irclog2html for #brlcad on 20060705

01:17.33 *** join/#brlcad IriX64 (n=IriX64@toronto-HSE-ppp4310781.sympatico.ca)
01:17.34 IriX64 blugh...disconnected.
01:17.57 IriX64 ill see you people tommorow.
03:07.54 *** join/#brlcad Twingy (n=justin@c-69-250-236-111.hsd1.md.comcast.net)
03:24.21 *** join/#brlcad Twingy (n=justin@c-69-250-236-111.hsd1.md.comcast.net)
05:27.13 *** join/#brlcad PKMOBILE (n=Apathy@c-68-33-243-45.hsd1.md.comcast.net)
08:44.05 *** join/#brlcad SWPadnos_ (n=Me@dsl245.esjtvtli.sover.net)
08:57.53 *** join/#brlcad SWPadnos_ (n=Me@dsl245.esjtvtli.sover.net)
10:21.04 *** join/#brlcad anz (n=kvirc@johvi-ypk.rescue.ee)
10:24.28 *** part/#brlcad anz (n=kvirc@johvi-ypk.rescue.ee)
15:02.49 CIA-9 BRL-CAD: 03lbutler * 10brlcad/sh/ (prior.psql prior.sh): scripts to collect data from gforge tasks
15:54.55 CIA-9 BRL-CAD: 03lbutler * 10brlcad/sh/prior.psql: expanded select so that names of the developers are included with the tasks
17:14.45 CIA-9 BRL-CAD: 03brlcad * 10brlcad/HACKING: gah, the horrirs. spell sentance correctlay
17:28.12 ``Erik "special" :)
17:36.54 ValarQ ?
17:40.18 *** join/#brlcad IriX64 (n=IriX64@toronto-HSE-ppp4310781.sympatico.ca)
17:41.28 IriX64 was trying to ftp send to ftp.brlcad.org got only 12kbps and it aborted anyway, go figure.
17:46.36 IriX64 whoever put that geometry browser in mged knew what they were doing :)
17:49.11 IriX64 the enterprise and shuttlecraft are beautiful.
17:55.52 ValarQ how usable is 'sketch'?
18:16.22 ValarQ and how stable is brlcad considered to be on amd64?
18:19.10 IriX64 on my amd64 it is very stable.
18:19.21 IriX64 what os?
18:19.31 SWPadnos lee-nucks :)
18:19.42 IriX64 nyuk :)
18:19.43 SWPadnos how about SMP?
18:19.54 ValarQ gentoo GNU/Linux
18:19.59 IriX64 symetrical multiprocessing? ;)
18:20.05 SWPadnos oh, thanks ;)
18:20.07 ValarQ yes
18:20.15 ValarQ i got 2cpus
18:20.22 SWPadnos btw - do you actually have an Irix system?
18:20.33 IriX64 my secret :)
18:20.38 SWPadnos heh
18:20.41 IriX64 linux64?
18:20.49 SWPadnos I have an Indigo2 / R10k
18:21.06 IriX64 i have the enterprise system :)
18:21.15 SWPadnos lucky you
18:21.28 IriX64 now if it would only fly :)
18:21.32 ValarQ IriX64: yeah, my entire system is 64bit
18:21.36 SWPadnos and transport :)
18:21.41 IriX64 heh
18:21.50 IriX64 is linux64 stable?
18:21.54 SWPadnos very
18:21.58 IriX64 for you?
18:22.03 SWPadnos more stable than any other common 64-bit OS :)
18:22.05 SWPadnos yes
18:22.18 SWPadnos not on my Indigo, on my dual opteron
18:22.25 IriX64 brlcad64 coming up ;)
18:22.42 SWPadnos I wonder how long it would take to compile on my Opteron rig
18:23.07 IriX64 on my system it take 45 minutes with another ten for make install.
18:23.49 SWPadnos ok - is that the giant server with 32 R14k CPUs, or something that mere mortals might have?
18:24.54 IriX64 i *am a mere mortal.
18:24.59 SWPadnos oh ;)
18:25.06 IriX64 ;)
18:28.17 ValarQ is it possible and if so how do i make an extrude on an NMG shape or sketch?
18:31.12 IriX64 im not a sophisticated use yet :)
18:31.17 IriX64 user too
18:31.52 ValarQ ok :(
18:32.13 IriX64 these others tho get with it :)
19:18.41 brlcad ValarQ: yes to the 64bit question and SMP support, it's a fundamental on the geometry library
19:20.09 brlcad ValarQ: compile should take "less than an hour" depending on which Opteron -- my guess would be about 10-15 minutes for most chips
19:20.49 brlcad quickest way to make a sketch is the type "make sketch" in the command window, it will create an empty sketch (which is of course useless) ;)
19:21.36 brlcad more practical is to then open that sketch in the sketch editor which is accessed either via the misc menu iirc or on the Edit menu
19:22.10 brlcad if you already have a sketch, it's trivial to extrude .. "make extrude" will get you started
19:24.05 ValarQ ok
19:26.17 IriX64 multiple definition of ___do_sjlj_init ? *WHY?
19:27.00 IriX64 libz.so crtstuff.c no such fiole what gives?
19:27.16 IriX64 file too.
19:27.24 ValarQ brlcad: 'make foo.s extrude' just kills mged
19:27.45 ValarQ brlcad: then when i fire up mged again 'e foo.s' almost kills my machine
19:30.35 ValarQ and 'gdb mged core' doesn't give me much ("Cannot access memory at address 0x2b5f32dbd988")
19:43.49 brlcad hmmm
19:44.50 ValarQ exactly what i was saying
19:45.28 ``Erik and the bt?
19:46.14 ValarQ bt?
19:46.26 brlcad ahh, so it does crash.. *ahem*
19:46.35 ``Erik backtrace
19:46.41 brlcad that's not the right syntax for what it's worth
19:46.58 ``Erik eh?
19:47.11 brlcad infinite loops until the stack is blown
19:47.29 ValarQ ``Erik: i get no backtrace
19:47.45 ``Erik "no backtrace" as in lots of question marks, or?
19:48.22 brlcad i have a backtrace, looking at it
19:48.33 ``Erik aight *shrug*
19:49.57 ValarQ (gdb) bt
19:49.57 ValarQ #0 0x00002b7350302843 in ?? ()
19:50.09 ValarQ thats all
19:50.32 brlcad ValarQ: the make command was the wrong command to tell you, use the "in" command
19:51.02 brlcad make creates a "prototypical"/example extrusion object (and clearly has some bug), that wouldn't have been extremely useful anyways
19:52.23 ValarQ ok
19:52.35 ValarQ what is H, A and B?
19:52.43 ValarQ (and K for that matter)
19:52.48 brlcad the crash has to do with it creating an empty sketch.. tries drawing a wireframe of that
19:53.07 ValarQ ok, i can see how it fails to do that :)
19:53.11 brlcad height vector and uniform A/B cross vectors
19:53.28 brlcad K is the step size
19:53.51 brlcad that is, it lets you do more than simple extrusions, you can extrude in different scaling directions
19:54.08 brlcad simplest is to use 0 0 1 for H, and 1 0 0 and 0 1 0 for A and B
19:54.16 brlcad play with different K values
19:54.32 brlcad K is basically how much to extrude
19:55.07 ValarQ ok, thanks for the help
20:17.26 IriX64 Elapsed compilation time: 42 minutes, 32 seconds on an amd64 2.4ghz system ValarQ.
20:21.04 brlcad that's a little slow..
20:21.12 brlcad for that system
20:21.27 IriX64 --enable-everything.
20:21.32 brlcad optimized?
20:21.36 IriX64 yes
20:21.51 brlcad ah, optimized will slow it down a variable amount
20:22.00 IriX64 summat. :)
20:22.01 brlcad gcc's not the hottest on amd64
20:22.16 brlcad (at least not yet)
20:22.22 IriX64 not many are ;)
20:23.39 IriX64 am i breaking any law trying to upload binaries to ftp.brlcad.org? orrules or anything?
20:25.16 IriX64 ~400 meg file and i was only getting 12k/s and it aborted anyway.
20:30.38 brlcad that's insanely oversized
20:31.01 brlcad sounds like fully-static binaries or something
20:32.28 brlcad no, you're not breaking the law, it's our server -- if you were providing them on a website or something similar, you'd technically be obligated to provide source as well perhaps
20:34.27 ``Erik *yawn*
20:34.40 IriX64 thankyou
20:34.48 ``Erik pr for fbsd ports is submitted and binary packages uploaded to sf... booyeah
20:34.52 brlcad the 12k/s sounds like a problem on your end
20:35.05 IriX64 yah prolly.
20:36.09 IriX64 ``Erik how long usually to posting for public?
20:36.10 ``Erik you got anonftp upload on a machine with that tiny of a disk, dude?
20:36.22 ``Erik irix64: what do you mean? the binaries on sf? or the pr for ports?
20:36.39 IriX64 binaries.
20:36.45 ``Erik they're already available
20:36.53 IriX64 ty
20:37.08 brlcad ``Erik: surprisingly anon isn't used much at all
20:37.08 ``Erik http://sourceforge.net/project/showfiles.php?group_id=105292&package_id=113558
20:37.23 brlcad interesting to see some of the junk that gets uploaded
20:37.30 ``Erik erm... I hope you don't permitted real accounts to ftp plain...
20:37.37 ``Erik s/ed//
20:37.41 IriX64 do you know what i tried to upload?
20:37.45 brlcad uhm
20:37.49 brlcad of course not
20:38.06 IriX64 used my name in the email address password.
20:38.08 brlcad if you try, it gives you the boot
20:38.17 ``Erik before password?
20:38.28 brlcad ooh, IriX64 ...
20:38.36 brlcad heh, i might have firewalled you
20:39.12 IriX64 what can i do sympatico insists on using your name in e-mail address well they used to anyway and im a long time member.
20:39.40 IriX64 shouldnt that be stonewalled? :)
20:40.40 IriX64 beautifull tank car from geometric solutions, someday i hope to be able to draw this well :)
20:40.54 brlcad ahh, yes.. it was you
20:40.56 brlcad connection from toronto-HSE-ppp4310781.sympatico.ca (69.159.35.87)
20:41.03 IriX64 thats me.
20:41.21 brlcad there were a ton of failed ftp timeout attempts coming from there
20:41.37 brlcad looked like it was stuck in some app loop, so I killed it
20:42.00 IriX64 last night i tried to leave it transferring and i got disconnected and my client keeps trying. shouldnt have trusted it and gone to bed.
20:42.25 IriX64 ill be carefull now.
20:42.37 brlcad yeah, it tried a couple hundred connections
20:42.43 brlcad you're unblocked
20:43.11 IriX64 and thanks.
20:43.43 brlcad meh
20:43.49 brlcad news does
20:44.15 brlcad changelog is for devs, they can figure it out
20:44.56 ``Erik and the changelog overflows 80 and has LOTS of unnecessary newlines o.O
20:50.16 CIA-9 BRL-CAD: 03brlcad * 10brlcad/HACKING: fold the Changelog to column 80 as part of the release steps to make Sourceforge display less evil
20:52.11 IriX64 rendering to the tube how slick, you guys really got it right.
20:52.51 brlcad "to the tube"?
20:52.59 IriX64 screen.
20:53.05 brlcad ah
20:53.24 IriX64 old school :)
20:56.34 IriX64 i better not turn it upside down, it'll drain all over my desk. :)
20:56.59 IriX64 ah no it wont the hatch is closed.
20:57.51 IriX64 raytrace complete lets get back to coding, see you people later.
21:00.41 brlcad sometimes best to just not ask
21:12.21 ValarQ oh, ok then
21:16.47 brlcad ValarQ: that wasn't meant for you :)
21:17.22 ValarQ yay :)
21:18.44 ValarQ not that i have anything to ask, i still working to get comfortable with mged
21:25.01 CIA-9 BRL-CAD: 03brlcad * 10brlcad/src/librt/g_extrude.c: (log message trimmed)
21:25.01 CIA-9 BRL-CAD: prevent crashing when trying to draw extrude primitives that use empty sketch
21:25.01 CIA-9 BRL-CAD: primitives. simple statements like 'make x extrude' could cause this as well as
21:25.01 CIA-9 BRL-CAD: using any (default) empty sketch. the wireframe drawing routing was getting
21:25.01 CIA-9 BRL-CAD: stuck in an infinite loop allocating memory adding empty vlist items to the mged
21:25.04 CIA-9 BRL-CAD: display list due to the comparison against a negative decrementing index. the
21:25.06 CIA-9 BRL-CAD: code now checks for the empty sketch condition and informs the user. clean up
21:25.44 brlcad message slightly trimmed. "clean up the code a bit in the process, e.g. do something about a couple always-true if statements.. thanks ValarQ for finding this bug."
21:30.48 ValarQ is it posible to select primitives from the mged shell?
21:31.12 brlcad yes, sed primitivename
21:31.28 brlcad anything you do in the gui can be done on the command line
21:32.04 brlcad ValarQ: have you seen the MGED Quick Reference sheet? that might help learn some of the commands
21:33.13 ValarQ yes, didn't found sed thought, maybe i missed it
21:35.09 brlcad under the Editing Geometry sectino
21:35.44 ValarQ yep, only me not reading my documentation like i should
21:35.47 brlcad understandable, as mged's concept of selecting an object implies entering an edit mode
21:36.13 ValarQ like vi, i'll feel right at home :)
21:37.53 brlcad heh
21:38.37 brlcad if you like vi, mged has a vi-mode command line edit style (akin to bash's vi mode and other shells)
21:43.21 ValarQ it's not that critical, i actually use a quite unholy combination of editors
21:45.06 CIA-9 BRL-CAD: 03brlcad * 10brlcad/NEWS:
21:45.06 CIA-9 BRL-CAD: prevent crashing mged when trying to draw extrude primitives that use empty sketch primitives.
21:45.06 CIA-9 BRL-CAD: simple statements like 'make x extrude' could cause this as well as using any (default) empty sketch.
21:45.06 CIA-9 BRL-CAD: the wireframe drawing routing was getting stuck in an infinite loop allocating memory adding empty vlist items to the mged display list due to the comparison against a negative decrementing index.
21:45.09 CIA-9 BRL-CAD: the code now checks for the empty sketch condition and informs the user. thanks to ValarQ for finding this bug.
22:01.34 ValarQ ops, "ERROR: NULL struct resource pointer, file db5_comb.c, line 510"
22:01.52 brlcad yeesh
22:02.26 brlcad how'd you get there?
22:02.43 ValarQ i tried to clone an object
22:02.46 brlcad ahhhh
22:03.01 ValarQ maybe i didn't specify the name correctly
22:03.06 brlcad the clone command does not currently work on v5 geometry databases, that's a new command
22:03.12 ValarQ ok
22:03.13 brlcad it only works on v4 ones at the moment
22:03.21 brlcad unfortunately
22:03.24 brlcad it's a great command
22:03.30 ValarQ maybe you understand my amd64 question now :)
22:03.56 brlcad not sure, that problem isn't amd64-specific..
22:04.04 brlcad neither was the other :)
22:04.24 ValarQ no, i just started to suspect something like that
22:04.53 brlcad brl-cad is pretty pervasively multiplatform.. if something isn't working right it almost guaranteed to be a bonefide cross-platform logic bug of some sort
22:05.31 ValarQ what languages is it written in? only C and tcl?
22:06.14 brlcad at the moment, primarily C
22:06.33 brlcad the mged and archer modelers are written in tcl/tk and itcl/itk
22:06.59 brlcad there are several C++ projects under way as well, but not in the core atm
22:07.39 brlcad there's actually a couple resources, such as the benchmark suite, that are written in posix bourne shell
22:07.42 ValarQ ok, there isn't many C++ source files in the cvs head
22:08.08 brlcad as well as a simple geometry database server written in php for web apps, and a java JNI interface for interacting with the raytrace library
22:08.38 brlcad yeah, most of the c++ work is in other modules and a couple not in CVS yet
22:09.10 brlcad the next generation solid modeler interface will be written primarily in C++
22:09.39 ValarQ but the core will remain C?
22:10.29 brlcad that's the current plan
22:11.08 brlcad the long long term plan is to have a C++ library layer that will eventually subsume the C core where it makes sense
22:11.37 ValarQ ok, big move to C++ then
22:11.58 brlcad "eventually", that's very long term
22:12.00 ValarQ no plans to use any higher level languages then?
22:12.39 brlcad there are, that same core written in C++ is extensively plugin-based with support for multiple higher level execution cores
22:13.16 ValarQ sounds good
22:13.17 brlcad currently planning on python, tcl, and shell, possibly considering lisp and perl down the road
22:13.28 brlcad (from the onset)
22:13.51 ValarQ scheme would be nice :)
22:14.08 brlcad heh, everyone has a favorite :)
22:14.20 brlcad that'll be hugely driven by developer interest
22:14.47 brlcad if someone wants to pick up ML or Smalltalk and they're involved enough, I certainly wouldn't object ;)
22:15.26 brlcad I'm just not going to think about them much without getting the higher-impact ones done first
22:15.56 brlcad lisp/scheme would both be good since it could capture some of the autocad background
22:17.17 ValarQ i'm a bit concerned...
22:17.31 brlcad not looking to create an academic project for the sake of language coolness, the user community is coming in with specific experience that we'd like to leverage of course where possible
22:17.38 *** join/#brlcad Lapo (n=kvirc@81-208-74-176.ip.fastwebnet.it)
22:17.44 Lapo Hi all :-D
22:17.46 brlcad concerned with?
22:17.47 ValarQ i know many of those languages has good C bindings but poor bindings for OOP languages like C++
22:18.22 ValarQ but that might be an effect of using C++ OOP-tools in the API
22:18.40 brlcad ValarQ: from a plugin-system perspective, all the higher level languages can be treated as scoped/classed/namespaced command sets
22:18.57 brlcad howdy Lapo
22:19.06 ValarQ i know for example that it's extremely easy to write Haskell wrappers for C libraries but not so easy for C++
22:19.20 Lapo hey op ;-)
22:19.49 brlcad ValarQ: true, though that is really mitigated by the architecture in other ways
22:20.00 Lapo this is a very "intelectual" conversation :-D
22:20.20 ValarQ well, i can start worrying about it when i start coding it :)
22:20.34 brlcad plain C wrappers would work just fine too as most all of the commands are striving to be transaction-based and mostly stateless
22:20.41 ValarQ Lapo: nah, not "very" :)
22:21.01 Lapo yes , to me... ;-)
22:21.39 Lapo Guys I have a little problem about brlcad
22:21.48 Lapo maybe you can help me...
22:21.55 Lapo I'm a big nood -.-
22:22.14 brlcad some services such as geometry services will be client/server based so the higher-level commands won't really need to worry about C++ per-se for a large portion of operations
22:23.45 ValarQ using the old trusted unixmethod called pipes then :)
22:24.03 brlcad for a lot of functionality yes ;)
22:24.34 brlcad brl-cad already has over 400 commands that do a huge variety of functionality (that is not exposed by mged even now)
22:24.56 Lapo ehm...
22:24.59 brlcad those become plugins in the new system almost unmodified so they can be leveraged without rewriting a million lines of code
22:25.12 brlcad Lapo: best to just ask your question instead of asking if you can ask..
22:25.12 ValarQ Lapo: just throw the question at brlcad
22:25.21 Lapo ok
22:25.23 brlcad ~ask
22:25.24 ibot Questions in the channel should be specific, informative, complete, concise, and on-topic. Don't ask if you can ask a quesiton first. Don't ask if a person is there, just ask what you intended to ask them. Better questions more frequently yield better answers. We are all here voluntarily. See also http://catb.org/~esr/faqs/smart-questions.html
22:25.41 Lapo ouch sorry guys
22:25.46 brlcad no problem ;)
22:25.58 brlcad everyone's allowed a freebie or two
22:26.16 Lapo this is my problem when I start mged:
22:26.47 Lapo Initializing and backgrounding, please wait...mged: symbol lookup error: mg
22:27.01 Lapo ed: undefined symbol: Tk_CreateCanvasBezierType
22:27.08 ValarQ better get some sleep, it's getting rather late over here...
22:27.10 brlcad ah
22:27.14 Lapo (please don't kill me)
22:27.16 ValarQ bye folks
22:27.21 brlcad ValarQ: nice chatting with you
22:27.22 brlcad cheers
22:27.27 Lapo bye ValarQ
22:27.56 brlcad Lapo: you're linking against a system Tk (or at least lacking libtclcad)
22:28.12 brlcad you compiled brl-cad yourself?
22:28.42 Lapo no
22:28.56 brlcad odd, where'd you get the binary from?
22:29.39 Lapo I got the tar.gz from the web site , then unpacked it
22:30.05 Lapo then compiled , and then I've copied in /usr/brlcad
22:30.42 Lapo probably this is not the right way X-)
22:30.44 brlcad which tar.gz
22:31.03 brlcad so you did compile brl-cad yourself..
22:31.42 Lapo brlcad-7.6.0_linux_ia32.tar.bz2
22:32.42 brlcad okay, so you didn't compile, you unpacked.. installed to /usr/brlcad
22:32.48 brlcad do you have a LD_LIBRARY_PATH set?
22:32.50 Lapo right....
22:33.27 Lapo It seems it does not exist
22:33.41 brlcad echo $LD_LIBRARY_PATH
22:33.58 Lapo yes, but it return an empty string
22:34.05 Lapo so I think it doesn't exist
22:34.20 brlcad k
22:34.26 brlcad hmm..
22:34.51 brlcad any particular reason you're useing 7.6.0 and not 7.8.0?
22:35.35 Lapo I got that file some month ago but I can download newer version
22:35.49 brlcad yeah, that would help
22:35.58 brlcad a ton of stuff has changed and improved
22:36.20 Lapo I hope installation is easier now :-D
22:36.42 Lapo I'm a mechanical engineer and wish to use the program
22:37.57 Lapo thanks for the help, see you ;-)
22:38.38 *** part/#brlcad Lapo (n=kvirc@81-208-74-176.ip.fastwebnet.it)
22:58.20 CIA-9 BRL-CAD: 03brlcad * 10brlcad/src/librt/g_extrude.c: unbreak the wireframe drawing logic.. something wierd is still going on. also, try even harder to avoid crashing on a bad index by making sure we don't index past the bn_vlist points array.
23:25.07 CIA-9 BRL-CAD: 03brlcad * 10brlcad/src/librt/g_extrude.c: oop, remove the ifdef block since testing is done
23:29.51 *** join/#brlcad PrezKennedy (n=Apathy@c-68-33-243-45.hsd1.md.comcast.net)

Generated by irclog2html.pl by Jeff Waugh - find it at freshmeat.net! Modified by Tim Riker to work with blootbot logs, split per channel, etc.