IRC log for #brlcad on 20091203

00:00.06 brlcad still calls it carbon :)
00:00.23 jack2 hehe, np
00:00.44 jack2 wonder if i should do an x11 variant too maybe
00:00.53 brlcad variant?
00:01.03 jack2 yeah, 2 packages
00:01.10 jack2 brlcad and brlcad-x11
00:01.11 brlcad what's the non-x11 variant?
00:01.31 jack2 trying to use as little of x11 as possible ;)
00:02.06 brlcad brlcad without x11 is interesting and should certainly be possible (I had a clean non-x11 build verified a couple months back) but you get little/no GUI then
00:02.20 jack2 mmh
00:02.25 jack2 mged, right?
00:02.33 brlcad right, you'd get a console-only mged
00:02.48 brlcad you'd still get mged, just command-line only
00:02.49 jack2 ok, that's pointless indeed
00:02.58 jack2 almost pointless at least
00:03.17 *** join/#brlcad talcite (n=matthew@dhcp-143-147.mcme-students.carleton.ca)
00:03.26 brlcad if you're already familiar with brl-cad, it's not pointless, but that's less than 1% of our usual users :)
00:03.34 jack2 exactly
00:03.35 brlcad most are new
00:03.56 brlcad otherwise, there's a helluvalot you can do with the other 400 tools
00:04.11 jack2 sure :)
00:04.31 jack2 wonder if i should hide all the stuff in a private bindir...
00:04.47 jack2 people might hate me for spamming %p/bin ;)
00:05.34 brlcad probably a good idea ;)
00:05.34 brlcad better yet, install the entire root to a private dir
00:05.51 jack2 what's wrong with macosx' zlib? (why does it build an own one?)
00:06.30 brlcad nothing should be wrong with it
00:07.17 ``Erik for fbsd, I use system provided zlib. libpng, tk and tcl
00:07.17 brlcad if you added --disable-all, it should force everything off
00:07.30 ``Erik um, and jama and uhhh
00:07.32 ``Erik other stuff :D
00:07.32 jack2 libpng works
00:07.43 jack2 but it builds an own zlib and tcl+tk
00:08.16 ``Erik hm, /usr/lib/libz.dylib might be a little fruity, might need finks zlib
00:08.28 brlcad tcl/tk make sense, os x doesn't ship an incrTcl so it has to use ours and ours requires 8.5 which is newer than the os x default
00:08.31 jack2 we don't provide zlib anymore
00:08.41 jack2 system one is good enough :)
00:08.51 jack2 (normally, that is)
00:09.08 ``Erik investigate the config.log and stuff to see why zlib fails to autodetect?
00:09.24 ``Erik our check might be too aggressive
00:09.28 jack2 brlcad: yup, i don't mind getting a fresh tcltk8.5 with itcl/itk of course
00:09.35 jack2 only zlib is odd
00:09.36 brlcad but that's still assuming you're letting it auto-detect -- you shouldn't be getting autodetection behavior if you used --disable-almost-everything
00:09.48 jack2 ok...will check
00:10.05 brlcad or --disable-all or whatever aliases
00:10.15 jack2 brlcad: nope, i did --enable-a-e and --disable-jove-build
00:10.18 brlcad should say 'no' down the summary list
00:10.22 brlcad a-e ?
00:10.27 brlcad all
00:10.29 jack2 almost everything
00:10.40 ``Erik hm, "Linux OpenRD 2.6.22.18-dirty #1 Sat Aug 29 15:18:06 IST 2009 armv5tejl armv5tejl armv5tejl GNU/Linux"
00:10.42 brlcad enable-almost-everything forces everything to build
00:10.49 brlcad all external dependencies
00:11.08 brlcad you want the opposite for fink, disable everything
00:11.17 jack2 d'oh, but why does it like my external libpng then... :)
00:11.23 brlcad "disable our compilation of [whatever]"
00:11.46 brlcad yeah, that's not right..
00:11.55 jack2 just odd
00:12.16 brlcad sounds like a build system bug, but would need to see your config.log
00:12.18 brlcad can you post it?
00:12.23 jack2 i don't really mind, zlib is really small
00:12.26 jack2 but funny
00:12.40 jack2 brlcad: ok, hold on a sec
00:13.01 brlcad well zlib makes more sense given you used --enable-all
00:13.05 brlcad it's enabled
00:13.11 brlcad they mystery is libpng
00:13.26 brlcad or anything else that was a 'no' in the summary
00:14.12 ``Erik the blue power ld is bright as hell, I might need to stick a piece of masking tape over it or solder a resister in there
00:14.47 jack2 brlcad: http://85.180.98.251/config.log
00:15.52 jack2 the CPPFLAGS are nonsense of course, ignore that -framework OpenGL
00:16.57 brlcad 404
00:17.41 jack2 indeed...wtf
00:17.45 jack2 one sec
00:18.26 ``Erik heh, / on that site is 404, too
00:19.38 jack2 ok, now it works
00:19.41 jack2 sorry
00:21.53 brlcad don't understand something, your config.log says libpng will be built
00:22.05 brlcad everything will be built, except jove
00:22.09 brlcad which is right
00:23.23 jack2 hrm
00:25.02 brlcad --enable-aquatk-build shouldn't do anything useful, and you shouldn't --enable-rtgl just yet unless you plan on coding :)
00:25.25 brlcad it works great, but not something to spring on the fink masses just yet, few issues to work out
00:27.03 jack2 hehe ok
00:29.52 jack2 woop, failure
00:30.04 brlcad orly?
00:30.07 jack2 linking .libs/libstepeditor.19.dylib is missing some symbols
00:30.18 brlcad o.O
00:30.31 jack2 ld: Undefined symbols:
00:30.31 jack2 __ZN18SDAI_DAObject_SDAIC2Ev
00:30.31 jack2 __ZN18SDAI_DAObject_SDAID2Ev
00:30.31 jack2 __ZTI18SDAI_DAObject_SDAI
00:31.47 brlcad where was it in the build?
00:32.07 brlcad in src/other/step or in src/conv/step ?
00:32.37 jack2 other/step
00:32.54 brlcad those SDAI_DAObject_SDAI symbols are in a diff lib
00:33.31 brlcad so the question is, what was it linking that it'd care and was that other lib on the link line
00:34.48 jack2 other lib? hold on, i'll lisppaste the whole link line
00:35.35 brlcad ah.. hm, found a problem I think
00:35.58 brlcad looks like some vars were renamed, bizarre that it wasn't caught on other mac systems
00:36.23 jack2 http://paste.lisp.org/+1YL5
00:38.15 *** join/#brlcad Ralith (n=ralith@69.90.48.97)
00:38.48 CIA-41 BRL-CAD: 03brlcad * r36778 10/brlcad/trunk/src/other/step/src/cleditor/Makefile.am: curiously, still using the CL* vars instead of what configure.ac is providing. link failure generating resolved dylib.
00:38.55 brlcad ahhhh.... hrm
00:39.02 brlcad step has it's own configure.ac
00:39.04 ``Erik heh, a lisp weenie
00:39.05 brlcad oops
00:39.16 jack2 brlcad: maybe that stupid libtool default sneaked in on other macs?
00:39.39 jack2 (-Wl,-undefined,dynamic_lookup or suppress)
00:40.46 CIA-41 BRL-CAD: 03brlcad * r36779 10/brlcad/trunk/src/other/step/src/cleditor/Makefile.am: oops, SCL has its own configure.ac, use the CL-prefixed vars (most importantly adding/keeping libdai)
00:41.51 brlcad these other systems are stock, we re-repair the libtool script if it detects badness but the error in this case is warranted for fully resolved libs.. but I bet there will be other issues
00:42.35 brlcad was the -all_load dumbness in the libtool script
00:43.02 jack2 yeah, i'm glad you remove that crap yourself meanwhile :)
00:44.09 brlcad hm, but you know I bet that's the issue.. SCL has it's own configure, it's own libtool script
00:45.16 jack2 something not propagating right?
00:45.42 brlcad maybe
00:45.48 jack2 modularity isn't bad...as long as the parts know how to talk to each other ;)
00:47.33 brlcad if you diff libtool src/other/step/libtool, what do you see?
00:47.45 brlcad just a generated and temp_rpath line or more?
00:48.41 brlcad regardless, the fix "should" be to edit src/other/step/Makefile.am and add ${CLDAI} \ to the LIBADD section
00:49.13 brlcad er, src/other/step/cleditor/Makefile.am
00:49.57 brlcad cd src/other/step/cleditor && make LIBS='${CLDAI}' might work.. :)
00:50.03 brlcad as an alternative
00:52.52 CIA-41 BRL-CAD: 03brlcad * r36780 10/brlcad/trunk/src/librt/primitives/ars/ars.c: more quellage via data validation
00:58.02 ``Erik looks for a serial cable O.o
01:07.42 CIA-41 BRL-CAD: 03brlcad * r36781 10/brlcad/trunk/src/librt/primitives/ars/ars.c: quell the rest of the ARS warnings. log shadow and unused parameter validation.
01:09.24 CIA-41 BRL-CAD: 03brlcad * r36782 10/brlcad/trunk/include/raytrace.h: quell unreachable code warnings for callers of RT_HIT_NORMAL() when either the normal is NULL or the flip flag is a constant. we squash the warnage by setting to a locally scoped temp var.
01:12.48 jack- brlcad: ok, thx
01:13.51 brlcad hits the road for a bit
07:45.54 *** join/#brlcad d_rossberg (n=rossberg@BZ.BZFLAG.BZ)
08:24.08 *** join/#brlcad _clock_ (n=_sushi_@80-218-244-105.dclient.hispeed.ch)
08:24.59 *** join/#brlcad _clock_ (n=_sushi_@80-218-244-105.dclient.hispeed.ch)
08:48.35 d_rossberg i had the following problem:
08:48.50 d_rossberg - created a (empty) group
08:49.15 d_rossberg - selected this group (rt_gettree)
08:49.47 d_rossberg - created a region with a solid and put it into the group
08:50.49 d_rossberg - now rt_shootray won't find anything
08:52.08 d_rossberg ist this behavior intended? (i.e. rt_gettree has to be called after all has been created)
10:14.14 *** join/#brlcad Elrohir (n=kvirc@p5B149CAF.dip.t-dialin.net)
11:17.28 d-lo Mornin all!
11:52.35 *** join/#brlcad cosurgi (n=cosurgi@atak.bl.pg.gda.pl) [NETSPLIT VICTIM]
11:52.35 *** join/#brlcad b0ef (n=b0ef@157.26.202.84.customer.cdi.no) [NETSPLIT VICTIM]
11:52.35 *** join/#brlcad poolio (n=poolio@BZ.BZFLAG.BZ) [NETSPLIT VICTIM]
11:52.35 *** join/#brlcad starseeker (n=starseek@BZ.BZFLAG.BZ) [NETSPLIT VICTIM]
11:52.35 *** join/#brlcad alex_joni (n=alex_jon@emc/board-of-directors/alexjoni) [NETSPLIT VICTIM]
11:52.35 *** join/#brlcad yukonbob (i=1000@s142-179-54-198.bc.hsia.telus.net) [NETSPLIT VICTIM]
11:52.35 *** join/#brlcad Maloeran (n=maloeran@glvortex.net) [NETSPLIT VICTIM]
11:52.35 *** join/#brlcad archivist (n=archivis@81.149.119.172) [NETSPLIT VICTIM]
11:54.27 *** join/#brlcad d_rossberg (n=rossberg@BZ.BZFLAG.BZ) [NETSPLIT VICTIM]
11:54.42 *** join/#brlcad b0ef (n=b0ef@157.26.202.84.customer.cdi.no) [NETSPLIT VICTIM]
11:58.50 *** join/#brlcad poolio (n=poolio@BZ.BZFLAG.BZ) [NETSPLIT VICTIM]
12:00.31 *** join/#brlcad cosurgi (n=cosurgi@atak.bl.pg.gda.pl) [NETSPLIT VICTIM]
12:03.16 *** join/#brlcad erik__ (n=erik@c-69-140-109-104.hsd1.md.comcast.net)
12:34.30 d-lo Mornin ``Erik
12:34.39 d-lo get your lil puter last night?
13:30.00 CIA-41 BRL-CAD: 03davidloman * r36783 10/rt^3/trunk/geoserve.config: Add basic configuration file.
13:30.14 CIA-41 BRL-CAD: 03davidloman * r36784 10/rt^3/trunk/ (include/GS/GeometryService.h src/GS/GeometryService.cxx): Override ::exec() from QCoreApplication to allow object level init prior to entering main event loop.
13:33.49 ``Erik yup
13:33.59 ``Erik pasted uname, even
13:34.18 d-lo 'pasted uname' ?
13:34.50 ``Erik yeah, scroll up... I lost it due to a poweroutage overnight :/
13:37.09 d-lo Suckage. When I got up at 0400, the wind was howling and power was flickering... but it stayed on. How long was yours out for?
13:38.49 ``Erik long enough to sap the UPS, but I had 3 machiens on it
13:39.20 ``Erik fortunately, I wake up at the same time every day, alarm or not :)
13:40.33 d-lo Ditto about the alarm. Sucks on weekends though. I don't think I have slept in since High School.
13:40.54 d-lo Now when the power went out, did the OpenRD machine get messed up?
13:43.08 ``Erik nah, it was messed up before :D
13:43.21 d-lo lol
13:43.49 d-lo Messed up upon arrival, or did it get messed up by operator? ;)
13:43.49 ``Erik rendered it unbootable before I went to sleep, uses a fruity uboot thing that isn't geared to do anything but boot linux off of the flash NAND memory
13:44.18 ``Erik so I was trying to coerce it with a different boot command to attempt to bring up the USB subsystem and try passing control off to a drive
13:44.29 ``Erik have to do it all via serial console to boot, fun stuff
13:44.45 d-lo Ah, hence the serial cable search.
13:44.48 d-lo didja find one?
13:44.55 ``Erik might take a week or two before I coerce it into booting bsd
13:45.03 ``Erik um, usb to the jtag faking it
13:45.45 d-lo I have a box full o junk and I am sure I have a handful of subd-9's in there if you need one.
13:46.43 ``Erik the usb solution works pretty well, just have to remember to force to program to go to 115200 baud, 8 bit, 1 stop bit and no flow control
13:46.46 ``Erik dang settings
13:47.19 ``Erik at the moment, it spools around, starts the usb subsystem, scans, looks for usb storage devices, complains about bad magic and reboots
13:47.51 d-lo did it come with some 'software' to 'flash' it back to normal via serial?
13:48.00 ``Erik <-- half wondering if he could stomach using linux in that role
13:48.19 ``Erik um, there's a recovery image on the dvd, not sure what exactly it does
13:49.06 d-lo I was just thinking that establishing the ability to restore it to 100% bootable again would be a good backup plan to have.
13:49.10 ``Erik or how to use it o.O it's slightly bigger than I estimated, too, but still quite small
13:49.22 ``Erik where's the adventure in that???
13:49.32 d-lo isnt it like 6x8x1 inches?
13:49.50 ``Erik porbably about, I was thinking it was more like 4x6x1 for some reason
13:50.04 ``Erik from looking at the end panel and trying to guess... shoulda looked at the numbers :)
13:50.12 d-lo Heh, well if adventure gives me a soildstate paperwieght, I'll take safe and sound anyday :) But then again, I pinch pennies.
13:51.35 ``Erik I haven't tried writing a new uboot image yet, trying to do this with adjusting settings, and I'm writing down the old settings as I go *shrug*
13:52.51 d-lo how much Ram does it have 512?
13:53.18 ``Erik yup, 512 ram, 4g NAND flash, 1.2ghz sheeva (roughly like an 800mhz p3), no fpu
13:53.41 d-lo whoa, 4GB of NAND flash? neato.
13:54.23 ``Erik and the dvd has all sorts of neat shit... schematics, PCB layout diagrams, ...
13:54.32 ``Erik this ain't no dell :D
13:54.54 d-lo I wonder if you could get windows to boot off it, lol
13:55.23 ``Erik given that windows only supports x86 (used to support alpha, but tha went away with nt4), I'd say...... no? :)
13:55.54 d-lo You killjoy you.
13:56.15 d-lo What about MacOS X? Could you make a mini-mini?
13:56.28 ``Erik nope, osX only rusn on ppc and intel
13:56.31 ``Erik http://en.wikipedia.org/wiki/ARM_architecture
13:57.29 d-lo didnt some of the older flavors of Mac OS support RISC procs?
13:57.38 ``Erik ppc is a risc
13:58.06 ``Erik 68k was cisc
13:58.42 ``Erik <-- keeps entertaining the notion of buying a 68040 and building a small machine out of it, then writing an oldschool *nix os on it
13:59.03 d-lo ...but.... why?
13:59.07 d-lo =D
13:59.19 ``Erik fun... and to jog my memory of the fun old days
14:00.46 d-lo Oh, did you find any docs on power consumption?
14:01.05 ``Erik um, didn't erally look
14:01.32 d-lo Just curious as to what Idle->Full load power ramp up looks like.
14:01.51 ``Erik Average power consumption is around 7.2W. Maximum power consumption measured with all 7 USB
14:01.54 ``Erik ports, HDD and Ethernet operational is 29.9W. In idle state system draws power of
14:01.57 ``Erik around 4.8W.
14:02.34 d-lo 30W on 12Vdc = ~ 2.5 amps. Notbad.
14:02.53 ``Erik I imagine a significant portion of that is the hard drive
14:03.15 d-lo There is a HD in the lil box too???
14:03.16 d-lo wow
14:03.27 ``Erik or perhaps the usb devices being used... my thumbdrive draws 200mw, where the external drive has a power supply ad draws 2mw
14:03.35 ``Erik well, not in i t, just the flash, but it has esata
14:03.57 ``Erik I haven't taken it apart yet, supposedly there's not enough space to fit a 2.5" drive
14:04.16 d-lo not too familiar with eSata, actually. I take it eSata cables also provide power to the device, akin to USB?
14:04.24 ``Erik no clue
14:05.32 d-lo Great. Some contractors outside my window are putting in ear plugs... this can't be a good sign. :/
14:05.41 ``Erik looks out O.o
14:05.55 ``Erik they have a circular saw of some kind
14:06.08 d-lo ...and a score to settle...
14:37.23 brlcad d_rossberg: that sounds like expected behavior
14:37.35 brlcad you probably just need to call rt_gettree again
14:37.50 brlcad you should be able to call it repeatedly
14:38.37 brlcad (man librt for some additional details)
14:39.00 brlcad gettree loads the geometry, so if it's modified, it needs to be reloaded
15:02.37 *** join/#brlcad Ralith (n=ralith@69.90.48.97)
15:12.40 d_rossberg brlcad: ok, i've already changed my code, i only wanted to be sure that it's ok
15:13.04 brlcad okay
15:14.15 jack- ``Erik: you could even buy an old amiga3000 and put an 68060 card in
15:14.28 jack- those were pretty fast back then ;)
15:14.29 brlcad depending on how you created/added the new prim/region, it's probably possible to avoid the rt_gettrees again
15:15.07 brlcad but at some point the data has to be read (whether from disk or inmem).. and that's presently gettrees job
15:15.49 d_rossberg i used my c++ interface from rt^3
15:16.23 d_rossberg it is very low-level internally
15:16.30 brlcad really wanting to get bu, bn, pkg, and rt fully cleaned up as stand-alone APIs with cleanly generating doxygen reports, so questions like this are answered more clearly
15:16.35 brlcad yeah, I figured
15:17.01 brlcad which is great, keeps things honest
15:17.22 brlcad plan on hitting that section of code up next, for the GE
15:17.26 brlcad after all this cleanup
15:18.16 brlcad and annotations, which I'm heartily behind on :/
15:18.38 d_rossberg yeah, and solving the brlcadversion.h-question ;)
15:18.47 brlcad yep!
15:19.00 brlcad 200 message in my mailbox that *require* a response.. that's one of them :)
15:19.30 brlcad (not all brl-cad related, of course) :)
15:19.58 brlcad oh, right, I got through a bunch.. down to just 87 today
15:22.32 d_rossberg i have 5 more primitives half-finished ... since April
15:25.58 brlcad which five?
15:28.00 brlcad and what does "half-finished" mean? they load but don't ray-trace or the API is missing or ...?
15:29.17 d_rossberg EllipticalTorus, HyperbolicCylinder, Hyperboloid, ParabolicCylinder and Paraboloid
15:29.26 d_rossberg i.e. the easy ones
15:30.37 d_rossberg "half-finished" means they are programed by an intern and i hadn't time to look at them
15:31.38 d_rossberg it is usually not so much work, however i've other preferences at the moment
16:59.07 *** join/#brlcad mafm (n=mafm@249.Red-88-11-185.dynamicIP.rima-tde.net)
17:03.00 CIA-41 BRL-CAD: 03davidloman * r36785 10/rt^3/trunk/ (include/GS/Config.h src/GS/CMakeLists.txt src/GS/Config.cxx): Beginings of Configuration loading system.
17:05.36 CIA-41 BRL-CAD: 03davidloman * r36786 10/rt^3/trunk/geoserve.config: Add address/port to config file.
17:44.33 *** join/#brlcad Elrohir (n=kvirc@p5B149CAF.dip.t-dialin.net)
18:48.27 CIA-41 BRL-CAD: 03indianlarry * r36787 10/brlcad/trunk/src/conv/step/ (FaceBound.cpp FaceBound.h): Added function to test if FaceBound is oriented.
18:59.39 CIA-41 BRL-CAD: 03indianlarry * r36788 10/brlcad/trunk/src/conv/step/ (PullbackCurve.cpp PullbackCurve.h): Added PBC_SEAM_TOL definition to explicitly test trim point sample closeness to surface seam. Also using BREP_EDGE_MISS_TOLERANCE as tolerance input to SurfaceTree::getSurfacePoint().
20:05.43 *** join/#brlcad IriX64 (n=IriX64@bas2-sudbury98-1178014899.dsl.bell.ca)
20:42.24 *** join/#brlcad louipc (n=louipc@archlinux/trusteduser/louipc)
21:18.23 CIA-41 BRL-CAD: 03indianlarry * r36789 10/brlcad/trunk/src/conv/step/OpenNurbsInterfaces.cpp:
21:18.23 CIA-41 BRL-CAD: Use the cues from FaceBound "Orientation" setting to determine loop direction consistency instead of openNURBS BREP LoopDirection() function.
21:18.23 CIA-41 BRL-CAD: Unitized ON_Line::Direction() used in intersectLines() function, large line magnitudes were causing some tolerence problems with bn_isect_line3_line3().
21:18.25 CIA-41 BRL-CAD: Gutted Circle::LoadONBrep(), now subdivides arc based on angular extents and builds NURB directly. (Will need to do the same for Ellipse)
21:18.28 CIA-41 BRL-CAD: Cleanup start/end point code in all the Conic curve loading.
21:45.13 *** join/#brlcad PrezKennedy (i=Matthew@whitecalf.net)
23:12.52 *** join/#brlcad R0b0t1 (n=Enigma@unaffiliated/r0b0t1)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.