| 00:02.21 | CIA-40 | BRL-CAD: 03brlcad * r34047 10/brlcad/trunk/NEWS: |
| 00:02.21 | CIA-40 | BRL-CAD: annotate keith bowman's first commit. he fixed mged's qorot command to properly |
| 00:02.21 | CIA-40 | BRL-CAD: rotate around a vector. this fixes sf bug 2663014 (qorot does not rotate around |
| 00:02.21 | CIA-40 | BRL-CAD: vector) reported by dwayne. was previously rotating around the origin instead |
| 00:02.22 | CIA-40 | BRL-CAD: of the provided xyz point. |
| 00:04.55 | CIA-40 | BRL-CAD: 03brlcad * r34048 10/brlcad/trunk/AUTHORS: additional info about indianlarry |
| 00:21.23 | CIA-40 | BRL-CAD: 03brlcad * r34049 10/brlcad/trunk/src/ (archer/archer.bat mged/mged.bat util/rtwizard.bat): bah, these weren't updated for release. *really* should not have version numbers in these files, nor should it be necessary to set BRLCAD_DATA even on Windows. |
| 00:58.51 | *** join/#brlcad dreeves (n=dreeves@67.130.253.14) |
| 01:15.40 | madant | Kernigan was a smart chap :) |
| 01:16.04 | madant | so was Gould : "We pass through this world but once. Few tragedies can be more extensive than the stunting of life, few injustices deeper than the denial of an opportunity to strive or even to hope, by a limit imposed from without, but falsely identified as lying within." |
| 01:16.34 | madant | "I am somehow less interested in the weight and convolutions of Einsteins brain than in the near certainty that people of equal talent have lived and died in cotton fields and sweatshops." |
| 01:34.21 | CIA-40 | BRL-CAD: 03brlcad * r34050 10/brlcad/trunk/NEWS: |
| 01:34.21 | CIA-40 | BRL-CAD: keith changed the rpp vertex ordering affecting typein and display with rpp's |
| 01:34.21 | CIA-40 | BRL-CAD: now matching arb8 vertex ordering. this addresses an unexpected behavior bug |
| 01:34.21 | CIA-40 | BRL-CAD: reported by dwaynelk in sf bug 2663183 (arb8/rpp vertex order different) |
| 01:37.42 | *** join/#brlcad dreeves_ (n=IceChat7@67.130.253.14) |
| 01:39.51 | CIA-40 | BRL-CAD: 03brlcad * r34051 10/brlcad/trunk/NEWS: keith fixed another bug reported by dwayne where the mirror command was incorrectly handling an rcc's radii resulting in non-perpendicular height vectors. this was reported in sf bug 2679693 (mirror error). |
| 01:46.26 | CIA-40 | BRL-CAD: 03brlcad * r34052 10/brlcad/trunk/NEWS: |
| 01:46.26 | CIA-40 | BRL-CAD: keith updated the clone command to utilize the same naming convention as |
| 01:46.26 | CIA-40 | BRL-CAD: build_region. this implements a long standing request from dwayne to change |
| 01:46.26 | CIA-40 | BRL-CAD: clone's naming conventions. implements sf bug/feature request 2663102 (clone |
| 01:46.26 | CIA-40 | BRL-CAD: does not name correctly) |
| 02:24.50 | CIA-40 | BRL-CAD: 03brlcad * r34053 10/brlcad/trunk/TODO: need a bot-nmg capability |
| 02:26.28 | brlcad | ``Erik: irix posted a patch if you're interested in reviewing it. quick read looks like a reject, but maybe you want to do something with it. |
| 02:29.34 | *** join/#brlcad ashishrai (i=d2d43dfb@gateway/web/ajax/mibbit.com/x-68e0482d73c93ae6) |
| 02:30.27 | *** join/#brlcad bjorkintosh (n=bjork@ip72-204-40-138.fv.ks.cox.net) |
| 02:49.32 | *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6) |
| 03:52.21 | *** join/#brlcad dtidrow (n=Don@c-68-62-76-34.hsd1.mi.comcast.net) [NETSPLIT VICTIM] |
| 03:52.22 | *** join/#brlcad Maloeran (n=maloeran@glvortex.net) [NETSPLIT VICTIM] |
| 03:52.26 | *** join/#brlcad ChanServ (ChanServ@services.) |
| 03:52.26 | *** mode/#brlcad [+o ChanServ] by irc.freenode.net |
| 05:25.25 | *** join/#brlcad dreeves (n=IceChat7@67.130.253.14) |
| 05:48.39 | *** join/#brlcad piksi (i=piksi@pi-xi.net) [NETSPLIT VICTIM] |
| 05:49.18 | *** join/#brlcad alex_joni (n=juve@emc/board-of-directors/alexjoni) [NETSPLIT VICTIM] |
| 06:11.36 | *** join/#brlcad cad96 (n=4382fd0e@bz.bzflag.bz) |
| 06:13.49 | *** join/#brlcad madant (n=madant@117.196.146.134) |
| 07:56.28 | *** join/#brlcad mafm (n=mafm@223.Red-83-49-86.dynamicIP.rima-tde.net) |
| 08:19.28 | *** join/#brlcad madant (n=madant@117.196.149.233) |
| 08:28.49 | mafm | hi |
| 08:40.04 | brlcad | howdy! |
| 08:45.37 | madant | hi mafm |
| 08:45.48 | madant | thesis over ? |
| 10:39.54 | *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6) |
| 10:40.06 | d-lo | Mernin all! |
| 11:00.57 | *** join/#brlcad BigAToo (n=BigAToo@pool-96-230-124-61.sbndin.btas.verizon.net) |
| 11:58.50 | brlcad | hola |
| 12:02.31 | d-lo | mornin brlcad, how goes it? |
| 12:13.14 | brlcad | great |
| 12:59.56 | *** join/#brlcad samrose (n=samrose@c-24-11-214-181.hsd1.mi.comcast.net) |
| 13:29.23 | ``Erik | hrm |
| 13:29.53 | ``Erik | looking at it now, brlcad, I think the data involved is either redundant or irrelevant |
| 13:30.09 | ``Erik | I'll put together a semi-polite reject message |
| 13:32.14 | brlcad | nods |
| 13:34.53 | ``Erik | ok, closed |
| 13:35.22 | *** join/#brlcad BigAToo (n=BigAToo@64.74.225.82) |
| 13:37.35 | ``Erik | hrm, shoulda probably said something about vetting intended changes by me before submitting, oh well |
| 13:37.53 | ``Erik | he did dot the t's and cross the i's, though, I'm kinda impressed |
| 13:39.18 | *** join/#brlcad _sushi_ (n=_sushi_@84-72-93-63.dclient.hispeed.ch) |
| 13:43.44 | *** join/#brlcad elite01 (n=omg@unaffiliated/elite01) |
| 13:45.00 | ``Erik | 4 lines of code, but he put in a copyright header (a little wrong with his starting date) and documented the function |
| 13:45.40 | ``Erik | needs to think of a good right-sized task for the dude |
| 14:00.46 | ``Erik | thought the mged -v woulda done it, but 'svn diff' is elusive, I guess |
| 14:17.16 | *** join/#brlcad andrecastelo (n=chatzill@201008160086.user.veloxzone.com.br) |
| 15:13.30 | *** join/#brlcad samrose (n=samrose@c-24-11-214-181.hsd1.mi.comcast.net) |
| 15:16.13 | brlcad | mged -v should have done it, but was too much |
| 15:16.33 | brlcad | wasn't just the diff, couldn't make the mod itself either |
| 15:22.55 | *** join/#brlcad dreeves (n=IceChat7@67.130.253.14) |
| 15:24.31 | dreeves | I have removed all the memory allocation from the extude solid it would appear on my machine that I got about 521% increase in speed |
| 15:24.43 | dreeves | after some clean up of the code I will submit a patch |
| 15:26.33 | brlcad | awesome! |
| 15:26.43 | brlcad | that's closer to what I would have expected :) |
| 15:27.07 | dreeves | yeah that makes a little more since |
| 15:27.27 | brlcad | actually was expecting about an order of magnitude, but 5x is good too :) |
| 15:27.28 | dreeves | That took me way longer than it should have a little rusty on the math |
| 15:27.54 | dreeves | It maybe on a better pc |
| 15:28.52 | dreeves | On my machine it is only about 13% slower than raytracing a rcc |
| 15:29.58 | dreeves | Also maybe a little more improvement I have only addressed the memory stuff in the root finder there is still a one array being allocated in shoot which I will get rid of before I submit the patch |
| 15:31.26 | brlcad | ah, yeah, there be more gains to be had then |
| 15:32.22 | dreeves | Yeah I don't expect to out perform raytracing an rcc though, do you? |
| 15:33.24 | dreeves | everything outside of the root finder only addressed about 7.5% of the total time before but of course that percentage is higher now |
| 15:36.47 | ``Erik | no,he got the mod, he kept trying to do other crap, though |
| 15:37.08 | brlcad | shouldn't outperform it, rcc should be one quadratic polynomial |
| 15:37.43 | ``Erik | hm, the root finder needs tlc, there're known issues with it |
| 15:37.49 | ``Erik | iirc |
| 15:38.13 | brlcad | interesting, http://gram.eng.uci.edu/~bobrow/papers_files/CAD_1985_NC_toolpath.pdf |
| 15:38.22 | dreeves | plus the end caps but still though I don't think it is going to get faster |
| 15:39.25 | dreeves | Well the root finder that extrude is using appears that only extrude is using it. I might be missing something though but I couldn't find anything else in brlcad that called it |
| 15:39.28 | ``Erik | btw, hiya dreeves *wave* nice to meet an old developer, I'm one of the newer ones :) |
| 15:39.43 | dreeves | hi there |
| 15:39.44 | ``Erik | hrm, there was a semi-generalized one that I believe tor used |
| 15:40.02 | brlcad | it is specific to extrude |
| 15:40.16 | dreeves | is that question? |
| 15:40.37 | ``Erik | ok, why would extrude not use the generalized one? |
| 15:40.41 | brlcad | it should have been moved up into src/librt/primitives, but was left out during the move |
| 15:41.43 | brlcad | ``Erik: arbitrary order, libbn's is hard-coded to quintic order iirc |
| 15:42.09 | ``Erik | huh, I thoguth the bn one was A) in bu and B) fully gneralized (thus the fp fuzz errors) |
| 15:42.44 | ``Erik | obviously, I'm not up to speed :) and aint' gonna be cuz I'm on vacation, later, ya'llz |
| 15:43.04 | ``Erik | missoura representin' |
| 15:43.18 | brlcad | it is generalized, just limited by array bounds (for performance and memory usage reasons) |
| 15:43.58 | madant | :) i remember it has good optimizations for quintic.. |
| 15:44.12 | madant | but i thought there was no generalized rootsolver in bn |
| 15:44.54 | dreeves | yeah I didn't see one |
| 15:45.03 | dreeves | not in bn |
| 15:45.31 | brlcad | ah right, it's still in librt |
| 15:45.42 | brlcad | bu has a few of the special cases |
| 15:45.56 | brlcad | src/librt/roots.c |
| 15:46.13 | ``Erik | that's wrong. I'll migrate it in a week or so if you don't wanna deal with it |
| 15:46.23 | brlcad | rt_poly_roots() |
| 15:46.43 | ``Erik | fails the test of least surprise |
| 15:47.13 | ``Erik | but I |
| 15:47.14 | brlcad | it wasn't put in libbn because it wasn't/isn't fully generalized (as seen by various possible instabilities when the guesses are poor or the polynomial "hard") |
| 15:47.27 | brlcad | similar to the bezier solver |
| 15:47.32 | ``Erik | I'm ignorin' all of this for a bit, got life to live :) |
| 15:47.40 | madant | :D |
| 15:47.44 | ``Erik | SPRING BREAK! *takes off shirt and runs around* |
| 15:48.12 | ``Erik | happy st pattys day, ya'll |
| 15:48.43 | brlcad | ``Erik: don't worry about migrating it, madant already did it |
| 15:49.20 | brlcad | still waiting to be verified |
| 15:49.34 | madant | what about my last years patch :) regression tests over ? |
| 15:49.56 | brlcad | hm? |
| 15:50.35 | brlcad | it's not been regression tested yet, that's why it's still there .. it needs to be soon though |
| 15:50.42 | ``Erik | aight, I may try to build a test suite when I get back, then |
| 15:50.49 | ``Erik | remind me, though |
| 15:51.02 | ``Erik | like put a postit on my monitor |
| 15:51.51 | ``Erik | I'm gonna bury myself in lithp web app stuff now and ignore irc |
| 15:52.04 | madant | hmm.. i could take another look at it if there is a problem after regression tests :) grr.. i will have to right :D |
| 15:58.05 | madant | going to sleep at 10 pm after a long long time :) |
| 16:00.03 | brlcad | :) |
| 16:47.52 | dreeves | So got rid of the remaining allocs and now it is a little over 6x the speed of before |
| 16:49.28 | dreeves | Could probably optimize the math to improve speed but I think I will focus on getting things cleaned up so I can submit as a patch now... |
| 16:50.33 | dreeves | BTW I was wrong it is only about half the speed of the rcc I think I was remembering numbers when I rt'ed at a higher res |
| 16:55.44 | brlcad | I wouldn't be too suprised by rcc being 2x-10x.. big range, but not too horribly different (at least for simple sketch/extrusions) |
| 16:56.27 | brlcad | the main difference would be that adding more rccs just increases time nearly linearly whereas more complex sketches increases non-linearly |
| 16:56.40 | brlcad | multiple sketches is probably still just nearly linear though |
| 16:57.11 | brlcad | either way pretty cool -- that would make bezeir's actually kinda practically useful now performance-wise |
| 16:57.29 | dreeves | definitly with some optimization that would be very true |
| 16:58.02 | dreeves | But I think for now it has crossed the line performance-wise |
| 16:58.41 | dreeves | When I get it clean up I will put it on sf |
| 16:59.09 | brlcad | yeah, I wouldn't work too hard on it simply because it's a pita to create them interactively still -- the interactive sketcher sucks, you can make them more easily using code than you can interactively |
| 16:59.16 | brlcad | sounds good |
| 17:08.17 | *** join/#brlcad ``Erik_ (i=erik@c-76-111-12-116.hsd1.md.comcast.net) |
| 18:03.14 | *** join/#brlcad BigAToo (n=BigAToo@64.74.225.82) |
| 18:32.43 | *** join/#brlcad schwinn434 (n=schwinn4@cpe-75-81-198-192.we.res.rr.com) |
| 19:47.46 | *** join/#brlcad dreeves (n=IceChat7@67.130.253.14) |
| 19:50.58 | *** join/#brlcad _sushi_ (n=_sushi_@77-58-239-253.dclient.hispeed.ch) |
| 19:55.37 | CIA-40 | BRL-CAD: 03starseeker * r34054 10/brlcad/trunk/db/ (Makefile.am goliath.asc): Add Goliath Tracked Mine model by Stephen Kennedy and Stephanie Weaver |
| 20:20.58 | brlcad | woot |
| 20:21.23 | brlcad | (NEWS) |
| 20:22.17 | brlcad | should also include another attribute on _GLOBAL that says that it's public domain and credit the museum |
| 20:23.08 | brlcad | or at least reference them saying the original used as a measurement reference was located at the apg museum |
| 21:45.18 | dreeves | brlcad do I need to perserve the oringinal function signatures in bezier_2d_insect in particular FindRoots and Bezier? |
| 21:45.50 | dreeves | I don't think anyone else is using those functions |
| 21:53.02 | brlcad | nope, those aren't public functions |
| 21:53.07 | brlcad | they can be ripped a new one if need be |
| 21:53.24 | brlcad | goes to embib a bit |
| 22:04.52 | Ralith | we've got a goliath model now? |
| 22:04.54 | Ralith | awesome! :D |
| 22:05.07 | Ralith | syncs svn |
| 22:43.20 | bjorkintosh | Goliath, on line. |
| 22:59.43 | *** join/#brlcad ``Erik_ (i=erik@c-76-111-12-116.hsd1.md.comcast.net) |