IRC log for #brlcad on 20090701

00:46.30*** join/#brlcad schwinn434_ (n=schwinn4@cpe-75-81-202-25.we.res.rr.com)
00:51.49*** join/#brlcad schwinn434 (n=schwinn4@75.81.202.25)
01:58.21*** join/#brlcad schwinn434_ (n=schwinn4@cpe-75-81-202-25.we.res.rr.com)
02:04.53*** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-196.sbndin.btas.verizon.net)
02:42.36*** join/#brlcad schwinn434 (n=schwinn4@cpe-75-81-202-25.we.res.rr.com)
02:44.53*** join/#brlcad schwinn434_ (n=schwinn4@75.81.202.25)
02:52.30*** join/#brlcad Ralith (n=ralith@216.162.199.202)
04:07.09CIA-30BRL-CAD: 03brlcad * r34910 10/brlcad/trunk/BUGS: elena found an odd bug in mged's command-mode parsing. semicolons are the separator but for some reason output of mged commands is being suppressed for all but the last command.
04:07.55*** join/#brlcad IriX64 (n=WarLock@bas2-sudbury98-1178015175.dsl.bell.ca)
04:12.43brlcadper elena's questions earlier: mged -c whatever.g "tree \[string trim \[ tops -n\] \]"
04:13.16brlcadhave to quote the globbing characters [], {}, & ?
04:14.27``Erikahhhh, that's right, forgot about that
04:15.14brlcadsemicolon should / does work .. but there's some sort of bug there
04:15.54brlcadmged -c test.g "puts hello ; puts world"  <= works for example
04:16.19brlcadeven interlaced with mged commands .. but for some reason mged commands aren't printing unless they're last
04:16.33``Erikwhat about the stdout/stderr issue?
04:16.35brlcadhave to trace it
04:17.08brlcadthat's mostly legacy, all textual logging goes through bu_log, which has always written to stderr
04:17.29brlcadwould be a pretty significant change so it's not been attempted yet
04:19.37brlcadthat was originally so the streamable tools could still log and send their data to a stream (e.g. for piping input to output on things like "pixhalve -s1024 some.pix | pix-png > file.png", and still seeing logged output)
04:19.55brlcadeverything uses the same logging mechanism so it goes to stderr
04:20.50brlcadnotes elena probably needs to read http://brlcad.org/wiki/SGI_Cube
07:24.49*** join/#brlcad _clock_ (n=_sushi_@77-58-151-159.dclient.hispeed.ch)
07:48.01*** join/#brlcad IriX64 (n=IriX64@bas2-sudbury98-1178015175.dsl.bell.ca)
08:02.24*** join/#brlcad jdoliner (n=jdoliner@c-68-51-75-169.hsd1.il.comcast.net)
08:09.14CIA-30BRL-CAD: 03d_rossberg * r34911 10/rt^3/trunk/ (4 files in 2 dirs):
08:09.15CIA-30BRL-CAD: change the data-type for color values from unsigned int to double (0 <= r/g/b <= 1)
08:09.17CIA-30BRL-CAD: although color values are stored as unsigned int this implementation detail should be hidden from the user
08:18.38*** join/#brlcad _clock_ (n=_sushi_@77-58-151-159.dclient.hispeed.ch)
08:47.52CIA-30BRL-CAD: 03Jdoliner 07http://brlcad.org * r1525 10/wiki/User:Jdoliner:
09:49.01*** join/#brlcad mafm (n=mafm@165.Red-81-35-69.dynamicIP.rima-tde.net)
10:25.35*** join/#brlcad Elrohir (n=kvirc@p5B14FC4C.dip.t-dialin.net)
13:22.50*** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-196.sbndin.btas.verizon.net)
14:18.52*** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-196.sbndin.btas.verizon.net)
15:40.11*** join/#brlcad elena (n=elena@89.136.118.141)
15:48.31*** join/#brlcad d-lo (n=claymore@bz.bzflag.bz) [NETSPLIT VICTIM]
15:48.31*** join/#brlcad roberthl (n=robert@silentflame/member/roberthl) [NETSPLIT VICTIM]
16:01.01*** join/#brlcad jdoliner (n=jdoliner@c-68-51-75-169.hsd1.il.comcast.net)
16:02.12*** join/#brlcad alecs1 (n=alex@193.186.164.88)
16:02.59alecs1hi, does BRL-CAD support tolerancing? I couldn't find it in the docs
16:03.31d-lodepends on what you mean.  To what are you wanting to apply a tolerance to?
16:03.33alecs1(tolerancing as in gd&t, that's the more known buzzword)
16:04.11louipcmake this hole 1" dia +- .005"
16:04.54alecs1well, to components in the tree; but I'm no BRL-CAD expert either, just evaluating and asking if it's not in the docs
16:05.13louipcbrl-cad doesn't have that feature as far as I know
16:05.26louipcbut I'm no authority on it :P
16:06.03d-loI would say no, it doesn't support GD&T.
16:06.41alecs1thanks
16:06.46alecs1do you know of other open-source programs that support it
16:12.03CIA-30BRL-CAD: 03starseeker * r34912 10/brlcad/trunk/ (11 files in 7 dirs): Add bbsize command per user request - reports a quick summary of how big the bounding box of an object is.
16:12.34louipcnope
16:12.50louipcqcad maybe... that's 2d though
16:18.16alecs1I see, thank you :)
16:23.23CIA-30BRL-CAD: 03starseeker * r34913 10/brlcad/trunk/src/tclscripts/lib/ (Db.tcl Mged.tcl): Don't need the bbsize command support on obsolete tcl code.
17:02.38brlcaddefinitely do not support tolerancing
17:06.00brlcadstarseeker: unless I read the commit wrong, you missed the bbsize.c file
17:06.58starseekeraaah, nuts
17:07.00starseekersorry
17:07.40CIA-30BRL-CAD: 03starseeker * r34914 10/brlcad/trunk/src/libged/bbsize.c: Yeah, it might help to add the key bbsize file.
17:07.56brlcadthx
17:08.04starseekernp :-)
17:38.13CIA-30BRL-CAD: 03brlcad * r34915 10/brlcad/trunk/NEWS: fixed compilation support for mac os x 10.5, problem compiling with ogl support enabled (mac ld issue with /usr/X11/lib/libGL.dylib cyclicly referring to framework dylib)
17:45.39*** join/#brlcad alecs1 (n=alex@budweis.scch.at)
17:54.31*** join/#brlcad BigAToo (n=BigAToo@64.74.225.82)
17:56.03*** join/#brlcad BigAToo1 (n=BigAToo@64.74.225.82)
17:58.13CIA-30BRL-CAD: 03homovulgaris * r34916 10/brlcad/trunk/src/libpc/pcMathGrammar.h: fleshing FunctionGrammar 1/2
18:03.28jdolinerindianlarry whenever you read this, I think I've worked out the kinks in my algorithmic woos. But would love to talk about it with you, just to do some sanity checks.
18:08.23``Eriksanity check: you have none :D
18:08.43``Erik(which is why ya fit in here)
18:08.54jdoliner:D
18:09.09jdolinerit certainly feels that way these days
18:09.22indianlarrywhat you come up with Joe
18:09.23jdolinerI close my eyes and see triangles intersecting each other
18:09.58jdolinerk so we wanted to have a convention for which way the edge we return should point
18:10.39jdolinerand this convention should depend on the normals of both of the triangles we intersect
18:11.01indianlarryokay
18:11.23jdolinerso we decide that the edge will be parallel to (norm1Xnorm2)
18:11.56CIA-30BRL-CAD: 03irpguardian * r34917 10/brlcad/trunk/src/proc-db/human.c:
18:11.56CIA-30BRL-CAD: Added poseable support for the torso. Also made the arms and legs now properly position themselves and attach to torso.
18:12.01CIA-30BRL-CAD: Legs and arms are still not properly positioned when the body is rotated however.
18:12.01indianlarryfollow so far
18:12.03jdolinerso if we do this
18:12.22jdolinerthen when we have an edge and the triangle it came from
18:14.00jdolinerthen in triangle1 the (edgeXNorm) points to the internal side, in triangle2 it points to the external side
18:14.26jdolinerjust a fact of life that this is going to be reversed
18:14.41jdolinerbut the important part is that it'll be consistent
18:15.57jdolinerso if we take our corner jabbing into a triangle
18:17.27jdolinerwe would get a the trimming curve back with the lines s.t. the normal points the opposite direction of the faces normal
18:17.39jdolinerwhich signifies an internal face
18:18.20jdolineroh forgot to mention abut the output that we need to reverse the line for mesh1
18:18.50jdolinerbut yeah I'm confident in that case
18:19.36jdolinernow the trickiness comes when it hits an edge
18:20.23jdolinerso it now passes up which edge the intersection came from
18:20.33jdolinerso suppose we get back the line pq
18:21.06jdolinerand we know that point p came from line ab
18:21.10``Erikyuh oh, libpc issue, no boost lazy.hpp
18:21.40jdolinerthen we know that the lines to include are:
18:21.58jdolinerap and bp
18:22.37jdolinerbecause those are consistent with the direction of pq
18:23.11``Eriksrc/libpc/pcMathGrammar.h:42:41: error: boost/spirit/dynamic/lazy.hpp: No such file or directory
18:23.37jdolinerso really the only tricky part is that ab might have more than 1 point on it
18:25.13jdolinerso we basically need to record where these points are in a slightly more sophisticated way
18:25.36jdolinerwhich is just going to be a polyline that keeps them all in order and then we take the segments out of it at the end.
18:26.39indianlarryjust thinkin that basically any loop that contains an outside edge would in fact be an an outside edge and be ordered as such
18:27.50indianlarryare you making loops from your intersections yet
18:28.11indianlarryor just the list of intersecting segments
18:29.31jdolinersort of, I have the could all ready to go to make loops, but I need to implement the stuff I just mentioned to be able to pass it the right segments
18:30.40jdolinerbefore this I was just grabbing the segments however they came out and it worked great on them
18:33.02indianlarrymaybe not visualizing correctly
18:33.49indianlarrybut seems like you could lose ap or pb depending on the boolean logic of your combination
18:35.15jdolineryeah one of them won't be in the final result
18:35.24jdolinerone is internal and the other is external
18:36.07indianlarryyeah okay, do you keep seperate loop list for each face
18:36.35CIA-30BRL-CAD: 03brlcad * r34918 10/brlcad/trunk/src/other/boost/spirit/dynamic/lazy.hpp: missing lazy.hpp, now used by pcMathGrammer.h though might still be missing some other files
18:36.41jdolineryes
18:37.10jdolinerfor each face in the original mesh there's actually possible multiple face loops
18:37.52indianlarryyeah and an inner can probably turn into an outer after the combination...
18:38.18jdolinerhow do you mean?
18:39.07jdolineroh, well if we do Mesh A - Mesh B then the inners of B will be an outer of the result
18:39.27indianlarryis you cut a face in falf and there was an inner loop -- yeah
18:39.39jdolinerk yeah
18:39.46indianlarrysorrry slow typer
18:39.58jdoliner:) young nimble fingers
18:40.13indianlarrysounds similar to the nurbs loops
18:40.30jdolineryeah it's actually really elegant after we have all the faces pointing the right way
18:40.37indianlarryis the first loop the outer and the rest inner loops
18:41.01jdolinerwell couldn't there be multiple outers?
18:41.35jdolineran outer loop is just one that intersects a triangle edge correct?
18:41.51indianlarrynot sure how the meshes are handles but for the nurbs there's only one outer
18:42.12indianlarryi think multiple outers would be treated as seperate faces
18:43.39jdolinerokay I think I see what it means for nurbs
18:44.17jdolineris it just the curve along the outside of the defined surface?
18:44.18indianlarryare you looking at bots or meshed faces
18:44.50indianlarryyes the outer boundry of the face
18:45.18jdolinerI believe bots
18:45.46jdolineryeah it just has an array of faces which could occur in any order
18:46.26jdolineralthough there can be quads in the bag as well
18:47.13CIA-30BRL-CAD: 03brlcad * r34919 10/brlcad/trunk/src/other/boost/spirit/ (5 files in 3 dirs): add a few more missing boost files to unbreak the build after recent pcMathGrammer changes
18:58.42indianlarryif bots then probably don't need to worry about loops and such right
18:59.58jdolineryeah, I deal with things on a triangle by triangle basis
19:01.45indianlarrysorry joe i think i'm leading you astray
19:03.05indianlarryguess you still need to maintain proper facet orientations
19:03.51jdolineryes
19:04.51indianlarryi guess a polyline accross a facet -> many facets
19:05.32jdolineryes
19:06.30jdolinerI mean I can't do any sort of geometry internal to my faces, any detail I want to include has gotta be done with triangles
19:07.48indianlarryyep the corner into a face of a triangle turns into a handfull of facets
19:08.41indianlarrydid the BOOLE folks have walking algrithm for this already?
19:15.59indianlarryor were they just doing bezier ?
19:20.29CIA-30BRL-CAD: 03erikgreenwald * r34920 10/isst/ (24 files in 4 dirs): import of ISST
19:22.55elenabrlcad, are you here?
19:25.32CIA-30BRL-CAD: 03erikgreenwald * r34921 10/isst/trunk/src/gui.c: acronymize
19:32.04elena``Erik, did you feed the squirrel again?
19:45.27brlcadhi elena
19:45.35elenahi.
19:45.38brlcadfew comments about your work yesterday
19:46.07elenadid you see the log of the chat with erik?
19:46.13brlcadsemicolon is the command separater -- what you were trying should have worked
19:46.16brlcadyes
19:46.23elena:)
19:46.25elenaok
19:46.38brlcadnever leaves IRC .. that's what 'screen'+'irssi' are for :)
19:46.52elena:)
19:47.13brlcadyou can see the result with something like this:  mged -c test.g "puts hello ; puts world"
19:47.38brlcadthere is some odd issue causing output to be quieted (though I believe it does run the commands still)
19:47.58brlcadhaven't confirmed yet though
19:48.11brlcadsecondly, I presume you've seen http://brlcad.org/wiki/SGI_Cube  ?
19:48.19brlcadthat shows three specific ways to script mged
19:48.21elenabut "tops ; tops" doesn't work :(
19:48.29brlcadit does and it doesn't
19:48.37brlcadI believe it's just not printing the result
19:48.38elenai haven't.
19:48.42elenathanks for the tip.
19:48.49brlcadbut actually runs the commands
19:49.05elenai spent all day reading about tcl :)
19:49.19elenaok.
19:49.25brlcadto do the tree of tops that you suggested, it was simply a matter of escaping the globbing characters
19:49.27elenaand about the stderr output?
19:50.10brlcadfor example:  mged -c whatever.g "tree \[string trim \[ tops -n\] \]"
19:51.17CIA-30BRL-CAD: 03erikgreenwald * r34922 10/isst/trunk/src/ (gui.c main.c sql.c): fix headers
19:51.32brlcadby default, mged has a globbing evaluation mode enabled so that glob characters (*[],?{}.) are specially recognized so you can do pattern matching on objects, e.g. "draw *" or "draw [a-f]*"
19:52.04brlcadso you either have to turn globbing off (set glob_compat_mode 0) ... or ... escape the glob characters ala \[ \]
19:53.25brlcadbtw, if you had multiple top-level objects, that example won't work -- you'd need to iterate over each object and call tree
19:53.55elenai've noticed. this is why i started to read more about tcl.
19:54.16elenait seems to me that it will help with the drupal integration.
19:54.24brlcadsomething like: foreach obj \{ \[ string trim \[ tops -n \] \] \} \{ tree $obj \}
19:54.57brlcadsince you're in php land, it may very well be easier to just run one command at a time and do your string parsing on the php side
19:55.26elenaless fun.
19:55.32elenamaybe less efficient, too.
19:55.57brlcadit should be fairly insignificant compared to what all you're doing and how often this is called
19:56.09brlcadbut fun is a good motivator
19:56.10elenaok. i'll think about it.
19:56.11brlcad:)
19:56.43brlcadnote that the http://brlcad.org/wiki/SGI_Cube example shows those three ways I mentioned, but is actually just a posix shell script driving mged
19:57.15brlcadthe cat <<EOF section is one of the three I mentioned, where it pipes multiple commands to mged
19:57.53elenai'll study it.
19:57.59elenathank you.
19:58.28brlcadif you want to capture stderr more easily, you can feed stderr to stdout with 2>&1
19:58.42elenai did that.
19:58.48brlcadk
19:58.52``Erikwonders if turning off globbing in the .mgedrc would be good for an account that just runs mged batch-mode
19:59.27elena@Erik: doesn't matter, i could turn it off at the begining of the script.
19:59.29brlcadpossible, but it really just saves a couple keystrokes, which are embedded in a script anyways
19:59.40brlcadnow if your script gets big/unreadable ...
20:00.07``Erikor at the beginning of the script, yes... escaping all the glob characters in a nontrivial script can quickly become ugly
20:13.33*** join/#brlcad _sushi_ (n=_sushi_@80-219-43-170.dclient.hispeed.ch)
20:39.32*** join/#brlcad alecs1 (i=alex@193.170.134.247)
20:51.54CIA-30BRL-CAD: 03indianlarry * r34923 10/brlcad/trunk/ (12 files in 5 dirs):
20:51.56CIA-30BRL-CAD: added near hit, near miss processing to shotline
20:51.58CIA-30BRL-CAD: also started to pull out debugging code and wrap into
20:52.00CIA-30BRL-CAD: 'brep' utility command in libged
21:16.46*** join/#brlcad docelic (n=docelic@78.134.203.84)
21:20.05CIA-30BRL-CAD: 03irpguardian * r34924 10/brlcad/trunk/src/proc-db/human.c: Added poseable support for the head and neck, so now the entire body has poseable support.
21:22.44*** join/#brlcad Patmcc19 (n=chatzill@71-223-27-160.phnx.qwest.net)
21:25.22CIA-30BRL-CAD: 03ebautu * r34925 10/web/trunk/htdocs/more/sites/all/themes/fireflystreamcom/template.php: Fixed error in theme comment wrapper
21:26.26CIA-30BRL-CAD: 03ebautu * r34926 10/web/trunk/htdocs/more/sites/all/modules/brlcad/scripts/ (. 7.14/ 7.14/metadata.txt): Scripts for integration with Drupal
21:27.45CIA-30BRL-CAD: 03ebautu * r34927 10/web/trunk/htdocs/more/sites/all/modules/brlcad/brlcad.module: Added suport for multiple BRL-CAD versions (script not yet ready). Multistep submission is incomplete, too.
21:30.11CIA-30BRL-CAD: 03Ebautu 07http://brlcad.org * r1526 10/wiki/More_Changelog: /* June, 17 - today */
21:49.09*** join/#brlcad Elrohir (n=kvirc@p5B14FC4C.dip.t-dialin.net)
22:09.06*** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-196.sbndin.btas.verizon.net)

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