IRC log for #brlcad on 20080327

00:01.37 *** join/#brlcad Twingy (n=justin@74.92.144.217)
00:02.05 nmh_2Grajw oh, hi
00:06.47 *** join/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1178014867.dsl.bell.ca)
00:23.33 brlcad nmh_2Grajw: have a particular question, just lurking? :)
00:24.05 nmh_2Grajw brlcad: if it is not a problem, both :)
00:24.15 brlcad not a problem in the least
00:31.38 *** join/#brlcad Daytona (n=jra@c-68-55-36-65.hsd1.md.comcast.net)
00:51.54 CIA-33 BRL-CAD: 03brlcad * r30570 10/brlcad/trunk/include/rtserver.h: remove duplicate decls, sort lines
00:52.31 louipc brlcad: no problemo
00:57.38 nmh_2Grajw hmm.. I am trying to export a single shape database to ascii, but I can't seem to find the file. Is there anything special I should know about this?
00:59.15 Daytona the mged export asks for a file name, is that how you did it?
01:00.21 nmh_2Grajw Daytona: yeah (the menu and dialog box). I tried "test.asc" , "/tmp/test.asc" and others
01:01.28 Daytona another way to do it is from the command line: g2asc file.g file.asc
01:03.33 nmh_2Grajw Daytona: hmm, now I feel dumb. I tried g2asc and (thought it didn't work) I just went to doublecheck and I have test.asc sitting right there in the directory.
01:04.54 Daytona I just tried saving from mged and it worked for me. The asc file appeared in thwe same directory where I started mged
01:05.42 nmh_2Grajw Daytona: yeah, that is what I see now. Thanks!
01:05.52 Daytona you're welcome
01:07.57 nmh_2Grajw continuing with the easy stuff - when I click the mouse button in the graphics window, the view zooms out - changing stuff in the mouse behavior menu doesn't change this. Is there an easy way for me to fix this?
01:10.30 CIA-33 BRL-CAD: 03brlcad * r30571 10/brlcad/trunk/TODO: dbconcat verified, seems to work like a charm with default, -s, -p with or without affix
01:10.52 Daytona the settings->Mouse Behavior menu mostly applies to middle button clicks
01:13.39 Daytona In fact, I think it only applies to middle button
01:14.27 nmh_2Grajw Daytona: I guess I will just be careful, since I only have one mouse button.
01:15.12 Daytona Yes, the interface is designed for a 3 button mouse
01:15.27 brlcad heya Daytona
01:15.45 Daytona yo
01:16.16 brlcad nmh_2Grajw: the mouse bindings as they are are called the "shift grips", you can get different behaviors depending on whether you're clicking with various modifiers pressed as well
01:17.02 Daytona Yes, perhaps brlcad can provide advice for users with a one button mouse?
01:17.05 brlcad I presume you're on a mac -- if you cmd-click, it should emulate the same event as a left-click
01:17.26 brlcad likewise, option-click will give you middle-mouse
01:17.51 Daytona I knew brlcad would know about those macs
01:17.55 brlcad (cmd == the command key, i.e. the key with the apple on it)
01:18.10 brlcad heh
01:19.21 nmh_2Grajw neato (getting something useful out of the mouse)
01:19.30 brlcad shift-control-click will give you dynamic zoom, shift-click will pan, ctrl-option-click will rotate
01:20.34 brlcad control-option-click is probably what you're used to, if anything
01:21.02 nmh_2Grajw yes, control-option-click == click, it seems
01:22.00 brlcad control-option-click(and hold depressed while moving the cursor)
01:23.57 brlcad ah, sorry, made that more complicated -- it's the same as just control-click(keep depressed and move cursor)
01:32.43 nmh_2Grajw ooc, how important is white space in the ascii database files? (does each put need to be on a single line, or could I have each vertex on its own line?)
01:33.32 Daytona The ascii file is just a Tcl script, so it follows the same syntax
01:35.49 Daytona you can extend to another line by escaping the newline
01:36.41 nmh_2Grajw ahh, good to know (both the Tcl and the newline)
01:58.40 nmh_2Grajw I am going to go off and poke at this some more, thank for the help!
01:59.11 brlcad nmh_2Grajw: sounds good, drop by any time
02:02.55 *** part/#brlcad nmh_2Grajw (n=nmh@gw.nomh.org)
03:21.24 *** join/#brlcad pacman87 (i=127@resnet-45-192.dorm.utexas.edu)
03:55.45 pacman87 after the source svn and compiling (25 minutes later), the last several lines are errors
03:55.56 pacman87 *getting the source
03:57.05 brlcad hey pacman87
03:57.12 brlcad pacman87: which os?
03:57.16 pacman87 slackware 12
03:57.29 brlcad hm, haven't seen that in a long while -- what's the error?
03:57.35 brlcad ~pastebin
03:57.36 ibot [~pastebin] A "pastebin" is a web-based service where you can paste anything over 3 lines without flooding the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://www.rafb.net/paste
03:57.49 brlcad er, paste.bzflag.bz
03:58.01 pacman87 http://pastebin.com/m4dfb88ac
04:01.34 brlcad need more than that
04:01.43 brlcad lines preceeding
04:01.46 pacman87 how far back?
04:01.52 brlcad to the compile line
04:01.56 brlcad gcc
04:02.03 pacman87 i ran make
04:02.22 brlcad yeah, but if you look up the preceeding lines, one of them mentions gcc
04:02.30 brlcad from there forward
04:02.32 pacman87 if gcc...
04:03.23 pacman87 http://pastebin.bzflag.bz/m8c1030c
04:04.01 brlcad ahh
04:04.05 brlcad what was your configure line?
04:04.19 brlcad looks like it auto-detected a system tcl
04:04.25 pacman87 ./configure --enable-optimized
04:04.31 brlcad but not a system incrTcl
04:04.44 brlcad 8.4 is incompatible with our incrTcl
04:04.56 brlcad make that ./configure --enable-optimized --enable-aa
04:05.00 brlcad er, --enable-all
04:05.26 pacman87 does --enable-optimized incease compile tim?
04:05.29 pacman87 time
04:07.01 brlcad heh
04:07.05 brlcad no
04:07.09 brlcad quite the contrary
04:07.31 brlcad er, sorry -- yes it does
04:07.47 brlcad read that backwards :)
04:07.52 pacman87 by how much?
04:08.03 brlcad about doubles it on most systems
04:08.23 brlcad likewise turning it on will give about a rough 2X runtime boost
04:08.30 brlcad (wrt ray-tracing)
04:08.37 pacman87 ok, so i'll take it off for now
04:08.44 pacman87 dont' want to wait another 25 min
04:08.53 brlcad yeah
04:09.11 brlcad sounds like you're machines not too spiffy too .. bwish is probably only about 1/3 through the compile
04:09.16 brlcad which is where it stopped
04:09.21 pacman87 wow
04:09.42 brlcad tis a big package, more than a million lines
04:09.57 brlcad lots of functionality
04:10.10 pacman87 it's a 2.8 GHz P4, no hyperthreading, 533MHz fsb, 1280 MB ram PC2700
04:10.56 brlcad yeah, mildly aged .. but that shouldn't take more than a half-hour I'd think
04:11.30 brlcad (with opt off, with it on depends on many factors)
04:12.13 pacman87 i'm interested in applying for GSoC
04:12.24 brlcad excellent
04:12.36 pacman87 3rd year ME student at UT austin
04:13.13 brlcad so where's your interest?
04:13.31 pacman87 i've been lookign at the new primitives section
04:13.45 brlcad cool
04:14.03 pacman87 even though that's not one of the 'high priority' ones on the list
04:14.11 pacman87 it's probably the one i'd be best at
04:14.18 brlcad a good application overrides the priorities
04:14.38 brlcad those priorities are project priorities, not applicant priorities
04:14.43 pacman87 i noticed your app guidelines "do
04:14.52 pacman87 's", says "play the game"
04:14.59 brlcad heh, it does?
04:15.02 pacman87 yeah
04:15.08 pacman87 second paragraph
04:15.09 brlcad I thought I got all those taken care of
04:15.11 brlcad thanks
04:15.29 pacman87 i suppose it could be taken more metaphorically
04:15.33 pacman87 the programming game
04:15.37 brlcad hehe
04:16.12 brlcad our project admin is also the project admin for bzflag
04:16.22 brlcad so the texts are ... really similar :)
04:16.32 brlcad nice that you noticed though, seriously
04:17.02 pacman87 finished! 6m20s
04:17.25 brlcad wow, that's one seriously expensive optimization loop
04:17.34 pacman87 no kidding
04:17.35 brlcad did you make clean or just make?
04:17.39 pacman87 make
04:17.41 brlcad okay
04:17.45 brlcad so that saved you some time
04:17.54 brlcad you'll have a mix of optimized and unoptimized
04:18.05 brlcad (which is not a problem)
04:19.08 brlcad means you probably only had about 15 minutes left, your optimized compile is probably about 30-40 minutes it sounds
04:19.16 pacman87 how are the current primitives handled?
04:19.20 hippieindamakin8 hey Sean
04:19.22 brlcad unoptimized is probably 15-25
04:20.03 brlcad pacman87: the librt library is the primary harbor for the primitives -- that's where they are predominantly defined
04:20.11 brlcad src/librt/g_*.c
04:20.28 pacman87 ok, i'll start my code reading there
04:20.41 brlcad there's a g_*.c for each primitive type with what should be fairly obvious callbacks
04:20.54 hippieindamakin8 hey Sean i wen through some files of jbrlcad
04:20.57 brlcad g_xxx.c is a stubbed version iirc
04:20.58 brlcad hey hippieindamakin8
04:21.05 hippieindamakin8 and i seem to get a hang of that problem statement
04:21.29 brlcad hippieindamakin8: after you read through jbrlcad, compare that to what's going on in src/librt
04:21.38 brlcad in the main brlcad module
04:21.40 hippieindamakin8 sure
04:21.59 brlcad you'll see that there is LOTS of overlap, it's designed off of the current C api (as we're already rather modular)
04:22.21 hippieindamakin8 i am rt now understand the optimised ray tracing stuff too :P
04:22.31 hippieindamakin8 ok..
04:22.55 brlcad pacman87: the other two places to look are src/libwdb where there is an mk_*() function for each primitive type to create them procedurally, and src/mged/typein.c and a few others have some of the editing/input methods for creating them
04:24.50 brlcad pacman87: you also considering applying to bz?
04:25.24 pacman87 i'm not really sure what i'd do there
04:26.12 brlcad k, just checking
04:26.27 brlcad now I realize why your nick was so familiar though :)
04:26.35 brlcad been there since last winter iirc
04:26.50 pacman87 right, that's when i started doing plugins
04:35.48 pacman87 hmm, running /usr/brlcad/bin/mged opens the window and takes 96%cpu to do nothing - no menus open
05:05.23 brlcad odd, not seen that before
05:05.33 brlcad what version are you using?
05:05.49 pacman87 7.12.1
05:06.00 brlcad probably something really recent
05:06.19 brlcad related to the non-backgrounding we do now
05:06.32 brlcad or it's actually receiving a constant stream of events
05:06.42 brlcad would have to profile/debug it
05:07.24 pacman87 what would you recommend using?
05:08.10 brlcad gprof, valgrind, oprofile
05:08.29 brlcad gprof is pretty simple, if installed, you could recompile with --enable-profile
05:08.43 brlcad and it'll profile it by default
05:09.17 pacman87 man gprof
05:11.40 brlcad basically involves adding -pg to the compile and link lines -- then you just run the binary once it's installed and it'll dump out a gmon.out file .. then you run gprof /usr/brlcad/bin/mged in the same dir as that gmon.out file and it'll generate a report
05:12.08 brlcad you could also run mged inside gdb, and just ctrl-c a few times to see where in the stack trace it is
05:13.08 pacman87 i've been telling myself to learn to use a debugger for a while now, but all of my programs have been small enough that printf() is sufficient
05:13.46 brlcad very useful skill to learn
05:14.02 hippieindamakin8 same here
05:16.57 pacman87 <PROTECTED>
05:16.57 pacman87 <PROTECTED>
05:40.10 brlcad so run, let it go for a couple seconds, then quit? :)
05:40.33 brlcad it should be exiting normally
05:41.15 pacman87 http://pastebin.bzflag.bz/m60fc5897
05:41.19 pacman87 from gdb
05:42.38 brlcad mged's not hung is it?
05:42.43 brlcad does it respond?
05:42.45 pacman87 no
05:42.55 pacman87 *no response
05:42.56 brlcad no to which?
05:43.01 brlcad ah, so it's hosed
05:43.12 brlcad it'll be hosed once you ctrl-c in mged
05:43.16 brlcad s/hosed/paused/
05:44.12 brlcad run "mged -c" .. and select 'nu' .. does that use cpu?
05:45.11 pacman87 no
05:45.57 pacman87 i get the standard "mged>" prompt, which responds normally
05:48.59 brlcad okay, that's good
05:51.29 brlcad next thing to try, same -c but select X
05:53.20 pacman87 standard "mged>" in terminal, plus a dm_X0 window, and cpu is fine
05:55.08 brlcad well that's enough to work on a new primitive *ahem* :)
05:55.33 pacman87 i first tried brlcad 3-4 years ago
05:55.43 pacman87 lookign for a free/open source CAD program
05:55.57 brlcad so mged is non-functional when you start it with no options, using full-cpu
05:56.03 brlcad do the menus respond at all?
05:56.08 pacman87 not at all
05:56.19 brlcad k
05:56.40 pacman87 but i got scared away by the command line (this was before i started teaching myself linux)
05:58.24 CIA-33 BRL-CAD: 03brlcad * r30572 10/brlcad/trunk/BUGS: mged seems to be locking up at least on slackware, taking up full cpu and being non-responsive. works find with -c and X dm. reported by pacman87 (thx).
05:58.52 brlcad yeah, our command-centric approach is really frightening to some
05:59.05 brlcad hence the ton of priority focus on revamping the gui
05:59.15 pacman87 UT sold their soul to SolidWorks
05:59.34 brlcad yeah
06:00.01 hippieindamakin8 i asked my professor here to use brlcad and he said NO :P it aint user friendly
06:00.16 pacman87 i'm thinking the revolve would be easiest to do
06:00.31 brlcad solidworks, unigraphics/nx, pro/engineer, and catia consume most of the cad domain with autocad dominating drafting
06:00.34 hippieindamakin8 its the same everywhere ppl use solidworks
06:01.07 hippieindamakin8 ya btw i used pythoncad before realising the existance of brlcad
06:01.15 pacman87 i use solidworks at my internship/part time job at ARLUT
06:02.21 hippieindamakin8 and i was asked everywhere to use.. autocad :P
06:04.21 hippieindamakin8 wat do u work on pacman ?
06:05.24 pacman87 autonomous surface vehicle for sonar testing
06:06.30 brlcad hippieindamakin8: and what did you think of pythoncad
06:07.59 hippieindamakin8 firstly it is good enough for drafting
06:08.35 brlcad really think so? I don't get that feeling
06:08.47 hippieindamakin8 very simple and good interface for the students
06:09.10 hippieindamakin8 that is wat i felt.. but yeah.. brlcad is most optimised
06:09.41 hippieindamakin8 it has so far the best real time ray tracing as i have read
06:10.25 hippieindamakin8 pythoncad is good enough for amatuers.. :) all my projects so far have been done on pythoncad after my freshman year in college
06:12.19 hippieindamakin8 the problem is that it cant export to any standard stuff
06:12.50 hippieindamakin8 i meant standard format... wat do u say brlcad
06:16.05 brlcad a very immature project with very little practical features, but a nice interface and reasonable features for the start of drafting purposes
06:17.00 hippieindamakin8 if u ask me to rate these i would rate pycad4.5 and brlcad 7
06:19.44 hippieindamakin8 cya guys.. time for the class.. i shall be back in 6 hrs
06:19.51 brlcad cya
06:47.14 brlcad waves g'night
06:48.34 brlcad pacman87: one more thing to try, you can run in gdb, set a breakpoint on main or some other early-executed function, and then single-step till things go awry
06:49.11 brlcad might get tricky as there's a tcl eval loop that kicks in, but with some patience you would eventually get to the line of interest
06:49.32 pacman87 i'll have to do some more research on debuggers first
06:49.44 brlcad you can use continue and breakpoints to get to the point in the code that is problematic
06:49.46 pacman87 thanks for your help
06:49.52 brlcad no problem
06:50.13 pacman87 definately a steep learning curve, though (cmd line)
06:51.37 brlcad useful gdb commands are "b file:line" or "b function" to set a breakpoint, "run" to run, "c" to continue, "l" to list the sources where you're at, "l line" or "l file:line" to list the sources in a specific place, "p variable" to print it's value, "up" and "down" to go up and down the call stack
06:52.37 brlcad "n" to execute the next statement, "s" to step 'into' a function
06:52.50 brlcad i think some of the common ones are summarized on the manual page iirc
06:53.08 pacman87 yeah, and it's got a pretty nice 'help' feature
06:53.35 pacman87 contemplates solving this as his 'bugfix'
06:53.50 brlcad would be impressive
06:54.28 brlcad and shouldn't be too hard .. might even be able to back up in time through the svn revisions and find a version that was working, find the commit that broke it
06:54.37 brlcad bets it was within the last two months
06:56.05 pacman87 i'm builing bzflag 2.0.11 and 2.99 now so i can run my servers from linux
06:56.42 pacman87 i dual boot winXP
06:57.15 pacman87 2am, goodnight
11:01.10 *** join/#brlcad Elperion (n=Bary@p54874A45.dip.t-dialin.net)
11:36.24 *** join/#brlcad d_rossberg (n=rossberg@bz.bzflag.bz)
11:53.42 *** join/#brlcad Elperion (n=Bary@p54874A45.dip.t-dialin.net)
12:35.50 *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch)
13:41.46 brlcad notes there's another submission
13:41.54 brlcad (and it needs more detail)
15:48.18 brlcad d_rossberg: have you tried the 7.12.0 installer?
15:53.41 *** join/#brlcad docelic (n=docelic@78.134.200.134)
15:55.07 d_rossberg brlcad: i can see the source download on sourceforge only, where can i find the installer?
15:58.06 brlcad ah, sorry
15:58.12 brlcad it's on the anonymous ftp site
15:58.19 brlcad ftp://ftp.brlcad.org/incoming
15:59.27 brlcad we weren't going to distribute binaries for this .0 release until it was better tested .. and with good reason seeing that wim encountered problems
16:01.16 d_rossberg i'm sorry but i'm using linux at home, on monday i'll be back in my office
16:02.57 d_rossberg but then i can do (probable) a complete test with a new visual studio
16:03.44 brlcad okay, no problem
16:33.43 brlcad wonders if Balakrishnan is the guy that was here last night
16:53.58 cosurgi brlcad: I got new GPU today, no more unplanned reboots (hopefully). My old GeForce 6600 had lost its fan one year after warranty expired, and my PC was instatenously rebooting once GPU temperature reached 110 C
16:54.49 cosurgi so today I bought fanless graphics card :) the cheapest one with dual-DVI, GeForce 8600 GT
16:55.45 cosurgi That GPU without a fan was a good method to prevent me from playing games. But it started to interfere with yade development (I'm using OpenGL there) so finally I had to buy a new one.
16:55.56 cosurgi (-:
17:00.06 *** join/#brlcad kannan (i=kannan@117.196.133.89)
17:01.13 brlcad cosurgi: cool, that's a nice card
18:51.46 *** join/#brlcad hippieindamakin8 (n=hippiein@203.200.95.130)
18:52.47 hippieindamakin8 hey Sean .. this OO stuff is awesome man.. i am too much interested in this project :) i shall send in a proposal as soon as i wake up
19:06.55 d_rossberg hippieindamakin8: yes, make a claim for this topic, it makes the discussion easier
19:08.45 d_rossberg there are some attempts you can use
19:16.42 hippieindamakin8 i have been going through the jbrlcad module and the librt libraries
19:17.42 hippieindamakin8 and i am seriously interested in implementing the OO geometry APIs. .. some of them have already been implemented :)
19:18.07 hippieindamakin8 d_rossberg : u are a mentor too ?
19:19.38 d_rossberg yes
19:19.48 hippieindamakin8 naice :)
19:20.44 hippieindamakin8 i have kind of figured out how the geometry has been rendered so far
19:22.47 hippieindamakin8 and i would like to code in JAVA itself to continue to build the module .. i am strongly familiar with both C++ and JAVA
19:24.49 d_rossberg i have written a c++ interface for my company's purpose, and now we want to implement a full interface as open source
19:25.18 hippieindamakin8 company as in ?
19:25.59 hippieindamakin8 at the same time u want to make it object oriented framework
19:28.03 d_rossberg i would recommend to implement a c++ api first and use this as basis for other languages (e.g. i build a COM interface to use BRL-CAD in MS Excel)
19:31.50 hippieindamakin8 COM interface what is it ?
19:32.42 hippieindamakin8 BRLCAD in MS Excel interesting.. wat exactly u gonna do .. send some csv s or smthing ?
19:32.52 hippieindamakin8 COM do u mean the serial port
19:33.22 hippieindamakin8 k :)
19:38.45 *** join/#brlcad hippieindamakin8 (n=hippiein@203.200.95.130)
19:45.01 *** join/#brlcad Z80-Boy (n=clock@77-56-79-120.dclient.hispeed.ch)
19:57.29 *** join/#brlcad Z80-Boy (n=clock@77-56-79-120.dclient.hispeed.ch)
19:57.31 louipc COM > http://en.wikipedia.org/wiki/Component_Object_Model
19:57.34 louipc methinks
19:58.24 hippieindamakin8 :( thanks
19:58.33 *** join/#brlcad clock_ (n=clock@77-56-79-120.dclient.hispeed.ch)
20:20.41 d_rossberg COM: this was a serious question?
20:23.18 hippieindamakin8 i knew that it was called DCOM in linux and the serial port was called COM port here :(
20:24.34 d_rossberg DCOM is the network version of COM (similar to CORBA)
20:26.21 hippieindamakin8 oh
20:29.00 *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6)
20:37.33 d_rossberg if you want to write MS Windows modules COM maybe a good choice: they can be used im Visual Basic (for applications too), some commercial CADs are offering a COM interface too
20:39.53 hippieindamakin8 ohh :)
21:28.44 *** join/#brlcad jdoliner (n=jdoliner@cpe-24-59-109-153.twcny.res.rr.com)
21:28.56 jdoliner hello
21:31.02 jdoliner anyone here
21:38.33 *** join/#brlcad pacman87 (i=127@resnet-45-192.dorm.utexas.edu)
21:38.54 hippieindamakin8 hello jdoliner
22:20.56 *** part/#brlcad louipc (n=louipc@bas8-toronto63-1096669639.dsl.bell.ca)
23:18.07 *** join/#brlcad vedge (n=vedge@205-237-251-209.ilesdelamadeleine.ca)
23:35.58 *** join/#brlcad Bariton (n=Bary@p54877B55.dip.t-dialin.net)
23:47.15 *** join/#brlcad Twingy (n=justin@74.92.144.217)

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