IRC log for #brlcad on 20120613

00:56.54CIA-55BRL-CAD: 03starseeker * r51083 10/brlcad/trunk/src/tclscripts/rtwizard/rtwizard.bat: More tweaks to rtwizard.bat
01:44.18CIA-55BRL-CAD: 03brlcad * r51084 10/brlcad/trunk/src/mged/mged.bat: there is no -g option to mged any more. plus, pass all arguments to mged if any (same as rtwizard.bat) instead of just the first one.
01:44.46CIA-55BRL-CAD: 03brlcad * r51085 10/brlcad/trunk/src/archer/archer.bat: follow suit with rtwizard.bat, pass all arguments to archer. not just the first.
01:53.29CIA-55BRL-CAD: 03brlcad * r51086 10/brlcad/trunk/src/librt/primitives/bot/tie.c: memset for simplicity and more future-proof
01:55.00brlcadstarseeker: did you notice bad behavior with the EXIT line in the bat file?
01:55.10brlcador just the principle
01:55.54CIA-55BRL-CAD: 03brlcad * r51087 10/brlcad/trunk/src/adrt/isst.bat: consistency with the other .bat scripts, pass all arguments instead of just the first
01:56.38brlcadhard to believe it's necessary, in which case it should be removed from all of our .bat files for consistency
02:04.45starseekerif I understand correctly, EXIT is used when we're launching something we know is graphical - that avoids leaving unwanted consoles up and running
02:05.27starseekeron an application that might be command line only (i.e. there's a decent chance of its getting launched from the command prompt in Windows) you don't want it, 'cause it will kill your console out from under you
02:05.32starseeker(happened to me)
02:06.56starseekerit may be moot - I'm not entirely sure, but I might have to make rtwizard into a full-fledged exe ala mged to allow libged's rtwizard call to work
02:08.06brlcadk
02:08.45starseekerneed to consult with Bob tomorrow
02:10.26starseekerworking fine on Linux and friends, but (as usual) Windows is being annoying
02:17.11starseekerhmm... spoke too soon
02:17.15starseekergrumble...
02:33.26*** join/#brlcad KimK (~Kim__@2001:470:1f0f:1042:4261:86ff:fe43:bcad)
03:10.25*** join/#brlcad KimK (~Kim__@2001:470:1f0f:1042:4261:86ff:fe43:bcad)
03:28.30*** join/#brlcad jbschw (~jbschw@ool-4355ee10.dyn.optonline.net)
03:35.37*** join/#brlcad xth1 (~thiago@187.106.53.3)
07:33.07*** join/#brlcad d_rossberg (~rossberg@BZ.BZFLAG.BZ)
09:21.06anrgmrtyd_rossberg: hi.. i can now shoot multiple rays per row.. as in for each row and parameter called raysPerVoxel i have raysPerVoxel x raysPerVoxel number of rays uniformly distributed on the face..
09:21.30anrgmrtyobviously raysPerVoxel will be renamed raysPerRow or something..
09:21.56anrgmrtyas the output i get a text file-- 0's and 1's..
09:26.24d_rossbergyou should prefer something like "level of detail" as this value isn't the number of rays
09:26.57anrgmrtyok.. I will use that as the variable name..
09:32.26anrgmrtycan you please give me pointers on what I could do next? maybe a text file is not the right thing.. a file with ascii characters instead??..
09:32.26anrgmrtyor probably having these voxels as a .g file?
09:32.55anrgmrtyi am not sure of how the second will be implemented though..
10:25.28*** join/#brlcad KimK (~Kim__@2001:470:1f0f:1042:4261:86ff:fe43:bcad)
10:32.29d_rossberganrgmrty: there are still some issues in your algorithm but i need some time to write them down
10:32.51d_rossberghowever, maybe you can do it by yourself too
10:35.02d_rossbergchange your program that it writes for every voxel its coordinated (indices), the regions in it and the fractions of these regions in the voxel
10:35.41d_rossberge.g. something like "(5,3,7): /all.g/box.r 76%"
10:36.25d_rossberg"         /all.g/sphere.r 40%"
10:36.53d_rossberg"(5,3,8): /all.g/box.r/90%"
10:36.59d_rossbergand so on
10:38.14d_rossbergafter you've developed the algorithm you should move it into a lib
10:39.13d_rossbergthis algorithm could have a call-back function as parameter
10:40.23d_rossbergand this call-back could be called for every (filled) voxel with the list of regions in this voxel together with the regions' fractions
10:41.35d_rossbergthen you could go to libged and implement a command for mged where the call-back creates arb8s (for example)
11:27.31CIA-55BRL-CAD: 03tbrowder2 * r51088 10/brlcad/trunk/TODO: add list of 'bad' man pages
11:30.42CIA-55BRL-CAD: 03tbrowder2 * r51089 10/brlcad/trunk/src/ (28 files in 10 dirs): correct errors found by Eric Raymond's doclifter, tidy some spacing and formats
11:49.26anrgmrtyd_rossberg: i am sorry i was away from desk..
12:45.14CIA-55BRL-CAD: 03starseeker * r51090 10/brlcad/trunk/src/tclscripts/rtwizard/rtwizard.tcl: Fix rtwizard.tcl for case where multiple .g args are specified (toplevel components may have a .g name)
13:01.41CIA-55BRL-CAD: 03starseeker * r51091 10/brlcad/trunk/src/libged/rtwizard.c: Fix perspective handling in rtwizard.c
13:03.07CIA-55BRL-CAD: 03bob1961 * r51092 10/brlcad/trunk/src/libged/rtwizard.c: Needs rtwizard.bat for windows.
14:14.41*** join/#brlcad KimK (~Kim__@209.248.147.2.nw.nuvox.net)
15:15.56CIA-55BRL-CAD: 03n_reed * r51093 10/brlcad/trunk/src/other/step/src/fedex_plus/ (classes.c classes.h classes_wrapper.cc multpass.c): break up generated strings; SCL git 212c114
16:09.01starseekerhmm - should look at this if I ever get to revisit the tire tool... http://www.dtic.mil/docs/citations/AD0702243
16:13.51CIA-55BRL-CAD: 03bob1961 * r51094 10/brlcad/trunk/src/libicv/fileformat.c: Added an O_BINARY flag to the call to open() in icv_image_save_open(). This fixes a breakage on windows.
16:27.42CIA-55BRL-CAD: 03starseeker * r51095 10/brlcad/trunk/NEWS: Bob fixed a bug in libicv that was causing pix files to be generated with corrupted data on Windows.
16:33.51CIA-55BRL-CAD: 03n_reed * r51096 10/brlcad/trunk/src/other/ (9 files in 4 dirs): Apply SCL git 3fd2222 which removes scl_char_str_list class, but std::vector makes a better replacement than std::deque.
16:57.38brlcadstarseeker: so is that the last issue?  what was the nmg problem?
17:17.01CIA-55BRL-CAD: 03n_reed * r51097 10/brlcad/trunk/src/other/step/src/fedex_plus/ (classes.c classes.h classes_wrapper.cc): eliminate a temp buffer; SCL git 1cc2774
17:18.51*** join/#brlcad cristina (~quassel@188.24.50.251)
17:18.56cristinahello
17:19.07brlcadhi cristina :)
17:19.14*** join/#brlcad cristina (~quassel@unaffiliated/cristina)
17:23.07cristinabrlcad: as you might know, I have roughly implemented something that generated a rectangle for each object in the database. I've used some source code from src/conv/g-dot.c file. My mentor advised me to not add it into the src/conv folder, instead, maybe, put it into src/gtools.
17:23.35cristinawhat would the proper place for this kind of implementation be?
17:24.25cristinawhat the file is meant to do is construct a structure based on adaptgrams that represents the graphical structure of the database
17:26.50brlcadcristina: so a couple cursory questions ... where did you discuss this with your mentor?? :)
17:27.25cristinabrlcad: by email :)
17:32.44brlcadcristina: okay, great -- I ask because it sounds a little curious that you implemented code to generate a rectangle for each object in the db
17:33.03brlcadthat sounds like something we already have code for, maybe a good case for reuse
17:34.27cristinahm, I meant to say that I implemented it using adaptagrams defined structures and then exported them in a .svg file. Finally, I visualized the data in the .svg file
17:35.27brlcadpics? :)
17:36.36cristinaok, I will show you but I can't right now because in the last two days I didn't have access to that computer :)
17:37.02brlcadcristina: well to answer your question, I do agree with your mentor that src/conv doesn't sound appropriate because that is where we convert between 3d geometry formats (and svg isn't a 3d geometry format)
17:38.00brlcadthat sounds more like a renderer, for which we have a variety of other tools in that category spread around
17:38.30brlcadprobably src/util or src/gtools would be fine
17:38.46brlcadgiven the adaptagrams dep, probably the latter
17:39.39cristinaI see, ok then. This means that I should also modify the library dependencies in the Makefile.am and CMakeLists.txt files
17:54.09brlcadcristina: so the question about where to put your work also depends on how you structure your development
17:54.51brlcadin a clean mvc paradigm, you'd end up with a clean separation between the model and the controller
17:55.34brlcadthat's basically the scope of our libged library, so that may well be the best place for a "core" model data structures component
17:56.04brlcadfrom there you could build up a src/gtools command-line interface and/or a src/tclscripts GUI component, or integration with mged/archer
18:25.33starseekerbrlcad: unfortunately, not the last issue
18:26.20starseekerbrlcad: I'm not quite sure what the NMG issue was/is - was talking to ``Erik yesterday and he suggested it might be something to do with how the metaball tessellation is interacting with the nmg code
18:26.59starseekerbrlcad: there are a couple more Windows issues with Archer
18:27.48brlcadhow'd you discover the metaball issue?
18:27.58starseekertried to facetize it
18:28.05brlcadsimple metaball?
18:28.13starseekeryep - make metaball.s metaball
18:28.15brlcadmore importantly, is that a new issue?
18:28.26starseekerit did work at one time, according to ``Erik
18:28.33brlcadwell, that's not saying much
18:28.34starseekerI couldn't confirm that trying earlier releases though
18:28.38brlcadhave a 7.20 to try?
18:28.42starseekeryep
18:28.58brlcadif that also fails, then it's not a release stopper, just an item for BUGS
18:29.01starseekerk
18:29.06starseekeradds it to BUGS
18:29.11brlcadif it does, then it's worth back-tracking
18:31.13starseekeryeah, looks like the same thing with 7.20.6
18:34.23CIA-55BRL-CAD: 03starseeker * r51098 10/brlcad/trunk/BUGS: Add note about metaball tessellation bug
19:00.42brlcadokay, "good"
19:03.10CIA-55BRL-CAD: 03Plussai 07http://brlcad.org * r3880 10/wiki/User:Plussai/GSoC_2012_log: /* 10 June 2012 */
19:05.40CIA-55BRL-CAD: 03Plussai 07http://brlcad.org * r3881 10/wiki/User:Plussai/GSoC_2012_log: /* 10 June 2012 */
19:06.12CIA-55BRL-CAD: 03Plussai 07http://brlcad.org * r3882 10/wiki/User:Plussai/GSoC_2012_log: /* 10 June 2012 */
19:07.07CIA-55BRL-CAD: 03Plussai 07http://brlcad.org * r3883 10/wiki/User:Plussai/GSoC_2012_log: /* 12 June 2012 */
19:07.33CIA-55BRL-CAD: 03Plussai 07http://brlcad.org * r3884 10/wiki/User:Plussai/GSoC_2012_log: /* 12 June 2012 */
19:07.51CIA-55BRL-CAD: 03Plussai 07http://brlcad.org * r3885 10/wiki/User:Plussai/GSoC_2012_log: /* 12 June 2012 */
19:08.15CIA-55BRL-CAD: 03Ksuzee 07http://brlcad.org * r3886 10/wiki/User:Ksuzee/Reports:
19:08.56CIA-55BRL-CAD: 03Ksuzee 07http://brlcad.org * r3887 10/wiki/User:Ksuzee/Reports:
19:10.26CIA-55BRL-CAD: 03bob1961 * r51099 10/brlcad/trunk/src/tclscripts/lib/RtImage.tcl: More shinola for Windows.
19:11.22CIA-55BRL-CAD: 03Plussai 07http://brlcad.org * r3888 10/wiki/User:Plussai/GSoC_2012_log: /* 27 May 2012 */
19:12.29CIA-55BRL-CAD: 03Plussai 07http://brlcad.org * r3889 10/wiki/User:Plussai/GSoC_2012_log: /* 27 May 2012 */
19:13.17CIA-55BRL-CAD: 03Plussai 07http://brlcad.org * r3890 10/wiki/User:Plussai/GSoC_2012_log: /* 27 May 2012 */
19:31.19``Eriklooks like the last confirmed working metaball tesselation was with 7.16.10, well before the coverity changes, all the nmg changes, and the cmake changes
19:33.35CIA-55BRL-CAD: 03starseeker * r51100 10/brlcad/trunk/src/tclscripts/lib/RtImage.tcl: Only pull image from framebuffer if we have to - seems to speed up rtimage script in simple color-only cases.
20:31.10cristinabrlcad: here is the link to the .svg file that has rectangles corresponding to the objects in the share/brlcad/7.21.0/db/axis.g database. It has just shapes, no arrangement and no connections between them. Note that my implementation doesn't yet avoid duplications of the same object.
20:31.16cristinahere is the link: http://dl.dropbox.com/u/3425790/axis-adaptagrams.svg
20:31.31cristinasorry for keeping you waiting
20:32.26brlcadno worries, thanks for sharing it ;)
20:33.03cristina:) I have to go now, I've got an exam in the morning
20:33.08brlcadgood luck
20:33.13cristinathank you
20:33.18brlcadnice progress too
20:33.48cristinahm, thanks :)
20:42.12CIA-55BRL-CAD: 03starseeker * r51101 10/brlcad/trunk/src/tclscripts/rtwizard/rtwizard.tcl: Handle port number being suppled as argument to -F
20:48.19*** mode/#brlcad [+o brlcad] by ChanServ
21:10.46CIA-55BRL-CAD: 03starseeker * r51102 10/brlcad/trunk/src/tclscripts/rtwizard/rtwizard.tcl: taskkill appears to need some extra arguments to wipe out the headless fbserv on Windows.
21:13.10CIA-55BRL-CAD: 03brlcad * r51103 10/brlcad/trunk/src/libbu/ (argv.c basename.c bitv.c color.c ctype.c lex.c): since the ctype API is often implemented as macros where the parameter becomes an array index, avoid a gcc 4.1.3 warning about using a char for the array index (auto-promoted). cast to int.
21:14.02CIA-55BRL-CAD: 03n_reed * r51104 10/brlcad/trunk/src/other/step/ (23 files in 3 dirs): use scl_memmgr in express; SCL git 82c84e7 and 4d6d249
21:47.10starseekermakes a note to take a closer look at this later: http://www.fpx.de/fp/Software/tclkill/
21:49.25starseekermight be worth tossing into btclsh/bwish
22:28.37*** join/#brlcad andrei_ (~andrei@5-12-76-135.residential.rdsnet.ro)
22:31.54CIA-55BRL-CAD: 03Popescu.andrei1991 07http://brlcad.org * r3891 10/wiki/User:Popescu.andrei1991: /* Daily development log */
22:33.19andrei_hello
22:33.43andrei_I edited the development logs on brlcad wiki and it seems not to update itself
22:33.56andrei_if I click edit again it seems fine, otherwise it just looks normal
22:35.05CIA-55BRL-CAD: 03Popescu.andrei1991 07http://brlcad.org * r3892 10/wiki/User:Popescu.andrei1991: /* Daily development log */
22:35.20andrei_nevermind , apparently I needed to have 2 \n to accept it as a newline.
22:42.55andrei_tpkg is fixed, should be up in ~ 15 mins
22:43.08andrei_the shell script in ~ 2-3 hours at most. Still tuning some stuff.
23:04.38CIA-55BRL-CAD: 03n_reed * r51105 10/brlcad/trunk/src/other/step/src/ (21 files in 2 dirs): use scl_memmgr in exppp and fedex_plus; SCL git 7ce8bcb and 4e0d73d
23:09.52*** join/#brlcad crdueck (~cdk@d173-238-127-19.home4.cgocable.net)
23:38.45andrei_http://pastebin.ca/2160995 I have two issues with the performance test script.  Firstly, I didn't manage to "silence" the server so that the total number of sent packages is more visible
23:39.01andrei_&>> doesn't work for me, or I am using it wrong.
23:40.05andrei_secondly, I don t know why but at some point ( I have tried several combination of itterations) the script crashes. It doesn't crash every time and neither on specific arguments
23:40.12andrei_this is how the crash report looks like
23:40.15andrei_Connecting to 127.0.0.1, port 2000
23:40.16andrei_pkg_open: client connect: errno=111
23:40.17andrei_Connection to 127.0.0.1, port 2000, failed.
23:40.19andrei_ERROR: Unable to open a connection to the server
23:40.22andrei_Saving stack trace to (BRL-CAD)-10124-bomb.log
23:40.24andrei_Listening on port 2000
23:41.18andrei_to me, it looks like the client gets called before the server. However, that doesn't really make any sense given the script.

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