| 00:23.50 | yukonbob | hello, cadheads |
| 00:24.06 | pacman87 | hi yukonbob |
| 00:24.15 | yukonbob | what's happening, pacman87? |
| 00:24.24 | pacman87 | um, not my raytracing |
| 00:24.36 | yukonbob | feh... |
| 00:24.44 | yukonbob | try solidworks ;) |
| 00:24.59 | pacman87 | i mean my primitive's raytracing |
| 00:25.06 | yukonbob | knows -- is joking |
| 00:25.14 | pacman87 | i dont thing solidworks does hyperboloids, anyway |
| 00:25.26 | yukonbob | hits google |
| 00:26.06 | yukonbob | my Net connection is working about as well as your raytracing :P |
| 00:29.05 | *** join/#brlcad Twingy (n=justin@74.92.144.217) |
| 00:29.50 | yukonbob | sees something about developing hyperboloid drill bits in solidworks in summaries... |
| 00:46.39 | pacman87 | progress: https://webspace.utexas.edu/trv82/www/hyp_rt3.png |
| 00:46.51 | pacman87 | still not sure why the bottom plate is showing, though |
| 01:20.40 | *** join/#brlcad iraytrace (n=iraytrac@c-69-250-125-121.hsd1.md.comcast.net) |
| 01:36.16 | brlcad | pacman87: woot, nice progress! |
| 01:36.29 | brlcad | my guess would be sort order |
| 01:36.45 | brlcad | the in-out segment points you add to the result should be sorted |
| 01:37.40 | brlcad | if you have the segment inverted so you present the back-facing bottom plate exit point before the side-wall entry point, you could see that kind of effect |
| 01:38.33 | brlcad | course could be something wrong with the middle surface.. those normals are whack |
| 01:46.54 | andrecastelo | hey brlcad .. the rt applications consist of view[app].c + worker.c + do.c + opt.c + main.c, right? what functions are specific to the view file, in this case viewmlt.c? i've reorganized the declarations and the mlt project file to be compliant with this system |
| 01:50.18 | brlcad | andrecastelo: there's about a dozen view*.c example rt apps that show the various call-backs you have to hook |
| 01:51.27 | andrecastelo | brlcad: ok, I'll look for them, thanks :) |
| 01:53.57 | brlcad | I believe a lot of it is covered in http://brlcad.org/w/images/3/3d/Application_Development.pdf too |
| 02:10.02 | andrecastelo | brlcad: yeah but it kind of confused me, because it has main() and etc. But i already solved the problem :D |
| 02:13.43 | brlcad | yeah, the part that is relevant is the "RTUIF" that it refers to , the view callbacks it refers to |
| 02:13.56 | brlcad | those are what are in your viewmlt.c file |
| 02:14.44 | brlcad | the tutorial starts from one level deeper, and also describes how to implement the rt front-end in addition to view interface |
| 02:18.57 | CIA-21 | BRL-CAD: 03andrecastelo * r31288 10/brlcad/trunk/ (misc/win32-msvc9/rtmlt/rtmlt.vcproj src/rt/viewmlt.c): |
| 02:18.57 | CIA-21 | BRL-CAD: Merged definitions from mlt_defs.h into viewmlt.c. Added functions to make |
| 02:18.57 | CIA-21 | BRL-CAD: viewmlt.c a rt compatible application. Added missing files (worker.c, do.c, |
| 02:18.57 | CIA-21 | BRL-CAD: main.c and opt.c) to win32-msvc9/rtmlt/rtmlt.vcproj. It should build correctly |
| 02:18.57 | CIA-21 | BRL-CAD: now. |
| 02:19.11 | andrecastelo | mlt_defs.h is useless now :S |
| 02:20.38 | brlcad | so remove it? |
| 02:20.51 | brlcad | nothing wrong with adding/removing as needed as it evolves :) |
| 02:20.58 | brlcad | don't forget the Makefile.am if you remove it ;) |
| 02:21.06 | andrecastelo | okay ;} |
| 02:23.21 | CIA-21 | BRL-CAD: 03andrecastelo * r31289 10/brlcad/trunk/src/rt/ (Makefile.am mlt_def.h): |
| 02:23.30 | andrecastelo | :S |
| 02:24.18 | andrecastelo | brlcad: sorry, can i revert that to add a decent commit message? |
| 02:51.53 | *** join/#brlcad Twingy (n=justin@74.92.144.217) |
| 03:15.35 | brlcad | andrecastelo: heh, nope -- you can never go back! |
| 03:15.48 | andrecastelo | :O |
| 03:16.40 | brlcad | if you really really want to leave a note, make a ws commit and annotate in the commit message that it is for r31289, etc |
| 03:16.41 | andrecastelo | ws commit? |
| 03:16.46 | brlcad | ~ws |
| 03:16.49 | ibot | ws is probably short for workstation. White Space, or the country code for Western Samoa |
| 03:18.10 | andrecastelo | ok, i'll do a western samoa commit ;) |
| 03:20.18 | brlcad | :) |
| 03:22.01 | CIA-21 | BRL-CAD: 03andrecastelo * r31290 10/brlcad/trunk/src/rt/viewmlt.c: WS Commit. In the last commit (r31289), removed mlt_defs.h and edited src/rt/Makefile.am accordingly. Definitions in mlt_defs.h are now in viewmlt.c. |
| 03:30.26 | poolio | brlcad: so is src/librt restructured as you wanted now or is there still more? |
| 03:42.32 | pacman87 | so i fixed my bottom plate, but the normals are definately off |
| 03:48.15 | louipc | ~pacman |
| 03:48.49 | *** join/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1177726039.dsl.bell.ca) |
| 03:48.54 | pacman87 | louipc: ?? |
| 03:50.15 | louipc | just wondering what ibot had for that |
| 03:50.29 | louipc | pacman is my package manager :D |
| 03:50.43 | pacman87 | ~pacman87 |
| 03:51.19 | louipc | besides most being most popularly known as a video game |
| 03:51.30 | pacman87 | ~pacman87 |
| 03:51.30 | ibot | well, pacman87 is not a package manager |
| 03:51.38 | louipc | errr - a most |
| 03:52.09 | louipc | nope |
| 03:55.09 | brlcad | poolio: pretty much |
| 03:55.45 | brlcad | not that any restructuring should hold anyone back from coding... :) |
| 03:56.09 | IriX64 | code blue, code blue :) |
| 03:56.24 | brlcad | the idea for the brep routine, though, will be to add a cxx file to each primitive for a brep routine (similar to the nurb routine) |
| 03:56.30 | brlcad | hello IriX64 |
| 03:56.36 | IriX64 | regards |
| 03:56.40 | louipc | obama wins |
| 03:56.55 | brlcad | did you ever talk to your friend about licensing on his bolt program? |
| 03:56.58 | poolio | brlcad: jah. I just wanted to commit either the sph or cyl brep stuff, just to get an idea of where things would be going :) |
| 03:57.17 | IriX64 | is it still there? |
| 03:57.19 | brlcad | getting the subdir was the main point |
| 03:57.27 | IriX64 | haven't looked at that site lately |
| 03:57.32 | brlcad | so maybe something like xxx_brep.cpp for each |
| 03:57.40 | brlcad | IriX64: has nothing to do with the site |
| 03:57.52 | poolio | Ok. and you just want the brep() routine? |
| 03:57.54 | brlcad | you showed me a bolt creation program that he made |
| 03:58.15 | IriX64 | i *thought i got the code from clock, can't find him now |
| 03:58.22 | brlcad | no, it wasn't clock |
| 03:58.31 | IriX64 | yes it was |
| 03:59.11 | IriX64 | mmm maybe another clock? |
| 03:59.59 | brlcad | why would you insist it was? |
| 04:00.09 | brlcad | you don't even seem to remember what I'm talking about |
| 04:00.23 | IriX64 | remeber some guy going by that nick sending it to me |
| 04:00.40 | brlcad | well, clock's name isn't Jamie |
| 04:00.44 | IriX64 | pastebined it |
| 04:01.03 | IriX64 | now im confused |
| 04:01.32 | brlcad | forget it, you apparently didn't care enough to follow up |
| 04:01.34 | brlcad | i'll just delete it |
| 04:01.41 | IriX64 | sorry man |
| 04:02.25 | IriX64 | gotta test my client, want it to msg nickserv when i start up |
| 04:02.31 | *** part/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1177726039.dsl.bell.ca) |
| 04:05.19 | *** join/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1177726039.dsl.bell.ca) |
| 04:25.04 | *** part/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1177726039.dsl.bell.ca) |
| 04:49.56 | *** join/#brlcad alex_joni (n=juve@emc/board-of-directors/alexjoni) |
| 05:28.26 | CIA-21 | BRL-CAD: 03brlcad * r31291 10/brlcad/trunk/configure.ac: add a tcl+itcl sanity check given how we can't build our version of incrTcl with a system 8.4 .. it's both or neither. |
| 05:30.33 | *** join/#brlcad prasad1 (n=psilva@h-72-245-122-226.mclnva23.covad.net) |
| 05:30.52 | brlcad | evenin prasad1 |
| 05:34.08 | pacman87 | success! https://webspace.utexas.edu/trv82/www/hyp_rt6.png |
| 05:47.09 | pacman87 | and here's a shell: https://webspace.utexas.edu/trv82/www/hyp_rt7.png |
| 05:55.34 | pacman87 | unfortunately, i had to pull all of the generalized code, so now it only works if your main axis is aligned with Z :( |
| 05:56.01 | pacman87 | but at least it works |
| 05:56.13 | pacman87 | goodnight all |
| 07:09.51 | *** join/#brlcad d_rossberg (n=rossberg@bz.bzflag.bz) |
| 10:15.52 | *** join/#brlcad mafm (n=mafm@elnet-111.lip.pt) |
| 10:21.59 | mafm | allo |
| 10:32.14 | *** join/#brlcad thing0 (n=ric@123.208.175.190) |
| 11:25.20 | *** join/#brlcad elite01 (n=elite01@dslb-088-071-046-242.pools.arcor-ip.net) |
| 11:52.58 | *** join/#brlcad andrecastelo (n=chatzill@189.71.68.244) |
| 12:07.22 | *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6) |
| 12:16.52 | *** join/#brlcad clock_ (n=clock@zux221-122-143.adsl.green.ch) |
| 12:35.31 | brlcad | pacman87: awesome! |
| 12:36.06 | brlcad | new primitives are always exciting :) |
| 12:36.56 | brlcad | howdy mafm, how goes it? |
| 12:40.19 | mafm | tired, but other than that OK :) |
| 12:53.08 | mafm | do you support "make distclean"? |
| 12:53.26 | mafm | it's starting to configure things before removing them |
| 13:05.27 | brlcad | yes, of course |
| 13:06.09 | brlcad | whether it has to configure before cleaning depends on the dependency tracking and edit states of files (timestamps) |
| 13:06.39 | brlcad | if you did an svn update, it would have to regenerate and rerun configure before it could perform the clean |
| 13:07.33 | brlcad | automake design, not our doing |
| 13:07.38 | alex_joni | that "bites" sometimes.. it's always best to do an distclean before updating |
| 13:08.22 | alex_joni | (if there was a long period of time since the last update.. and lots of infrastructure changes) |
| 13:09.25 | brlcad | the faster more general rule I usually suggest is if you see configure.ac or any Makefile.am get updated, run: sh autogen.sh && ./configure --enable-all && make clean && make |
| 13:09.55 | brlcad | that'll generally complete more quickly and reliably than trying to force a distclean, and should give the same result |
| 13:10.19 | brlcad | since distclean can actually fail under some update/edit situations |
| 13:15.48 | brlcad | once you get more familiar with what the Makefile's are doing and how the timestamp updates work, you can get away with the localized depcomp rebuilds and automatic reconfigurate |
| 13:16.06 | brlcad | but few get to the level of understanding ;) |
| 13:16.32 | mafm | not very likely, yep :D |
| 13:16.41 | mafm | was cmake considered? |
| 13:16.45 | brlcad | sure |
| 13:16.55 | brlcad | it's used for some of the windows build |
| 13:17.13 | brlcad | to redo the entire build system with it though is a pretty big effort |
| 13:18.06 | ``Erik | cmake adds a new end user dependancy, no? |
| 13:18.14 | brlcad | with pretty minimal payoff -- what we have works and works rather well, robust, familiar, etc, just painful to the autotool newbies (which tend to be in the minority) |
| 13:18.32 | brlcad | yeah, it's not nearly as prevalent |
| 13:19.04 | mafm | I meant mostly as it seems to reduce compilation times, not sure if that is an issue for devels |
| 13:19.10 | ``Erik | yeah, I think we have two "guru" grade automake weenies (if I'm allowed to claim that for myself, heh) who'll fix the newbs mistakes if they're hacking |
| 13:19.24 | brlcad | nah, that's a petty benefit frankly |
| 13:19.31 | ``Erik | good structuring of an automake system will be pretty damn quick |
| 13:19.33 | brlcad | compilation time is dominated by the machine |
| 13:19.52 | ``Erik | I'm tempted to flatten the src/adrt directory for the half second speed boost O>o |
| 13:19.53 | ``Erik | O.o |
| 13:20.26 | ``Erik | (also; we have "make fast" which speeds things up a fair amount... gmake -sj13 fast) |
| 13:21.20 | mafm | well dunno, but it takes something like 5 minutes to run "configure" on my system |
| 13:21.37 | ``Erik | what is your system? the only thing I've seen close to that slow is windows O.o |
| 13:21.38 | brlcad | we'd have to do those same checks with cmake |
| 13:21.43 | brlcad | that problem doesn't go away |
| 13:21.55 | mafm | amd64, 1gb ram, 2 years old |
| 13:21.57 | brlcad | it's just not called a "configure" script |
| 13:22.29 | ``Erik | OS is a bigger impact, I think |
| 13:22.59 | ``Erik | windows is ass slow, fbsd is smoking fast, linux is almost as fast as fbsd, mac is somewhere between bsd/linux and windows |
| 13:23.04 | mafm | Debian GNU/Linux, random :P version |
| 13:23.38 | brlcad | it's I/O intense, if you have slow OS or slow disks, it'll be teh suck |
| 13:23.39 | ``Erik | weird, does it have any obvious stall points? or are you feeding it funny options? mebbe if you "export CFLAGS=-pipe" ? |
| 13:24.47 | brlcad | suspects a bit of exaggeration too, probably closer to 3 minutes on that system |
| 13:25.18 | brlcad | it reports the time, how long exactly? |
| 13:25.21 | ``Erik | "time sh configure" :) |
| 13:25.29 | brlcad | it's in the summary |
| 13:25.41 | ``Erik | how's the conf, burly? just one big win paper so far? |
| 13:26.32 | ``Erik | hrm, hendrix (manic depressive) into old soundgarden (jesus christ pose)... my computer wants me to be active today O.o |
| 13:30.42 | brlcad | there have been several really good papers and posters |
| 13:31.39 | mafm | I'm not doing anything special... it might not even be 5 minutes but instead 2, dunno |
| 13:32.26 | mafm | some of the projects where I worked recommended a full clean build every morning, and things like these are time-consuming, but that's not the case for all projects of course ;) |
| 13:32.29 | brlcad | there was another good one yesterday that dealt with extracting an implicit model from CT scan data for capturing blood vessels, neural connectivity, and other "tube networks" from a friend down at Texas A&M |
| 13:33.01 | ``Erik | <-- tries to avoid touching configure.ac or primary headers as much as possible to avoid that step |
| 13:33.08 | brlcad | mafm: those instructions are recommended only to those that don't learn how the tools work but just use them as tools |
| 13:33.17 | ``Erik | that's why I tend to be conservative about changing or adding directories :) |
| 13:33.24 | brlcad | you can get away with much much faster rebuild times if/when you know when you really need to do what |
| 13:34.33 | ``Erik | (also; 3/(24*60) isn't that big) |
| 13:35.13 | mafm | Elapsed configuration time ...........: 1 minute, 4 seconds <- plus unknown time generating makefiles |
| 13:35.18 | brlcad | heh |
| 13:35.37 | brlcad | that includes makefile generation time |
| 13:35.58 | mafm | hmm? it starts to generate them after that |
| 13:36.19 | ``Erik | then try "time sh configure" ? |
| 13:36.56 | mafm | http://rafb.net/p/xYG7fn39.html |
| 13:37.07 | mafm | ok there I go ``Erik |
| 13:37.41 | ``Erik | <-- thinks it's a non-issue *shrug* |
| 13:38.11 | ``Erik | better integration with system tcl/tk/[incr]/etc would be a more useful build "optimization" |
| 13:45.24 | *** join/#brlcad Elperion (n=Bary@p54874E03.dip.t-dialin.net) |
| 13:46.45 | brlcad | or realizaing it's less than 1% of a given workday and instead optimizing something that is more likely going to make a more significant impact ;) |
| 13:47.02 | ``Erik | neat http://media.threadless.com//product/282/zoom.gif |
| 13:47.28 | brlcad | you probably spend more time going to the bathroom, you'd save more time "holding it" ;) |
| 13:47.44 | ``Erik | uhm, 0.06944% :) |
| 13:47.58 | ``Erik | of a 24 hour period, anyways |
| 13:48.46 | brlcad | ~3.0/(8*60) |
| 13:48.46 | ibot | 0.00625 |
| 13:48.56 | brlcad | even for an 8 hour |
| 13:49.30 | ``Erik | wonders if he can concoct a horrible command to embed a simple lisp interpreter in ibot O.o :> |
| 13:49.42 | brlcad | this discussion was enough time to run configure more than 20 times over :) |
| 13:49.59 | ``Erik | that'd be a hell of a hack, and I imagine I have things to do that other people consider far more important O:-) |
| 13:50.45 | *** join/#brlcad louipc (n=louipc@76-10-146-181.dsl.teksavvy.com) |
| 13:50.52 | brlcad | multitasks listening to a guy talk about extracting the hippocampus automatically from brain scan data |
| 13:51.36 | ``Erik | wants to keep his hippopatumus in his brain for the time being O.o not ready for mgmt training (matter removal) |
| 13:52.37 | ``Erik | laptops can be dangerous at conferences, I got bored at vis/vast and wrote a trivial raytracer... I got bored at javaone and, uh, read lots of webcomics I think |
| 13:53.49 | archivist | hmm brain removal is that CAZ computer aided zombies |
| 13:53.58 | clock_ | lol |
| 13:54.10 | clock_ | And Jay Adams is DAZ. Drug Aided Zombies. |
| 13:54.25 | ``Erik | heh, amusingly, I saw "shaun of the dead" last weekend :D must just be a zombie week |
| 13:55.10 | ``Erik | so it looks like clinton isn't going to make it to the primaries |
| 13:55.38 | clock_ | ``Erik: how can the mankind appreciate a trivial raytracer when the Military-Industrial Complex has brought us the ultimate non-trivial raytracer, BRL-CAD? |
| 13:55.45 | clock_ | I hope Obama makes it |
| 13:56.00 | clock_ | He's attractive and he's a minority. |
| 13:56.31 | ``Erik | uhm, I can appreciate generating images from a model file after an hour of coding from scratch? |
| 13:56.44 | louipc | I hears something about hilary wanting to be VP |
| 13:57.02 | clock_ | Who has better chances Obama or McCain? |
| 13:57.03 | ``Erik | I wrote a pong game (in one day), too, is that completely useless because other people have written better ones? |
| 13:57.05 | ``Erik | :) |
| 13:57.26 | louipc | that might be a smart move if what they say about people dropping Hilary for McCain is true |
| 13:57.34 | louipc | or dropping Obama |
| 13:57.50 | clock_ | they shouldn't drop Obama Obama has nice thick lips |
| 13:57.52 | louipc | ``Erik: fun exercise? |
| 13:58.00 | clock_ | Sorry for being superficial but aren't the US voters as well? |
| 13:58.10 | louipc | dunno |
| 13:58.38 | clock_ | We need to destroy the terror with all available force |
| 13:59.07 | clock_ | Who's not with us is against us (oh yeah isn't the guy in the TV right?) |
| 13:59.09 | ``Erik | with the choice between a woman and a black man on the democratic ticket, a lot of bigots are going to cement their vote on the white guy :/ (but it seems to me that the bigots tent to vote republican, anyways, so mebbe it won't be a big factor) |
| 13:59.33 | ``Erik | s/tent/tend/ |
| 13:59.39 | clock_ | what does bigot mean? I know bigotry is something bad, but not sure what it actually is |
| 14:00.17 | clock_ | Obama's trousers should accidentally drop in some TV show |
| 14:00.23 | clock_ | Then he would get the women |
| 14:00.29 | clock_ | the women's votes |
| 14:00.48 | louipc | haha |
| 14:00.50 | ``Erik | the US is far too repressive and anti-sex for that |
| 14:01.06 | ``Erik | O.o |
| 14:01.06 | clock_ | McCain would then try as well but he would lose all the women |
| 14:01.17 | louipc | wardrobe malfunction eh? |
| 14:01.35 | ``Erik | well, no matter what, it'll be a step up... how can you go DOWN from bush? :> |
| 14:01.37 | ``Erik | *duck* |
| 14:01.41 | mafm | times: http://rafb.net/p/Re3Sxc27.html |
| 14:01.58 | louipc | we could dig up stalin |
| 14:02.27 | clock_ | They say Obama spent his childhood and adolescence in Honolulu |
| 14:02.31 | clock_ | I wonder if he can surf |
| 14:02.32 | ``Erik | heh, sibera, guantanamo, whatever :) |
| 14:02.56 | clock_ | I hope if he's a president he requests the CIA torture manuals |
| 14:03.09 | louipc | mafm: how about the make? |
| 14:03.10 | clock_ | and changes "waterboarding" to "wakeboarding" |
| 14:03.45 | ``Erik | mafm: I don't think you're taking the cache abilities into consideration with your timing... also; you've spent way more time worrying about it than a month of just doing it in regular dev mode :) |
| 14:07.05 | ``Erik | http://rafb.net/p/Re3Sxc27.html |
| 14:07.11 | ``Erik | bleh, mouse puke |
| 14:07.25 | ``Erik | http://icanhascheezburger.files.wordpress.com/2008/06/funny-pictures-anteater-debugged-laptop.jpg |
| 14:07.27 | ``Erik | there we go :) |
| 14:10.16 | mafm | louipc: make is in the making |
| 14:10.47 | louipc | holy anteater |
| 14:13.11 | ``Erik | organic debugger :) |
| 14:13.31 | mafm | ``Erik: I'm not worrying about that... I just asked if you had though about cmake or something similar alternative, for me it has a lot of advantages but that's not the point :) |
| 14:13.42 | mafm | and I'm running the time tests just because you asked |
| 14:15.25 | ``Erik | we've talked about cmake, a couple other "new" build systems, if you check out an older version, it uses one called "cake", but automake was picked up because it was better than cake for handling diverse platforms with minimal developer concern, ubiquitous and "standard", and we had solid knowledge of it *shrug* BRL-CAD only went auto* 4.5 years ago, converting it was my second task :) |
| 14:20.10 | mafm | I see |
| 14:27.48 | mafm | same old error :S http://rafb.net/p/8K7VU536.html |
| 14:28.56 | ``Erik | sh configure --enable-all --enable-optimized --prefix=/place/it/goes |
| 14:29.43 | brlcad | the refree error is not fixed without --enable-all yet |
| 14:29.53 | ``Erik | afaik, the issue is with funky cpp paths, it's picking up $(top_srcdir)/src/other/tcl/generic/regex.h when it SHOULD be getting /usr/include/regex.h |
| 14:40.57 | brlcad | mafm: yeah, I'd agree that it has several advantages, just performances is a minor one in the big scheme of things .. the biggest is a unified build environment that spans to windows |
| 14:41.47 | brlcad | i wouldn't mind cmake if someone actually fully duplicated our current build system in it .. but I just don't see that happening anytime soon, it'd be weeks of full-time effort if not more |
| 14:42.14 | mafm | same error, maybe I should make clean before reconfiguring with --enable-all? |
| 14:42.26 | brlcad | yeah, have to make clean |
| 14:42.50 | brlcad | Tcl_ReFree is an unresolved symbol.. meaning you have a .o object file somewhere with references to it |
| 14:43.09 | brlcad | so just running make again doesn't invalidate that .o |
| 14:43.14 | ``Erik | (if you know which files, you can just remove the specific .o and .lo files, but it's easier to just make clean) |
| 14:43.35 | brlcad | from make's perspective, there's nothing to do |
| 14:45.33 | mafm | it's librt, but after removing it complained about the file not being present, so I just issued the make clean |
| 14:46.14 | brlcad | that all boils down to learning what the tools are doing :) |
| 14:46.25 | brlcad | there actually is deterministic behavior, but it takes a while to learn the tools |
| 15:29.31 | CIA-21 | BRL-CAD: 03erikgreenwald * r31292 10/brlcad/trunk/src/adrt/adrt.h: bump (c) info |
| 15:31.25 | CIA-21 | BRL-CAD: 03erikgreenwald * r31293 10/brlcad/trunk/src/adrt/slave/ (load.c load.h slave.c): use a struct instead of indexing a void* |
| 15:34.04 | CIA-21 | BRL-CAD: 03erikgreenwald * r31294 10/brlcad/trunk/src/adrt/adrt_struct.h: use a struct instead of indexing a void* |
| 15:34.06 | ``Erik | yeah, helps if I commit the struct I'm using, too |
| 15:55.21 | pacman87 | getting there: https://webspace.utexas.edu/trv82/www/hyp_rt8.png |
| 15:55.35 | pacman87 | time to go fix my normals again |
| 15:57.01 | starseeker | nice! |
| 15:57.36 | starseeker | should finish up mucking with nirt and do something cool like a new primitive... |
| 16:07.49 | pacman87 | https://webspace.utexas.edu/trv82/www/hyp_rt9.png |
| 16:09.00 | pacman87 | now i have to figure out the new primitive file structure so i can commit |
| 16:09.08 | starseeker | Cooooool :-) |
| 16:10.07 | pacman87 | i had to pull all of the speed enhancements and start over, so prep() is pretty useless atm |
| 16:21.59 | brlcad | ~pacman87++ |
| 16:22.01 | *** join/#brlcad andrecastelo_ (n=chatzill@189.71.54.154) |
| 16:23.10 | brlcad | pacman87: how about elliptical caps? |
| 16:24.09 | brlcad | howdy andrecastelo_ |
| 16:29.01 | brlcad | pacman87: there was a pretty decent paper on sweeps about an hour ago, I'll see if I can get it to you :) |
| 16:40.16 | PrezKennedy | howdy brlcad |
| 16:48.01 | *** join/#brlcad dtidrow (n=dtidrow@c-69-255-182-248.hsd1.va.comcast.net) |
| 16:58.02 | mafm | I guess that I should put my code in a different directory... any suggestion? |
| 17:14.19 | pacman87 | brlcad: elliptical caps: https://webspace.utexas.edu/trv82/www/hyp_rt10.png |
| 17:18.49 | pacman87 | one question: how are tangent hits counted? is the same point used for seg_in and seg_out? |
| 17:22.23 | *** join/#brlcad andrecastelo__ (n=chatzill@189.71.54.154) |
| 17:30.36 | andrecastelo | howdy brlcad, what's up? |
| 17:32.09 | *** join/#brlcad dtidrow (n=dtidrow@c-69-255-182-248.hsd1.va.comcast.net) |
| 17:32.14 | andrecastelo | howdy ``Erik :) |
| 17:38.37 | brlcad | mafm: yours is probably the most "distinct" given the external dependencies it's going to require |
| 17:38.59 | brlcad | pacman87: hah, you're on a roll :) |
| 17:39.14 | brlcad | that's pretty f'n sweet |
| 17:39.49 | pacman87 | only at that view though, other angles are giving me flipped normals |
| 17:41.33 | mafm | so... src/weirdo then? :) |
| 17:44.31 | brlcad | so... maybe not in the same module, gimme a min to rummage |
| 17:45.54 | starseeker | spots what might be another nirt bug and gets out his sledgehammer... |
| 17:49.00 | CIA-21 | BRL-CAD: 03brlcad * r31295 10/brlcad/trunk/src/adrt/adrt.h: screw the 'version year'? how about just say what adrt means |
| 18:00.15 | starseeker | looks at direct_output, decides it is raising too many questions , and decides to add it to the bug list... |
| 18:01.59 | brlcad | nirt cmd line example against one of the db g's would be useful |
| 18:03.59 | starseeker | of the bug? |
| 18:04.01 | starseeker | can do... |
| 18:05.41 | brlcad | finally! .. found a way to reproduce the _TclReFree error from here for non-enable-all |
| 18:06.46 | mafm | w00t |
| 18:09.22 | starseeker | http://pastebin.bzflag.bz/d64313679 |
| 18:09.36 | starseeker | that's the first issue... |
| 18:09.48 | starseeker | the second didn't reproduce, so maybe it's not an issue after all... |
| 18:10.06 | brlcad | you mean the "file test.cs is" ? |
| 18:10.12 | starseeker | yep |
| 18:10.20 | starseeker | it ate the last charater |
| 18:10.20 | brlcad | off-by-one somewhere |
| 18:10.28 | starseeker | oh, it gets better |
| 18:10.32 | ``Erik | <-- pats his belly |
| 18:10.34 | brlcad | probably strcpy |
| 18:10.41 | starseeker | if you type dest default (no space) you get back stdout |
| 18:10.44 | brlcad | er strlcpy |
| 18:11.12 | starseeker | if you do it WITH the space, you get a file called default |
| 18:11.54 | brlcad | same problem |
| 18:11.56 | starseeker | could it be the original authors intended that you would have to have a space after a name? |
| 18:12.17 | starseeker | thought dest default without space would create a file called defaul |
| 18:12.28 | brlcad | default == defaul vs "default " == default |
| 18:12.43 | brlcad | yeah, so it's probably a copy happening between those two checks |
| 18:12.43 | starseeker | but default doesn't == defaul |
| 18:12.48 | starseeker | it equals stdout |
| 18:13.06 | brlcad | so it does the strcmp before it gets trimmed |
| 18:13.13 | starseeker | ah |
| 18:13.31 | starseeker | digs a bit... |
| 18:13.36 | brlcad | otherwise, no .. no way needing a space was intended |
| 18:13.59 | brlcad | in face the command interpreter should trim all whitespace |
| 18:14.07 | brlcad | (should but probably doesn't) |
| 18:15.05 | prasad1 | brlcad, where are u? |
| 18:15.13 | brlcad | new york |
| 18:15.26 | brlcad | long island |
| 18:16.03 | prasad1 | sig conference? |
| 18:22.57 | brlcad | acm solid and physical modeling (spm) just ended, now starts ieee international conference on shape modeling (smi) |
| 18:23.13 | brlcad | joint conference |
| 18:24.27 | brlcad | spm was pretty cool .. the second best paper was a fast implementation of nurbs surface-on-surface evaluation as well as fast xyz to uv mapping |
| 18:24.43 | brlcad | (which is what I'm currently working on) |
| 18:24.55 | starseeker | sweeet |
| 18:25.57 | brlcad | will need a few days to get through just the top half-dozen most interesting/useful papers |
| 18:28.42 | CIA-21 | BRL-CAD: 03erikgreenwald * r31296 10/brlcad/trunk/src/adrt/libtie/ (tie_define.h tie_kdtree.c): Move some macros around, eliminate some others... |
| 18:30.10 | CIA-21 | BRL-CAD: 03starseeker * r31297 10/brlcad/trunk/src/nirt/parse_fmt.c: nirt's dest command was chopping off the last character in the filename string |
| 18:30.13 | starseeker | got it |
| 18:30.33 | starseeker | looks like it isn't trimming whitespace at the end, just at the beginning... should probably fix that |
| 18:32.14 | brlcad | yep |
| 18:32.45 | brlcad | if you have a vls, there is a bu_vls_trimwhitespace |
| 18:32.55 | brlcad | er, bu_vls_trimspace() |
| 18:33.10 | starseeker | wishes he had vls... |
| 18:33.22 | starseeker | nirt didn't seem to like vls much |
| 18:33.34 | starseeker | I suppose it didn't exist when it was written or some such? |
| 18:34.03 | brlcad | possible, they're both pretty old |
| 18:34.22 | brlcad | sometimes just devs that don't know about the available facilities |
| 18:34.30 | starseeker | could be |
| 18:34.48 | brlcad | it is a big package after all, you only know what you encounter unless you go looking for something specific |
| 18:35.29 | brlcad | and most resist at first because it's a simple operation to "just do it" yourself |
| 18:35.40 | brlcad | only recent that some of the facilities were even documented in HACKING |
| 18:35.50 | brlcad | just knowledge one acquires |
| 18:35.55 | CIA-21 | BRL-CAD: 03erikgreenwald * r31298 10/brlcad/trunk/src/rt/viewmlt.c: fix mlt_hit prototype |
| 18:36.28 | starseeker | yep. It hurts most in the file-reading side of things - this I can probably throw in a quick hack for now and vls-ify later |
| 18:38.05 | starseeker | checking for whitespace isn't too hard, fortunately... |
| 18:41.08 | brlcad | yeah, something like while(isspace(*cp)) cp++; while(isspace(cp[strlen(cp)-1])) cp[strlen(cp)-1] = '\0'; |
| 18:41.27 | CIA-21 | BRL-CAD: 03bob1961 * r31299 10/brlcad/trunk/src/tclscripts/archer/ArcherCore.tcl: Add code to prevent raytrace control panel being destroyed by user. Decactive panel instead of destroy. |
| 18:42.04 | starseeker | ah nice |
| 18:42.26 | starseeker | oh yeah - heh |
| 18:42.35 | starseeker | managed to hard crash his mac yesterday with archer |
| 18:42.46 | brlcad | neat |
| 18:44.02 | starseeker | really likes archer as compared to mged - maybe libged will help bring it closer to feature parity with mged... |
| 18:47.27 | starseeker | crud - what's the opposite of Ctrl-s on the command line |
| 18:47.57 | starseeker | At, nevermind - Ctrl-Q |
| 18:48.11 | starseeker | Ctrl-q ratjer |
| 18:48.13 | starseeker | rather |
| 18:50.26 | brlcad | interesting conference .. every single attendee has a laptop at the moment |
| 18:51.11 | pacman87 | brlcad: take a picture |
| 18:55.05 | brlcad | heh, alas no camera for the lighting or good view to capture everyone |
| 19:00.54 | CIA-21 | BRL-CAD: 03starseeker * r31300 10/brlcad/trunk/src/nirt/parse_fmt.c: Add quick fix to avoid whitespace at the end of a name - this whole function (OK, most of nirt) should be reworked to use vls at some point. |
| 19:07.46 | starseeker | notes the problem with documenting tools is that you keep having to fix the tool to do what it is supposed to do... |
| 19:10.18 | *** join/#brlcad docelic (n=docelic@78.134.196.26) |
| 19:10.35 | starseeker | good for the tool though... |
| 19:13.15 | mafm | going home now, see you tomorrow |
| 19:22.23 | brlcad | yep, maintenance burden .. that's (part of) why there is a lot of consideration when adopting a new external dependency |
| 19:22.36 | brlcad | those are things you can't easily fix or maintain |
| 19:23.11 | brlcad | and the importance of refactoring the bigger the project gets (for BRL-CAD it's huge never-ending requirement) |
| 19:29.50 | CIA-21 | BRL-CAD: 03bob1961 * r31301 10/brlcad/trunk/src/tclscripts/archer/ArcherCore.tcl: If parentClass is not set, set it to Toplevel regardless of platform. |
| 19:39.08 | starseeker | decides to see if the archer changes avoid the hard crash... |
| 19:39.45 | brlcad | starseeker: is it user-visible? |
| 19:40.14 | starseeker | you mean the dest fix? I suppose so in the sense that if someone tries it they'll get the correct filename now... |
| 19:40.27 | brlcad | i.e. did a release go out with that bug in the last release |
| 19:40.38 | starseeker | almost certainly |
| 19:40.51 | brlcad | the it "almost certainly" .. |
| 19:41.02 | starseeker | the it? |
| 19:41.10 | starseeker | parse error ;-) |
| 19:41.23 | brlcad | then it "almost certainly" .. deserves a note |
| 19:41.28 | starseeker | gotcha |
| 19:41.50 | brlcad | would be good quick sanity check to be certain |
| 19:42.22 | starseeker | k. I've been needing to build a release tarball for comparision purposes anyway... |
| 19:53.07 | *** join/#brlcad Elperion (n=Bary@p54874E03.dip.t-dialin.net) |
| 19:53.44 | brlcad | or find a machine with a /usr/brlcad installed |
| 19:54.26 | brlcad | probably did make the release, the string clamping changes happened recently, but before the last release |
| 19:56.53 | CIA-21 | BRL-CAD: 03brlcad * r31302 10/brlcad/trunk/include/bio.h: account for badness in the Tcl regex header. for this to work, bio.h has to come after the include for regex.h of course, but then regex.h is a system header so it should be before bio.h regardless. |
| 19:57.20 | CIA-21 | BRL-CAD: 03brlcad * r31303 10/brlcad/trunk/src/librt/regionfix.c: regex.h is a system header, include it before bio.h |
| 20:01.30 | brlcad | there, that should fix the tcl_refree error |
| 20:01.41 | starseeker | does happy dance |
| 20:03.26 | starseeker | Yep, it's in 7.12.4 |
| 20:03.46 | starseeker | updates news file, then gets back to attempting to crash the mac... |
| 20:07.03 | CIA-21 | BRL-CAD: 03starseeker * r31304 10/brlcad/trunk/NEWS: Fixed nirt's dest command to use the full supplied filename string instead of chopping off the last character - whitespace is also handled after the string as well as before. |
| 20:08.53 | *** join/#brlcad archivist (n=archivis@host81-149-119-172.in-addr.btopenworld.com) |
| 20:08.59 | CIA-21 | BRL-CAD: 03brlcad * r31305 10/brlcad/trunk/configure.ac: check for strcasecmp so BLT has it properly defined, avoid duplicate symbols at link time |
| 21:16.21 | *** join/#brlcad pacman87 (n=timothy@71.170.63.120) |
| 21:39.50 | pacman87 | another pretty picture: https://webspace.utexas.edu/trv82/www/hyp_rt12.png |
| 21:46.45 | *** join/#brlcad clock_ (n=clock@77-56-94-220.dclient.hispeed.ch) |
| 21:46.51 | pacman87 | hi clock_ |
| 21:48.46 | clock_ | gu] |
| 21:48.48 | clock_ | hi |
| 21:52.50 | andrecastelo | pacman87: looking really good! nearly finished with the hyperboloid? |
| 21:53.03 | pacman87 | well, rt from within mged still fails |
| 21:53.12 | pacman87 | and i still need tess() and curve() |
| 21:53.19 | pacman87 | but it's getting there :) |
| 21:54.56 | andrecastelo | in this pace you will finish the whole gsoc project before the end of june! :D |
| 21:55.47 | pacman87 | it's been a week and a half, and the revolve/sweep will be a lot more math-heavy |
| 22:03.33 | CIA-21 | BRL-CAD: 03pacman87 * r31306 10/brlcad/trunk/src/librt/primitives/hyp/hyp.c: prep(), shot(), and norm() updated: moved most of the calculations from prep() to shot() in order to insure correctness, and updated norm() to account for the changed private data stored in the hitpoint |
| 22:03.42 | pacman87 | now everyone can play with it! |
| 22:15.58 | pacman87 | rt is wrong in the very specific case when you're viewing from 45 degrees off of the main axis in the direction of of the major axis of the ellipse |
| 22:18.07 | andrecastelo | (typo at line 22?) |
| 22:28.22 | ``Erik | O.o |
| 22:29.58 | andrecastelo | howdy ``Erik |
| 22:30.25 | ``Erik | how's it going, andre? |
| 22:30.53 | andrecastelo | i'm happy viewmlt.c is now working together with main.c and the other rtuif files :) |
| 22:31.51 | ``Erik | yeah, gotta get a grid set up and start doing your bidirectional work now :) |
| 22:32.05 | andrecastelo | yup |
| 22:32.46 | ``Erik | uhm, for saving files, I added nifty capabilities in libbu (image.c), but the only rt to use it right now is uhm, viewedge.c... or you can dump straight pix and it'll get cleaned up later |
| 22:35.52 | andrecastelo | hmm i see.. interesting.. it will be nice if i can test the bidirectional work as i go, instead of going blindly, right? |
| 22:36.14 | pacman87 | andrecastelo: good catch on the typo, thanks. |
| 22:36.49 | andrecastelo | pacman87: np, didn't want to mess with the file you're working on |
| 22:37.12 | andrecastelo | (that's why i didn't committed immediately etc) |
| 22:37.15 | CIA-21 | BRL-CAD: 03pacman87 * r31307 10/brlcad/trunk/src/librt/primitives/hyp/hyp.c: change group to primitives and file to hyp.c |
| 22:37.23 | pacman87 | the group needed changing too |
| 22:37.55 | pacman87 | wait, i still didnt fix the typo :( |
| 22:38.34 | andrecastelo | ``Erik: the grid you're talking about, is the final image matrix? |
| 22:38.47 | ``Erik | yeah, it IS nice to have immediate feedback |
| 22:38.48 | CIA-21 | BRL-CAD: 03pacman87 * r31308 10/brlcad/trunk/src/librt/primitives/hyp/hyp.c: typo fix, thanks andrecastelo |
| 22:39.31 | ``Erik | every pixel in the output image has at least one primary ray shot from it, ya gotta set up the initial ray information (with appropriate matrix-fu) |
| 22:43.51 | andrecastelo | ``Erik: i see.. the ray information will be set in the application structure, depending on the run, something like that? taking turns between a pixel and a light source? |
| 22:44.24 | CIA-21 | BRL-CAD: 03andrecastelo * r31309 10/brlcad/trunk/src/rt/viewedge.c: Fixed typo in a comment section (algorythm -> algorithm). |
| 23:02.05 | ``Erik | that's up to you, andre |
| 23:02.46 | ``Erik | heh, I just updated and did a double-take at viewedge.c being updated before looking at irc :D rtedge needs some major reworking |
| 23:05.58 | andrecastelo | ``Erik: hm, what about something like an adapted view_pixel()? it would do something like this - pick the first pixel, set the origin of the ray to the camera, shoot the ray, record the points; set the origin of the ray to the light source, shoot a ray and record hit points; then test if this is an acceptable path |
| 23:06.35 | andrecastelo | (like not going through a wall or not connecting properly, still need to put a little more thought in this) |
| 23:07.32 | andrecastelo | then, if it is a good path, save the path, and mutate the pixels, adding the contribution of this path to the image |
| 23:07.48 | andrecastelo | (iterating through all the pixels) |
| 23:08.14 | andrecastelo | after that, mutate a point in the path.. then calculate the contribution of this path to all pixels |
| 23:08.23 | andrecastelo | what do you think? |
| 23:09.33 | *** join/#brlcad Twingy (n=justin@74.92.144.217) |
| 23:13.02 | andrecastelo | brb |
| 23:23.41 | ``Erik | oi, twingy |
| 23:24.43 | ``Erik | andre, I actually don't know the algorithm well enough to comment on that... I can look it up tomorrow, but what I'm primarily providing for you is knowledge of BRL-CAD, C and general project management :) |
| 23:32.44 | *** join/#brlcad IriX64 (n=mariodot@bas2-sudbury98-1177725958.dsl.bell.ca) |
