| 01:20.44 | brlcad | vladbogo: so ... basically it SHOULD be working how you wrote it |
| 01:21.16 | brlcad | perhaps Tk is not creating a real "Window" |
| 01:21.39 | brlcad | try creating a third X11 window (manually) |
| 01:22.01 | brlcad | find an X11 tutorial on the web, create a simple empty window, pass that Window to Qt and see if you can get it to use it |
| 01:22.27 | brlcad | or ask on the Qt and/or Tk forums and/or stackoverflow what you might be doing wrong |
| 01:23.06 | brlcad | unfortunately, the answer isn't obvious but your logic seems sound so the next step is to either walk the logic in a debugger or ... ask for help ;) |
| 03:03.33 | *** join/#brlcad evgeny (~Miranda@109.194.34.184) |
| 03:36.07 | Notify | 03BRL-CAD:brlcad * 56044 brlcad/trunk/src/libbrep/opennurbs_ext.cpp: initialize usplit and vsplit to the midpoint, otherwise there's potential for them to get used without being initialized. |
| 03:37.12 | Notify | 03BRL-CAD:brlcad * 56045 brlcad/trunk/src/libbrep/opennurbs_ext.cpp: what the heck happened to the style??? |
| 03:41.01 | *** join/#brlcad caen23 (~caen23@92.81.215.44) |
| 03:42.31 | Notify | 03BRL-CAD:phoenixyjll * 56046 brlcad/trunk/src/libbrep/intersect.cpp: Normalize the curves, and check tangent events. |
| 04:06.18 | Notify | 03BRL-CAD:phoenixyjll * 56047 brlcad/trunk/src/libbrep/intersect.cpp: Check the domain parameters. |
| 04:42.36 | Notify | 03BRL-CAD:brlcad * 56048 brlcad/trunk/src/rt/do.c: should not be using raw codes with open(), use the symbolic form for rw |
| 04:50.54 | Notify | 03BRL-CAD:brlcad * 56049 brlcad/trunk/src/rt/do.c: improve the error message when we cannot write to the output file. the fdopen failure was masking the open failure causing an unhelpful 'bad file descriptor' message instead of the original permissions denied message. |
| 04:52.40 | Notify | 03BRL-CAD:phoenixyjll * 56050 brlcad/trunk/src/libbrep/intersect.cpp: Add seaming curves to handle intersecting intersection curves. |
| 04:56.27 | Notify | 03BRL-CAD:brlcad * 56051 brlcad/trunk/src/rt/do.c: let the user know that we read the existing output data into the framebuffer (even though we still call shootray for all pixels) |
| 05:00.25 | Notify | 03BRL-CAD:brlcad * 56052 brlcad/trunk/src/rt/do.c: ordering for pre c99 |
| 05:03.16 | Notify | 03BRL-CAD:brlcad * 56053 brlcad/trunk/src/rt/do.c: put new log message behind the existing verbosity flag |
| 05:09.09 | Notify | 03BRL-CAD:phoenixyjll * 56054 brlcad/trunk/src/libbrep/intersect.cpp: Sometimes we can merge the new seaming curves - don't need to generate new ones. |
| 05:25.54 | Notify | 03BRL-CAD:phoenixyjll * 56055 brlcad/trunk/src/libbrep/intersect.cpp: Eliminate compiler warnings. |
| 05:34.57 | vladbogo | brlcad: I've tried creating a third X11 window and pass that window to Qt but I get the same result (2 windows) |
| 05:36.03 | vladbogo | I will also ask on the qt forum/stackoverflow and hopefully find a solution. |
| 06:29.52 | *** join/#brlcad caen23 (~caen23@92.81.215.44) |
| 07:04.23 | Notify | 03BRL-CAD:phoenixyjll * 56056 brlcad/trunk/src/libbrep/intersect.cpp: Should consider 2D distances. And if there are already intersections, we don't generate seaming curves. |
| 07:32.20 | *** join/#brlcad merzo (~merzo@user-94-45-58-138-1.skif.com.ua) |
| 07:46.34 | Notify | 03BRL-CAD Wiki:Phoenix * 5695 /wiki/User:Phoenix/GSoc2013/Reports: /* Week 4 */ |
| 07:47.59 | Notify | 03BRL-CAD Wiki:Phoenix * 5696 /wiki/User:Phoenix/GSoc2013/Reports: /* Test Results */ |
| 07:48.18 | Notify | 03BRL-CAD Wiki:Phoenix * 5697 /wiki/User:Phoenix/GSoc2013/Reports: /* Test Results */ |
| 07:49.48 | *** join/#brlcad merzo (~merzo@user-94-45-58-138-1.skif.com.ua) |
| 09:39.29 | *** join/#brlcad kimz (~AndChat56@110.225.245.211) |
| 11:04.17 | *** join/#brlcad merzo (~merzo@user-94-45-58-138-1.skif.com.ua) |
| 11:51.18 | Notify | 03BRL-CAD:brlcad * 56057 brlcad/trunk/src/librt/primitives/rec/rec.c: overhaul the rec shot routine. eliminates a recently introduced magic constant (1.0e10) that wasn't correct (a big root does not indicate a need to check the end plates). completely restructure the logic to eliminate the gotos and the associated spagetti. now we always check the cylinder and end plates, collapsing any duplicate points. these |
| 11:51.20 | Notify | changes were tested extensively with a default rec from all possible orientations showing no change other than face normal on grazing edge. performance was actually faster for most views (but within 5% noise). |
| 11:52.53 | *** join/#brlcad d_rossberg (~rossberg@66-118-151-70.static.sagonet.net) |
| 13:40.03 | *** join/#brlcad hickoryknoll (~hickorykn@66-118-151-70.static.sagonet.net) |
| 14:40.05 | *** join/#brlcad evgeny (~Miranda@77.106.106.192) |
| 14:40.11 | *** part/#brlcad evgeny (~Miranda@77.106.106.192) |
| 16:50.38 | zero_level | brlcad : I want to discuss aout three issues |
| 16:50.50 | zero_level | a) About Implementation of filters |
| 16:51.03 | zero_level | b) About histograms in double images |
| 16:52.22 | zero_level | c) About converting the image utilities to use implemented functions of libicv |
| 16:53.19 | zero_level | about A) |
| 16:54.29 | zero_level | BRL-CAD has three different utilities for filttering bwfilter, pixfilter, pix3filter |
| 16:55.12 | zero_level | all these three use a 3X3 (bwfilter, pixfilter) amd 3x3x3 (pix3filter) |
| 16:55.51 | zero_level | I am converting them as two functions |
| 16:56.05 | zero_level | icv_filter() for bwfilter and pixfilter |
| 16:56.17 | zero_level | icv_filter3() for pix3filter |
| 16:56.33 | zero_level | My questions is : |
| 16:58.09 | zero_level | I am not sure about the size of the image BRL-CAD predominately uses (My guess 512x512). But do you think I should include a functionality for |
| 16:58.17 | zero_level | variable size filter ? |
| 16:59.57 | zero_level | If the image size is small (<512^2) My gut feeling is that 3 dimensional size filter will sufice |
| 17:00.22 | zero_level | about B) |
| 17:01.44 | zero_level | Or do u think i shld put them in mail ? |
| 17:03.18 | *** join/#brlcad vladbogo (~vladbogo@188.25.237.173) |
| 17:32.31 | *** part/#brlcad hickoryknoll (~hickorykn@66-118-151-70.static.sagonet.net) |
| 18:00.28 | *** join/#brlcad cstirk (~quassel@c-71-56-216-45.hsd1.co.comcast.net) |
| 18:17.06 | brlcad | zero_level: we can discuss here and then you can summarize via e-mail for others |
| 18:17.21 | brlcad | default image size for all tools should be 512x512 |
| 18:17.30 | brlcad | but you shouldn't assume that size |
| 18:18.04 | brlcad | is there a way to combine all three into one generalized function? |
| 18:18.43 | brlcad | i'm not familiar with the 3x3x3 kernel you speak of for pix3filter |
| 18:19.18 | brlcad | having a variable size filter is VERY desirable, but that would obviously be new functionality so you certainly don't have to accommodate it |
| 18:19.51 | brlcad | but if you happen to accommodate it in the process of generalizing those three into one or two functions, great |
| 18:21.51 | vladbogo | brlcad: hi. I've asked on sourceforge and qt forums and hopefully I will get a response soon. |
| 18:22.07 | vladbogo | on stackoverflow* |
| 18:23.17 | zero_level | brlcad: right. The only reasoni asked for variable lenth filter was because of implementation issue. |
| 18:24.11 | zero_level | Now i will modify the current function as a generalization of filter length |
| 18:26.53 | zero_level | currently it is hard coded so that only 3-Dim Filter are used. |
| 18:27.43 | brlcad | interesting new site: http://choosealicense.com/ |
| 18:27.57 | brlcad | vladbogo: excellent, let us know what you find out |
| 18:28.39 | zero_level | pix3filter is there in src/util/pix3filter.c |
| 18:28.49 | brlcad | zero_level: the trick/danger is dealing with larger filters correctly on the boundaries |
| 18:29.13 | brlcad | I know where the sources are, just haven't needed to use it or have read the sources in a very long time |
| 18:29.21 | brlcad | can you summarize? |
| 18:30.51 | zero_level | alright |
| 18:31.13 | zero_level | Its basicaaly for three images |
| 18:31.59 | zero_level | and besides the normal low, high, and average u could create a smear amination using this |
| 18:32.20 | zero_level | and the combined is put to a single image |
| 18:34.57 | zero_level | and yes boundaries ar the issue. n/2 lines in both sides might look awful |
| 18:35.16 | zero_level | but then if the images are large. It wouldnt matter. |
| 18:36.10 | zero_level | about b) although i will be implementing this latter in the week. |
| 18:36.34 | zero_level | How do we written for histogram functions. |
| 18:36.47 | zero_level | What should be the hisogram for double data. |
| 18:37.16 | zero_level | The only way I got is ask the user the number of bins and written a long int array of that size. |
| 18:37.34 | zero_level | Just wanted to make sure I was write abt that |
| 18:37.40 | zero_level | *if |
| 18:39.14 | zero_level | so this looks long int* icv_histogram(icv_image_t* bif, int num_bins) |
| 18:39.45 | zero_level | and also there is an issue if the imgage is rgb data. ? |
| 18:41.26 | zero_level | earlier i thought it to be int icv_histogram(icv_image_t*bif, int num_bins, long int* bin1, long int *bin2, long int *bin3 ) |
| 18:41.50 | zero_level | but i am worried for different channel images (1,4) |
| 18:47.30 | *** join/#brlcad cstirk (~quassel@c-71-56-216-45.hsd1.co.comcast.net) |
| 18:57.50 | brlcad | zero_level: so then merging the three sounds unhelpful |
| 18:57.56 | brlcad | just the two like you had planned |
| 18:59.06 | brlcad | if you're familiar with the generic filter in photoshop or gimp, that might be a good example to follow |
| 19:03.18 | brlcad | as for histogram, you could specify the image, num_bins, start channel #, channel count, and output array (should be size_t ... NOT long int) |
| 19:14.25 | vladbogo | brlcad: in order to process Qt events the QApplication::processEvents function should be called in mged's main loop. Should I add a new processEvent callback to struct dm which does nothing for the other dms or should I put an #ifdef DM_QT call process events #endif? |
| 19:16.10 | brlcad | vladbogo: hehe, you clearly KNOW the right way already :) |
| 19:16.28 | brlcad | callback ftw |
| 19:17.35 | vladbogo | brlcad: that's what I thought but just wanted to be sure |
| 19:19.15 | vladbogo | brlcad: also offtopic I saw that Notify does not print updates |
| 19:22.43 | brlcad | notify sometimes gets stuck |
| 19:25.14 | Notify | 03BRL-CAD:vladbogo * 56058 brlcad/trunk/src/libdm/dm-qt.cpp: Remove trailing ws and other minor changes. |
| 19:25.19 | Notify | 03BRL-CAD:vladbogo * 56059 (brlcad/trunk/include/dm-qt.h brlcad/trunk/src/libdm/dm-qt.cpp): Implemented qt_reshape and added the QtTkMainWindow to priv_vars. |
| 19:25.22 | Notify | 03BRL-CAD:vladbogo * 56060 brlcad/trunk/src/libdm/dm-qt.cpp: Implemented qt_configureWin_guts. |
| 19:25.25 | Notify | 03BRL-CAD:vladbogo * 56061 brlcad/trunk/src/libdm/dm-qt.cpp: Set window background color. |
| 19:25.27 | Notify | 03BRL-CAD:vladbogo * 56062 brlcad/trunk/src/libdm/dm-qt.cpp: Remove trailing ws and other small cleanups. |
| 22:15.25 | Notify | 03BRL-CAD Wiki:Level zero * 5698 /wiki/User:Level_zero/GSOC13/logs: /* Week 4 */ |
| 22:15.27 | Notify | 03BRL-CAD Wiki:Harman052 * 5699 /wiki/User:Harman052/GSoc2013/Logs: |
| 22:15.29 | Notify | 03BRL-CAD Wiki:Vladbogolin * 5700 /wiki/User:Vladbogolin/GSoC2013/Logs: |
| 22:15.32 | Notify | 03BRL-CAD Wiki:49.249.1.63 * 5701 /wiki/User:KeshaSShah/GSoC13/Reports: |
| 22:15.33 | Notify | 03BRL-CAD Wiki:Izakkayems * 5702 /wiki/User:Izak/GSOC_2013_logs: /* From July 15th to July 20th */ |
| 22:15.35 | Notify | 03BRL-CAD Wiki:NyahCh3ck20 * 5703 /wiki/User:NyahCh3ck20/GSoc2013/Coding_Repor: /* 15 July - 21 July */ |
| 22:15.43 | Notify | 03BRL-CAD:vladbogo * 56063 (brlcad/trunk/include/dm-Null.h brlcad/trunk/include/dm.h and 10 others): Added the processEvents callback to struct dm. |