IRC log for #brlcad on 20090325

00:04.05 *** join/#brlcad madant (n=madant@117.196.132.90)
00:35.04 *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-59.sbndin.btas.verizon.net)
02:51.56 *** join/#brlcad madant (n=madant@117.196.132.139)
02:53.36 yukonbob hello, cadheads
03:34.16 pacman87 howdy, yukonbob
03:37.41 *** join/#brlcad madant (n=madant@117.196.138.147)
03:55.18 starseeker is surprised by how much info the gsoc mentor form requires
03:58.26 yukonbob hey pacman87
04:09.48 *** join/#brlcad jdoliner (n=jdoliner@c-68-51-76-57.hsd1.il.comcast.net)
04:31.57 *** join/#brlcad dreeves (n=dreeves@67.130.253.14)
04:33.44 starseeker ah, there's pixie's rib parser: http://pixie.svn.sourceforge.net/viewvc/pixie/trunk/src/rndr/rndr.cpp?revision=1193&view=markup
04:37.51 brlcad notes that task is listed: http://brlcad.org/~sean/ideas.html
04:39.05 brlcad another integration idea was to test out pixie itself as a "shader" or as an rt lighting model
04:41.55 pacman87 brlcad: is the "cylindrical wedge" on that list (under primitives) just a cylinder with base/top at an arbitrary angle?
04:43.57 brlcad pacman87: http://mathworld.wolfram.com/CylindricalWedge.html
04:44.15 brlcad pretty simple primitive
04:44.32 pacman87 looks like it
04:45.25 brlcad basically a tgc/rcc but sliced through the side and one base at an arbitrary orientation
04:47.35 brlcad input params are somewhat different
04:47.55 pacman87 yeah, that's what i was thinking about
04:48.12 pacman87 same thing with the hyp
04:48.27 pacman87 a few different ways to specify dimensions
04:48.41 brlcad not sure if it could be easily generalized to a generalized conical wedge instead of right circular
04:49.26 pacman87 radius, height vector, and two normal vectors for the surfaces?
04:50.38 brlcad cylindrical wedge starts getting down into the weeds of being an interesting starter primitive for someone to learn with
04:51.07 brlcad somewhat limited uses though
04:51.23 pacman87 yeah, but it being on the TODO list struck me as a "why hasn't it happened yet?"
04:52.32 brlcad nah, I'm thinking actually something like tgc where top ij vectors are maybe 2:1 and bottom vectors are 1:2 (looks sorta twisted), then slice through side and base
04:52.58 brlcad that's the ideas page, not quite the same as the TODO file
04:53.17 brlcad though both contain items that are just "would be nice to have" regardless of priority
04:53.25 pacman87 it'd be nice if each primitive had all of its dimensions editable, and the user selects which set are driving, and leaving the rest driven
04:54.25 brlcad not necessarily things that we are actively pursuing
04:54.25 pacman87 ah, i should read the url and/or the page title
04:54.27 brlcad heh, or the big paragraph at the beginning ;)
04:54.29 *** join/#brlcad dreeves_ (n=IceChat7@67.130.253.14)
04:55.02 yukonbob that cylindrical wedge is perdy
04:55.03 pacman87 one of my big questions with the sweep is how to manage the 'orientation along the path'
05:04.48 brlcad you could limit the orientation to make it piecewise -- make the sweep only have a starting orientation/azimuth and an ending azimuth
05:05.03 brlcad then linearly interpolate
05:05.48 brlcad or would take the spline input curve and allow the specification of N orientations along that path ranged from 0.0 to 1.0 distance along the sweep curve (and then linearly interpolate)
05:10.45 brlcad autocad allows the sketch to be perpendicular or at an angle, along a spline path, with twisting angles (azimuths), and scaling factors
05:11.08 brlcad twists and scales are both linearly interp'd, though twists can be more than 360
05:11.44 brlcad http://www.dailyautocad.com/autocad/3d-modeling-8-using-sweep-command/
05:12.11 pacman87 most of my experience is from solidworks
05:13.45 brlcad similar flexibility
05:13.49 brlcad http://www.solidsmack.com/how-the-heck-is-this-made-creating-solidworks-spirals-sweeps-and-helixes/2008-10-01/
05:16.17 brlcad not sure how they did the hourglass other than by tying multiples together or if solidworks has keypoints
05:18.43 brlcad aha, guide curves
05:18.48 brlcad http://www.youtube.com/watch?v=3OUCd8Fp0cE
07:54.27 *** join/#brlcad madant_ (n=madant@117.196.151.246)
08:17.39 *** join/#brlcad starseeker (n=starseek@bz.bzflag.bz)
08:17.46 *** join/#brlcad brlcad (n=sean@bz.bzflag.bz)
08:21.43 *** join/#brlcad poolio (n=poolio@bz.bzflag.bz)
08:23.38 *** join/#brlcad MinuteEl1ctron (n=MinuteEl@bz.bzflag.bz)
08:30.52 *** join/#brlcad d-lo (n=claymore@bz.bzflag.bz)
08:37.38 *** join/#brlcad madant_ (n=madant@117.196.138.104)
08:41.01 madant_ to anybody interested :) theoretical foundations of geometric constraint solving.. the references are the good part ;) http://www.cs.purdue.edu/homes/cmh/electrobook/theory.html
08:49.17 CIA-40 BRL-CAD: 03Homovulgaris 07http://brlcad.org * r1291 10/wiki/Libpc: Basic Structure
08:49.39 CIA-40 BRL-CAD: 03Homovulgaris 07http://brlcad.org * r1292 10/wiki/Developer_Documents: /* Development Projects */
08:50.38 *** join/#brlcad _sushi_ (n=_sushi_@84-72-93-63.dclient.hispeed.ch)
09:53.42 *** join/#brlcad mafm (n=mafm@223.Red-83-49-86.dynamicIP.rima-tde.net)
09:55.22 *** join/#brlcad madant__ (n=madant@117.196.140.211)
09:57.21 *** join/#brlcad madant (n=madant@117.196.140.211)
11:26.11 *** join/#brlcad hippieindamakin8 (n=hippiein@202.3.77.38)
12:36.00 mafm brlcad: any candidate for the GUI framework?
12:37.02 d-lo Why are *YOU* so interested, eh? =D
12:37.16 mafm because I might have to do some mentoring
12:37.56 mafm or hide behind the armchair in the case that s/he embarrases me with his/her skills :)
12:43.03 mafm d-lo: are you going to be a mentor this year? does it depend on the projects elected, I guess
12:43.27 d-lo mafm: could be one!
12:47.26 mafm d-lo: I don't if I should say congrats, or I feel for you :)
12:50.12 *** join/#brlcad elite01 (n=omg@unaffiliated/elite01)
13:23.44 *** join/#brlcad madant (n=madant@117.196.144.175)
13:49.25 *** join/#brlcad d_rossberg (n=rossberg@bz.bzflag.bz)
13:50.25 *** join/#brlcad SWPadnos (n=Me@dsl107.esjtvtli.sover.net)
14:04.46 *** join/#brlcad hippieindamakin8 (n=hippiein@202.3.77.38)
14:06.13 ``Erik_ "csg ray trace optimisations" sounds insanely vague O.o
14:25.00 *** join/#brlcad pacman87 (i=500@resnet-46-40.dorm.utexas.edu)
14:53.23 starseeker sees the article on the Texas school board and winces
15:07.24 *** join/#brlcad ``Erik_ (i=erik@c-76-111-12-116.hsd1.md.comcast.net)
15:24.27 *** join/#brlcad ``Erik__ (i=erik@c-76-111-12-116.hsd1.md.comcast.net)
15:57.07 *** join/#brlcad ``Erik___ (i=erik@c-76-111-12-116.hsd1.md.comcast.net)
15:57.10 *** join/#brlcad hippieindamakin8 (n=hippiein@202.3.77.38)
16:00.04 *** join/#brlcad MinuteElectron (n=MinuteEl@unaffiliated/minuteelectron)
16:07.32 CIA-40 BRL-CAD: 03d_rossberg * r34073 10/brlcad/trunk/src/conv/CMakeLists.txt: included g5-g4 downgrade tool
16:11.35 CIA-40 BRL-CAD: 03d_rossberg * r34074 10/rt^3/trunk/ (include/brlcad/Object.h src/coreInterface/Object.cpp): the return value of Copy() is always ignored
16:13.43 CIA-40 BRL-CAD: 03d_rossberg * r34075 10/rt^3/trunk/ (8 files in 2 dirs): revised error handling
16:52.14 *** join/#brlcad cad38 (n=c752f34a@bz.bzflag.bz)
17:11.35 *** join/#brlcad elite01 (n=omg@unaffiliated/elite01)
17:25.05 CIA-40 BRL-CAD: 03Adminssmith 07http://brlcad.org * r1293 10/wiki/Main_Page:
17:46.39 starseeker brlcad: hmm, the db5.h header has a DB5_MINORTYPE_BRLCAD_COMBINATION entry, but the spec lists combinations as a non-geometry type
18:06.39 d-lo www.nbos.com
18:17.31 ``Erik nifty, I'm back online
18:17.36 ``Erik yay, comcast!
18:17.45 d-lo They finally fix your connection?
18:17.50 ``Erik no
18:17.53 ``Erik yay, they suck. :D
18:22.13 CIA-40 BRL-CAD: 03Sean 07http://brlcad.org * r0 10/wiki/Special:Log/block: blocked [[User:Adminssmith]] with an expiry time of infinite (account creation disabled): Spamming links to external sites
18:22.42 CIA-40 BRL-CAD: 03Sean 07http://brlcad.org * r1294 10/wiki/Main_Page: Undo revision 1293 by [[Special:Contributions/Adminssmith|Adminssmith]] ([[User talk:Adminssmith|Talk]])
18:23.23 ``Erik huh, nutty, one of their distributors is in bellevue, wa
18:23.36 ``Erik drove there to buy cd's and blue hair dye back in the day
18:26.34 d-lo blue? heh. heheh.. ....ROFL
18:26.51 ``Erik O.o what? :D
18:27.01 ``Erik it was the mid 90's in the seattle area
18:27.06 ``Erik I was COOL, yo! (or somethin')
18:27.38 hippieindamakin8 ``Erik, mid 90s and Seattle woah wat a place to be :)
18:28.17 hippieindamakin8 but i would prefer going back to the times of woodstock and me a hippie ;)
18:28.29 hippieindamakin8 68 i meant
18:34.23 starseeker mutter mutter... grip and joint have geometry minor types too, despite having minor types in major type 1
18:36.09 starseeker not clean
18:43.50 starseeker wishes these geometry minor types could have been assigned in an order based on mathematical grouping... oh, well
18:46.01 CIA-40 BRL-CAD: 03starseeker * r34076 10/brlcad/trunk/doc/docbook/BRL-CAD_db_format.xml: Add in content up to geometry minor types table.
18:55.08 CIA-40 BRL-CAD: 03starseeker * r34077 10/brlcad/trunk/doc/docbook/BRL-CAD_db_format.xml: Add geometry attributes minor type section.
18:55.36 starseeker whoops, that was a major type session
19:42.49 *** join/#brlcad jdoliner (n=jdoliner@c-68-51-76-57.hsd1.il.comcast.net)
20:00.13 *** join/#brlcad ``Erik (i=erik@c-76-111-12-116.hsd1.md.comcast.net)
20:40.20 *** join/#brlcad tina__ (n=tina@f053144017.adsl.alicedsl.de)
20:42.24 *** part/#brlcad tina__ (n=tina@f053144017.adsl.alicedsl.de)
20:44.35 *** join/#brlcad BigAToo (n=BigAToo@66.103.88.193)
20:48.45 *** join/#brlcad BigATo1 (n=BigAToo@66.103.88.193)
21:07.04 *** join/#brlcad andrecastelo (n=chatzill@189.71.33.19)
21:45.18 CIA-40 BRL-CAD: 03starseeker * r34078 10/brlcad/trunk/doc/docbook/BRL-CAD_db_format.xml: Add spec contents up to Object Length.
21:51.02 starseeker <snort> I like the part were it says "detailed specifications of each of these are necessary..."
22:06.19 *** join/#brlcad _sushi_ (n=_sushi_@77-58-245-172.dclient.hispeed.ch)
22:07.13 CIA-40 BRL-CAD: 03Paulcs 07http://brlcad.org * r1295 10/wiki/Animation: /* Compositing an animation */
22:11.54 CIA-40 BRL-CAD: 03Paulcs 07http://brlcad.org * r1296 10/wiki/Animation: /* With Mencoder */
22:12.20 *** join/#brlcad csanyipal (n=csanyipa@252-144-85-95.dynamic.stcable.net)
22:12.28 csanyipal howdy!
22:13.47 csanyipal I want just to say that that I wrote down a small addon to http://brlcad.org/wiki/Animation.
22:14.16 csanyipal One can use Mencoder to create an animation from png files.
22:14.33 *** join/#brlcad hippieindamakin8 (n=hippiein@202.3.77.38)
22:53.23 Ralith wonders if there are any version control systems that don't suffer horrible lag when dealing with codebases this large
22:55.21 Ralith huh, this is weird
22:55.45 Ralith 'svn diff' is showing lots of changes that I didn't make; it's like instead of applying some revisions it just skipped them and assumed I'd deleted them.
22:56.43 Ralith is tempted to revert his changes, but that'd mean starting over from scratch.
22:56.50 Ralith that's what I get for going idle for a long period, I guess
23:07.45 Ralith looks like something to do with lineendings.
23:09.03 Ralith someone wiped out all the ^Ms and for some reason svn thinks I put them back?
23:17.15 Ralith begins making extensive use of revert --recursive
23:20.32 Ralith looks like it's mostly in .dsp files.
23:21.39 Ralith okay, that halved the diff size.
23:23.56 Ralith .dsp files are huge.
23:26.27 Ralith hey, I wonder how hard it would be to write a tool to evaluate a single line of cpp syntax in context.
23:27.39 ``Erik "cpp syntax in context"?
23:27.39 Ralith okay, I'm back down to my actual changes! yay! :D
23:28.08 Ralith ``Erik: as in, considering other cpp expressions in the file, such as includes and defines, as would be necessary for useful behavior in most cases.
23:28.21 Ralith basically, run cpp on the file and extract the corresponding line.
23:28.47 Ralith does cpp offer any way to determine the corresponding line, or would that have to be hacked on?
23:28.52 ``Erik ah, I usually just do gcc -E | less and search for something close
23:29.03 Ralith yeah, but wouldn't it be awesome to have it bound in emacs?
23:29.30 Ralith perhaps a minibuffer message showing the expansion, and a keybind to perform the replacement.
23:30.12 ``Erik I'm sure a little elithp could handle that fairly easily, provided you can pull the actual command to compile the object
23:30.54 Ralith wouldn't it just be a matter of running cpp on it?
23:31.04 Ralith still not sure how to extract the relevant line(s) though.
23:31.16 ``Erik post-cpp code has #1234 springled through it to help the compiler get line numbers correct, be a matter of finding the greatest line indicator before your line of interest, then counting (mebbe doing 5 or so lines to help with context)
23:31.29 Ralith oo
23:31.31 Ralith that's helpful
23:31.44 ``Erik gcc -E myprog.c > myprog.cpp ; grep '^#[0-9]' myprog.cpp
23:31.57 Ralith is gcc -E perferable to cpp?
23:31.58 ``Erik (or .e if you don't want to confuse the more retarded c++ developers)
23:32.07 Ralith lol
23:32.09 ``Erik um, they should execute the same code
23:32.17 Ralith hm, I suppose one problem would be getting the include paths right
23:32.33 ``Erik yup, thus my comment about getting the actual command to compile the object :)
23:32.38 Ralith ahh, that's what you meant.
23:32.41 ``Erik then you could s/-c -o blah.o/-E/
23:32.47 Ralith yeah.
23:32.58 Ralith perhaps have it parse Makefiles?
23:33.03 Ralith that's getting kind of specialized though
23:33.39 Ralith ...weird, I appear to have finished these changes.
23:33.56 ``Erik iirc, emacs has the ability to execute make with some combination of 18 keys held down simultaniously
23:34.24 ``Erik finding how it does that, and changing it to do "rm thisfile.o && make thisfile.o" to crib it
23:34.31 Ralith only three, actually. I used to know it, but I've been using enough arcane toolchains lately that I've just been running a terminal instance in the same desktop and switching to that.
23:34.43 ``Erik or, if you use bsd make instead of gnu make, make -V might be able to do it
23:34.55 Ralith well, that's unportable
23:35.10 ``Erik yeah, but expecting to use gcc is unportable, as well
23:35.46 ``Erik would imagine a couple big honkin' (cond) statements
23:35.48 Ralith point.
23:35.56 Ralith yeah that's just inelegant
23:36.11 Ralith at least assuming gcc works across probably 95% of installations
23:36.31 Ralith certainly in our case.
23:36.35 ``Erik hrm
23:36.45 ``Erik 'cept we test compile on like 5 different compilers, dude
23:36.53 Ralith but what do we all *use*?
23:36.58 ``Erik mipspro, sunw, ...
23:37.04 ``Erik a couple msvc's
23:37.28 ``Erik mebbe we should use tendra :D
23:37.32 Ralith besides, we'd have to assume makefiles to get anywhere
23:37.47 Ralith and that's probably less common than gcc in the world at large
23:37.54 Ralith well, who knows
23:38.06 ``Erik aaanyways, it's all shades of grey, where to draw the line is an engineering decision :)
23:38.10 Ralith ultimately I think gcc's a reasonable assumption for the purposes of a convenience tools.
23:38.13 Ralith tool*
23:38.21 ``Erik primarily developes on FBSD, which has been itching to dump gcc for quite a while
23:38.33 ``Erik some argued tendra, now there's another one that's getting attention
23:38.34 Ralith pcc does look shiny.,
23:38.41 bjorkintosh in favor of what?
23:38.44 Ralith see above
23:38.49 ``Erik izzit pcc?
23:38.52 Ralith pretty sure.
23:39.03 Ralith that's the one that, iirc, OpenBSD has been giving lots of attention to
23:39.09 ``Erik sounds about right *shrug* I d'no, I got tired of all the bikeshedding around that topic
23:39.17 Ralith and OpenBSD seems to have a habit of starting trends.
23:39.33 Ralith pcc is interesting to me because last I heard of it it was much faster than gcc.
23:39.52 Ralith not that that's *too* valuable in the long run, but handy in a dev environment, and on source-based systems.
23:40.05 ``Erik fast compiling, or faster generated code?
23:40.08 Ralith the former
23:40.22 Ralith I'd be surprised if they could match gcc's slew of optimization options.
23:41.04 ``Erik the bsd's feed off of eachother a lot, they're all radically different than eachother (far more different than linux distros), but have a lot of community overlap and listen to eachother well, it's neat
23:41.22 ``Erik uhm, until recently, gcc was pretty shitty at optimizing, even microsoft was wiping the floor with gcc3
23:41.22 Ralith yeah, BSD's awesome.
23:41.29 Ralith right, until recently.
23:41.33 Ralith not so anymore.
23:41.36 Ralith (or so I hear.)
23:41.42 ``Erik the new, uh, TSA code in 4 made gcc40 far slower, but it sped up a lot pretty fast
23:42.01 Ralith I have to confess I've installed linux on my desktop, though
23:42.13 ``Erik used to have linux machines, now only has them at work :)
23:42.13 Ralith sometimes you just gotta run some flash :P
23:42.22 ``Erik <-- pets his hackintops
23:42.43 Ralith and I've also been toying with graphics stuff lately, and nvidia hasn't released a Cg compiler for FreeBSD yet.
23:43.16 Ralith I really wish the hardware industry was more open.
23:43.17 ``Erik the first thing I did when I bought my old g3 ibook was watch strongbad email. all of it. :D
23:44.02 ``Erik nvidia drags its feet a LOT with the fbsd stuff, I had to email them to get a release to allow a 3 year old card to work :(
23:44.20 Ralith I'm just thankful they support it at all.
23:44.37 ``Erik that was a fight and a half heh
23:44.47 Ralith well, less thankful, more amazed.
23:44.56 ``Erik they didn't move until people were reverse engineering their hardware out in public
23:45.09 Ralith it's even somewhat more stable than the nvidia/X I have set up on this linux box
23:45.21 ``Erik I imagine it was just to throw a carrot to prevent their "vital secrets" from being figured out by angsty fbsd fools
23:45.22 ``Erik :)
23:45.25 Ralith though that may be more due to FreeBSD than nvidia.
23:45.29 Ralith lol
23:45.33 Ralith that reminds me
23:45.38 Ralith how's ATI doing with that whole releasing the docs thing?
23:45.40 ``Erik quickest way to make a fbsd unstable - nvidia driver. :D
23:45.45 ``Erik <-- hasn't been following
23:46.28 Ralith 'kay
23:46.35 Ralith cuz that was really exciting when I first heard about it
23:46.41 Ralith but didn't really seem to go anywhere.
23:49.30 ``Erik "marylin manson???" "yeah, it's all his or her fault! I'm going to give him or her a piece of my mind or penis" hah
23:51.41 Ralith ``Erik: on a more topical subject: I'm a bit unsure what proper procedure is for me with respect to committing code (in this case, a removal of the redundant vecmath.h from lgt). I've got write access, or did last I checked, but I'm uncertain if I should be requesting review (after testing to the best of my ability, of course) before checking changes in.
23:53.23 Ralith (said removal is nontrivial as it required swapping out macros with different syntax, and I know of no clear way to verify that it all still works the same once compiled)
23:55.12 ``Erik the benchmark suite? :)
23:55.20 Ralith covers lgt?
23:55.28 ``Erik commits can always be undone, but if you want review, you can always post the patch somewhere
23:55.34 ``Erik oh, lgt, um
23:55.37 ``Erik d'no
23:56.04 Ralith yeah, I'm just wondering if I should make a point of requesting review.
23:57.26 ``Erik depends on how confident you are :) it'll get reviewed either way
23:58.13 ``Erik the commit bit is the "we trust you enough to do what's right, or suffer the horrible reprecussions"
23:58.28 Ralith kk, thanks for clearing that up
23:58.42 ``Erik w00t, clear as mud! :D

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