| 00:18.45 | *** join/#brlcad kesha_ (~kesha@49.249.8.110) | |
| 00:29.17 | brlcad | Izak__: haha, that animation is ridiculous! |
| 00:32.15 | kanzure | brlcad: thoughts about that problem with brlcad.git ? |
| 00:32.27 | kanzure | brlcad: i have a brlcad.git repo sitting on a box but all of the commit ids have changed.. which is bad. |
| 00:35.26 | *** join/#brlcad jrayhawk (~jrayhawk@nursie.omgwallhack.org) | |
| 00:39.59 | brlcad | Ch3ck: there is no way of knowing the "original centerpoint" because no such concept exists at the primitive level unless the primitive defines such a concept |
| 00:40.33 | brlcad | Ch3ck: but again, IT DOES NOT MATTER ... you can just assume the bounding box center, and translate it to 0,0,0 if it's not there pretending that's the origin |
| 00:41.34 | brlcad | Izak__: that's why the animation wiki page was very specific, so that you end up with a simple spinning orbital animation ... FAR easier to see shape that way |
| 00:43.37 | Notify | 03BRL-CAD:tbrowder2 * 57537 brlcad/trunk/src/util/CMakeLists.txt: add more tmp files (NOINSTALL) for testing TCLAP implementation |
| 00:43.42 | brlcad | kanzure: you'll have to ask ``Erik as he is the caretaker of the git-svn bridge, but I can certainly poke at it if you need something looked at |
| 00:44.17 | kanzure | brlcad: basically what happened is that all of the commits in the history changed |
| 00:44.24 | kanzure | brlcad: so now we need to decide to keep the old commits or the new commit ids |
| 00:45.12 | Notify | 03BRL-CAD:brlcad * 57538 brlcad/trunk/src/util/CMakeLists.txt: the keyword is NO_INSTALL |
| 00:46.34 | kanzure | ``Erik: ping, see above about the git-svn bridge |
| 00:51.11 | jrayhawk | Or switch to commit messages that only mention the revision and not the repository URL. |
| 00:53.13 | brlcad | yep, that's all him |
| 00:53.39 | brlcad | unless you can point me directly at exactly one thing to run or edit |
| 01:00.30 | jrayhawk | It would be best to preserve whatever commit history has the most public inertia and start building subsequent revisions (with somewhat more futureproof commit messages) on top of that, so it will probably involve a judgement call from someone familiar with its users along with some logic changes. |
| 01:03.38 | jrayhawk | e.g. if(revision > 55160){commitmessage = new_format} else {commitmessage = old_format} |
| 01:17.03 | *** join/#brlcad FLOSSrookie (~brian@107-200-34-111.lightspeed.tulsok.sbcglobal.net) | |
| 01:25.10 | brlcad | the repo server migrated around r55208, does that have anything to do with this change you're seeing? |
| 01:47.05 | jrayhawk | Yeah. |
| 01:51.09 | kanzure | brlcad: what does migrated mean? |
| 01:51.39 | kanzure | sourceforge change or brlcad.git change? |
| 01:51.47 | kanzure | oh i see, i bet it was a sourceforge change |
| 01:51.55 | brlcad | sourceforge had an entire infrastructure replacement back in march, url changed, backend changed |
| 01:52.07 | brlcad | still same repo, but everything around it changed |
| 01:55.11 | kanzure | ah, that's very helpful actually |
| 03:02.44 | Notify | 03BRL-CAD:phoenixyjll * 57539 brlcad/trunk/src/libbrep/boolean.cpp: The number of surfaces should be used for the surf_tree. |
| 03:03.43 | *** join/#brlcad kesha__ (~kesha@49.249.0.69) | |
| 03:20.28 | Notify | 03BRL-CAD:phoenixyjll * 57540 brlcad/trunk/src/libbrep/boolean.cpp: IsFaceInsideBrep() returns -1 for unknown inside/outside status. |
| 03:24.39 | *** join/#brlcad FLOSSrookie (~brian@107-200-34-111.lightspeed.tulsok.sbcglobal.net) | |
| 03:26.01 | Notify | 03BRL-CAD:phoenixyjll * 57541 brlcad/trunk/src/libbrep/boolean.cpp: Enlarge the bounding boxes with INTERSECTION_TOL. |
| 03:43.38 | *** join/#brlcad kesha__ (~kesha@49.249.0.143) | |
| 03:54.19 | *** join/#brlcad archivist_ (~archivist@host81-149-189-98.in-addr.btopenworld.com) | |
| 03:56.30 | *** join/#brlcad kesha__ (~kesha@49.249.0.143) | |
| 04:02.49 | *** join/#brlcad kesha__ (~kesha@49.249.0.60) | |
| 04:04.03 | *** join/#brlcad archivist (~archivist@host81-149-189-98.in-addr.btopenworld.com) | |
| 04:04.37 | *** join/#brlcad archivist_ (~archivist@host81-149-189-98.in-addr.btopenworld.com) | |
| 04:13.04 | *** join/#brlcad Ch3ck (~Ch3ck@66-118-151-70.static.sagonet.net) | |
| 04:13.12 | *** join/#brlcad Izak_ (~Izak@66-118-151-70.static.sagonet.net) | |
| 04:13.25 | *** join/#brlcad maths22 (~gcimaths@66-118-151-70.static.sagonet.net) | |
| 04:48.28 | brlcad | Ch3ck: been quiet lately |
| 04:48.41 | brlcad | you've disappeared for several mondays now.. |
| 04:49.25 | brlcad | let me know how the leaf bb calculation goes (you can leverage the bbox() callback defined in the rt_functab table) |
| 05:13.29 | *** join/#brlcad kesha (~kesha@49.202.231.160) | |
| 05:27.57 | *** join/#brlcad kesha_ (~kesha@49.202.238.42) | |
| 06:38.10 | *** join/#brlcad kesha_ (~kesha@49.249.0.9) | |
| 06:57.14 | kesha_ | What are the revision stamps for 7.22 and trunk ? |
| 07:13.26 | *** join/#brlcad vladbogo (~vlad@188.25.237.111) | |
| 07:17.54 | *** join/#brlcad kesha__ (~kesha@49.249.17.97) | |
| 07:23.57 | *** join/#brlcad kimzzzz (~AndChat31@1.38.31.251) | |
| 07:39.53 | Notify | 03BRL-CAD:vladbogo * 57542 (brlcad/trunk/include/dm-qt.h brlcad/trunk/src/libdm/dm-qt.cpp): Release dm resources when close called. |
| 07:41.14 | *** join/#brlcad cogitokat (~kat@ip70-171-0-190.ga.at.cox.net) | |
| 08:04.30 | Notify | 03BRL-CAD:vladbogo * 57543 (brlcad/trunk/src/libdm/dm-generic.c brlcad/trunk/src/libdm/dm-qt.cpp brlcad/trunk/src/mged/mged.c): Use Tcl_CreateEventSource to call QApplication::processEvents instead of calling it by hand. |
| 08:47.54 | *** join/#brlcad kesha (~kesha@49.249.0.47) | |
| 08:58.29 | *** join/#brlcad kesha (~kesha@49.202.231.167) | |
| 09:07.58 | *** join/#brlcad kesha (~kesha@49.202.231.167) | |
| 09:21.51 | *** join/#brlcad cogitokat (~kat@ip70-171-0-190.ga.at.cox.net) | |
| 09:39.50 | Notify | 03BRL-CAD:phoenixyjll * 57544 brlcad/trunk/src/libbrep/test_curve_intersect.cpp: Fix memory leak. |
| 09:43.46 | Notify | 03BRL-CAD:phoenixyjll * 57545 brlcad/trunk/src/libbrep/test_point_intersect.cpp: Fix memory leak. |
| 09:51.28 | Izak__ | brlcad: What do you mean by ridiculous ? |
| 09:52.12 | Izak__ | Animation page lead me to deliver it in .mpg format which is better than the slideshow |
| 09:57.27 | *** join/#brlcad Ch3ck_ (~Shadownet@195.24.220.16) | |
| 10:08.41 | Izak__ | How can I color an object ? ``Erik: |
| 10:11.44 | *** join/#brlcad kesha (~kesha@49.249.16.181) | |
| 10:52.10 | *** join/#brlcad kesha (~kesha@49.249.16.181) | |
| 11:09.10 | *** join/#brlcad kesha (~kesha@49.249.16.181) | |
| 11:24.58 | *** join/#brlcad kesha (~kesha@49.249.16.181) | |
| 11:30.53 | brlcad | Izak__: animation page covered a LOT more than just outputting something in .mpg format |
| 11:34.23 | brlcad | and you didn't get that point until I said that ... |
| 11:35.12 | brlcad | another point is to see the object spinning |
| 11:36.00 | Izak__ | brlcad: Did you watch the movie? |
| 11:36.03 | brlcad | for that, you need animation frames for enough angles and to compose them together (unstylized) |
| 11:36.06 | brlcad | yes |
| 11:36.25 | brlcad | why else would I have said it was ridiculous? |
| 11:36.38 | brlcad | it doesn't show an object spinning |
| 11:37.17 | brlcad | and it seems to have very few frames |
| 11:37.31 | brlcad | and there are stylzed transition effects |
| 11:37.54 | brlcad | and ... music ... |
| 11:38.09 | Izak__ | Which i presume you don't need... |
| 11:38.13 | brlcad | solidly missing the point |
| 11:39.06 | brlcad | if you don't want to do it, just say so but don't half-ass it please |
| 11:39.30 | brlcad | :) |
| 11:39.47 | Izak__ | I want to do it brlcad |
| 11:39.49 | brlcad | I can't see what that shape looks like from that, so to me it's kinda useless |
| 11:40.19 | brlcad | I think the problem is perhaps you don't realize what the end-result looks like? |
| 11:40.35 | brlcad | you've certainly seen something spinning before |
| 11:40.38 | Izak__ | Sorry, i did not know it was so so important. I just thought it was an aside to the project. I will focus on this soon. |
| 11:40.59 | Izak__ | I thought the point was to show that raytracing was showing better results |
| 11:41.47 | Izak__ | thats after the corections to rt_hrt_shot |
| 11:41.51 | brlcad | it's not that is so so important |
| 11:42.25 | brlcad | but there was a useful point ... and nothing you've done has been useful yet |
| 11:42.45 | brlcad | it was to see what this shape actually looksl ike |
| 11:43.04 | Izak__ | I thought that movie shows the heart at different inclinations ? |
| 11:43.34 | Izak__ | I am sure i did not understand you |
| 11:44.00 | Izak__ | Please could you explain the point |
| 11:44.24 | Ch3ck_ | brlcad: successfully computed the bounding box using the bbox() routine in the functab table simplifies alot of the work. Well while pulling translations using the centrepoint I was under the assumption that all objects begin a point zero while during the buildup process, they are moved to different positions relative to the origin(0,0,0) which I believe based on this premise I could easily compute the translation in relation to the origin. but it seems |
| 11:44.24 | Ch3ck_ | you don't agree? could you please elaborate? |
| 11:45.16 | brlcad | Izak__: the point of doing an orbital animation is to get a sense for the 3D shape of an object |
| 11:45.27 | Izak__ | I simply ra the Fly-around animation script and it produced images which i used to make the movie |
| 11:45.36 | brlcad | it's the difference between taking 3 or 4 pictures of the object, and taking 360 pictures, one for each angle |
| 11:45.50 | brlcad | yes, that part is fine |
| 11:45.59 | brlcad | so then you should compose them into a movie as-is |
| 11:46.04 | brlcad | they ARE the animation *frames* |
| 11:47.00 | brlcad | there is no orbit because you used some software that stylizes the output |
| 11:47.18 | brlcad | they need to just be composed together so we can just see the shape |
| 11:49.52 | brlcad | this is a 4d animation, but it has the same basic concept: http://www.youtube.com/watch?v=BqfwPQvb7KA |
| 11:50.03 | brlcad | you're spinning around the object so you can see the shape |
| 11:50.43 | brlcad | I can clearly see that shape even with a 4th dimension added that makes it constantly changing |
| 11:50.59 | brlcad | its visibile because "it's spinning" |
| 11:51.25 | brlcad | now does your look at all like that? :) |
| 11:53.19 | brlcad | following the animation script wiki page would have resulted in something like that and the shape would have been more apparent |
| 11:54.05 | brlcad | here's a page filled with real brl-cad examples based around that, including a ton of orbital animations: http://ronja.twibright.com/3d/ |
| 11:54.47 | brlcad | e.g., http://ronja.twibright.com/3d/front_cap_mpeg4_asp.avi |
| 11:58.02 | Izak__ | Well the Fly around script didn't quite run as was on the Animation page so I edited it a little |
| 11:58.05 | *** join/#brlcad kesha_ (~kesha@49.249.16.50) | |
| 11:59.12 | brlcad | Izak__: what didn't run? should fix the page if there's something wrong there |
| 11:59.27 | brlcad | and it's a teaching page, to understand the concepts so you can do your own thing |
| 11:59.34 | brlcad | not to be a strict recipe |
| 11:59.52 | brlcad | but still, it should have worked as-is, so what didn't work? |
| 12:00.59 | Izak__ | `loop 000 359 10` |
| 12:01.36 | brlcad | what didn't work about that? |
| 12:02.35 | Izak__ | Forgotten the error message |
| 12:02.43 | Izak__ | But I used http://paste.kde.org/p87a670cd/ instead |
| 12:03.56 | brlcad | well considering loop is one of our tools, I'd like to know what that error is so it can be fixed if it's something wrong with loop |
| 12:04.38 | brlcad | your manual loop is fine, but we should fix the issue if you encountered an issue |
| 12:06.49 | Izak__ | I had to remove that 'loop'. I can't retrieve the error messages |
| 12:07.55 | brlcad | had to? why? |
| 12:08.02 | brlcad | this is all non-information |
| 12:08.06 | brlcad | if the error was "loop: command not found" then that would have been your own fault, and you wasted time working around it instead of understanding what mistake |
| 12:08.10 | brlcad | run: loop 000 359 10 |
| 12:08.12 | brlcad | what does it do? |
| 12:09.36 | Izak__ | It prints 000 010 up to 340 |
| 12:09.55 | Izak__ | brlcad: Note that This did not work before |
| 12:10.13 | Izak__ | So it's not like I never actually tried runing that script |
| 12:10.19 | brlcad | I know that |
| 12:10.32 | brlcad | the point is to understand errors when encountered |
| 12:10.46 | brlcad | if the error was yours, then it was a point to learn |
| 12:10.56 | brlcad | if the error was in a tool, it was worth fixing |
| 12:11.04 | brlcad | at this point, there's no information as to which it was |
| 12:11.40 | brlcad | if it works now, it was almost certainly something you did wrong, perhaps didn't have your PATH set to find loop or something, perhaps a typo |
| 12:12.28 | brlcad | you'll discover if you have not already, I have NO patience for ignoring errors ... it doesn't help you learn, it doesn't help the project improve |
| 12:12.53 | brlcad | what does this produce: for i in `for i in `loop 000 359 10` ; do rt -a $i -e 35 -w 640 -n 480 -o image$i.png database.g top_level_object |
| 12:13.02 | brlcad | oops, not that |
| 12:13.28 | brlcad | for i in `loop 000 359 10` ; do echo "rt -a $i" ; done |
| 12:13.43 | brlcad | that |
| 12:14.04 | brlcad | should print a bunch of rt -a # lines |
| 12:14.53 | Izak__ | rt: BRL-CAD geometry database not specified.... |
| 12:15.33 | brlcad | you got that by running "for i in `loop 000 359 10` ; do echo "rt -a $i" ; done" ? |
| 12:15.37 | brlcad | I doubt it |
| 12:17.45 | Izak__ | I get rt -a 000 up to rt -a 350 as output |
| 12:18.35 | Izak__ | brlcad: When i worked with this script before, I threw questions to IRC as to the that impediments and i get no response |
| 12:19.36 | brlcad | from what I'm seeing, there are no impediments |
| 12:20.00 | brlcad | the page as written is working for you right now exactly as written |
| 12:20.18 | brlcad | the only variable is you! :) |
| 12:22.42 | Izak__ | This is so NOT a good day |
| 12:23.04 | brlcad | you seem to be really resisting the notion that you did something wrong |
| 12:23.24 | brlcad | if that's a cultural issue, you're going to have to get over it ... that's how we learn |
| 12:23.49 | brlcad | I suggest giving it another try, maybe with 1 degree ... and using one of the three animation composers |
| 12:23.55 | brlcad | so it's smooth |
| 12:24.22 | brlcad | 640 480 is to big, a 512x512 default is fine or even smaller so it renders faster |
| 12:24.48 | Izak__ | I am already correcting the movie brlcad: |
| 12:25.13 | brlcad | really look forward to seeing it spinning |
| 12:25.33 | Izak__ | Okay . /me working on that |
| 12:28.17 | Ch3ck_ | is waiting on brlcad, to clarify on the math on pulling transformations from current centrepoint based on the AABB |
| 13:28.07 | *** join/#brlcad kesha_ (~kesha@49.249.200.58) | |
| 14:35.01 | Notify | 03BRL-CAD Wiki:Phoenix * 6120 /wiki/User:Phoenix/GSoc2013/Reports: /* Week 13 */ |
| 14:38.52 | *** join/#brlcad AndChat|317009 (~AndChat31@1.38.31.251) | |
| 14:53.24 | kesha_ | brlcad: I have a doubt - http://paste.kde.org/p542855c0/ |
| 15:01.30 | kesha_ | I found a good site - http://www.steptools.com/support/stdev_docs/stpfiles/ap203/ testing step samples from here. |
| 15:35.56 | brlcad | kesha_: you can use a precompiled 7.22.0, that's fine |
| 15:37.00 | brlcad | I suggest running from your build directory for svn checkouts, though (don't install them, just run them after "make" with "bin/step-g" |
| 15:39.00 | kesha_ | okay. |
| 16:35.04 | kanzure | hello cruel world |
| 16:35.08 | kanzure | ``Erik: ping again? see backlog. |
| 16:35.58 | brlcad | kesha_: is that 7.22.0 import/rendering of d.g you e-mailed correct or not? |
| 16:36.15 | kanzure | brlcad: https://github.com/cogitokat/brlcad-python-ctypes-demo/blob/master/globe.py |
| 16:39.30 | brlcad | kanzure: hah, that's awesome |
| 16:39.46 | kanzure | brlcad: hmm line 152 seems very wrong |
| 16:39.48 | brlcad | it's a terrible proc-db, but awesome conversion |
| 16:39.54 | brlcad | (i wrote it) |
| 16:39.58 | kanzure | because it was already defined above |
| 16:40.01 | kanzure | i see :) |
| 16:40.40 | kanzure | i still need to clean up the wrappers a lot |
| 16:40.56 | kanzure | and figure out how to make it work out-of-the-box on the major platforms. hrm. |
| 16:42.08 | kanzure | since the core libraries don't change that often, i think writing the wrappers manually will be better. the generated junk is unacceptable. |
| 16:42.45 | Notify | 03BRL-CAD Wiki:IIIzzzaaakkk * 6121 /wiki/Animation: /* Fly-around animation */ |
| 16:45.59 | kesha_ | brlcad: There was no image file of it. I think it is correct .Isn't it ? |
| 16:53.03 | zero_level | brlcad, ``Erik : I would like to discuss the current TODO items in libicv. |
| 16:53.05 | brlcad | kanzure: so the challenge then is how to reduce it so the wrappers are actually fewer lines of code |
| 16:53.13 | Notify | 03BRL-CAD:mohitdaga * 57546 (brlcad/trunk/include/icv.h brlcad/trunk/src/libicv/fileformat.c brlcad/trunk/src/libicv/operations.c): Convert all void return functions to int return. This is done in a process to have better error contralability and to bring a unison. |
| 16:53.44 | zero_level | is trying to create a unified aproach for input argument validation. (in the meanwhile) |
| 16:53.47 | brlcad | kesha_: you're asking me if that's correct -- I have no idea, I do not know what was in that step file |
| 16:53.47 | kanzure | brlcad: agreed, i think there's a lot of common idioms used everywhere that can be collapsed into some core helper methods |
| 16:54.25 | brlcad | kesha_: that gets at understanding the converter, you need something that will tell you whether it's right or wrong (an image, another cad program, a screenshot, a nod from cliff, something) |
| 16:54.41 | brlcad | kanzure: some of that can/should certainly happen on the C side too |
| 16:54.55 | brlcad | but I know there are some python-specific features that could probably be leveraged |
| 16:54.56 | kesha_ | But, looking at the model, it smells bad |
| 16:55.19 | kanzure | brlcad: i definitely want things like python lists to work by default, instead of having to use ctypes.POINTER(bu.bu_list) etc. |
| 16:58.28 | kesha_ | brlcad: Isn't there any other way ? Can't we infer from .stp / .g file in any way or the output on terminal during conversion(step-g) or anything else ? |
| 17:03.19 | brlcad | kesha_: that was a half dozen ways |
| 17:03.30 | brlcad | inferring from the file itself requires reading the file's contents |
| 17:06.13 | kesha_ | I dont think it is necessary for testing for me to learn that. Better, I will choose models carefully. Right ? |
| 17:35.27 | brlcad | kesha_: no it's not necessary, but you need to have SOME way to know whether something is right or not |
| 17:36.16 | brlcad | and that's an image, a video, verifying the model in another program, seeing some screenshot of what it looks like, asking someone else that might know, reading the file's contents in detail ... SOMETHING |
| 17:36.41 | brlcad | if it "smells" bad like you said, all the more reason to check it in something else |
| 17:39.08 | kesha_ | I have freecad installed. Do you know how to check stp models there or shall I google it ? Any other than freecad, you suggest to install ? |
| 17:42.07 | brlcad | give them a try |
| 17:42.13 | brlcad | the point is just cross validation |
| 17:42.27 | brlcad | if you see a whole lot more 3D geometry, then something may be wrong |
| 17:42.41 | brlcad | if you basically see the same thing, it's a validation of sorts |
| 17:43.04 | brlcad | next step would be to check trunk, then to check an even earlier version |
| 17:44.12 | kesha_ | ok. Continuing .. |
| 17:44.13 | brlcad | kesha_: note this is what it looked like at one point in the past: http://brlcad.org/gallery/screenshots/Via_OpenBook_part_d |
| 17:44.26 | brlcad | screenshot says 7.14.7 |
| 17:45.05 | brlcad | so you'll want to test a version before and/or near that to try and reproduce that view |
| 17:45.36 | brlcad | at a glance, it's definitely worse |
| 17:45.42 | brlcad | (now) |
| 17:45.51 | brlcad | so the goal is to find exactly when it got worse |
| 17:48.19 | kesha_ | For that, I will need multiple versions . My question was that - how to go back and forth from one version to another when dealing with many versions. |
| 18:05.45 | ``Erik | kanzure: when sf changed their repo, I completely regenerated the git-svn bridge... srry |
| 18:06.35 | kanzure | ``Erik: okay. i'm wondering about regenerating it again to not include the sourceforge urls because sourceforge is (apparently) unreliable about these urls. |
| 18:06.40 | kanzure | jrayhawk: ping |
| 18:07.03 | kanzure | ``Erik: the reason i'm bothering is because i had brlcad.git from a year ago (or whatever) and all of the commit ids changed on me.. |
| 18:08.57 | ``Erik | kanzure: with any luck, this is a once in a decade event and BRL-CAD will have caught up to the mid 00's by 2023 ;) |
| 18:09.06 | kanzure | 00's ? |
| 18:09.14 | ``Erik | git is '05 |
| 18:09.31 | kanzure | you have much more trust in sourceforge than i do :) |
| 18:09.42 | kanzure | well anyway, it's just really problematic because it breaks all of the brlcad git repos out there |
| 18:10.31 | kanzure | redoing the commit messages to not mention the sourceforge url will prevent sourceforge from causing this problem in the future |
| 18:11.10 | kanzure | alternatively i could setup my own git-svn bridge but then the commit ids will be different between people that clone the git repo from brlcad.org/brlcad.git and whatever i setup. |
| 18:11.24 | brlcad | there is actually a need to upgrade our backend repo soon |
| 18:11.37 | brlcad | dont' know if that'll affect anything gitwise |
| 18:11.52 | brlcad | won't change urls |
| 18:11.58 | kanzure | could you elaborate on the change? |
| 18:12.08 | ``Erik | going from svn 1.7 to 1.8 won't change anything re git |
| 18:12.08 | brlcad | kanzure: or you could just use an svn checkout too ;) |
| 18:12.28 | brlcad | the repo very well may still be 1.4 |
| 18:12.37 | brlcad | i couldn't find anything authoritative |
| 18:12.43 | brlcad | svnsync might be upgrading on the fly |
| 18:12.44 | kesha_ | waiting .. |
| 18:12.46 | kanzure | sourceforge allows you to upgrade the svn version of the repo server? |
| 18:12.53 | ``Erik | migrating from svn 1.4 to, say, git 1.8 would be keen |
| 18:12.57 | ``Erik | or darcs, even |
| 18:13.10 | ``Erik | just not mercurial O.o |
| 18:13.13 | brlcad | kesha_: never just "wait", you don't have time to wait .. at least keep trying things and learning while "waiting" ;) |
| 18:13.56 | brlcad | kesha_: there are many ways you can go about doing what you asked |
| 18:14.36 | brlcad | you can "svn up -rREV ." to change a checkout to a different revision |
| 18:15.05 | brlcad | you can create separate build trees per revision or just delete your build dir each time |
| 18:16.09 | kesha_ | deleting every time will be too tedious :/ |
| 18:16.39 | brlcad | you're trying to find a specific single commit that made it worse |
| 18:16.48 | ``Erik | still more productive than sitting on irc and bitching about how hard it is *cough* |
| 18:17.12 | kesha_ | is there any link to different versions and its REVISIOn stamp ? |
| 18:17.17 | brlcad | nope |
| 18:17.41 | brlcad | best is probably to run "svn annotate NEWS" and look at the revision near each release header |
| 18:18.39 | brlcad | so you'll then test.. you know 57500: broken |
| 18:18.50 | brlcad | 50000:broken |
| 18:18.57 | brlcad | 40000:still broken |
| 18:19.04 | brlcad | 30000:no step importer |
| 18:19.11 | brlcad | 35000:working |
| 18:19.21 | brlcad | 38000:broken |
| 18:19.31 | brlcad | 37000:working |
| 18:19.33 | brlcad | etc |
| 18:19.45 | brlcad | until you narrow down on exactly what made it worse |
| 18:20.08 | brlcad | note that not every revision will necessarily compile, so you might need to jump a few revisions forward or backwards |
| 18:21.08 | kesha_ | cool. Got it :) |
| 18:21.16 | ``Erik | bisection ftw |
| 18:21.21 | brlcad | how long does a compilation take you? |
| 18:21.48 | brlcad | WRITE DOWN each version you try and the result (even if it didn't compile) |
| 18:21.48 | kesha_ | 5-7 mins |
| 18:21.54 | brlcad | okay, that's good |
| 18:22.25 | brlcad | you should be able to isolate what made it worse within an hour or two at most, presuming you FIRST confirm it looking better |
| 18:22.50 | brlcad | jump to the revision near 7.14.6 first |
| 18:23.08 | brlcad | make sure you can find a version that looks like the screenshot |
| 18:23.08 | kesha_ | okay |
| 18:23.14 | Notify | 03BRL-CAD:starseeker * 57547 brlcad/trunk/src/conv/step/g-step/g-step.cpp: Use the output file name in the step header |
| 18:23.59 | brlcad | svn log src/conv/step and svn log src/other/step will give you a hint at what revisions might be important |
| 18:28.20 | Notify | 03BRL-CAD Wiki:IIIzzzaaakkk * 6122 /wiki/User:Izak/GSOC_2013_logs: /* September 9th to September 14th */ |
| 18:29.54 | Notify | 03BRL-CAD:starseeker * 57548 brlcad/trunk/src/conv/step/g-step/g-step.cpp: Require object name rather than hardcoding brep.s - not much to argument handling at the moment, hopefully will be able to use the advanced argument parsing setup currently being developed later. |
| 18:40.33 | Notify | 03BRL-CAD Wiki:Vladbogolin * 6123 /wiki/User:Vladbogolin/GSoC2013/Logs: /* Week 13 */ |
| 18:40.43 | Izak__ | brlcad: Please can you down this video, watch and give me any corrections http://youtu.be/lXAvX-Mjb24 |
| 18:41.00 | Izak__ | ``Erik: Please can you down this video, watch and give me any corrections http://youtu.be/lXAvX-Mjb24 |
| 18:43.14 | ``Erik | Izak__: single hrt primitive using something like sh/orbit.sh ? |
| 18:45.58 | Izak__ | ``Erik: No. Used the script on the Animation wiki to produce it |
| 18:46.10 | Izak__ | Watched it already ? |
| 18:47.28 | ``Erik | hm, either way, top looks like an uneven number of hit points, bottom looks like bad hit points (boxy)... are there any reference images for what it's supposed to look like? |
| 18:47.49 | Notify | 03BRL-CAD Wiki:IIIzzzaaakkk * 6124 /wiki/User:Izak/GSOC_2013_logs: /* September 9th to September 14th */ |
| 18:48.12 | ``Erik | does the wireframe look good? maybe fire up mged, pull up a wireframe, raytrace it, then nirt where the raytraced image is 'wrong' and start looking for clues there? |
| 18:50.16 | Izak__ | ``Erik: I am not yet done wit the wireframe. Still working on it |
| 18:51.31 | ``Erik | aight, that'll help I think... otherwise, good progress! |
| 18:52.24 | Izak__ | ``Erik: The reference image is http://en.wikipedia.org/wiki/File:Heart3D.png |
| 18:52.54 | ``Erik | this is a more complex surface than I'd imagined :D |
| 18:53.14 | Izak__ | What is shown in the video is from images produced only after rt, that is, prep and shot |
| 18:54.04 | ``Erik | well, prop, shot, norm, bbox, import5, ... there are a lot of moving parts to get data into rt_shootray() :) |
| 18:54.06 | Izak__ | Still have to work on the explicit equation |
| 18:56.28 | ``Erik | s/prop/prep/ |
| 19:01.00 | Notify | 03BRL-CAD:mohitdaga * 57549 (brlcad/trunk/include/icv.h brlcad/trunk/src/libicv/bw.c and 9 others): Add macros for image validation. These reduces code repetetion. |
| 19:02.23 | zero_level | brlcad, ``Erik please have a look at 57549. |
| 19:02.42 | zero_level | Added macros as per (brlcad's) suggestion. |
| 19:03.36 | Izak__ | Off to bed |
| 19:11.27 | Ch3ck | brlcad: waiting.. |
| 19:15.30 | Notify | 03BRL-CAD Wiki:NyahCh3ck20 * 6125 /wiki/User:NyahCh3ck20/GSoc2013/Coding_Repor: /* Sept 09 - Sept 15 */ |
| 19:15.44 | Notify | 03BRL-CAD:tbrowder2 * 57550 (brlcad/trunk/src/util/CMakeLists.txt brlcad/trunk/src/util/bu_opt_parse.cpp and 2 others): have a buildable wrapper for TCLAP; more work needed for a working demo |
| 19:17.53 | Notify | 03BRL-CAD:mohitdaga * 57551 (brlcad/trunk/src/libicv/fileformat.c brlcad/trunk/src/libicv/filter.c and 2 others): Trailing WS |
| 19:24.39 | jrayhawk | kanzure: not clear on why you want to build a third history |
| 19:27.24 | kanzure | jrayhawk: one that will be immune to sourceforge |
| 19:27.49 | kanzure | jrayhawk: obv. they are unreliable. they should have at least kept the subdomains working or something.. they broke *all* git-svn mirrors that use them. they might do it again. |
| 19:28.45 | kanzure | dunno. probably an extreme response from me. |
| 19:29.01 | kanzure | but if i have 10 people forking it and i have to use git grafts in the future, i'm gonna have a lot of work to do. |
| 19:29.34 | kanzure | as far as i know, using a git graft is the only workaround if upstream changes under your feet |
| 19:53.50 | jrayhawk | the current histories are already immutable |
| 20:15.27 | kanzure | huh? if i have a pile of a few hundred commits on top of the last git repo, the only way to recover it is through a git graft. i just don't want to have to git graft again in the future.. |
| 20:27.25 | Notify | 03BRL-CAD:mohitdaga * 57552 brlcad/trunk/src/libicv/color_space.c: Validation and Log messages for color_space utilities. |
| 20:30.34 | kesha_ | brlcad: make gives http://paste.kde.org/p4a9bfa90/ |
| 20:37.35 | Notify | 03BRL-CAD:mohitdaga * 57553 (brlcad/trunk/include/icv.h brlcad/trunk/src/libicv/CMakeLists.txt): Rename decimate.c to size.c. Originally decimate.c was supposed to contain rountines to scale down image size. Later enlarging routines were also added to create a single icv_resize api. |
| 20:42.26 | zero_level | kesha_ : are you trying to build from brlcad main src code ? or STEP ? |
| 20:42.53 | zero_level | looks at the error code |
| 20:43.11 | kesha_ | zero_level: brlcad main src code |
| 20:43.26 | zero_level | kesha_ Just figured :) |
| 20:43.28 | kesha_ | version 7.14.6 |
| 20:43.35 | zero_level | kesha_ what commands did you run ? |
| 20:44.19 | kesha_ | sh autogen.sh -> this created configure script and then ./configure --enable-optimized |
| 20:44.32 | zero_level | kesha_ : I believe you did svn up -rX , where X = V22.0 tag ? |
| 20:45.29 | zero_level | do we need autogen ? |
| 20:46.09 | kesha_ | X=V14.6 tag |
| 20:46.29 | kesha_ | yes , because there is no CMAkeLists.txt |
| 20:46.34 | zero_level | ok. |
| 20:46.40 | kesha_ | And that is what instructed in INSTALL file |
| 20:48.10 | kesha_ | http://paste.kde.org/pfd7e0b7d/ |
| 20:51.45 | zero_level | kesha_ : are u building inplace ? or in some directory ? |
| 20:54.14 | kesha_ | inplace .. in build directory, its again giving error. |
| 20:54.45 | kesha_ | http://paste.kde.org/pbe1ceccb/ |
| 21:02.58 | kesha_ | starseeker: can you look into why 7.14.6 version is not able to compile? |
| 21:04.15 | Notify | 03BRL-CAD:starseeker * 57554 brlcad/trunk/src/conv/step/g-step/Assembly_Product.cpp: Make some notes on what we need to look for when dealing with combs. |
| 21:06.54 | Notify | 03BRL-CAD:mohitdaga * 57555 brlcad/trunk/src/libicv/encoding.c: Add input argument validation. Also reduce a variable. |
| 21:14.16 | Notify | 03BRL-CAD:mohitdaga * 57556 (brlcad/trunk/src/libicv/ppm.c =================================================================== and 37 others): Adding ppm.c This file will contain reading and writting ppm formats. |
| 21:23.52 | starseeker | kesha_: not at the moment - can you try a newer one? |
| 21:24.55 | starseeker | or just try configure from the release tarball without using autogen? |
| 21:25.21 | starseeker | newer autogen tools and older autogen files may not be a happy combination |
| 21:26.01 | kesha_ | starseeker: basically I want to check where the d.stp model from your collection viaopenbook is getting worse. |
| 21:26.14 | kesha_ | Screenshot has 7.16.7 version |
| 21:28.52 | Notify | 03BRL-CAD:mohitdaga * 57557 brlcad/trunk/src/libicv/filter.c: Input argument validation in icv_fade(). Also API function should not exit an app instead show a log message and return Error. |
| 21:30.18 | Notify | 03BRL-CAD:mohitdaga * 57558 brlcad/trunk/src/libicv/filter.c: WS |
| 21:32.30 | *** join/#brlcad kesha_ (~kesha@49.249.200.58) | |
| 21:36.26 | Notify | 03BRL-CAD:mohitdaga * 57559 brlcad/trunk/src/libicv/filter.c: TYPO |
| 21:44.21 | Notify | 03BRL-CAD:mohitdaga * 57560 (brlcad/trunk/src/libicv/bw.c brlcad/trunk/src/libicv/pix.c): Remove unwanted macros and headers. These are not needed due to usage fopen instead of open. |
| 21:46.00 | Notify | 03BRL-CAD:mohitdaga * 57561 (brlcad/trunk/src/libicv/CMakeLists.txt brlcad/trunk/src/libicv/fileformat.c brlcad/trunk/src/libicv/ppm.c): Move ppm_write and image_flip function to ppm.c |
| 21:49.58 | *** join/#brlcad pawleeq_ (~pawleeq@static-cl031028150026.unet.cz) | |
| 21:50.03 | pawleeq_ | hello |
| 21:51.04 | Notify | 03BRL-CAD:mohitdaga * 57562 brlcad/trunk/src/libicv/ppm.c: use fopen instead of open. Also remove unwanted headers and macros. |
| 21:51.55 | pawleeq_ | I am trying to build brlcad from svn and need to set the location of tk.h file, however, if I pass TCL_BIN_PREFIX="/usr/include/tcl8.5/" to cmake it ignores it, what I am doing wrong? |
| 21:54.34 | Notify | 03BRL-CAD:mohitdaga * 57563 brlcad/trunk/src/libicv/ppm.c: FIX fixed size buffer. |
| 21:59.28 | *** join/#brlcad mpictor (~mark@c-67-177-102-131.hsd1.in.comcast.net) | |
| 22:12.44 | Notify | 03BRL-CAD:tbrowder2 * 57564 brlcad/trunk/src/util/dsp_add_t.cpp: move TCLAP customization into file bu_opt_parse_private.h |
| 22:12.49 | pawleeq_ | Found it: the variable is -DBRLCAD_TCL_BIN_PREFIX="path" |
| 22:14.57 | Notify | 03BRL-CAD:tbrowder2 * 57565 brlcad/trunk/src/util/bu_opt_parse.h: rearrange struct; add union; neaten |
| 22:15.41 | Notify | 03BRL-CAD:tbrowder2 * 57566 brlcad/trunk/src/util/bu_opt_parse.h: remove two unneeded decls |
| 22:16.35 | Notify | 03BRL-CAD:tbrowder2 * 57567 brlcad/trunk/src/util/bu_opt_parse.cpp: correct func signature |
| 22:17.59 | Notify | 03BRL-CAD:tbrowder2 * 57568 brlcad/trunk/src/util/dsp_add2.c: add arg defs same as dsp_add_t; remove old code; change to C90 comments |
| 22:20.47 | Notify | 03BRL-CAD:tbrowder2 * 57569 brlcad/trunk/src/util/dsp_add2.c: add comment; build is good; still need to work on guts of wrapped function |
| 22:37.57 | Notify | 03BRL-CAD:mohitdaga * 57570 brlcad/trunk/src/libicv/ppm.c: Add ppm read function. This function also posses special power to read from stdin. |
| 22:39.36 | Notify | 03BRL-CAD:mohitdaga * 57571 brlcad/trunk/src/libicv/fileformat.c: Add case for ppm in icv_read |
| 22:40.00 | brlcad | Izak_: OMG that's perfect |
| 22:40.11 | brlcad | (the youtube video) |
| 22:43.01 | Notify | 03BRL-CAD:mohitdaga * 57572 brlcad/trunk/src/libicv/fileformat.c: Trailing WS |
| 22:46.05 | Notify | 03BRL-CAD:tbrowder2 * 57573 brlcad/trunk/src/util/dsp_add_t.cpp: move non-TCLAP code out of try block |
| 22:47.29 | Notify | 03BRL-CAD:mohitdaga * 57574 brlcad/trunk/src/libicv/fileformat.c: remove double checking of image. |
| 22:48.50 | brlcad | Izak_: now THAT I can seriously see the shape and how it's evaluating in prep+shot ... that's really fantastic |
| 22:50.22 | Notify | 03BRL-CAD:mohitdaga * 57575 brlcad/trunk/src/libicv/fileformat.c: Input Argument Validation |
| 22:52.04 | brlcad | Izak_: that also implies you have quite a challenge ahead of you... that looks like a variety of errors |
| 22:53.27 | brlcad | looks like the bounding box might be wrong, the solver is surprisingly stable |
| 22:53.56 | Notify | 03BRL-CAD:mohitdaga * 57576 brlcad/trunk/src/libicv/fileformat.c: TYPO |
| 22:55.37 | brlcad | the equations look like they may be off (normal doesn't seem to match the hit points) and that's going to be very tedious to verify |
| 22:55.54 | zero_level | hii brlcaD |
| 22:56.14 | brlcad | nice progress zero_level |
| 22:56.50 | brlcad | noticed you elminated the block in bw.c, no ICV_IMG_VAL_INT(img) |
| 22:56.59 | brlcad | looks like maybe a mistake? |
| 22:57.42 | zero_level | brlcad : bw_write is hidden |
| 22:57.53 | zero_level | I checked in icv_write instead. ;) |
| 22:58.04 | zero_level | Thanks for reviewing though. |
| 22:58.06 | zero_level | :) |
| 22:58.09 | brlcad | okay, cool |
| 22:58.54 | brlcad | kesha_: did you get a comile? |
| 22:59.46 | Notify | 03BRL-CAD:mohitdaga * 57577 brlcad/trunk/src/libicv/pix.c: Dont need image validation in pix_write. Image is validated in icv_write instead. |
| 22:59.48 | zero_level | but brlcad : I didnt like those macros. :) |
| 22:59.54 | brlcad | forgot about the build system switch, going back that far -- nice to see you found and followed the old build instructions |
| 23:00.02 | mpictor | brlcad: have you heard of a new static analyzer called STACK? it looks for code that may get optimized away due to undefined behavior http://blog.regehr.org/archives/970 |
| 23:00.15 | brlcad | zero_level: alternative solution? |
| 23:00.21 | zero_level | It makes the code reading messy. |
| 23:00.23 | brlcad | mpictor: nope |
| 23:00.36 | zero_level | WE just let them as it is :) |
| 23:00.37 | brlcad | zero_level: so better name? |
| 23:00.53 | zero_level | can u suggest me some. I bad at naming. |
| 23:00.59 | zero_level | ^am |
| 23:01.00 | mpictor | one drawback is that it requires a bleeding-edge version of clang |
| 23:01.36 | brlcad | how's that a drawback? :) |
| 23:01.40 | brlcad | (just kidding) |
| 23:01.49 | mpictor | lol |
| 23:02.06 | brlcad | loves the latest versions of clang and gcc, they are both breaking new ground crazy fast with their static analysis abilities |
| 23:02.16 | brlcad | nice competition |
| 23:02.55 | zero_level | brlcad : I will also need your feedback on how to proceed for the TODO items you have listed. |
| 23:03.00 | brlcad | we need to run ourselves through coverity again, it's been over a year .. I bet they've got a slew too |
| 23:03.45 | zero_level | In particular for "Support dynamically loading formats (e.g., define a plug-in API)." |
| 23:03.48 | brlcad | kesha_: i'm not sure about that autogen.sh error, but you could run it again as "VERBOSE=1 ./autogen.sh" to hopefully see .. the real test is whether you can run ./configure and make |
| 23:04.17 | zero_level | As per today I am signing off. |
| 23:04.38 | brlcad | zero_level: okay |
| 23:04.43 | brlcad | however |
| 23:04.49 | brlcad | the way to proceed with dynamic loading |
| 23:04.54 | brlcad | is to first encapsulate |
| 23:05.11 | brlcad | put everything needed for bw support, for example, into one file or one subdirectory |
| 23:05.42 | zero_level | brlcad : Even if that needs repetition of crtain functions ? |
| 23:05.54 | brlcad | depends, that's what has to be inspected |
| 23:06.06 | zero_level | the current status is all the formats are in one file. |
| 23:06.06 | brlcad | what might belong as libicv plugin-API |
| 23:06.12 | zero_level | bw.c pix.c ppm.c |
| 23:06.17 | brlcad | yeah, but not everything needed for that format |
| 23:06.24 | brlcad | there are typedefs in the header |
| 23:06.31 | brlcad | there are enum avlues |
| 23:06.36 | brlcad | /avlues/values/ |
| 23:06.43 | zero_level | I amm planning the development such that we will need functions from encoding.c |
| 23:07.02 | brlcad | literally *everything* needed for a format needs to end up in one subdir |
| 23:07.43 | zero_level | ok. If that is strict notion then we might end up repiting codes from encoding.c |
| 23:09.18 | brlcad | well that's the challenge |
| 23:09.30 | brlcad | how to do it without repeating *anything* |
| 23:09.57 | brlcad | usually this is achieved with registration functions and private API |
| 23:10.06 | brlcad | but we can talk more later |
| 23:10.15 | brlcad | good progress today on a lot of the little things |
| 23:10.32 | zero_level | ok. brlcad |
| 23:10.53 | brlcad | Ch3ck: what's up? |
| 23:11.56 | brlcad | waiting on ... ? you have a notion that the bounding box center is the "origin" ... primitives do not have a consistently defined origin, so that's the best you have to work with |
| 23:15.57 | brlcad | create an translation matrix that will move the bounding box to that position and then move the primitive by the inverse translation (so it ends up at the origin), and you pull the original matrix up |
