IRC log for #brlcad on 20081120

00:27.41 starseeker sighs as he realizes all the hyp.c math will need a rework
00:28.30 brlcad sighs as he realizes the g-vrml exporter will need a rework
00:29.29 ``Erik_ sighs as he realizes that computers are hard and his brain will need a rework
00:31.07 starseeker ``Erik_: I thought you were rewriting your brain in Lisp?
00:36.38 Ralith brlcad, g-vrml? Neat!
00:40.54 brlcad Ralith: hm? we've had that converter for ages
00:40.55 starseeker brlcad: Cool, market below 8000
00:41.00 brlcad yep
00:41.31 Ralith brlcad, I'm sure we've had lots of things I don't know about for ages :P
00:41.36 Ralith doesn't make it less neat.
00:41.56 brlcad :)
00:42.13 brlcad finds new things I either didn't know about or had long since forgotten all the time ;)
00:42.52 brlcad it seems to be very unhappy with dxf-imported bots for some reason, though .. very annoying
00:43.47 starseeker falls over laughing at the big 3 automaker CEOs flying to D.C. in private jets to ask for money
00:44.08 starseeker not that it makes any practical difference, but the symbolism is great
01:03.28 Ralith yeah
03:30.53 CIA-62 BRL-CAD: 03brlcad * r33224 10/brlcad/trunk/ (Makefile.am bench/Makefile.am):
03:30.53 CIA-62 BRL-CAD: hook in make check to make test and make benchmark so that distcheck will also
03:30.53 CIA-62 BRL-CAD: have to pass those two in order to succeed. still don't want distclean wiping
03:30.53 CIA-62 BRL-CAD: out the benchmark 'summary' file, so have to add a manual exemption to the
03:30.53 CIA-62 BRL-CAD: top-level, but we do need to clean up after the various log and pix files.
04:00.20 CIA-62 BRL-CAD: 03brlcad * r33225 10/brlcad/trunk/TODO: per a forum discussion thread about g-vrml that has come up before, it'd be useful if g-vrml had an option to output v1 or v2 format. this would make a great easy little project for someone looking to get involved.
04:02.12 CIA-62 BRL-CAD: 03brlcad * r33226 10/brlcad/trunk/BUGS:
04:02.12 CIA-62 BRL-CAD: g-vrml seems to ignore bots that are created with dxf-g. a simple workaround
04:02.12 CIA-62 BRL-CAD: seems to be to pass the imported bot through the stl exporter/importer then
04:02.12 CIA-62 BRL-CAD: it'll export as vrml (but also ends up doubling the number of faces).
07:50.38 *** join/#brlcad Axman6_ (n=Axman6@pdpc/supporter/student/Axman6)
08:02.09 CIA-62 BRL-CAD: 03starseeker * r33227 10/brlcad/trunk/ (7 files in 5 dirs):
08:02.09 CIA-62 BRL-CAD: Rework the hyp primitive to use input parameters provied by user as primary
08:02.09 CIA-62 BRL-CAD: storage of dimensions. Compiles and make seems to throw up a hyp - next up is
08:02.09 CIA-62 BRL-CAD: more extensive testing of the change and continuing with the editing support.
08:02.29 starseeker woo-hoo!
08:02.39 starseeker <crash> zzzzzz.....
09:22.21 *** join/#brlcad clock_ (n=clock@84-72-91-240.dclient.hispeed.ch)
11:12.47 *** join/#brlcad Ralith_ (n=ralith@216.162.199.202)
11:47.49 claymore Morning all.
12:06.48 CIA-62 BRL-CAD: 03davidloman * r33228 10/rt^3/trunk/src/geometryService/cpp/docs/BME.eap: Continuing Architecture Work: Segregated projects into appropriate namespaces. Begun coding work load divisions.
12:21.54 CIA-62 BRL-CAD: 03davidloman * r33229 10/rt^3/trunk/src/iBME/ (. superceded_GS/): Organizational Changes. Creating directories and moving code snippets and libraries into their appropriate directories.
12:22.02 CIA-62 BRL-CAD: 03davidloman * r33230 10/rt^3/trunk/src/ (geometryService/cpp/docs/ iBME/docs/): Organizational Changes. Creating directories and moving code snippets and libraries into their appropriate directories.
12:24.43 CIA-62 BRL-CAD: 03davidloman * r33231 10/rt^3/trunk/src/geometryService/ (cpp/ java/): Organizational Changes. Creating directories and moving code snippets and libraries into their appropriate directories.
12:26.36 CIA-62 BRL-CAD: 03davidloman * r33232 10/rt^3/trunk/src/iBME/superceded_GS/ (cpp/ cpp/docs/ java/): Organizational Changes. Moved first attempt at the Geometry Service into a superceded_ directory under iBME.
12:27.47 CIA-62 BRL-CAD: 03davidloman * r33233 10/rt^3/trunk/src/geometryService/: Organizational Changes. Deleted empty src/geometryService directory.
12:38.47 brlcad mornin'
12:39.46 claymore howdy. Just waking or getting ready to crash?
12:40.12 brlcad neither, I've been up for a while
12:40.47 claymore fun stuff! How you been? Read through the emails at work... looks like someone's memory stick was naughty lol.
12:47.26 claymore brlcad: So I am working on getting a 'Roast of the GeometryService' meeting lined up. When are you available?
13:15.16 brlcad okay, cool -- but if this is an arl thing, then can discuss it there
13:18.43 claymore no discussion needed. Just need to know what days you will be available. ;)
13:29.14 CIA-62 BRL-CAD: 03davidloman * r33234 10/rt^3/trunk/src/iBME/docs/BME.eap: Continuing Architecture Work: Added ResourceManifest class to enhance robustness of Resource transfers.
13:31.01 *** join/#brlcad d_rossberg (n=rossberg@bz.bzflag.bz)
13:32.41 brlcad d_rossberg: moin!
13:39.20 d_rossberg hallo brlcad
13:39.50 claymore d_rossberg: heya. how goes things?
13:43.00 d_rossberg hi claymore, which of the many things on my desktop do you mean?
13:43.23 claymore d_rossberg: things in general.
13:45.25 CIA-62 BRL-CAD: 03davidloman * r33235 10/rt^3/trunk/src/iBME/ (86 files in 7 dirs): Continuing Architecture Work: Initial import of class skeleton code.
13:45.52 d_rossberg well, the end of the year is near and there are some things which have to be finished ... but i shouldn't complain
13:46.09 starseeker boots brain...
13:46.57 *** join/#brlcad elite01 (n=omg@unaffiliated/elite01)
13:47.17 d_rossberg btw, i had a look at the iBME section in the wiki
13:48.07 claymore d_rossberg: Yeah, I got about 5% into my work when I had to take a few days off. But I am back at it. More to come on the documentation.
13:48.16 clock_ I got an idea
13:48.22 claymore d_rossberg: input always welcome! Questions too.
13:48.28 clock_ Since BRL-CAD uses an efficient short format to store 3D databases
13:48.41 clock_ I have developed a very dense 2D barcode with error correction
13:48.53 clock_ One could make a sticker that would contain a 3D model of an object and stick it on an object
13:49.10 clock_ If someone needed to interface that object mechanically he would scan the sticker and would have a model available
13:49.23 claymore clock_: neat idea :)
13:49.51 claymore clock_: what physical dimensions? what storage capacity?
13:49.52 clock_ It fits 200kB per page A4 with a laser printer
13:50.06 clock_ A6 then 50kB etc.
13:50.25 clock_ But now I see a usual .g file is like 19kB or so maybe it would be too big
13:50.35 clock_ maybe compressed
13:51.04 claymore heh, um, most of the .g's I have worked with are 50-400MB ;)
13:51.08 clock_ lol
13:51.10 clock_ no way then :)
13:51.34 claymore I got excited when you tossed out the words 'High density'
13:51.36 clock_ Or it could be a tiny stamp containing only the URLs with the information
13:52.06 clock_ It has higher density than the patented PARC Dataglyph which has allegedly only 50kB per A4
13:52.09 claymore those ginormous files consisted of mostly bots though...
13:53.19 claymore well, back to work.
13:53.24 claymore :/
14:03.47 brlcad claymore: btw, if it's not been said.. the commit messages are much better :) been able to follow much better
14:04.44 claymore brlcad: I have been SOOOO tempted to run them through a Ebonics translator just for fun. But thanks, I'm learning still :)
14:04.44 brlcad especially the ones today, that's the 'it tells a story' aspect that is great to have in the revision history
14:05.29 brlcad hey, you're talking to other devs, so jokes are sometimes appreciated :)
14:05.30 claymore d_rossberg: do you have any immediate feedback from the docs on the wiki?
14:05.49 starseeker <snort> particularly insults to Windows
14:06.15 brlcad now just wishes there was a better text-diffable format for diagrams
14:06.37 claymore starseeker: I got a brother-in-law that works for MS and kicks back nearly ALL the perks to his family... so I need to be somewhat careful :)
14:06.46 starseeker ah :-)
14:07.12 starseeker is only likely to get a poison dart arrow from MS, considering my history with open source
14:07.24 claymore starseeker: Nothing like legit copies of XP Pro for $10 :)
14:07.25 clock_ I wonder what an output from Ebonics translator looks like.
14:07.57 brlcad claymore: sure there is, copies of BSD or Linux for $0 :)
14:09.21 claymore brlcad: Not when you run your own whitebox bidness out of your house and the customer wants Windows. :)
14:09.32 starseeker brlcad: there's matplotlib: http://matplotlib.sourceforge.net/gallery.html
14:10.40 starseeker probably a bit too function/dat plotting specific though
14:10.49 brlcad starseeker: someone should build an architexture/project diagram tool on top of that then :)
14:10.57 brlcad by itself, though, it's too low-level
14:11.18 starseeker nods
14:11.33 starseeker plus, introduces python as a requirement
14:11.37 brlcad hm, what was the tool that one of the gsoc students used..
14:11.47 starseeker what specific application did you have in mind?
14:12.41 starseeker claymore's docs?
14:12.50 brlcad something like visio
14:13.04 starseeker Ah, yes...
14:13.31 starseeker was thinking we could make claymore re-implement his diagrams in graphviz ;-)
14:13.44 brlcad not even something code-specific, omnigraffle is probably an even better example
14:14.34 starseeker googles omnigraffle
14:15.27 starseeker oh, it is graphviz based
14:15.33 claymore starseeker: You and what army pal! *protects his diagrams*
14:15.48 starseeker uh - the one outside the window? :-P
14:16.02 claymore starseeker: Thats your verizon network silly.
14:16.51 starseeker brlcad: pity that's commercial
14:16.59 brlcad yeah, it is
14:17.16 brlcad omni's tools are all outstanding, but mostly non-free
14:17.43 starseeker hmm... tcldot + tk widgets + lots of ugly hackery...
14:18.22 starseeker + graphviz as a requirement...
14:18.34 starseeker ~= trouble probably
14:25.51 starseeker humph - looks like the only Tk+graphviz work was done in Perl
14:26.21 starseeker must investigate tcldot... potential for evil here :-)
14:27.43 *** join/#brlcad mafm (n=mafm@193.136.2.121)
14:27.58 claymore
14:29.00 mafm hi pplz
14:29.46 claymore hai mafm!
14:30.17 starseeker alrightie, as the song says "back on the road again"
14:36.35 CIA-62 BRL-CAD: 03brlcad * r33236 10/brlcad/trunk/src/librt/primitives/hyp/hyp.c: style/consistency cleanup. format up comments, remove embedded tabbage.
14:38.57 claymore imagines what Starseeker Nelson would look like..... *shiver*
14:39.02 claymore :D
14:41.11 d_rossberg claymore: now i had a closer look to your docs ...
14:42.16 d_rossberg e.g.: BrlcadDbResource::getGeometryByObjectName() can never return a GeoRegion
14:43.53 claymore Firstly, the BrlcadDbResource class is under my spotlight right now... there is something not right about how I implemented it.
14:44.37 claymore but as for not being able to return a GeoRegion... why not? GeoRegion isa GeoCollection isa AbstractResource....
14:45.46 d_rossberg that's true, but getGeometryByObjectName always returns an AbstractResource (i.e. the base class)
14:48.28 claymore what you said is also true. Since a GeoRegion isa AbstractResource, getGeometryByObjectName can return ANY AbstractResource or subclass...
14:48.40 claymore .... so I don't see how it can't.
14:53.23 claymore not that I couldn't be wrong, but that's how I though OO inheritance worked.
14:53.48 claymore though = thought
14:53.58 d_rossberg at the end of the execution of getGeometryByObjectName the copy constructor of AbstactResource will be called which can create an AbstractResource only
14:54.42 claymore no copies, just references/pointers.
14:55.21 d_rossberg there is no & or *
14:56.09 d_rossberg returning the value means using the copy constructor
14:56.26 claymore righto. Design docs are still a work in progress. Assume ALL object passing is either reference or pointer. Never by value.
14:56.59 d_rossberg but then: who owns these objects? how long do they live?
14:58.00 d_rossberg will be back in some minutes
14:58.09 claymore Many different things can have a handle on AbstractResource objects. Primarily, they will reside in the ResourceCache.
15:00.10 claymore As various commands are recieved, the GE will perform UUID lookups first in the ResourceCache, then secondly in the ResourceSvnLink, then in anyother AbstractResourceSources registered with the ResourceManager.
15:01.25 claymore Once (if) the AbstractResoruce is located by the ResourceManager, a handle can be passed into the appropriate Function in the GED object and the recieved 'command' is then executed.
15:02.53 claymore Their Lifespan is determined by the Caching Methodology that is employed in the ResourceCache. That methodolgy is TDB.
15:02.59 claymore lol, TBD even.
15:17.28 d_rossberg aha, i didn't recognized the cache object
15:19.01 claymore it makes perfect sense in my head... now the REAL project is making it make sense on paper :)
15:55.06 *** join/#brlcad Elrohir (n=kvirc@p5B14D5A9.dip.t-dialin.net)
16:21.16 ``Erik_ takes a moment to self detonate
16:21.52 Kool[a]id ``Erik_: LOL
16:21.54 claymore ``Erik_: How ya feeling?
16:22.01 ``Erik_ aplosive
16:22.20 ``Erik_ much better, still a deep cough, but I think it's residual
16:22.26 ``Erik_ hungry
16:22.27 ``Erik_ :D
16:22.32 claymore lol car just for ``Erik_ : http://icanhascheezburger.com/2007/10/31/leeeeeeeeeroy1/
16:22.36 claymore car = cat
16:22.50 ``Erik_ LEEEERRROOOOYYYY jenkins!
16:23.09 ``Erik_ that was an effin' hilarious video
16:23.27 claymore still nearly pees himself laughing at that vid.
16:26.05 claymore brings back some very norrid EverCrack memories......
16:26.06 ``Erik_ I was up at 4 this morning (fell asleep on the couch early last night), logged into that stupid game, but just wasn't feeling it :/ even after spending $70 on the 'collectors edition' update
16:26.33 claymore that Frozen Lich King exp pack thingy?
16:26.43 ``Erik_ I think I spent more time arguing with strangers than playing... I'm very internet special olympics like that
16:26.44 ``Erik_ yes
16:27.30 ``Erik_ http://www.ifilmdb.com/Crunchyroll/Pictures/ArguingOnTheInternet-Special%20Olympics.jpg
17:18.30 CIA-62 BRL-CAD: 03davidloman * r33237 10/jbrlcad/trunk/ (.classpath .project): Deleted .classpath and .project since they are IDE/Computer specific files and shouldn't really be in the Repository.
17:28.53 CIA-62 BRL-CAD: 03davidloman * r33238 10/jbrlcad/trunk/: Added .classpath and .project files to the svn:ignore property of the root directory.
17:45.16 CIA-62 BRL-CAD: 03davidloman * r33239 10/rt^3/trunk/src/iBME/ (87 files in 8 dirs): Updated Architecture to reflect object References and Pointers instead of 'pass by value' where applicable. Generated new UML images.
19:01.20 *** join/#brlcad marko1 (n=mark@cpe-74-72-195-92.nyc.res.rr.com)
19:01.43 marko1 there IS an IRC for BRL CAD!
19:02.51 marko1 is anyone there?
19:04.03 marko1 is it possible to apply transforms and rotations of a region as a whole?
19:04.28 marko1 looks like arced is the command to use
19:04.44 marko1 can't get it working
19:11.40 mafm there is, but sometimes everybody is... erm... sleeping? :)
19:12.01 marko1 it's the oed command
19:12.10 marko1 reading the oed tutorial
19:12.11 marko1 now
19:22.43 marko1 do you HAVE to B before you enter oed?
19:44.50 *** join/#brlcad mafm (n=mafm@193.136.2.121)
19:59.50 marko1 mafm: so I got the basic oed moves down
19:59.59 marko1 now I'm trying to use them from command line
20:00.04 marko1 mged -c style
20:00.59 marko1 I keep getting
20:00.59 marko1 no solids in view
20:00.59 marko1 Unable to do <Object Translation> from VIEWING state.
20:01.00 marko1 Expecting OBJ EDIT state.
20:01.00 marko1 no solids in view
20:01.00 marko1 Unable to do <Object Translation> from VIEWING state.
20:01.01 marko1 Expecting OBJ EDIT state.
20:04.16 marko1 NM
20:04.20 marko1 got it
20:14.21 *** join/#brlcad elite01 (n=omg@unaffiliated/elite01)
20:33.12 claymore Sorry! Many of us have been off inmeeting land.
20:33.29 marko1 I sorta got it
20:33.52 marko1 but now it seems like it's only moving the right hand solid
20:33.57 claymore problems with the OED command?
20:33.59 marko1 not the whole regoion
20:34.11 marko1 yeah
20:34.44 claymore check this and see if it helps: http://brlcad.org/w/images/3/36/Object_Editing_-_the_oed_Command.pdf
20:34.49 marko1 ha
20:34.54 marko1 reading that now
20:35.20 marko1 does oed behave differently for combinations vs regions
20:35.43 claymore A simple way to think about oed is that there is a path from root to solid for every solid
20:35.51 claymore marko1: No it shouldn't...
20:35.56 marko1 k
20:36.38 claymore if you take that full path and split it in two halves, that split will represent the editing point and everything from that split on down to the solid will be affected (all branches of the tree)
20:36.52 marko1 whoa
20:37.03 marko1 let me read that 3 more times
20:37.14 marko1 ah
20:37.18 marko1 I think I see
20:37.21 marko1 let me try
20:37.41 claymore think of it like your arm: shoulder/upperarm/elbow/lowerarm/wrist/hand.s
20:37.52 claymore (pardon the horrid spelling)
20:38.18 claymore if you: mged>oed /shoulder/upperarm/ elbow/lowerarm/wrist/hand.s
20:38.58 claymore then you are editing the 'elbow' and lowerarm, wrist, and hand.s will all be affected (as will all the fingers, but i didnt want to go to THAT level of complexity)
20:39.04 marko1 what if you wanted to move the shoulder
20:39.35 claymore try mged>oed / shoulder/upperarm/elbow/lowerarm/wrist/hand.s
20:39.48 claymore note the single slash that indicates 'root' or 'toplevel'
20:39.57 marko1 bingo!
20:39.59 marko1 got it
20:40.09 claymore excellent!
20:40.15 marko1 claymore: very helpful
20:40.24 marko1 subtle thing
20:40.30 marko1 hard to get from docs
20:40.47 claymore brlcad is wickedly powerful... and wickedly hard to learn :)
20:41.08 marko1 I'm using ruby to emit mged commands
20:41.10 marko1 it's great
20:41.42 CIA-62 BRL-CAD: 03davidloman * r33240 10/rt^3/trunk/src/iBME/docs/ (BME.eap GeometryEngine_UML.png): Updated Architecture to reflect a new more object References and Pointers instead of 'pass by value' where applicable. Generated new UML images.
20:42.22 claymore using mged classic mode?
20:45.55 CIA-62 BRL-CAD: 03davidloman * r33241 10/rt^3/trunk/: IDE specific svn:ignore .project and .cproject
20:46.41 claymore great. snow on the way home :/
20:46.48 claymore welp, see yall later.
21:11.25 marko1 is there a way to force order of operations?
21:11.30 marko1 seems parens don't work
21:11.54 marko1 r bobbin_solid u (support_plate u bobbin_torus + bobbin_half) - bobbin_negativ
21:12.53 *** join/#brlcad cad64 (n=52a9e334@bz.bzflag.bz)
21:32.25 brlcad marko1: yes there is, but it'll take a sec to explain and I gotta run :)
21:33.39 marko1 ok
21:34.04 marko1 one word clue?
21:34.10 marko1 maybe I can look itup
21:59.43 *** join/#brlcad louipc (n=louipc@archlinux/trusteduser/louipc)
22:22.04 brlcad marko1: the order of operations is that of a stack of operations so you can apply the operations in iteration
22:22.47 brlcad to do what you're implying above, the easiest thing you could do is either put the paren'd section into a comb or distribute the subtraction
22:23.34 brlcad c bobbin_pos u support_plate u bobbin_torus + bobbin_half ; r bobbin_solid u bobbin_pos - bobbin_negativ
22:23.57 marko1 ok
22:24.04 marko1 that's basically what I'm doing
22:24.19 brlcad otherwise: r bobbin_solid u support_plate - bobbin_negativ u bobbin_torus - bobbin_negativ + bobbin_half
22:24.25 marko1 just want to use parens like a normal boolean
22:25.04 brlcad there is a tool that lets you construct the boolean using parens, but it's not been tested/used in a while so I would recommend using it
22:25.05 marko1 constantly bumping into order of op issues
22:25.56 brlcad they're preorder operators, not infix
22:26.18 brlcad your example is a mix of pre and infix
22:27.27 brlcad that is to say that the operator applies to the word in following and the previous result to the left
22:28.22 brlcad that's why you start with a "u something" .. you union that object with the previous (null) result .. giving you "something"
22:30.15 marko1 ok
22:33.23 brlcad not the best explanation, but sticking stuff in combinations usually makes things simpler to understand regardless, even if you have to make a few extra combinations :)
22:35.38 brlcad marko1: also to answer your earlier question -- no you don't have to B before running oed -- B means "Blast" which means 'erase everything and draw this'
22:40.32 marko1 thanks!
22:40.44 marko1 I'm getting it bit by bit
22:42.13 marko1 I really like how I can scrip it into larger workflows
22:45.09 *** join/#brlcad alex_joni (n=juve@emc/board-of-directors/alexjoni)
22:54.25 *** join/#brlcad PrezKennedy (i=Matthew@whitecalf.net)
23:07.56 *** join/#brlcad louipc (n=louipc@archlinux/trusteduser/louipc)
23:33.12 *** join/#brlcad louipc (n=louipc@archlinux/trusteduser/louipc)

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