was rare. |
| 15:28.40 | ``Erik | a couple states ago, I had a buddy that lived out in a community with no high speed uplink, he was planning on getting a t1 and setting up a repeated ethernet deal with his neighbors to cost share |
| 15:28.51 | Axman6 | bastard |
| 15:29.01 | Axman6 | i couldn't do that gere though |
| 15:29.05 | Axman6 | here* |
| 15:29.16 | ``Erik | highspeed co-op... this was before adsl and cable modems started getting traction |
| 15:29.27 | Axman6 | canberra's just too spread out... and even the backbone ain't that fast here :P |
| 15:29.41 | Z80-Boy | where in australia is canberra? |
| 15:30.04 | ``Erik | in the tiny little strip with people, dur :D *duck* |
| 15:30.17 | CIA-28 | BRL-CAD: 03brlcad * 10brlcad/src/libwdb/strsol.c: quell empty compilation unit warning |
| 15:30.26 | Z80-Boy | which coast? |
| 15:31.08 | Axman6 | east, inland from sydney.. it happens to be the nation's capitol ;) |
| 15:32.24 | ``Erik | http://maps.google.com/maps?ll=-35.281501,148.683472&spn=21.622459,54.777832&z=6 |
| 15:33.30 | Z80-Boy | Is it good for surfing>? |
| 15:33.54 | Axman6 | yeah, if you don't ind a 200KM drive ;) |
| 15:33.58 | Z80-Boy | omg |
| 15:33.59 | Axman6 | mind* |
| 15:34.16 | Z80-Boy | is it 200km from the coast? |
| 15:34.27 | Z80-Boy | or on the coast but 200km to the next usable surfing spot? |
| 15:34.30 | ``Erik | but it's australia, you can do that in what, an hour? a little less? :D |
| 15:34.40 | Z80-Boy | or on the cost but 200km are occupied by rowdy local surfers? |
| 15:34.43 | CIA-28 | BRL-CAD: 03brlcad * 10brlcad/src/libfb/ (if_X.c if_X24.c if_ogl.c if_tk.c if_wgl.c): quell empty compilation unit warning |
| 15:34.50 | Axman6 | ah 200Km from the coast |
| 15:35.02 | Axman6 | nah* damn keyboard |
| 15:35.36 | Z80-Boy | ``Erik: that means you need to go >200 km/h |
| 15:35.51 | ``Erik | uh huh? it's australia! there's nothin' there, you can open it up :D *duck* |
| 15:35.53 | Axman6 | ``Erik: we have sane speed limits, fastest in this part of Aus is 110Kph. the NT used to have unlimited roads |
| 15:36.03 | Z80-Boy | what if you hit a kangaroo, koala, wombat or elephant-sized spider in that speed? |
| 15:36.11 | Z80-Boy | It could have arepercussions for your vehicle |
| 15:36.21 | Axman6 | kangaroo's does serious damage to cars man |
| 15:36.27 | ``Erik | dur (and I think kangaroos are the big ones, as bad as deer here in the US) |
| 15:36.33 | Axman6 | like, you can write off your car if you hit one |
| 15:36.37 | brlcad | Axman6: thanks for the hint, though I'm pretty adept in parallel compilation :) |
| 15:36.50 | Axman6 | lol, my bad :P |
| 15:36.59 | brlcad | ``Erik: heh |
| 15:37.24 | Axman6 | yeah... if you are going -j it, don't do it until i say it's ok, ok? don't want to burn my lap ;) |
| 15:37.27 | ``Erik | -j2 is a waste of hw, we do at least -j8 on our workstations, axman :) |
| 15:38.24 | ``Erik | and brlcad wrote a 'fast' target to do big parallel compiles to take advantage of the 12 and 16 core boxen floating around |
| 15:39.39 | Axman6 | and 512 too? |
| 15:40.17 | ``Erik | I haven't dorked with compile speeds on machines with multiple cores sharing a cache like the duos do, but back in the day of each core having its own die and cache, you wanted more threads than cores... like on a two processor Usparc (like a usparc2 or e450), I'd do -j3 |
| 15:40.50 | ``Erik | um, we have I think 2048 core "machines", but recursive make means the most you can parallelize is the number of compilation units in a given directory |
| 15:41.09 | Axman6 | yeah, Apple are pretty good at threading support i believe, so that might improve things a little |
| 15:41.57 | ``Erik | I d'no, multithreading on g5 has serious kernel flaws for performance... balancing instead of tuning affinity for cache coherence, and no user knobs back in the day... the shared cache machines alleviate that |
| 15:42.58 | ``Erik | the mac pros probably have some of that issue, since they have two l2 caches (two 4-core dies) |
| 15:44.33 | Axman6 | yeah.. lost me now |
| 15:45.27 | brlcad | the 'fast' target also links in parallel unlike default all target, so it can speed things up on a cluster or machines with wide I/O to spare |
| 15:45.35 | *** join/#brlcad Elperion (n=Bary@p54877079.dip.t-dialin.net) | |
| 15:46.04 | brlcad | Axman6: I have compiled on a 512 processor box (single image SMP) before, quite nice |
| 15:46.28 | Z80-Boy | brlcad: then you use -j 1024 |
| 15:46.30 | Z80-Boy | ? |
| 15:46.31 | brlcad | limited more by the make architecture at that point, the per-directory linking phases |
| 15:46.32 | Axman6 | heh :) |
| 15:55.08 | alex_joni | http://miller.emu.id.au/pmiller/books/rmch/ :) |
| 15:56.23 | brlcad | alex_joni: yes, that's old old stuff (and mostly crap) |
| 15:56.46 | alex_joni | brlcad: yeah, I know.. that's why I smiled |
| 15:57.02 | alex_joni | however, it does seem to provide advantages sometimes to have a central Makefile |
| 15:57.35 | brlcad | I've been meaning to write a Nonrecursive Make Also Considered Harmful for a couple years |
| 15:57.48 | brlcad | there are advantages and disadvantages just like most things |
| 15:57.52 | alex_joni | yup |
| 15:59.01 | brlcad | about as fruitful as trying to categorically declare centralized revision control systems as bad (or good) |
| 16:00.07 | brlcad | you give up quite a lot of features by going non-recursive, features that are a pita to implement and maintain in a non-recursive make build system |
| 16:01.00 | brlcad | i still end up waiting more on gcc for the actual compiles and disk I/O than I end up waiting on make |
| 16:01.49 | brlcad | configure and the installation phase actually takes the longest on the faster systems these days than the actual compilation |
| 16:02.12 | Axman6 | could configure be made faster? |
| 16:02.23 | brlcad | mostly due to I/O .. the disks just can't read/write the data fast enough |
| 16:02.52 | Axman6 | probably not then |
| 16:02.53 | brlcad | configure could be streamlined to perform fewer tests, but it's biggest problem is that it's non-trivial to parallelize |
| 16:03.12 | brlcad | and it is massively i/o bound anyways so yeah, wouldn't gain a *whole* lot |
| 16:03.23 | alex_joni | right.. configure is not a place for optimizing |
| 16:03.43 | alex_joni | (you only run it seldom anyways.. <- from the devel point of view) |
| 16:04.14 | alex_joni | the biggest issues I had with recursive make was depenencies from one subfolder to another.. that caused some extra compiling |
| 16:04.18 | Axman6 | it's 50% of the compile process for average users though ;) |
| 16:08.40 | brlcad | alex_joni: most of the cases of "extra compiling" that I've heard of were either incorrectly structured source files or actual valid (i.e. necessary) recompilation |
| 16:09.41 | brlcad | Axman6: what's 50%? the configure phase? |
| 16:10.09 | Axman6 | well, not time wise. but can be pretty major art times |
| 16:12.05 | brlcad | on an old 12 proc altix, it takes about 1 minute to run configure, 2 minutes to compile everything, and 4 minutes to install |
| 16:14.20 | alex_joni | brlcad: I guess it was a combo CVS + incorrectly structured files.. :) |
| 16:14.42 | alex_joni | (I only said CVS because you know what pain it is to move files around..) |
| 16:14.48 | brlcad | on MBP it's about 1 min configure, 9 min to compile (on 2-core), and about 4 min to install ;) |
| 16:17.21 | Axman6 | that's while playing james bond ;) |
| 16:18.21 | CIA-28 | libirc: 03brlcad * r355 10/trunk/libirc/botlib/ (inc/Makefile.am inc/botlib.h src/Makefile.am): add build files to compile the new noinst botlib |
| 16:18.47 | CIA-28 | libirc: 03brlcad * r356 10/trunk/libirc/botlib/Makefile.am: ws |
| 16:19.22 | CIA-28 | libirc: 03brlcad * r357 10/trunk/libirc/NEWS: itemize news for a 0.2 release |
| 16:19.40 | CIA-28 | libirc: 03brlcad * r358 10/trunk/libirc/Makefile.am: have to traverse botlib before src |
| 16:20.00 | CIA-28 | libirc: 03brlcad * r359 10/trunk/libirc/src/Makefile.am: link in botlib |
| 16:20.15 | CIA-28 | libirc: 03brlcad * r360 10/trunk/libirc/configure.ac: generate the makefiles for botlib |
| 16:21.30 | ``Erik | or mebbe tclscripts, since there's nothing to do in 'all', just 'install' |
| 16:22.23 | Axman6 | brlcad: any of those Leopard fixes by any chance?? |
| 16:22.38 | brlcad | Axman6: nope |
| 16:22.48 | brlcad | my compile actually worked out of the box on the first try :) |
| 16:22.51 | brlcad | (on your machine) |
| 16:22.54 | Axman6 | heh, oh well |
| 16:23.00 | Axman6 | that is odd : |
| 16:23.03 | brlcad | but has the cg_png bug on install |
| 16:23.03 | Axman6 | :| |
| 16:23.17 | brlcad | you ran distclean at some point |
| 16:23.35 | brlcad | so you blew away the already-generated tclIndex files .. so when it went to regenerate them, it failed |
| 16:23.50 | brlcad | that problem still exists but since I've no need to distclean, the build completely succeeds |
| 16:23.58 | brlcad | just fails after it's installed |
| 16:24.01 | Axman6 | since i copied my .bash_profile over for you so you had the same PATH etc. i did |
| 16:24.17 | brlcad | it's not anything path-related |
| 16:24.22 | Axman6 | rightp |
| 16:24.27 | Axman6 | o* |
| 16:25.01 | Axman6 | yeah i didn't think it would be, but lis are found usinf PATH in a way right? maybe i'm making that up... |
| 16:26.08 | Axman6 | libs* beh |
| 16:26.18 | brlcad | nope |
| 16:26.59 | brlcad | they're find by traversing a set of system-default dyld search paths, and can be overridden with various compilation/linker flags and environment variables |
| 16:27.01 | CIA-28 | libirc: 03brlcad * r361 10/trunk/libirc/ (31 files in 11 dirs): credit where credit is due, assign copyright to the Jeff guy that actually wrote most of it all |
| 16:27.11 | brlcad | DYLD_LIBRARY_PATH for example |
| 16:28.34 | Axman6 | yeah |
| 16:29.03 | ``Erik | neato, symbolics was the first to register a .com domain name (they made lithp machineth) |
| 16:29.20 | CIA-28 | libirc: 03brlcad * r362 10/trunk/libirc/AUTHORS: yet he is not me |
| 16:29.30 | CIA-28 | libirc: 03brlcad * r363 10/trunk/libirc/autogen.sh: and he didn't write this |
| 16:30.08 | CIA-28 | libirc: 03brlcad * r364 10/trunk/libirc/configure.ac: this is libIRC project.. not 'that other project' |
| 16:36.15 | CIA-28 | libirc: 03brlcad * r365 10/trunk/libirc/ (5 files in 3 dirs): accept pelya's sf patch [ 1843257 ] "MSVC 6 project files" -- thanks\! |
| 16:39.19 | CIA-28 | libirc: 03brlcad * r366 10/trunk/libirc/AUTHORS: thank pelya for the vc6 build system addition |
| 16:41.10 | CIA-28 | libirc: 03brlcad * r367 10/trunk/libirc/Makefile.am: add vc6 to the source dist |
| 16:47.03 | brlcad | well now that's really interesting |
| 16:47.17 | brlcad | once installed, our compiled-in rpaths do seem to kick in and take over |
| 16:47.28 | brlcad | no dyld issues |
| 16:47.33 | brlcad | Axman6: try running mged -f |
| 16:48.06 | Axman6 | X Error of failed request: BadMatch (invalid parameter attributes) |
| 16:48.09 | brlcad | do you get the X error, the cg err, or no err |
| 16:48.11 | brlcad | okay |
| 16:48.12 | Axman6 | etc. |
| 16:48.15 | brlcad | well that's good |
| 16:49.11 | brlcad | that means a base install will actually work just fine compilation-wise, so long as you don't distclean |
| 16:54.04 | brlcad | hm, I do get the X11 bug even remotely (talking to your X11 server) |
| 16:54.28 | brlcad | initial thoughts look like this might be an Xorg bug |
| 16:58.37 | Axman6 | very very possible |
| 16:58.57 | Z80-Boy | X11 is one big bug |
| 16:59.00 | Axman6 | Apple seem to have broken X pretty badly with the move to Xorg |
| 16:59.03 | Axman6 | agreed |
| 16:59.50 | Z80-Boy | I compile Xorg 6.9 on Linux From Scratch and it works fine except every time I shut it down, it prints a white rectangle in the place of cursor, a dark strip in the bottom and the screen freezes. |
| 16:59.55 | Z80-Boy | I have to reboot the computer then. |
| 17:00.32 | Axman6 | haha |
| 17:00.53 | brlcad | Axman6: try this: XLIB_SKIP_ARGB_VISUALS=1 mged -f |
| 17:01.09 | Axman6 | i'm so glad Apple didn't go with X and went their own way. get a great UNIX system, without the X crap |
| 17:01.28 | Axman6 | same error |
| 17:10.08 | *** join/#brlcad prasad_ (n=psilva@static-70-108-244-218.res.east.verizon.net) | |
| 17:27.10 | brlcad | huh, one of the suggestions I ran into was to roll back to Tiger's X11, which apparently works |
| 17:27.26 | brlcad | so maybe just a matter of time, waiting for the fix |
| 17:27.45 | brlcad | still looking for details as to what exactly is *causing* the problem to see if it can be avoided |
| 17:31.12 | Axman6 | i think i'm going to cry, just watched james bond's latest Aston get written off |
| 17:33.57 | brlcad | where he flipped it a few dozen times? |
| 17:34.20 | brlcad | to avoid the girl they tossed into the road |
| 17:34.22 | alex_joni | brlcad: ever looked at u3d ? |
| 17:35.11 | Axman6 | yeah |
| 17:35.13 | ``Erik | like http://my2iu.blogspot.com/2005/04/u3d-is-half-baked.html ? :) |
| 17:35.16 | brlcad | alex_joni: the geometry format standard from intel? |
| 17:35.26 | alex_joni | yeah, that one |
| 17:35.27 | brlcad | http://www.intel.com/technology/systems/u3d/ |
| 17:35.32 | brlcad | no, never heard of it :P |
| 17:35.39 | alex_joni | lol @ brlcad |
| 17:36.02 | alex_joni | I'm specifically looking at it because it *should* be easy to integrade U3D objects into pdf's |
| 17:36.54 | ``Erik | depends on whos pdf library you use, I'd imagine |
| 17:37.32 | alex_joni | I think people succeeded with libharu (if I recall the name correctly) |
| 17:38.01 | alex_joni | ``Erik: but it seems Adobe integrated U3D into pdf's version 7 or later |
| 17:38.15 | ``Erik | yeah, if you happen to use adobes pdf libraries... :D |
| 17:39.22 | alex_joni | well.. umm .. yeah |
| 17:39.40 | Axman6 | :P |
| 17:39.59 | ``Erik | good idea, axman, start codin' that up :D |
| 17:40.08 | brlcad | ahh, now I see why Apple made the s/X11R6/X11/ move |
| 17:40.15 | brlcad | looks like Xorg is pushing out an R7 finally |
| 17:40.30 | Axman6 | i thought it was R7 |
| 17:40.40 | Axman6 | not that they eans much to me |
| 17:40.48 | Axman6 | that means* wow |
| 17:41.01 | brlcad | I think I new that, but forgot |
| 17:41.26 | brlcad | and knew it too |
| 17:42.17 | ``Erik | fbsd has been using xorg 7.3 for a bit |
| 17:42.19 | brlcad | ah, R7.3 just in september |
| 17:42.27 | brlcad | .2 in feb |
| 17:42.45 | ``Erik | ah, those silly beta using fools |
| 17:43.41 | ``Erik | rug |
| 17:44.02 | brlcad | speaking of which.. beta3 |
| 17:44.11 | ``Erik | saw it a few days ago |
| 17:44.24 | ``Erik | if they're cycling betas that fast, I'm hoping they're getting close to a release O.o |
| 17:44.39 | brlcad | "8.5 is nearing the end of feature development." |
| 17:45.16 | brlcad | they're concept of a beta isn't apparently "fix things" but last minute additions |
| 17:45.24 | ``Erik | odd |
| 17:45.39 | ``Erik | wonder what 'alpha' means to them |
| 17:45.50 | brlcad | just getting started ;) |
| 17:46.11 | ``Erik | some camps call that pre-alpha O.o :D |
| 17:46.35 | curious | them slackers |
| 17:46.39 | brlcad | that just might fix/hide this 10.5 startup bug |
| 17:47.25 | Axman6 | calling it a beta? :P |
| 17:47.27 | brlcad | from what I've seen, the bug seems to be provoked by ati driver and the new freetype code (as it pertains to tcl/tk) along with a couple Xorg bugs |
| 17:48.06 | brlcad | howdy yukonbob |
| 17:49.07 | CIA-28 | BRL-CAD: 03erikgreenwald * 10brlcad/src/mged/dodraw.c: silly coder, snprintf needs a length, too |
| 17:49.09 | Axman6 | brlcad: well wouldn't be an ATI problem here |
| 17:49.18 | Axman6 | oww :( |
| 17:49.24 | Axman6 | 74C |
| 17:49.33 | ``Erik | uhmmmm, uhhhh |
| 17:49.46 | ``Erik | get, uh, CoreDuoTemp and smcSomething, uhhhh |
| 17:50.05 | yukonbob | Axman6: good -- I'm ready for breakfast -- two eggs, basted + brown toast, please. |
| 17:50.22 | Axman6 | 82C |
| 17:50.30 | Axman6 | not even 100% cpu either |
| 17:50.32 | ``Erik | there're some nifty utilities for controlling the fan speed envelopes |
| 17:50.41 | ``Erik | smcFanControl |
| 17:50.57 | Axman6 | yeah i have SMCFanController... it'a what's telling me the temp :P |
| 17:51.09 | Axman6 | 86C |
| 17:51.17 | *** part/#brlcad jhujhiti (n=jhujhiti@64.252.100.253) | |
| 17:51.19 | Axman6 | fan's haven't kicked in yet |
| 17:51.23 | ``Erik | if you push up the low and tweak the envelope, you can keep the temp a lot lower |
| 17:51.24 | Axman6 | 87 |
| 17:51.36 | Axman6 | there we go |
| 17:51.59 | brlcad | Axman6: yep, 'cause of that I/O limit |
| 17:52.09 | ``Erik | I mostly did it because I was getting annoyed at the fans going on and off a lot, so I pushed the low up to like 2000 or 2500 rpm (not very audible) and they greatly reduced the fan cycling |
| 17:52.17 | CIA-28 | BRL-CAD: 03brlcad * 10brlcad/src/mged/dodraw.c: double oops, snprintf doesn't need %s size limits |
| 17:52.20 | brlcad | it really wouldn't go must faster if you have 20 cpus or a cpu that was 10x faster |
| 17:52.20 | Axman6 | brlcad: make? |
| 17:52.24 | brlcad | s/must/much/ |
| 17:52.32 | brlcad | no, the autogen.sh and configure steps |
| 17:53.00 | ``Erik | the %Ns limit doesn't hurt nothin', but pushing the format string as the limit does... :D |
| 17:53.04 | Axman6 | ok. yeah because make cn easily use 100% on moth cores |
| 17:53.33 | brlcad | ``Erik: i know, it was just pointless |
| 17:54.18 | brlcad | would have mattered had buffer had more than 32 chars space, but not at same |
| 17:59.24 | Axman6 | bah, there is no way bond would use a sony vaio! |
| 18:03.44 | brlcad | I was right, it is related to the freetype2 X11 processing code in Tk |
| 18:03.50 | brlcad | Axman6: should be working now |
| 18:04.44 | Axman6 | mged -f? |
| 18:05.13 | Axman6 | same error still |
| 18:05.56 | Axman6 | hang on |
| 18:07.07 | Axman6 | how oddwell, no error, but i don't get the mged prompt |
| 18:07.40 | Axman6 | so i can't seem to actually do anything with it :\ |
| 18:20.07 | Axman6 | and mged -f gives the same error again |
| 18:25.52 | Axman6 | 8ok, sleep time. g'night all. brlcad if you have anything you need me to test, or news, just hilight me and i'l get it in the morning |
| 18:29.54 | ``Erik | heh, cvs -z3 commit -m 'terrible crimes against humanity' |
| 18:29.56 | ``Erik | *cough* O.o |
| 18:31.31 | brlcad | hm, Axman6 yours answers were rather confusing to me :) |
| 18:31.46 | brlcad | Axman6: did no windows pop up when you ran mged -f? |
| 18:43.39 | brlcad | Axman6: because it works here.. I successfully ran the mged installed in /usr/brlcad/bin/mged and actually have it displaying all the way from the continent you're on to mine over X, and seems to be working just fine |
| 18:44.49 | brlcad | at least.. where working is a 5 min startup delay and each mouse click takes about 1 minute to respond :) |
| 18:48.37 | CIA-28 | BRL-CAD: 03brlcad * 10brlcad/configure.ac: disable the new freetype font system in Tk 8.5 for now until they're final -- this should 'fix'/avoid a bug reported by dwayne regarding the mged font manager not working right in 7.10.4 |
| 21:47.19 | *** join/#brlcad shprotius (n=shprotiu@c-76-106-99-152.hsd1.md.comcast.net) | |
| 21:55.54 | CIA-28 | BRL-CAD: 03brlcad * 10brlcad/src/conv/proe-g.c: be more consistent on the buffer lengths and limits (needs testing on the name generation) using MAX_LINE_LENGTH for the names too |
| 21:58.18 | yukonbob | brlcad: scan.coverity still pending? |
| 22:08.03 | CIA-28 | BRL-CAD: 03brlcad * 10brlcad/src/external/ProEngineer/proe-brl.c: move the format string to the (single) use |
| 22:25.13 | CIA-28 | BRL-CAD: 03erikgreenwald * 10brlcad/src/adrt/libtie/ (6 files): build execution paths for both single and double precision floats in the same lib |
| 22:25.26 | brlcad | yukonbob: yep |
| 22:28.00 | CIA-28 | BRL-CAD: 03brlcad * 10brlcad/src/fb/fbstretch.c: don't need the HAVE_STDARG_H checks any more, keep it simple |
| 22:28.22 | CIA-28 | BRL-CAD: 03brlcad * 10brlcad/src/fb/fbfade.c: const labelification |
| 22:28.50 | CIA-28 | BRL-CAD: 03brlcad * 10brlcad/src/fb/cell-fb.c: make sure we don't go too far on the buffer |
| 22:30.45 | CIA-28 | BRL-CAD: 03brlcad * 10brlcad/src/fb/fbstretch.c: const lables |
| 22:31.01 | CIA-28 | BRL-CAD: 03erikgreenwald * 10brlcad/src/adrt/libtie/ (tie.c tie.h): make triangle degeneracy testing a runtime selectable feature |
| 22:31.03 | yukonbob | brlcad: :P |
| 22:43.00 | *** join/#brlcad louipc (n=louipc@bas8-toronto63-1177613911.dsl.bell.ca) | |
| 22:57.48 | *** join/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1177593139.dsl.bell.ca) | |
| 22:58.20 | IriX64 | http://www3.sympatico.ca/7.11.0.png <--- finally |
| 22:58.40 | CIA-28 | BRL-CAD: 03brlcad * 10brlcad/src/ (17 files in 8 dirs): c89 simplification. assume stdarg is a given. |
| 23:01.00 | IriX64 | that makefile.am thing is one thing, -lXft in X_libs is needed in bwish,mged,and rttherm |
| 23:01.37 | louipc | IriX64: what's in the image? |
| 23:01.45 | IriX64 | ktank :) |
| 23:01.55 | louipc | oh |
| 23:02.14 | IriX64 | you said you were tired of helicopters ;) |
| 23:02.21 | louipc | hahha |
| 23:02.50 | louipc | IriX64: have you tried playing with any linux/bsd yet? |
| 23:03.10 | IriX64 | not yet, waiting for Christmas |
| 23:03.17 | louipc | ah |
| 23:03.42 | louipc | you know, I think I'd rather just use one of those in a VM rather than cygwin heh |
| 23:03.56 | IriX64 | vm costs money :) |
| 23:04.34 | IriX64 | altho,like all else once spent it's spent :) |
| 23:04.57 | louipc | http://www.virtualbox.org/wiki/Downloads |
| 23:05.10 | louipc | that one is free |
| 23:06.19 | IriX64 | downloading it thanks man |
| 23:06.36 | alex_joni | there is also a free one from VMware |
| 23:06.45 | louipc | yeah I thought so too |
| 23:06.46 | alex_joni | VMware server, if my memory still serves me |
| 23:06.57 | IriX64 | checked vmware,didn't see the free one |
| 23:07.10 | louipc | haha it's on the front page |
| 23:07.11 | alex_joni | IriX64: what platform? |
| 23:07.16 | louipc | http://www.vmware.com/ |
| 23:07.17 | alex_joni | (for the host..) |
| 23:07.21 | IriX64 | amd64 |
| 23:08.10 | IriX64 | vista64 actually |
| 23:08.18 | alex_joni | http://register.vmware.com/content/download103.html |
| 23:08.21 | louipc | oh you have to register though |
| 23:08.41 | louipc | hey you bypassed it? |
| 23:08.57 | IriX64 | i registered the evaluation workstation copy |
| 23:09.08 | alex_joni | louipc: no, you still have to register after the download |
| 23:09.14 | louipc | ah |
| 23:09.26 | alex_joni | s/after/during/before/ |
| 23:09.37 | alex_joni | whatever suits you :) |
| 23:09.51 | IriX64 | heh whatever suit fits :) |
| 23:10.14 | alex_joni | well.. so far I didn't either.. but lately I got 2-3 examples that really worked as advertised |
| 23:10.35 | alex_joni | vmware was one of them |
| 23:11.30 | louipc | you eat your visitors? |
| 23:13.25 | alex_joni | louipc: if they're foolish enough to come without food |
| 23:13.49 | louipc | aha! |
| 23:20.04 | louipc | qemu does |
| 23:22.51 | louipc | oh yeah bochs too |
