| 00:54.50 | *** join/#brlcad cstirk (~quassel@c-71-56-216-45.hsd1.co.comcast.net) |
| 01:57.21 | starseeker | wonders if the following would be interesting for generating nmg shapes procedurally: http://www.antiprism.com/ |
| 01:58.14 | starseeker | or csg in some cases: http://www.antiprism.com/examples/100_demonstration_models/geodesic_pack_Med.jpg.1.html |
| 03:11.28 | brlcad | starseeker: yeah, that'd make for some proc-db output too |
| 03:11.32 | brlcad | some fun |
| 03:14.20 | brlcad | is reminded to look into http://www.qhull.org |
| 03:55.43 | brlcad | antiprism looks like it might have some useful routines for torus and sphere/ellipsoid/torus tessellation, wireframes, surface sampling, and similar |
| 03:57.18 | *** join/#brlcad evgeny (~Miranda@77.106.106.192) |
| 04:03.06 | brlcad | repel looks useful |
| 04:27.04 | zero_level | brlcad: The following are the difference bwhisteq and METHOD2 for lenna, barbara and mandrill images respectivel |
| 04:27.33 | zero_level | pixdiff bytes: 170022 matching, 92120 off by 1, 1 off by many |
| 04:27.34 | zero_level | pixdiff bytes: 207757 matching, 54386 off by 1, 0 off by many |
| 04:27.36 | zero_level | pixdiff bytes: 177579 matching, 84564 off by 1, 0 off by many |
| 04:38.37 | zero_level | its just a one pixel difference due to the rand01() function |
| 06:08.48 | brlcad | hm, okay |
| 06:09.08 | brlcad | I had a slight glimmer of hope that it might actually be some sort of standard definition that would get us even closer |
| 06:09.17 | brlcad | looks like that's not the case though, so probably worth removing |
| 06:09.28 | zero_level | alright |
| 06:10.24 | zero_level | so do i submit a patch for this ? |
| 06:12.00 | Notify | 03BRL-CAD:brlcad * 56035 brlcad/trunk/src/util/bwhisteq.c: remove METHOD2 since zero_level's testing shows that it doesn't appear to be significantly any better, just introduces an off-by-many on matlab comparisons due to the random weighting. |
| 06:15.54 | zero_level | thats fine I guess ;) |
| 07:54.31 | *** join/#brlcad caen23_ (~caen23@92.81.215.44) |
| 12:48.15 | *** join/#brlcad vladbogo (~vlad@188.25.237.173) |
| 13:03.22 | *** join/#brlcad vladbogo (~vladbogo@188.25.237.173) |
| 14:02.31 | starseeker | brlcad: if you're OK with the qhull license, there's probably several interesting things that can do - it's quite widely used (Meshlab, PCL, etc...) |
| 14:04.02 | starseeker | seems to pretty much be *the* solution when people are looking for a convex hull... |
| 14:06.44 | *** join/#brlcad vladbogo (~vladbogo@188.25.237.173) |
| 14:22.02 | *** join/#brlcad kesha (~kesha@110.225.10.69) |
| 14:49.35 | Notify | 03BRL-CAD:vladbogo * 56036 brlcad/trunk/src/libdm/CMakeLists.txt: Quiet float-equal warnings when Qt is enabled. |
| 15:24.40 | Notify | 03BRL-CAD:vladbogo * 56037 (brlcad/trunk/include/CMakeLists.txt brlcad/trunk/src/libdm/dm-qt.cpp): Created the Qapplication and the Qt main window. |
| 15:28.04 | Notify | 03BRL-CAD:vladbogo * 56038 brlcad/trunk/src/libdm/dm-qt.cpp: remove trailing ws |
| 15:30.54 | *** join/#brlcad evgeny (~Miranda@77.106.106.192) |
| 15:50.44 | Notify | 03BRL-CAD:vladbogo * 56039 (brlcad/trunk/include/dm-qt.h brlcad/trunk/src/libdm/dm-qt.cpp): Added qt_vars structure to dm-qt.h and use it as struct dm priv_vars. |
| 16:07.11 | brlcad | starseeker: I wouldn't integrate it just for the sake of integrating it -- when someone gets to the point of needing a convex hull, we can evaluate |
| 16:58.33 | *** join/#brlcad cstirk (~quassel@c-71-56-216-45.hsd1.co.comcast.net) |
| 17:35.32 | Notify | 03BRL-CAD:vladbogo * 56040 brlcad/trunk/src/libdm/dm-qt.cpp: Implemented qt-close so that the display manager can be gracefully released. |
| 17:49.08 | vladbogo | hi all. I am trying to embed Qt in Tk window and until now I haven't been successful. I am currently trying to use the QWidget::create function to set the widget's window to be the Tk one. I hope you could give me some hints in order to complete the embedding. |
| 18:56.09 | Notify | 03BRL-CAD:vladbogo * 56041 brlcad/trunk/src/libdm/CMakeLists.txt: Do not assume that the compiler is gcc or compatible when setting the -Wno-float-equal flag. |
| 19:01.45 | brlcad | vladbogo: I saw your commit and got excited thinking you'd figured it out ;) |
| 19:03.25 | vladbogo | brlcad: I've tried different approaches all past two weeks and this one I think is the closest |
| 19:03.55 | Notify | 03BRL-CAD:brlcad * 56042 brlcad/trunk/include/raytrace.h: provide a zero-initializer for hit structs |
| 19:04.52 | Notify | 03BRL-CAD:brlcad * 56043 brlcad/trunk/include/raytrace.h: init the hit struct magic number too |
| 19:05.10 | brlcad | vladbogo: it's the approach I would have thought to try |
| 19:05.43 | brlcad | qt provides hooks into lower-level objects, it's whether you can pass low-level constructs successfully |
| 19:06.03 | brlcad | so what's it currently do if that doesn't work |
| 19:06.36 | vladbogo | brlcad: I don't know why but it simply creates two windows instead setting the widgets parent to the tk one |
| 19:08.00 | vladbogo | I guess that the WId that I give to QWidget::create it's wrong or at least Qt interprets it as wrong |
| 19:08.25 | Notify | 03BRL-CAD Wiki:NyahCh3ck20 * 5692 /wiki/User:NyahCh3ck20/GSoc2013/Coding_Repor: /* 8 July - 14 July */ |
| 19:14.15 | Notify | 03BRL-CAD Wiki:NyahCh3ck20 * 5693 /wiki/User:NyahCh3ck20/GSoc2013/Coding_Repor: /* 8 July - 14 July */ |
| 19:14.24 | brlcad | vladbogo: the docs for winId() say that WId is very much platform and even compilation specific |
| 19:15.08 | brlcad | try looking at the value of winId() for the one created and the value of the Tk one |
| 19:16.13 | brlcad | it sounds like the right approach, but you might have to get a DIFFERENT handle derived/extracted from the Tk handle, not the handle itself |
| 19:17.22 | brlcad | like the ->xtkwin field |
| 19:18.05 | vladbogo | brlcad: I am using the Window returned by Tk_WindowId() |
| 19:18.59 | vladbogo | brlcad: the values for the Tk one and the window created are different |
| 19:19.14 | brlcad | you'll have to read the Qt sources to see what type it's expecting for an X11 build |
| 19:19.45 | brlcad | what platform are you on? |
| 19:20.26 | vladbogo | linux |
| 19:20.56 | brlcad | Tk_Screen could be something else Qt might be expecting, but not worth guessing ... just get their sources and find that create() method |
| 19:21.13 | vladbogo | the value for WId it's unsigned int if I remember correctly) |
| 19:21.57 | brlcad | if you cannot figure it out, it's entirely reasonable to go down a dramatically different path ... letting Qt create the window and NOT using Tk .. |
| 19:23.48 | vladbogo | brlcad: drossberg also suggested not using Tk so I researched a little bit and at a first sight the mged's refresh and event_check function should be changed |
| 19:42.14 | vladbogo | brlcad: just got the qt sources and WId in X11 case is Window |
| 20:23.10 | vladbogo | brlcad: also tried the same approach on a raw X11 window and I get the same output: 2 windows with different WIds so I am sure I am missing something but I don't know what |
| 20:32.26 | *** join/#brlcad cstirk (~quassel@c-71-56-216-45.hsd1.co.comcast.net) |
| 21:09.27 | Notify | 03BRL-CAD Wiki:Vladbogolin * 5694 /wiki/User:Vladbogolin/GSoC2013/Logs: |
| 21:39.25 | *** join/#brlcad vladbogo (~vlad@188.25.237.173) |
