| 00:01.39 | CIA-23 | BRL-CAD: 03starseeker * r51662 10/brlcad/trunk/src/librt/test_botpatches.cpp: bit of optimization |
| 00:14.54 | *** join/#brlcad Mahi (~Mahi@ec2-50-16-98-252.compute-1.amazonaws.com) |
| 01:04.24 | CIA-23 | BRL-CAD: 03crdueck * r51663 10/brlcad/trunk/src/libged/analyze.c: analyze_arbn_face and analyze_bot_face dont need a face centroid to compute volume, just any point on the face. use a vertice instead |
| 01:10.20 | brlcad | cristina: did you get things sorted out with your graph command? |
| 01:11.13 | brlcad | cristina: note that 'geometree' is a pure script command, not a libged command -- at least not yet |
| 01:12.00 | brlcad | of course the GeometryBrowser itcl class calls into several other commands (several of which are libged commands) |
| 01:16.33 | brlcad | the "proper" way to implement a command like geometree now is to have a libged 'graph' command call your DisplayGraph class (directly or indirectly through a graph.tcl source script that it loads) |
| 01:53.39 | CIA-23 | BRL-CAD: 03starseeker * r51664 10/brlcad/trunk/src/ (rt/viewarea.c util/lowp.c): Apply patches lowp.patch and viewarea.patch by Ksenija Slivko (ksuzee) from 3529861 |
| 02:02.47 | CIA-23 | BRL-CAD: 03starseeker * r51665 10/brlcad/trunk/src/fb/pl-fb.c: Apply patch pl-fb.patch from 3534954 by Ksenija Slivko (ksuzee) |
| 02:05.10 | CIA-23 | BRL-CAD: 03starseeker * r51666 10/brlcad/trunk/src/fb/pl-fb.c: Whoops - back out 51665 - not sure this preserves functionality |
| 06:59.58 | *** join/#brlcad cristina (~quassel@78.96.44.61) |
| 07:30.26 | andrei_ | brlcad, I have been working to finish all unit tests for libpkg ( since I didn't know what to do with the csv data ) |
| 07:31.03 | andrei_ | but there are some issues I just can't get over, I simply don't understand it. |
| 07:33.15 | andrei_ | http://slexy.org/view/s208s9eeL1 |
| 07:33.38 | andrei_ | fork() gets me over the blocking call, but it still won't output the printf's |
| 08:31.06 | *** join/#brlcad PrezKennedyX (~DarkCalf@173.231.40.99) |
| 08:41.36 | *** join/#brlcad andrei (andrei@188.25.161.217) |
| 09:01.06 | *** join/#brlcad stas (~stas@82.208.133.12) |
| 09:07.18 | CIA-23 | BRL-CAD: 03Phoenix 07http://brlcad.org * r4226 10/wiki/User:Phoenix/GSoc2012/Reports: /* Week 10 */ |
| 09:12.17 | CIA-23 | BRL-CAD: 03phoenixyjll * r51667 10/brlcad/trunk/src/librt/opennurbs_ext.cpp: Do not generate the whole surface tree - only a small useful subset is generated simultaneously with the intersection process. Much faster. |
| 09:30.26 | *** join/#brlcad ksuzee (~ksu@193.151.107.42) |
| 09:33.42 | ksuzee | starseeker: hello, Cliff |
| 10:06.19 | ksuzee | starseeker: thanks for your checking, I've answered you to mail-list |
| 10:27.36 | *** join/#brlcad flash_ (~flash@adsl-ull-36-47.49-151.net24.it) |
| 10:33.44 | *** join/#brlcad Andy_G (~chatzilla@77-253-129-223.adsl.inetia.pl) |
| 10:56.18 | *** join/#brlcad reuben (~reuben@93-97-69-251.zone5.bethere.co.uk) |
| 10:59.28 | Andy_G | Hi everyone :) My name is Jedrzej Gorski and I would like to ask few questions about open projects for ESA SOCIS program. |
| 10:59.58 | Andy_G | Anyone up for some conversation?? |
| 11:05.24 | flash_ | I'm up... but sadly i'm here for the same reason you are.. |
| 11:06.16 | reuben | That makes three of us. |
| 11:06.23 | flash_ | =) |
| 11:07.16 | flash_ | I think they all are sleeping now.. |
| 11:07.20 | reuben | Which projects are you guys looking at, I'm going for the celestial mechanics one but have a couple of questions. |
| 11:07.29 | flash_ | it's 4-5 o'clock in America |
| 11:07.44 | flash_ | I was thinking about "bending light' |
| 11:08.00 | ``Erik | 3-7, actually, and just ask the questions and lurk... someone will eventually answer :) |
| 11:08.33 | brlcad | go ahead and ask you questions, someone will always eventually reply if you're still here when they read it |
| 11:08.41 | reuben | Cool, just got to get my thoughts in order. |
| 11:08.44 | ``Erik | woops, 1-7, forgot about oconus bits |
| 11:15.13 | Andy_G | I guess there is no best time to catch up with you :) |
| 11:17.31 | Andy_G | In general I was looking at project ideas and "Application Format" |
| 11:20.28 | Andy_G | I never used your software before but I'm curious if I can contribute anything valuable. Some of the ideas looks challenging like "Non-vacume gravity simulator" |
| 11:21.43 | Andy_G | My question is what sort of skills are in place for successful collaboration with your group in area of EASY and MEDIUM projects ?? |
| 11:22.54 | Andy_G | I thinking about level of programming skills and etc. |
| 11:23.56 | ``Erik | I'm not sure exactly what ideas were posted for the esa thing, but the usual skills for open source development... we use C, subversion (on sourceforge), cmake, ... if you can check out the repo, compile and launch 'mged', you're halfway there |
| 11:26.22 | Andy_G | ok .. I'll check that :) |
| 11:26.36 | ``Erik | some projects may involve tcl (gui stuff) and/or c++ (nurbs/step, third party stuff like bullet) |
| 11:27.55 | ``Erik | but doing an svn checkout, compiling and running mged should explain quite a bit :) |
| 11:31.00 | Andy_G | I was working in a past for project GENSO (ESA Education Office) and I had to install all necessary software, set-up SVN and deal with typical developer issues but all project is based on Java |
| 11:31.36 | reuben | Hi everyone I've been looking to get involved in an open source project for a while and I figure the ESA summer of code is the perfect way to start. |
| 11:31.36 | reuben | I've downloaded brl cad and started learning to use the programming to give me an idea of where to start. I'm looking at adding celestial mechanics to brl cad as this fits well with my background, but would like to get some feedback on how to interface with the rest of brl cad, specifically whether you believe it is best implemented as a script or as a tool. Secondly what is the planned primary use for this, making pretty videos or hardco |
| 11:31.36 | reuben | re simulation. i.e. should I focus on the animation side of things, ability to track particles with the camera exctera, or the ability to easily get/plot particle data, and the accuracy and stability over long simulation periods. |
| 11:31.38 | reuben | Also interested in the best place to start looking at the code, it would be great to sort out a patch before finishing my application. |
| 11:32.30 | Andy_G | never the less I'm looking for new challenge so I'll try to do what you suggest :) |
| 11:36.47 | andrei | ``Erik, regarding libpkg unit tests. For example, is it possible to unit test a close connection function ? or one that sends a file ? |
| 11:37.36 | andrei | brlcad suggested to write unit tests for _all_ libpkg functions, I did write unit tests for some and I will put them on sourceforge, but on others I just don't understand how to do it |
| 11:37.53 | ``Erik | reuben: looking at the wiki page, I'd assume the simulation would be functionality in a library and hooked up to other tools (like archer). |
| 11:40.32 | ``Erik | andrei: some functions would be very difficult to write unit tests for and would require mocking other functions or system calls.. (I'd think for pure unit testing, few functions would not require mocking of some sort...). Could you clean up what you have so far and upload it as a patch to sourceforge? |
| 11:40.59 | andrei | upload it as a " big " patch or each unit test separately? |
| 11:41.24 | ``Erik | how many are there? |
| 11:43.12 | ``Erik | if more than, say, half a dozen, one big patch |
| 11:43.26 | andrei | ah, no, I had something like 3 or 4 that were fully working |
| 11:43.30 | andrei | and some others were I was stuck |
| 11:43.57 | andrei | so just one patch, right. Thanks |
| 11:44.57 | brlcad | reuben: it is a great way to get started (with open source) and welcome :) |
| 11:45.25 | brlcad | Andy_G: the projects will always be scaled to your abilities, so not much worry there |
| 11:45.43 | brlcad | the bigger idea is to find topics that sound challengingg but also doable and interesting |
| 11:46.10 | brlcad | they don't even have to be on our project lists, they just have to be part of brl-cad and something we're mutually interested in |
| 11:46.43 | *** join/#brlcad andrei (~andrei@188.25.161.217) |
| 11:47.01 | brlcad | I'd suggest starting with our virtual disk image distribution where everything is already downloaded, installed, and ready to go -- give that a try and see if you can run the tools in there, then see if you can compile the sources bundled there |
| 11:49.41 | reuben | Thanks, I was using MGED before, is Archer just an alternative GUI, it looks like Qt. |
| 11:50.21 | ``Erik | archer is an alternative gui that will hopefully replace mged in the future... it's written in tcl/tk |
| 11:51.13 | Andy_G | brlcad: Thanks for encourage and tips :) I'll give it a try today :) |
| 11:52.03 | starseeker | reuben: Archer uses (mostly) the new ttk widgets. Alas, not Qt ;-) |
| 11:52.33 | starseeker | but still, it will eventually consitute a significant upgrade to MGED's GUI abilities. |
| 11:54.29 | reuben | It's looking good already, looking forward to the completion. |
| 11:55.48 | CIA-23 | BRL-CAD: 03Erik 07http://brlcad.org * r4227 10/wiki/Building_from_SVN: |
| 11:57.32 | ``Erik | that page needs to be updated for cmake stuff, wonder if the example should use a seperate build directory O.o |
| 12:00.26 | flash_ | Hi, i'm thinking about the 'bending light' project (for SOCIS of course). I took a look to raytrace.h and optical.h and i'm wandering if i've got the necessary skills to work on it. I'm familiar with C, but i've never participated in a open source project.. |
| 12:02.47 | flash_ | i know nothing about raytracing too (besides the basic idea). |
| 12:03.22 | flash_ | i'm downloading the virtualbox distro now, and i'll give it a try |
| 12:03.50 | CIA-23 | BRL-CAD: 03Erik 07http://brlcad.org * r4228 10/wiki/User:Popescu.andrei1991: stub in dates |
| 12:11.47 | CIA-23 | BRL-CAD: 03Popescu.andrei1991 07http://brlcad.org * r4229 10/wiki/User:Popescu.andrei1991: /* Week 9 */ |
| 12:36.34 | *** join/#brlcad elf11 (~elf11@p5.eregie.pub.ro) |
| 12:39.04 | elf11 | Hello, I've been looking around the brlcad project and I think about applying fo the celestial mechanics particle system or the non vacuum gravity simulator |
| 12:39.50 | elf11 | I know that last year ESA participant worked on the Vacuum gravity simulator, and was wondering if I could still give it a shot |
| 12:40.58 | elf11 | I have good knowledge of C/C++ also some graphic design and working at the moment on a personal project where I try to integrate the bullet engine in a game |
| 13:22.10 | CIA-23 | BRL-CAD: 03128.63.32.62 07http://brlcad.org * r4230 10/wiki/Building_from_SVN: autotools -> CMake |
| 13:22.28 | *** join/#brlcad elf11_ (~elf11@p5.eregie.pub.ro) |
| 13:23.52 | starseeker | elf11_: certainly continuing work on the bullet-based gravity simulator is a viable project |
| 13:24.25 | starseeker | elf11_: to try it out, you need to have bullet installed on your system (we don't currently bundle it) and check out the simulate command |
| 13:26.23 | elf11_ | Okay, thank you for the info I will look further into it and will come back with more questions |
| 13:27.43 | starseeker | elf11_: previous work is described here: http://brlcad.org/wiki/User:Abhijit |
| 13:29.38 | elf11_ | Yes, I've read that and I will look more through the code :) |
| 13:39.30 | andrei | why is this a blocking call ? pkg_open(server, s_port, "tcp", NULL, NULL, NULL, NULL); |
| 13:41.34 | andrei | I can't use threading because it s not portable and I can't use select because I don't get " access" to the fd. What other ways to escape a blocking call can I use ? |
| 13:45.22 | flash_ | starseeker: what skills are required for 'bending light'? i've worked with C/C++, openGl and (if it's relevant) CUDA |
| 13:45.36 | flash_ | but have no experience in raytracing |
| 14:03.28 | andrei | ``Erik , I have a persistent issue in several unit tests. There are functions such as pkg_open or pkg_getclient that are blocking. They handle the fd internally so I can t get it with select() without modifying the pkg code. Can't use threading aswell, what can I do ? |
| 14:06.59 | starseeker | flash_: not completely sure on that one... |
| 14:07.14 | starseeker | flash_: better ask brlcad what he had in mind |
| 14:07.47 | flash_ | ok |
| 14:07.59 | flash_ | i'll wait for him to come online |
| 14:08.04 | flash_ | thanks |
| 14:08.30 | starseeker | presumably he's talking about simulating light interacting with gravity (gravational lensing, etc.) but I don't know how specifically he wants to deal with that in the raytracer |
| 14:09.17 | flash_ | Yeah, the fact is that i've no idea of how a raytracer works, so |
| 14:09.24 | flash_ | i would need some advices |
| 14:09.46 | starseeker | flash_: I'm pretty sure you'll need to know that - probably want to start with some basic reading on raytracing |
| 14:10.46 | flash_ | i know the idea of raytracing, but obiouvsly i'll need more knowldege |
| 14:11.11 | flash_ | i'm downloading the virtual machine distribution |
| 14:11.28 | flash_ | so i can start looking around |
| 14:14.26 | ``Erik | andrei: I don't see why you couldn't use the threading in libbu (bu_parallel) if you want to do that kind of integration testing... with pkg_open(), a 'pure' unit test would involve replacing things like socket() and connect() in the executables runtime symbol table (using hackery like dlsym or LD_PRELOAD), or the executable could fork() itself, run the server in one image and the client doing the test int he other |
| 14:19.28 | andrei | I ll take some time to document LD_PRELOAD and dlsym, as I have never heard of them before |
| 14:19.46 | ``Erik | they're unix specific things, no more portable than pthreads would be |
| 14:20.20 | ``Erik | um, but I did do some dlsym/dlopen wrapping for isst, I think I may've put it in bu, lemme look |
| 14:21.29 | ``Erik | yeah, src/libbu/dlfcn.c ... bu_dlopen, bu_dlsym, bu_dlclose... it's fairly advanced stuff and you can screw up the running program if you're not careful, which is very hard to debug |
| 14:22.07 | ``Erik | easiest would be to fork() and run a server and client pair in proc spaces |
| 14:24.20 | andrei | <PROTECTED> |
| 14:24.48 | ``Erik | if((ppid=fork())==0) { /* server */ socket(); listen(); global_ready=1; accept(); exit(); } else if (ppid>0) { while(global_ready==0)sleep(0); pkg_open(); /* do stuff */; wait(ppid); return status; } |
| 14:25.01 | ``Erik | something of that nature, roughly |
| 14:42.11 | andrei | one last question : should I start the server with a system call ( system(./tpkg -R -p < port > ) or by effectively calling the function run_server ? |
| 14:45.12 | ``Erik | system() should never be used, execv() family is the right way to start another executable... if the run_serve function is available, that'd be best |
| 14:45.42 | andrei | I might move some code around to avoid duplications, but I should be able to use run_server. |
| 14:46.23 | ``Erik | (as an aside, "system call" usually refers to any function that calls into the kernel, such was write(), open(), sleep(), etc... not the system() function) |
| 14:46.48 | andrei | ah, sorry |
| 14:47.05 | CIA-23 | BRL-CAD: 03starseeker * r51668 10/brlcad/trunk/src/librt/test_botpatches.cpp: Try to consolidate some functionality that's getting used in multiple places |
| 14:47.41 | ``Erik | and the reason system() is bad is because it assumes the environment and path are all 'sane', creating a slew of vulnerabilities for malicious attacks or tricky bugs |
| 15:20.42 | *** join/#brlcad xth1 (~thiago@187-40-88-226.user.veloxzone.com.br) |
| 15:48.38 | *** join/#brlcad flash (~flash@adsl-ull-209-20.49-151.net24.it) |
| 16:02.13 | *** join/#brlcad Mahi_ (~Mahi@ec2-107-22-114-188.compute-1.amazonaws.com) |
| 16:05.50 | pnjun | Hi, anyone up for some questions about SOCIS 'bending light' project? |
| 16:14.03 | *** join/#brlcad Mahi (~Mahi@ec2-50-16-47-142.compute-1.amazonaws.com) |
| 16:56.45 | *** join/#brlcad Andy_G (~chatzilla@77-253-129-223.adsl.inetia.pl) |
| 17:09.06 | *** join/#brlcad dtidrow (~dtidrow@c-68-84-167-135.hsd1.mi.comcast.net) |
| 17:23.39 | *** join/#brlcad pnjun (~flash@adsl-ull-229-57.49-151.net24.it) |
| 18:05.28 | *** join/#brlcad flash (~flash@adsl-ull-229-57.49-151.net24.it) |
| 18:22.58 | *** join/#brlcad elf11 (~elf11@p5.eregie.pub.ro) |
| 18:27.38 | pnjun | \list |
| 18:28.02 | pnjun | sorry, wrong slash.. =) |
| 19:02.06 | *** join/#brlcad Andy_G (~chatzilla@77-253-129-223.adsl.inetia.pl) |
| 19:58.52 | CIA-23 | BRL-CAD: 03starseeker * r51669 10/brlcad/trunk/src/libbn/tri_tri.c: Avoid floating point fuzz issues in coplanar test requiring non-zero area. |
| 20:06.12 | CIA-23 | BRL-CAD: 03starseeker * r51670 10/brlcad/trunk/src/librt/test_botpatches.cpp: |
| 20:06.12 | CIA-23 | BRL-CAD: Crude logic for patch building (just yank overlapping face and create a patch |
| 20:06.12 | CIA-23 | BRL-CAD: out of it) but take the first steps to dealing with problematic edge triangles |
| 20:06.12 | CIA-23 | BRL-CAD: overlapping with other edges. Probably not fully correct or optimal yet. |
| 20:12.55 | *** join/#brlcad cristina (~quassel@188.24.50.251) |
| 20:13.10 | *** join/#brlcad cristina (~quassel@unaffiliated/cristina) |
| 20:16.49 | cristina | brlcad: Hello, regarding your question about my "graph" command: I didn't get things totally sorted out. |
| 20:16.58 | cristina | <PROTECTED> |
| 20:17.04 | cristina | However, I wanted to pop up the window with the "graph" command. |
| 20:17.11 | cristina | As for the libged command, I used the "bot" command as an example and implemented my ged_graph() routine (doesn't that represent a libged command?) |
| 20:17.48 | *** join/#brlcad stas (~stas@188.24.50.251) |
| 20:59.01 | *** join/#brlcad xth1 (~thiago@187-40-88-226.user.veloxzone.com.br) |
| 21:03.16 | CIA-23 | BRL-CAD: 03starseeker * r51671 10/brlcad/trunk/src/librt/ (CMakeLists.txt test_botpatches.cpp): Add triangles to edge set based on overlap activity. |
| 21:31.08 | CIA-23 | BRL-CAD: 03starseeker * r51672 10/brlcad/trunk/src/librt/CMakeLists.txt: Whoops - don't uncomment the fit/bottest stuff yet |
| 23:19.40 | *** join/#brlcad yiyus (1242712427@je.je.je) |