| 00:36.44 | starseeker | phooey - FreeCAD didn't like the shape representation definition - guess I'll have to put both styles of shape definition in |
| 01:39.39 | brlcad | starseeker: for what it's worth, I've heard numerous commentaries about opencascade's STEP support being pretty terrible |
| 02:15.02 | Notify | 03BRL-CAD:starseeker * 57084 brlcad/trunk/src/other/stepcode/include/ordered_attrs.h: Take a stab at exporting the orderedAttrs functions for Windows. |
| 02:21.09 | starseeker | brlcad: fair enough, but it can load the Rhino export of the same geometry |
| 02:21.25 | starseeker | so it's clearly worth adding the extra header entries |
| 02:21.44 | starseeker | it's not hard, just a bit tedioius to hook up ;-) |
| 02:23.08 | starseeker | it's a bit tricky sorting out what the actually *requried* parts of the STEP files are from the informational entries we probably don't care about |
| 02:23.43 | brlcad | sure, not saying the course of action you mentioned wasn't worth it |
| 02:23.52 | starseeker | longer term it might be worth seeing about establishing some attribute conventions for _GLOBAL or something that map to the STEP bits I'm seeing |
| 02:24.20 | brlcad | just that freecad is unreliable in terms of correctness |
| 02:24.30 | starseeker | is currently mired in figuring out exactly why his step export doesn't import when Rhino's does |
| 02:24.49 | starseeker | there are geometry differences in that we are (currently) not splitting closed loops |
| 02:24.56 | starseeker | er closed curves rather |
| 02:25.25 | starseeker | but I can't tell whether I'm just not handling them correctly writing them out, or step-g isn't reading them in correctly |
| 02:26.13 | starseeker | agrees freecad isn't a benchmark of correctness |
| 02:27.07 | starseeker | they are the most logical open source export target for our geometry though (maybe even the only other realistic one) |
| 02:27.21 | starseeker | at least, as far as NURBS are concerned |
| 02:28.11 | starseeker | the OCE guys were intersted in looking at stepcode at one point, so once we've got our stuff polished up it may make them take another look |
| 02:29.02 | starseeker | takes another run at the bu_semaphore test as long as he's here... |
| 02:32.49 | Notify | 03BRL-CAD:starseeker * 57085 brlcad/trunk/src/libbu/tests/CMakeLists.txt: Whatever needs to go in the WINMM entry, it's gonna have to be part of the ADDEXEC line... |
| 02:41.22 | starseeker | blinks |
| 02:41.41 | starseeker | brlcad: that might have done it - I just got a successful build of tester_bu_semaphore on MSVC |
| 13:44.57 | *** join/#brlcad infobot (~infobot@rikers.org) | |
| 13:44.57 | *** topic/#brlcad is BRL-CAD || http://brlcad.org || logs: http://ibot.rikers.org/%23brlcad/ || GSoC 2013! http://brlcad.org/wiki/Google_Summer_of_Code | |
| 14:32.43 | Notify | 03BRL-CAD:carlmoore * 57103 brlcad/trunk/src/librt/primitives/hrt/hrt.c: remove trailing blanks/tabs; fix spelling; supply periods in 'e.g.' |
| 14:50.25 | brlcad | heh |
| 14:52.48 | brlcad | hickoryknoll: you around? like to get a copy of your various codes and talk about what's next :) |
| 14:53.08 | Ch3ck_ | sipping a bottle of coke! |
| 15:25.25 | zero_level_ | hey can anyone help me switch off [-Werror=array-bounds] flags ? |
| 15:28.55 | Ch3ck_ | brlcad: i've just submitted the patch for stubbing an empty pull into BRL-CAD could you please apply it so I could proceed with generating for pull_comb() subroutine? |
| 15:40.17 | zero_level_ | Izak__ : I am not sure why this is happening on my pc. |
| 15:40.31 | zero_level_ | smthing to do with hrt.c |
| 15:42.36 | zero_level_ | at line 502 and 517 |
| 15:45.29 | Izak__ | Hoping someone helps zero_level: turn off those flags |
| 15:46.56 | Izak__ | ``Erik: How do rt_xxx_plot() functions work? |
| 15:50.00 | *** join/#brlcad kesha (~kesha@14.139.122.114) | |
| 15:52.25 | *** join/#brlcad kesha (~kesha@14.139.122.114) | |
| 16:26.21 | *** join/#brlcad caen23 (~caen23@92.83.172.41) | |
| 16:45.03 | *** join/#brlcad vladbogo (~vlad@86.121.103.57) | |
| 17:07.29 | *** join/#brlcad vladbogo_ (~vlad@86.121.96.61) | |
| 17:20.11 | ``Erik | Izak_: they generate a list of line segments to approximate the wireframe of the geometry in 3d space, which something like the mged display window uses to draw the geometry using GL_LINES... like for a BoT, it just draws the edges of the triangles, for a sph it draws edges in a latitude/longitude style... |
| 17:21.49 | ``Erik | http://www.youtube.com/watch?v=T2DXrs0OpHU#t=20 quantum computing explained |
| 17:22.00 | ``Erik | (by the dude who does phdcomic) |
| 17:22.42 | ``Erik | starseeker: http://users.cms.caltech.edu/~mvanier/hacking/rants/cars.html |
| 17:31.48 | starseeker | ``Erik: heh, that's awesome - thanks! |
| 17:36.04 | starseeker | "Visual Basic - a car that drives you" |
| 17:36.29 | *** join/#brlcad Notify (~notify@66-118-151-70.static.sagonet.net) | |
| 17:36.36 | Notify | 03BRL-CAD:carlmoore * 57104 brlcad/trunk/src/conv/nmg/g-nmg.c: remove unneeded brace pairs, and re-implement -P argument in ***Usage*** statement (why did that revert from my earlier commit?) |
| 17:36.40 | Notify | 03BRL-CAD Wiki:IIIzzzaaakkk * 6028 /wiki/User:Izak/GSOC_2013_logs: /* August 19th to August 24th */ |
| 18:05.38 | zero_level_ | brlcad : Can you please help me regarding the "array-bounds" flag ? |
| 18:14.57 | *** join/#brlcad brlcad (~sean@66-118-151-70.static.sagonet.net) | |
| 18:15.37 | *** join/#brlcad hickoryknoll (~hickorykn@66-118-151-70.static.sagonet.net) | |
| 18:16.11 | *** join/#brlcad Izak_ (~Izak@66-118-151-70.static.sagonet.net) | |
| 18:16.24 | *** join/#brlcad n_reed (~molto_cre@66-118-151-70.static.sagonet.net) | |
| 18:16.25 | *** join/#brlcad Ch3ck (~Ch3ck@66-118-151-70.static.sagonet.net) | |
| 18:16.42 | *** join/#brlcad starseeker (~starseeke@66-118-151-70.static.sagonet.net) | |
| 18:16.42 | *** join/#brlcad zero_level (~mohit@66-118-151-70.static.sagonet.net) | |
| 18:16.44 | *** join/#brlcad ejno (~ejno@unaffiliated/kazaik) | |
| 18:17.20 | *** join/#brlcad kesha (~kesha@14.139.122.114) | |
| 18:17.55 | *** join/#brlcad maths22 (~gcimaths@66-118-151-70.static.sagonet.net) | |
| 18:22.40 | brlcad | Izak_: the answer is NOT to turn off the flags |
| 18:22.56 | brlcad | they are messages intended for you to fix the problem the compiler identified |
| 18:23.09 | *** join/#brlcad Notify (~notify@66-118-151-70.static.sagonet.net) | |
| 18:29.15 | Notify | 03BRL-CAD Wiki:Vladbogolin * 6029 /wiki/User:Vladbogolin/GSoC2013/Logs: /* Week 10 */ |
| 18:40.03 | *** join/#brlcad zero_level_ (0e8bf3a2@gateway/web/freenode/ip.14.139.243.162) | |
| 18:40.28 | ``Erik | :o it worked! |
| 18:45.44 | Notify | 03BRL-CAD:carlmoore * 57105 brlcad/trunk/src/libicv/bw.c: remove trailing blanks/tabs |
| 18:50.04 | *** join/#brlcad Izak (~Izak@195.24.220.16) | |
| 18:53.19 | *** join/#brlcad Izak__ (~Izak@195.24.220.16) | |
| 18:57.12 | Notify | 03BRL-CAD:starseeker * 57106 brlcad/trunk/src/conv/step/g-step/ON_Brep.cpp: Add more top-level shape definition logic. Seeing differences in how these are defined, so need to give a little thought as to what best practices would be... |
| 18:59.23 | Notify | 03BRL-CAD:brlcad * 57107 NIL: create a branch where indianlarry can commit nurbs work in progress without any interference to/from trunk development. |
| 19:00.43 | Izak__ | brlcad:My compiler shows me no errors. The sextic equation has 6 roots and 7 coefficients, just as the torus' final equation has atmost 4 roots and 5 coefficients. So I don't understand how to fix an error I am not experiencing |
| 19:02.14 | brlcad | Izak__: newer versions of gcc are better at warning |
| 19:02.47 | brlcad | if you are to believe the warning, what was it saying was the problem? |
| 19:03.11 | brlcad | (and you should believe the warning, it's a real issue) |
| 19:03.47 | Notify | 03BRL-CAD:starseeker * 57108 brlcad/trunk/doc/html/CMakeLists.txt: Add logo ico file to CMake logic |
| 19:03.54 | brlcad | ``Erik: suspense is killing |
| 19:04.01 | Izak__ | It says I provided more elements than an array could carry: BUT I never did that |
| 19:04.45 | ``Erik | ? |
| 19:04.45 | brlcad | 14:40 < ``Erik> :o it worked! |
| 19:04.45 | ``Erik | oh, notify came back with no action on my part |
| 19:04.45 | brlcad | ah, awesome |
| 19:04.51 | brlcad | Izak__: but you did |
| 19:05.01 | brlcad | again, believe the message .. now to understand it |
| 19:05.24 | brlcad | what exactly was the message and/or line of code |
| 19:05.41 | Izak__ | I am using gcc 4.4.6. |
| 19:05.54 | brlcad | yep, that's ancient in gcc terms ;) |
| 19:06.18 | brlcad | gcc 4.7 and 4.8 started adding and enabling some pretty advanced bug detection infrastructure |
| 19:06.42 | ``Erik | recent clang is pretty impressive with bug spotting and static analysis, as well |
| 19:06.52 | brlcad | yeah |
| 19:07.11 | brlcad | they both have a variant of the research that coverity is built around |
| 19:07.13 | Izak__ | Mohit says it is line 502 and 517 of hrt.c |
| 19:07.20 | ``Erik | (fwiw, fbsd is using a fork of gcc 4.2.1 due to license issues, so all this new shiney bug stuff is 'different' there) |
| 19:07.40 | brlcad | Izak__: what's 502 (and do you have a log to see the exact message?) |
| 19:08.56 | Izak__ | brlcad: This is what mohit showed me http://paste.kde.org/p4d4c4a66/ |
| 19:10.11 | zero_level_ | brlcad : in the meanwhile can you inform me how to switch off the flag. |
| 19:10.37 | zero_level_ | I searched using grep, add_definitions , dont see any settings ! O.o |
| 19:14.10 | Izak__ | What's happening ? |
| 19:17.29 | *** join/#brlcad DarkCalf (~DarkCalf@173.231.40.99) | |
| 19:20.49 | *** join/#brlcad n_reed (~molto_cre@66-118-151-70.static.sagonet.net) | |
| 19:21.17 | *** join/#brlcad brlcad (~sean@66-118-151-70.static.sagonet.net) | |
| 19:21.36 | *** join/#brlcad hickoryknoll (~hickorykn@66-118-151-70.static.sagonet.net) | |
| 19:21.46 | *** join/#brlcad zero_level (~mohit@66-118-151-70.static.sagonet.net) | |
| 19:21.46 | *** join/#brlcad maths22 (~gcimaths@66-118-151-70.static.sagonet.net) | |
| 19:21.51 | brlcad | sago is fuxed today |
| 19:21.59 | *** join/#brlcad ejno (~ejno@66-118-151-70.static.sagonet.net) | |
| 19:22.03 | *** join/#brlcad ejno (~ejno@unaffiliated/kazaik) | |
| 19:22.11 | *** join/#brlcad Izak_ (~Izak@66-118-151-70.static.sagonet.net) | |
| 19:22.11 | Izak__ | why? |
| 19:22.21 | *** join/#brlcad n_reed_ (~molto_cre@66-118-151-70.static.sagonet.net) | |
| 19:22.24 | brlcad | these network problems |
| 19:22.36 | brlcad | their entire network keeps going down |
| 19:23.03 | brlcad | somebody is probably attacking them, having some friday fun |
| 19:23.12 | *** join/#brlcad starseeker (~starseeke@66-118-151-70.static.sagonet.net) | |
| 19:23.23 | *** join/#brlcad Notify (~notify@66-118-151-70.static.sagonet.net) | |
| 19:23.38 | brlcad | Izak__: okay, hopefully our discussion doesn't get derailed again but lets start with the first one, hrt.c:502:9 |
| 19:24.00 | *** join/#brlcad Ch3ck (~Ch3ck@66-118-151-70.static.sagonet.net) | |
| 19:24.09 | brlcad | what is the error? |
| 19:24.12 | zero_level_ | Izak__ to smothen your discussion . |
| 19:24.15 | zero_level_ | cf is of dimension 5 |
| 19:24.34 | zero_level_ | fastf_t cf[BN_MAX_POLY_DEGREE+1]; |
| 19:24.44 | brlcad | zero_level_: the point is for him to figure this out, no? :) |
| 19:24.59 | zero_level_ | where BN_MAX)POLY_DEGREE = 4 |
| 19:25.08 | zero_level_ | brlcad : ok. |
| 19:25.08 | brlcad | appreciated, but opportunity lost |
| 19:25.16 | zero_level_ | oops. |
| 19:25.53 | brlcad | Izak__: so there, see if you understand the problem now |
| 19:26.06 | *** part/#brlcad n_reed_ (~molto_cre@66-118-151-70.static.sagonet.net) | |
| 19:26.32 | zero_level_ | brlcad : but by the look of the code i couldnt figure out if we needed that stuff from 502 to 525 in hrt.c |
| 19:28.07 | brlcad | well one has to know what "that stuff" means to figure that out, yes? |
| 19:28.26 | Izak__ | zero_level: Are you saying the sextic equation does not have roots or what? I dont follow |
| 19:28.51 | brlcad | Izak__: if you didn't follow what he just told you, lets continue :) |
| 19:29.02 | brlcad | he basically gave you the answer to what the problem is |
| 19:29.47 | brlcad | but if you don't see it, we can walk through the problem to make sure you understand exactly what the compiler is saying |
| 19:30.22 | zero_level_ | brlcad : sure. |
| 19:30.30 | zero_level_ | brlcad : let me help Izak_ here. |
| 19:30.43 | zero_level_ | I will try to guide him step by step. |
| 19:31.07 | brlcad | zero_level_: go for it, appreciated |
| 19:31.39 | zero_level_ | Izak__ : since compilers are right. (always). And I face this problem. So there is a problem. |
| 19:31.40 | brlcad | ~zero_level++ |
| 19:32.16 | Izak__ | listening |
| 19:32.29 | zero_level_ | Izak__ : If you look at the kde.paste link |
| 19:32.59 | zero_level_ | there is a error on line 502. at 9th col. |
| 19:33.18 | zero_level_ | do u see this ? |
| 19:33.38 | zero_level_ | why do u think that is a problem ? |
| 19:33.41 | Izak__ | seen it |
| 19:34.19 | Izak__ | cause 5 > BN_MAX_POLY_DEGREE |
| 19:34.28 | zero_level_ | no. |
| 19:34.41 | zero_level_ | do u use grep. |
| 19:34.46 | Izak__ | yes |
| 19:34.55 | zero_level_ | what is s ? |
| 19:35.10 | zero_level_ | replace by s by S. |
| 19:35.22 | zero_level_ | so what is S ? Which data type it is ? |
| 19:35.54 | Izak__ | a polynomial |
| 19:36.12 | zero_level_ | ok. what is S.cf ? |
| 19:36.52 | Izak__ | Array of polynomial coefficients |
| 19:36.54 | zero_level_ | also find out where is the polynomial data type defined ? |
| 19:36.58 | zero_level_ | ok. |
| 19:37.06 | zero_level_ | what is the length of the array ? |
| 19:40.29 | Izak__ | 5 |
| 19:41.39 | Izak__ | there? |
| 19:42.21 | zero_level_ | can u see find if any where in the complete src repository we have more than 4 degree of polynomial. ? |
| 19:43.37 | Izak__ | I think from bn.h the maxmum order of a polynomial is 4 so I don't think we have more than a 4 dgr poly |
| 19:44.04 | zero_level_ | i did grep -r "dgr = 6" |
| 19:44.18 | zero_level_ | and only got hrt.c |
| 19:45.00 | starseeker | yipe - gqa isn't happy (regression tests failing with a stack smash) |
| 19:45.00 | zero_level_ | was just wondering if there is a way to handle poly. with 5 or more degrees |
| 19:52.10 | brlcad | so that's the issue Izak__ |
| 19:52.14 | brlcad | there is a compile-time limit on the maximum size of polynomials |
| 19:52.32 | brlcad | that limit is set with BN_MAX_POLY_DEGREE |
| 19:52.52 | brlcad | do you see how the compiler message is right? |
| 19:53.00 | Izak__ | yes |
| 19:53.20 | brlcad | do you see how just turning off the message would be bad? :) |
| 19:53.37 | Izak__ | yeah |
| 19:54.02 | brlcad | you're indexing beyond the compile-time defined size of that cf array |
| 19:54.06 | brlcad | So... |
| 19:54.11 | brlcad | how to fix this |
| 19:54.20 | brlcad | what's the first idea that comes to mind? |
| 19:54.49 | Izak__ | my compiler has to be upgraded |
| 19:54.58 | brlcad | heh |
| 19:55.04 | brlcad | well all that will do is give you the error |
| 19:55.08 | brlcad | you have the error |
| 19:55.19 | brlcad | what's the first idea that comes to mind for fixing the error? |
| 19:55.40 | Izak__ | Making the root_solver accomodate sextic polynomials |
| 19:56.29 | brlcad | well, yes, you obviously need to accommodate this particular sextic polynomial |
| 19:56.53 | brlcad | but that implies the current implementation won't handle them, which you do not know |
| 19:57.23 | Izak__ | Increment BN_MAX_POLY_DEGREE to 6 |
| 19:57.41 | brlcad | that would be the first step |
| 19:57.52 | brlcad | that will increase the array size and at least make your lines not be errors |
| 19:58.18 | brlcad | won't mean it'll evaluate correctly or at all, but that definitely has to happen just from a C data structures perspective |
| 19:58.38 | brlcad | that, however, is a pretty major change... |
| 19:58.39 | Izak__ | Work on roots.c |
| 19:58.42 | brlcad | you'll need to test it |
| 19:58.56 | brlcad | the problem won't be compilation |
| 19:59.17 | brlcad | you have to make sure that increasing the size to 6 doesn't break any of the 4th order polynomial evaluations |
| 19:59.29 | brlcad | fortunately, we have unit and regression tests to confirm it |
| 19:59.31 | Notify | 03BRL-CAD:starseeker * 57109 (brlcad/trunk/src/other/stepcode/src/express/test/print_attrs.c brlcad/trunk/src/other/stepcode/src/express/test/print_schemas.c): Wrap unistd.h as well. |
| 19:59.49 | brlcad | start there, set it to 6, compile, then run "make test" and "make benchmark" and make sure both pass without any failures |
| 20:00.08 | brlcad | make sure they pass without failures BEFORE you set it to 6 too, obviously... |
| 20:00.16 | brlcad | to make sure it's in a working state before you "compare" |
| 20:00.25 | brlcad | Izak__: make sense what you need to do? |
| 20:06.08 | Izak__ | thankful to brlcad: and zero_level |
| 20:07.23 | brlcad | the next step will be to confirm whether the root solver can actually handle a stable sextic equation |
| 20:07.53 | brlcad | I suggest editing the unit test |
| 20:08.03 | brlcad | in src/libbn/tests is a roots tests |
| 20:08.33 | brlcad | you can add a 6th order test (after confirming that BN_MAX_POLY_DEGREE=6 doesn't break anything) |
| 20:08.38 | Izak__ | was thinking of that |
| 20:08.49 | brlcad | something with known input/output values like the real example in http://elib.mi.sanu.ac.rs/files/journals/tm/21/tm1124.pdf |
| 20:12.12 | *** join/#brlcad mpictor (~mark@2601:d:b280:b5:d63d:7eff:fe2d:2505) | |
| 20:13.53 | Izak__ | understanding what " that's what made your gsoc proposal interesting in the first place" means |
| 20:24.58 | zero_level_ | Izak__ , brlcad : in libbn/poly.c i dont see any function which solves sextic polynomial. |
| 20:25.06 | zero_level_ | looks like we have to implement one. |
| 20:25.18 | Ch3ck_ | you could get started. |
| 20:25.19 | Ch3ck_ | ;) |
| 20:25.39 | Ch3ck_ | since i know it's really complicated writing a root solver of order 6 |
| 20:26.36 | zero_level_ | Ch3ck_ : I understand the complexity. |
| 20:26.48 | zero_level_ | Just brought in because of the discussion. |
| 20:27.19 | zero_level_ | Izak__ : Its ok. n |
| 20:27.23 | Ch3ck_ | well will check on some recent math on resolvingthe problem. |
| 20:36.10 | starseeker | brlcad: looks like enabling the stack protection flags is giving gqa a fit - that's when the gqa regression test starts failing (r56758) |
| 20:36.15 | Ch3ck_ | brlcad: starseeker: waiting for the empty pull routine patch to be applied so I could create the one for pulling combinations in the pull |
| 20:39.50 | Notify | 03BRL-CAD Wiki:NyahCh3ck20 * 6030 /wiki/User:NyahCh3ck20/GSoc2013/Coding_Repor: /* 19 August - 25 August */ |
| 21:03.24 | *** part/#brlcad Ch3ck_ (~Ch3ck@195.24.220.16) | |
| 21:04.41 | Ch3ck | gotta home get some rest now probably take on Prison Break ;) |
| 21:09.26 | brlcad | zero_level_: there's no specific support for sextic, but it may or may not matter (for this particular primitive) |
| 21:09.41 | brlcad | the routine used is in src/librt/roots.c |
| 21:11.07 | brlcad | Izak_: the test program I was thinking of was actually src/util/roots_example.c |
| 21:11.35 | brlcad | that's basically a unit test, but it predates the current unit test infrastructure |
| 21:11.52 | ``Erik | a more generalized root solver would be nice to have in libbn |
| 21:12.04 | brlcad | if our solver doesn't work for sextic, we'll look at integrating something more generalized (that already exists, like gmp) |
| 21:12.21 | brlcad | but hopefully not gmp |
| 21:12.27 | brlcad | it's a pig |
| 21:12.48 | Notify | 03BRL-CAD Wiki:NyahCh3ck20 * 0 /wiki/File:Pull_Combination.png: this pulls an unpushed object |
| 21:14.01 | Notify | 03BRL-CAD Wiki:NyahCh3ck20 * 0 /wiki/File:Pushed_pulled_object.png: This restores the object to original state after push is applied |
| 21:15.46 | Notify | 03BRL-CAD Wiki:NyahCh3ck20 * 6033 /wiki/User:NyahCh3ck20/GSoc2013/Coding_Repor: /* Coding Log Report for GSoc 2013 */ |
| 21:17.35 | Notify | 03BRL-CAD Wiki:NyahCh3ck20 * 6034 /wiki/User:NyahCh3ck20/GSoc2013/Coding_Repor: /* = Test Results */ |
| 21:33.23 | Notify | 03BRL-CAD:brlcad * 57110 brlcad/trunk/src/libged/gqa.c: do not tokenize over spaces |
| 21:42.23 | *** join/#brlcad mpictor (~mark@c-67-177-102-131.hsd1.in.comcast.net) | |
| 22:06.46 | ``Erik | Neil deGrasse Tyson #@neiltyson If you need to invoke your academic pedigree or job title for people to believe what you say, then you need a better argument |
| 22:14.46 | Notify | 03BRL-CAD:starseeker * 57111 brlcad/trunk/src/libged/gqa.c: Fix stack smash where strtok was reading beyond the limits of optarg (Sean found it) |