| 00:07.15 | *** join/#brlcad roberthl (n=robert@rhl.me.uk) |
| 00:45.25 | CIA-30 | BRL-CAD: 03ralith * r35202 10/rt^3/trunk/src/g3d/ (OgreGLWidget.cxx OgreGLWidget.h): Added a slot to OgreGLWidget to set projection type (perspective or orthographic). |
| 00:48.39 | CIA-30 | BRL-CAD: 03ralith * r35203 10/rt^3/trunk/src/g3d/OgreGLWidget.cxx: Scrapped public access to OgreGLWidget's Ogre internals. |
| 00:48.49 | Ralith | wups |
| 00:48.59 | CIA-30 | BRL-CAD: 03ralith * r35204 10/rt^3/trunk/src/g3d/OgreGLWidget.h: Added header changes forgotten in previous commit. |
| 01:00.35 | *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6) |
| 01:07.41 | CIA-30 | BRL-CAD: 03ralith * r35205 10/rt^3/trunk/src/g3d/ (MainWindow.cxx OgreGLWidget.cxx mainwindow.ui): Attempted to hook up the camera projection mode menu to Ogre's relevant functionality. No visible effect. |
| 01:07.58 | Ralith | oh wait! |
| 01:13.20 | Ralith | yay |
| 01:13.25 | Ralith | now the GUI actually does something :] |
| 01:13.30 | Ralith | although it doesn't have any visible side effects... |
| 01:13.35 | CIA-30 | BRL-CAD: 03ralith * r35206 10/rt^3/trunk/src/g3d/ (MainWindow.cxx OgreGLWidget.cxx OgreGLWidget.h): Working (albeit slightly hackish) projection mode configuration! |
| 01:18.09 | Ralith | next is camera controls, I think. |
| 01:18.17 | Ralith | then I'll take another whack at seeing why Ogre is so weirdly misaligned |
| 01:36.43 | CIA-30 | BRL-CAD: 03Ralith 07http://brlcad.org * r1575 10/wiki/User:Ralith: Log for 2008-07-17 |
| 01:37.14 | Ralith | I have to say, producing visible results is quite fun. |
| 06:33.48 | *** join/#brlcad stevegt_ (n=stevegt@c-24-130-122-25.hsd1.ca.comcast.net) |
| 09:05.28 | *** join/#brlcad _sushi_ (n=_sushi_@84-72-10-209.dclient.hispeed.ch) |
| 13:34.40 | ``Erik | yeah, that's the money shot O.o |
| 14:52.05 | *** join/#brlcad Unregistered (i=DCEmuUse@89-201-241-153.dsl.optinet.hr) |
| 16:09.18 | *** join/#brlcad stevegt_ (n=stevegt@110.sub-70-210-150.myvzw.com) |
| 17:25.50 | *** join/#brlcad bobbens (i=bobbens@saw4ever.de) |
| 17:26.17 | bobbens | Has anyone compiled 7.14.8 with gcc 4.4.0? |
| 17:26.45 | louipc | yep |
| 17:27.03 | bobbens | got it working? it fails here with http://pastebin.com/d5c694624 |
| 17:27.26 | louipc | well it got caught on jove, but then I just disabled jove |
| 17:27.28 | louipc | no need for that |
| 17:27.42 | bobbens | mmm, was trying to use the Arch AUR package :) |
| 17:28.09 | bobbens | what's this jove you speak of? |
| 17:28.27 | bobbens | actually, it does seem to --disable-jove |
| 17:28.39 | louipc | it's an old text editor bundled with brlcad |
| 17:28.50 | louipc | notice I uploaded that PKGBUILD |
| 17:29.01 | bobbens | ah |
| 17:29.39 | louipc | hmm what arch are you on? |
| 17:29.44 | bobbens | x86_64 |
| 17:30.02 | louipc | :( |
| 17:30.31 | louipc | I can't debug that heh... I'm i686 |
| 17:31.01 | bobbens | old school :) |
| 17:32.56 | louipc | someone else had a config error too.. it didn't want to build ogl support even with --with-opengl |
| 17:49.52 | louipc | bobbens: hmm in my build step wasn't even configured |
| 17:55.41 | louipc | bobbens: I get the same error when trying to build it manually |
| 17:59.33 | bobbens | ah, ok |
| 17:59.40 | louipc | Here's a description of the issue: http://blog.flameeyes.eu/2009/07/02/how-_not_-to-fix-gcc-4-4-bugs |
| 18:00.58 | ``Erik | jove should be disabled in a normal 'configure' :/ |
| 18:03.17 | bobbens | louipc: mmm, blog isn't loading for me |
| 18:03.31 | bobbens | funny how I say that and it loads after 3 failed attemps |
| 18:03.37 | louipc | haha |
| 18:03.39 | bobbens | more proof that complaining on irc fixes your problems |
| 18:04.14 | louipc | so that needs to be patched... but it also seems something is wrong with configure |
| 18:07.39 | louipc | ``Erik: jove is on [auto] right now... |
| 18:07.58 | louipc | which means [yes] if you don't have it eh? |
| 18:08.14 | louipc | well... on 7.14.8 anyways |
| 18:08.41 | louipc | .. and SVN hehe |
| 18:08.55 | ``Erik | hm, I thought it was set up to only build if configure couldn't find a "sane" editor (vi, vim, ... even emacs) |
| 18:10.48 | louipc | that doesn't work very well for building packages |
| 18:10.58 | louipc | you don't really know what the target system will have installed |
| 18:11.23 | louipc | but everybody usually has vi :D |
| 18:13.14 | ``Erik | I explicitely --disable-jove for the fbsd package/port |
| 18:14.51 | louipc | shall we disable it in brl-cad proper? |
| 18:15.01 | louipc | I'm all for it |
| 18:15.35 | louipc | for disabling it |
| 18:16.44 | ``Erik | I've been saying that should be done, well... for 6 years now |
| 18:17.01 | ``Erik | I think it's on the deprecated list and will be gone for 8 or something |
| 18:17.19 | louipc | ... |
| 18:17.39 | ``Erik | grab a couple car batteries, jumper cables, and visit brlcad to discuss it *shrug* :D |
| 18:17.55 | louipc | hehe |
| 18:18.27 | louipc | so when 8 rolls around it'll be disabled |
| 18:18.36 | louipc | and then totally removed at 9? |
| 18:20.00 | ``Erik | gone in 8 I think? *shrug* |
| 18:42.12 | louipc | bobbens: hmm that const char/char bug is fixed in svn you might want to try svn |
| 18:46.43 | bobbens | louipc: might give it a shot tomorrow, thanks for the insight |
| 18:47.03 | louipc | cool no problem |
| 19:07.16 | *** join/#brlcad _sushi_ (n=_sushi_@84-72-10-209.dclient.hispeed.ch) |
| 20:08.20 | ``Erik | .deps retardedness. *sigh* |
| 20:08.35 | louipc | who did what? |
| 20:09.03 | ``Erik | I'm building on a box that hasn't been updated in about a year |
| 20:09.12 | louipc | nice |
| 20:09.24 | ``Erik | auto* seems to get fucktarded with the .deps entries when things move |
| 20:11.25 | ``Erik | svn makes moving things easy, so it's done without consideration for impact... cvs required care, so much more care was put into file location and moving stuff wasn't taken so lightly :( |
| 20:11.44 | ``Erik | if your tools make fucking up easier to recover from, you're gonna fuck up more. :( |
| 20:11.48 | ``Erik | </rant> |
| 20:12.08 | louipc | makes sense |
| 20:12.15 | bobbens | on svn it's not that bad since you don't keep a local copies of changes |
| 20:12.23 | bobbens | with git it's insanely easy to bloat a repo with binary stuff |
| 20:12.28 | bobbens | has that problem :( |
| 20:12.52 | ``Erik | svn keeps local copies, but it doesn't keep the "all of everything everywhere" that all distributed systems carry |
| 20:13.04 | ``Erik | git, darcs, mercurial, etc... |
| 20:13.24 | bobbens | well with the price of memory, diskspace and such going down, decentralized version control systems will dominate I'd say |
| 20:13.34 | bobbens | plus it's nice not to depend on a server |
| 20:13.42 | louipc | yeah definitely |
| 20:14.03 | bobbens | if only binary diffs got up to par :P |
| 20:14.16 | bobbens | maybe have some git stuff that actually does per-filetype patches or something |
| 20:14.29 | ``Erik | *shrug* drew in #ucw/#lisp/#tech.coop is into distributed systems, but he tends to take his boat out for long periods with no intarwebz access... |
| 20:15.14 | bobbens | well philosophically it's also very different |
| 20:15.21 | ``Erik | personally, I don't spend more than a day without intarwebz, and can destructure my patches for committing, so even RCS is doable, CVS has some nice bennies |
| 20:15.22 | louipc | I'm a fan of git too |
| 20:15.28 | bobbens | with svn I rarely ever work with branches since they're a pain in the ass with svn |
| 20:15.36 | bobbens | but with git I'm creating/destroying/merging branches all the time |
| 20:15.41 | louipc | yep |
| 20:15.55 | ``Erik | heh, the big point of svn was to make branches easy to work with :D |
| 20:16.07 | louipc | hah |
| 20:16.14 | bobbens | well compared to git... :) |
| 20:16.19 | bobbens | you do git checkout -b foo |
| 20:16.22 | ``Erik | people who don't... quite.. GET branches... really fail hard with CVS |
| 20:16.25 | bobbens | and instantly have a branch |
| 20:16.28 | louipc | why did they keep the silly $Id tag then? |
| 20:16.48 | bobbens | git merging is also very fancy |
| 20:16.53 | bobbens | when you start cherry-picking and doing funky stuff |
| 20:17.04 | bobbens | I'm a pretty die-hard git fanatic, so my views are slightly biased |
| 20:17.18 | louipc | bobbens: cool what kind of devel do you do? |
| 20:17.53 | bobbens | robotics (work), game (hobby) and everything else (documents - LaTeX, $HOME, etc...) -> git |
| 20:17.58 | bobbens | i'm still a student though :) |
| 20:18.14 | bobbens | just wanted to try out brlcad because my current UGV chassis design is epic failure |
| 20:18.27 | bobbens | found out that the support foundation doesn't let the encoders fit :P |
| 20:18.33 | bobbens | so I might as well remake it |
| 20:18.43 | bobbens | properly modelling the batteries, PCBs and such |
| 20:19.12 | louipc | neat |
| 20:20.25 | bobbens | I meant to try brlcad like 3 years ago |
| 20:20.29 | bobbens | but learning curve was too harsh |
| 20:20.40 | bobbens | I have a wee bit of free time so I'll give it another shot :P |
| 20:20.45 | louipc | it hasn't changed much |
| 20:20.53 | louipc | :P |
| 20:21.23 | louipc | hey and you're smarter than 3 yrs ago.. probably |
| 20:21.30 | bobbens | yeah |
| 20:21.40 | bobbens | well I'm used to solid works though from the few CAD classes we have to take |
| 20:21.57 | bobbens | and I actually have a project for it :) |
| 20:22.03 | bobbens | last time it was just to mess around |
| 20:22.28 | bobbens | but isn't some GSoC guy adding some stuff to do work in the 3d view directly with surfaces like solid works does? |
| 20:23.39 | louipc | umm there's work on a new GUI |
| 20:24.07 | bobbens | I just need to do parametric modelling basically |
| 20:24.28 | bobbens | well the only thing I can really work with is convex sheet metal |
| 20:24.39 | bobbens | so the design will be simple, mainly it's getting the holes and sizes right all over |
| 21:04.49 | brlcad | waves hello |
| 21:06.32 | bobbens | hello brlcad |
| 21:23.29 | brlcad | louipc: go ahead and default it to no |
| 21:24.16 | brlcad | the const char * to char* conversion was fixed |
| 21:26.02 | brlcad | bobbens: glad to hear about the interest -- latest svn should take care of that build problem |
| 21:26.05 | brlcad | ~cadsvn |
| 21:26.06 | ibot | To obtain BRL-CAD from Subversion: svn checkout https://brlcad.svn.sourceforge.net/svnroot/brlcad/brlcad/trunk brlcad |
| 21:27.25 | brlcad | we still don't provide parametric modeling, fwiw, until more infrastructure is in place on parametric and constraints support, but there is a lot of work going into the gui |
| 21:28.42 | bobbens | ah |
| 21:28.47 | bobbens | that's a bummer |
| 21:28.58 | bobbens | parametric modelling was the main thing I was looking forward to |
| 21:30.55 | brlcad | what do you mean when you say parametric modeling? |
| 21:31.15 | brlcad | (some folks mean different things unfortunately, gotta ask) |
| 21:31.15 | *** join/#brlcad Axman6 (n=Axman6@pdpc/supporter/student/Axman6) |
| 21:31.17 | bobbens | like you want the chassis to be 50 by 50, and then you want a hole 10 from the edge |
| 21:31.24 | bobbens | then you decide you want it to be 40 by 40 instead |
| 21:31.29 | bobbens | and the hole stays 10 from the edge |
| 21:32.32 | brlcad | okay, yeah |
| 22:21.33 | bobbens | make benchmark is pretty impressive :) |
| 22:22.48 | brlcad | what's yoru vgr count? |
| 22:23.52 | bobbens | vgr count? |
| 22:24.09 | bobbens | well I'm working on something similar to that for GSoC and SDL :) |
| 22:31.16 | brlcad | bobbens: there's a final number that's reported during the benchmark, that's called your vgr count |
| 22:31.30 | brlcad | it's a linear metric of overall computation performance, baselined |
| 22:31.53 | brlcad | which gsoc project? link? |
| 22:31.59 | bobbens | *vgr ghanima 5600.81 5339.51 6737.44 5472.48 4824.49 26.91 4666.94 |
| 22:32.01 | bobbens | hadn't gotten there yet :) |
| 22:33.00 | brlcad | 4666, not too bad |
| 22:33.06 | brlcad | I presume that wasn't optimized either |
| 22:33.18 | bobbens | http://socghop.appspot.com/student_project/show/google/gsoc2009/sdl/t124024854746 |
| 22:34.06 | brlcad | ah, sams your mentor, cool -- he's a great guy |
| 22:34.15 | bobbens | yeah, I was with SDL last year too |
| 22:34.26 | bobbens | only problem is he has too much real life :) |
| 22:34.30 | bobbens | which is a good thing for him though :P |
| 22:35.24 | brlcad | nods |
| 22:36.02 | brlcad | benchmark isn't exactly our testing suite, just one of many regression tests and a stand-alone performance measurement tool in itself |
| 22:36.21 | bobbens | well I just do regressions |
| 22:36.22 | bobbens | not performance |
| 22:36.34 | bobbens | the problem is the cross-platformness and nature of SDL |
| 22:36.35 | brlcad | as ray-tracing is at the core of what we do, using raytracing for our regressions is natural |
| 22:36.43 | bobbens | how does one test input and stuff in a cross-platform way? :) |
| 22:38.42 | brlcad | have per-platform regression tests, then higher-level integration tests that combine results |
| 22:39.06 | bobbens | well, that's the thing I'm trying to avoid |
| 22:39.18 | bobbens | I mean SDL runs on a lot of platforms, that would kill me :) |
| 22:39.28 | bobbens | many I don't have access to |
| 22:40.06 | brlcad | sure, but that's not exactly a testing question |
| 22:40.19 | brlcad | testing setup isn't the same as running the tests :) |
| 22:40.43 | bobbens | well if I have to write platform-specific tests, I need access to those machines :) |
| 22:41.10 | brlcad | you need someone with access to those machines, you only need access if you're also the tester |
| 22:41.20 | bobbens | I mean I'd have to write stuff for the iPhone, nintendo DS, playstastion 3, minix, ps2, etc... |
| 22:41.30 | bobbens | well I'd need access to write the input tests :) |
| 22:41.43 | bobbens | anyway, I'll just write the crossplatform parts |
| 22:42.06 | bobbens | and leave the platform-specific as non-automated |
| 22:42.12 | bobbens | force the user to do stuff |
| 22:43.05 | brlcad | my point was that you don't personally need access, you just need someone who does that you can work with and/or that can provide the required data |
| 22:43.29 | brlcad | i mean it certainly could help, but it's not genuine to say it's required |
| 22:43.33 | bobbens | well by personal access I meant ssh :P |
| 22:43.49 | bobbens | last year I did the haptic interface for mac os x over ssh :) |
| 22:43.54 | brlcad | and I mean you don't even need to touch a minix system in order to write a good minix test |
| 22:44.04 | bobbens | it helps :) |
| 22:44.10 | brlcad | it *can* help |
| 22:44.21 | bobbens | I rarely ever write code that compiles first pass unless it's under 100 lines |
| 22:45.50 | bobbens | anyway, this is offtopic :) |
| 22:45.57 | brlcad | given sdl's job is already one of abstraction, it would seem logical to leverage the end-users more as an actual part of the integrated testing infrastructure |
| 22:46.20 | bobbens | well ideally I would have a "make test" that would automatically test everything for regressions |
| 22:46.39 | bobbens | could be run after every couple of commits |
| 22:46.54 | bobbens | goes to sleep, robots don't code themselves in the morning if a bobbens doesn't wake up |
| 22:47.03 | brlcad | like having high-level tests of SDL_Window, for example, and say (just for example) that if you had a make test rule, and it was run on a checkout and it failed -- the a regression reporte would be sent in to a report accumulation system |
| 22:47.47 | brlcad | then cross-platform testing is leveraged across the community, or could be set up in a nightly compile farm or whatever suits |
| 23:20.46 | CIA-30 | BRL-CAD: 03johnranderson * r35207 10/brlcad/trunk/src/librt/primitives/bot/bot.c: |
| 23:20.46 | CIA-30 | BRL-CAD: "db adjust" and "db put" will now warn the user if an invalid vertex index |
| 23:20.49 | CIA-30 | BRL-CAD: is entered for a face. |
| 23:28.35 | *** join/#brlcad Patmcc19 (n=chatzill@71-223-26-58.phnx.qwest.net) |