| 01:01.15 | *** join/#brlcad clock (~clock@77-58-143-135.dclient.hispeed.ch) |
| 02:41.30 | brlcad | blessing: glad you got it working |
| 02:42.47 | brlcad | a note of irc etiquette .. if you have more than 5 lines of something to paste, you should use a pastebin service (e.g., http://pastebin.ca ) where you post a url instead of pasting so many lines directly into the channel ... pasting too many lines can actually get you automatically kicked off of the network |
| 02:44.28 | brlcad | ``Erik: sounds like an expert system .. surely there are dozens that could be used, but they're all long term pains in the butt if it's for processes that change regularly |
| 03:33.19 | Notify | 03BRL-CAD:brlcad * 63139 (brlcad/trunk/BUGS brlcad/trunk/TODO): knowing the issue exists makes this one a show-stopper. commit 42149 introduced the spaces, but can't be wholesale reverted. need to walk through the commit and inspect nearby commits to find all string literals that had commas within them changed. bug uncovered by ?\200?\152tefan-Gabriel Mirea. |
| 03:33.28 | brlcad | stefan's on a roll pinpointing issues |
| 04:48.15 | *** join/#brlcad FreezingCold (~FreezingC@135.0.41.14) |
| 05:02.46 | *** join/#brlcad maths22 (~maths22@66-118-151-70.static.sagonet.net) |
| 05:02.46 | *** join/#brlcad n_reed (~molto_cre@66-118-151-70.static.sagonet.net) |
| 07:33.59 | *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51) |
| 07:44.46 | *** join/#brlcad bhattigurjot (~scott@202.164.53.117) |
| 08:05.32 | *** join/#brlcad ries (~ries@D979EA84.cm-3-2d.dynamic.ziggo.nl) |
| 08:06.48 | *** join/#brlcad mihaineacsu (~mihaineac@92.81.135.185) |
| 08:07.13 | *** join/#brlcad bhattigurjot (~scott@202.164.53.117) |
| 08:08.10 | *** join/#brlcad mihaineacsu_ (~mihaineac@92.81.128.142) |
| 08:17.43 | *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51) |
| 08:18.50 | *** join/#brlcad mihaineacsu (~mihaineac@92.81.39.52) |
| 08:44.54 | *** join/#brlcad kesha (~kesha@14.139.122.114) |
| 08:51.34 | *** join/#brlcad kesha_ (~kesha@14.139.122.114) |
| 10:23.37 | *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51) |
| 10:54.19 | ``Erik | brlcad: I suppose you could look at it as an expert system, but one with a single end node... (webmd's graph has a bazillion end nodes and I'd assume they try to make it as shallow a tree as possible... ours would be more like a path with little detours that push you back to the main path) |
| 10:55.34 | ``Erik | most of the process has been pretty static for a while... 'install X headers' hasn't changed in a lllooonnggg time, the last real change afaik was switching to cmake? |
| 10:55.41 | ``Erik | *shrug* was an idea :) |
| 11:44.07 | *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51) |
| 11:50.52 | *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51) |
| 12:13.00 | *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51) |
| 12:38.53 | ``Erik | http://gcc.godbolt.org/ |
| 12:39.04 | ``Erik | (interactive compiler for c/c++...) |
| 13:26.29 | *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51) |
| 13:38.28 | Notify | 03BRL-CAD:starseeker * 63140 brlcad/trunk/TODO: p2t should avoid crashing now. |
| 13:43.04 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) |
| 14:05.13 | *** join/#brlcad teepee- (bc5c2133@gateway/web/freenode/ip.188.92.33.51) |
| 14:41.08 | *** join/#brlcad mihaineacsu_ (~mihaineac@92.85.202.2) |
| 14:55.23 | *** join/#brlcad kintel (~kintel@unaffiliated/kintel) |
| 14:59.53 | Notify | 03BRL-CAD:brlcad * 63141 brlcad/trunk/src/libbu/affinity.c: set the mask in advance using the right type, quellage |
| 15:04.32 | Notify | 03BRL-CAD:brlcad * 63142 (brlcad/trunk/include/bu/cv.h brlcad/trunk/src/libbu/b64.c): bad things happen if a negative length is specified, wrong type. using size_t instead. |
| 15:06.00 | Notify | 03BRL-CAD:brlcad * 63143 brlcad/trunk/src/libbu/b64.c: style consistency |
| 15:20.10 | *** join/#brlcad kamal__ (~kamal@59.161.69.248) |
| 15:34.11 | Notify | 03BRL-CAD:brlcad * 63144 (brlcad/trunk/misc/CMake/test_srcs/builddelta_end.c.in brlcad/trunk/misc/CMake/test_srcs/builddelta_start.c.in brlcad/trunk/misc/CMake/test_srcs/print_timestamp.c): define _CRT_SECURE_NO_WARNINGS for these build system helper apps as they a have no threading potential. shouldn't be using _CRT_SECURE_NO_WARNINGS anywhere in the main sources, but is fine here. |
| 15:45.34 | Notify | 03BRL-CAD Wiki:TobiasDM633 * 0 /wiki/User:TobiasDM633: |
| 15:46.41 | *** join/#brlcad Darshpreet (~Darsh@202.164.53.117) |
| 15:51.42 | *** join/#brlcad kamal__ (~kamal@59.161.69.248) |
| 15:53.34 | Notify | 03BRL-CAD:brlcad * 63145 brlcad/trunk/src/libbu/affinity.c: indicate our intent, dwordified. |
| 15:53.45 | kamal__ | Hi I have been playing with mged recently. I am a li'l confused with the fact that materials are assigned to regions unlike shapes. What is the basic difference between shape and region? |
| 15:54.32 | kamal__ | Is shape something like a wireframe? Since BRL-CAD is based on CSG, so it must be solid. I want to confirm this too. |
| 15:55.54 | Notify | 03BRL-CAD:brlcad * 63146 brlcad/trunk/include/vmath.h: protect against the redefinition of INFINITY if we happen to include bu headers before we get to including this math.h header. this quells warnings in the Windows compile but potentially affects any platform that defines INFINITY. |
| 16:07.18 | *** join/#brlcad blessing (~chick@41.205.22.41) |
| 16:15.48 | *** join/#brlcad Darshpreet (~Darsh@202.164.53.117) |
| 16:28.53 | *** join/#brlcad bhattigurjot (~scott@117.199.99.5) |
| 16:53.08 | *** join/#brlcad Darshpreet (~Darsh@202.164.53.117) |
| 17:25.42 | *** join/#brlcad blessing (~chick@41.205.22.41) |
| 17:28.22 | *** join/#brlcad kamal__ (~kamal@59.161.77.119) |
| 17:53.46 | *** join/#brlcad FreezingCold (~FreezingC@135.0.41.14) |
| 17:55.48 | *** join/#brlcad gaganjyot (~gaganjyot@124.253.225.90) |
| 18:57.01 | *** join/#brlcad stefanmirea (~stefan@p16.eregie.pub.ro) |
| 19:50.38 | Notify | 03BRL-CAD:brlcad * 63147 brlcad/trunk/include/vmath.h: bah you know this. cpp is pure symbolic substitution, these are not variables in a stateful language. need a different approach: leverage pragma push_macro() and pop_macro() to get the desired effect. supported by llvm, gcc, and msvc, so this might not be portable enough. |
| 19:57.01 | Notify | 03BRL-CAD:starseeker * 63148 (brlcad/trunk/CMakeLists.txt brlcad/trunk/misc/CMake/Distcheck.cmake and 5 others): Experiment with EXCLUDE_FROM_DEFAULT_BUILD to see if it can make F7 viable in Visual Studio. |
| 19:58.08 | Notify | 03BRL-CAD:brlcad * 63149 brlcad/trunk/include/vmath.h: still need to define it to trigger the pop. |
| 20:31.27 | *** join/#brlcad clock (~clock@77-58-143-135.dclient.hispeed.ch) |
| 21:05.32 | *** join/#brlcad clock (~clock@77-58-143-135.dclient.hispeed.ch) |
| 21:14.49 | *** join/#brlcad stefanmirea (~stefan@p16.eregie.pub.ro) |
| 21:28.48 | *** join/#brlcad xemdetia (xemdetia@nat/ibm/x-foalcfvpshkatmgs) |
| 21:29.00 | *** part/#brlcad xemdetia (xemdetia@nat/ibm/x-foalcfvpshkatmgs) |
| 21:49.09 | *** join/#brlcad clock (~clock@77-58-143-135.dclient.hispeed.ch) |
| 21:54.24 | brlcad | that interactive compiler is really cool... |
| 22:08.15 | Stragus | An interactive compiler? |
| 22:08.25 | Stragus | You can see the assembly produced as you type code? :) |
| 22:15.17 | teepee | Stragus: yep -> <``Erik> [14:38:53] http://gcc.godbolt.org/ |
| 22:27.13 | Stragus | Hum... I like the concept |
| 22:28.02 | Stragus | It's not quite usable in a browser, but an IDE showing the assembly right away for the code one is typing would be neat |
| 22:45.27 | ``Erik | that interactive compiler thingymabob is on github, I'm sure you could modify it to work with emacs, vim, or some ide... :) |
| 22:47.04 | Stragus | Yup, I like it |
| 22:48.07 | ``Erik | prods at cuda some, even though opencl smells technically better (yeah yeah, betamax was technically better than vhs, darcs is technically better than git, ...) |
| 22:48.40 | Stragus | CUDA feels good with the low-level driver API, it's close to the hardware |
| 22:48.47 | Stragus | I'm not fond of that higher level API |
| 22:49.05 | ``Erik | that's why you only code in asm and not C, right? ;> *duck* |
| 22:50.06 | ``Erik | and instead of a nice shimano shifter, you reach down and push the derailleur by hand :D |
| 22:51.02 | ``Erik | done anything with the 'ptx' assembly yet? |
| 22:51.37 | Stragus | Yup, I used some PTX to do big number arithmetic and stuff like that |
| 22:51.58 | Stragus | It feels nice doing assembly while managing variables, not registers |
| 22:53.48 | ``Erik | is pondering writing a terrain erosion program using cuda, but has a pretty low level card... 16 core, 256m, I think 'fermi' capabilities |
| 22:55.35 | Stragus | Are you sure that's Fermi and not Tesla? That's a very low count of cores |
| 22:55.48 | Stragus | My old desktop Fermi has 512 cores |
| 22:56.28 | Stragus | Terrain erosion, water flow and so on? I have done something like that once... :) |
| 22:58.39 | ``Erik | um, lemme check on the specs again... ack, yes, tesla... cuda 1.1 capabilities |
| 22:59.38 | Stragus | Tesla hardware can only do one kernel at a time... so when the GPU is busy with heavy long-lasting kernels, your user interface freezes |
| 22:59.40 | ``Erik | http://paste.lisp.org/display/143985 |
| 23:08.09 | ``Erik | will have to remember he's using a very hobbled card :) |
| 23:08.13 | ``Erik | :o https://github.com/takagi/cl-cuda |
| 23:13.07 | Stragus | Fermi-grade hardware is very nice |
| 23:13.48 | Stragus | Kepler is also good obviously, but you have to access memory through texture fetches for best performance, that's a little nasty |
| 23:14.12 | *** join/#brlcad clock (~clock@84-72-8-53.dclient.hispeed.ch) |
| 23:46.47 | *** join/#brlcad clock (~clock@84-72-8-53.dclient.hispeed.ch) |
| 23:48.03 | ``Erik | is getting build failure on osx.9, INFINITY_ not defined |