| 00:01.39 | CIA-50 | BRL-CAD: 03Sean 07http://brlcad.org * r2314 10/wiki/Google_Code_In/Project_Ideas: few more code tasks |
| 00:03.10 | CIA-50 | BRL-CAD: 03Sean 07http://brlcad.org * r2315 10/wiki/Google_Code_In/Project_Ideas: |
| 00:26.49 | CIA-50 | BRL-CAD: 03Sean 07http://brlcad.org * r2316 10/wiki/Google_Code_In/Project_Ideas: add doc tasks |
| 00:32.17 | CIA-50 | BRL-CAD: 03Sean 07http://brlcad.org * r2317 10/wiki/Google_Code_In/Project_Ideas: fix and separators |
| 00:46.15 | CIA-50 | BRL-CAD: 03Sean 07http://brlcad.org * r2318 10/wiki/Google_Code_In/Project_Ideas: add a couple user interface ideas |
| 00:53.22 | CIA-50 | BRL-CAD: 03brlcad * r41198 10/brlcad/trunk/src/librt/db_tree.c: remove dead code, and it's one of just two places in librt that directly calls nmg routines (should decouple fully) |
| 00:57.51 | CIA-50 | BRL-CAD: 03brlcad * r41199 10/brlcad/trunk/src/librt/db_tree.c: avoid using the 'new' c++ keyword for variable names |
| 03:25.05 | CIA-50 | BRL-CAD: 03Sean 07http://brlcad.org * r2319 10/wiki/Google_Code_In/Project_Ideas: expand on the documents section tasks |
| 03:47.48 | starseeker | brlcad: sorry, wasn't at computer - do you have what you need from me? |
| 03:55.35 | CIA-50 | BRL-CAD: 03Sean 07http://brlcad.org * r2320 10/wiki/Google_Code_In/Project_Ideas: outreach task |
| 03:56.37 | CIA-50 | BRL-CAD: 03Sean 07http://brlcad.org * r2321 10/wiki/Google_Code_In/Project_Ideas: |
| 04:05.15 | *** join/#brlcad velociostrich (~nsd@c-68-37-119-2.hsd1.nj.comcast.net) |
| 04:51.44 | louipc | brlcad: I wan't part of gsoc, but I did volunteer to consult for that usability thing |
| 06:08.06 | CIA-50 | BRL-CAD: 03Sean 07http://brlcad.org * r2322 10/wiki/Google_Code_In/Project_Ideas: add references |
| 06:24.26 | CIA-50 | BRL-CAD: 03Sean 07http://brlcad.org * r2323 10/wiki/Google_Code_In/Project_Ideas: expand quality assurance tasks |
| 06:47.41 | CIA-50 | BRL-CAD: 03Sean 07http://brlcad.org * r2324 10/wiki/Google_Code_In/Project_Ideas: expand Research |
| 07:07.32 | CIA-50 | BRL-CAD: 03Sean 07http://brlcad.org * r2325 10/wiki/Google_Code_In/Project_Ideas: expand Training |
| 08:10.04 | *** join/#brlcad mafm_ (~mafm@83.37.177.224) |
| 12:29.22 | starseeker | ah hah http://www.freebsdsoftware.org/multimedia/mpeg.html |
| 12:29.42 | starseeker | figured there had to be some kind of liberally licensed MPEG encoder somewhere |
| 12:42.02 | starseeker | ok, looks like the Berkeley MPEG tools are also viable |
| 12:43.00 | starseeker | heh - lotta patches to that code, and more needed for the warnings, but at least gentoo organized 'em |
| 13:01.44 | starseeker | stashes copies where they are handy, in addition to putting the archives all in one directory (hate the old ones that barf files all over toplevel) and converting the stanford doc.ps to pdf to it can be read |
| 13:01.53 | starseeker | http://bzflag.bz/~starseeker/mpeg/ |
| 13:30.30 | starseeker | hmm - ffv1 might be useful for saving and playing back things like MGED wireframe videos, although I suppose it can't be viewed by default by most commercial players |
| 13:38.17 | starseeker | might actually be smarter to strip libavcodec down to just the non-controversial parts and use that... |
| 15:27.53 | brlcad | starseeker: looking at developing a pix-mpg tool? |
| 15:29.24 | brlcad | or a pix-avi |
| 15:29.53 | brlcad | pix-flv |
| 15:36.32 | ``Erik | has just been saving to png and using ffmpeg, but would appreciate a decent lossless (or high quality) codec O.o |
| 15:36.58 | ``Erik | mpeg artifacts on simple phong shaded geometry is just.. ugly |
| 17:02.14 | *** join/#brlcad Zaebos (~irc@pd95b7f5e.dip0.t-ipconnect.de) |
| 18:07.04 | starseeker | brlcad: was thinking along the lines of defining an animation in archer, then having an "Export as MPEG" option |
| 18:07.53 | starseeker | not sure MPEG-1 is suitable for wireframe visualizations though - or if it is, it'll take some care |
| 18:08.05 | starseeker | wishes the MPEG-2 patents expired sooner |
| 18:08.53 | starseeker | the lossless codec in libavcodec looks like a good compromise, except of course for the problem of playing it back without a special purpose tool |
| 18:09.08 | starseeker | or mplayer/vlc/something else not likely to be there by default |
| 18:09.56 | starseeker | MJPEG with very minimal jpeg compression might be an option, but I don't know about playback on various platforms |
| 18:12.16 | starseeker | hmm... http://msdn.microsoft.com/en-us/library/dd390691%28VS.85%29.aspx |
| 18:31.20 | starseeker | auugh |
| 18:31.26 | starseeker | http://ffmpeg.arrozcru.org/wiki/index.php?title=Main_Page |
| 18:31.42 | starseeker | and here the whole point of the CMake effort is to build cleanly under MSVC++ |
| 18:31.44 | starseeker | blegh |
| 18:33.46 | starseeker | wonders how hard the parts we might want to use would be to backport |
| 18:38.37 | *** join/#brlcad R0b0t1 (~Enigma@64-136-219-55.dyn.everestkc.net) |
| 18:38.37 | *** join/#brlcad R0b0t1 (~Enigma@unaffiliated/r0b0t1) |
| 18:55.09 | brlcad | http://compression.ru/video/codec_comparison/lossless_codecs_2007_en.html |
| 18:56.21 | starseeker | hah, cool: http://www.digitalpreservation.gov/formats/fdd/fdd000063.shtml |
| 18:57.28 | *** join/#brlcad PrezKennedy (MK@whitecalf.net) |
| 18:57.36 | starseeker | brlcad: cool! |
| 18:58.00 | starseeker | now if we can just find a website that identifies what video formats are supported "out of the box" on various OSes... |
| 18:58.08 | brlcad | yeah, philosophically, we should stick to a lossless encoding if at all possible |
| 18:59.03 | starseeker | aside from the possible exception of MJPEG done with little compression, I'm drawing a blank on lossless compression formats that are supported out of the box and/or compatible with our licensing requirements |
| 18:59.39 | brlcad | it not being default installed isn't as much of an issue -- especially if they can feed the lossless video into mplayer/vlc/quicktime/whatever to convert to lossy |
| 19:00.02 | starseeker | in that case we're probably looking at ffv1 |
| 19:00.19 | starseeker | libavcodec is LGPL now, apparently |
| 19:00.50 | louipc | starseeker: so time for mingw? |
| 19:00.52 | starseeker | need to check if the default huffyuv encoder in libavcodec is lgpl or not - the original code is GPL |
| 19:01.11 | starseeker | louipc: I'd like to have mingw as an option, but it can't really be the default |
| 19:01.24 | starseeker | Windows devs expect to be able to use Visual Studio |
| 19:01.30 | louipc | I guess only if you want ffmpeg hehe |
| 19:02.17 | starseeker | to do what we would need, we actually would need a pretty tiny subset of ffmpeg - just the libavcodec pieces for our codec(s) of interest, and probably libavformat to be able to create files |
| 19:04.47 | starseeker | pix files have got to be close to a "raw" feed, although there are probably low-level formatting issues |
| 19:05.50 | starseeker | 'course, this comes under the "low priority" heading at the moment :-/ |
| 19:08.34 | starseeker | looks like http://ffdshow-tryout.sourceforge.net/ + http://virtualdub.sourceforge.net/ can enable conversion of lots of codecs on Windows |
| 19:09.38 | brlcad | starseeker: I wouldn't be surprised if there isn't some tiny academic encoder code that we could bundle |
| 19:10.05 | brlcad | libavcodec can probably do the trick, but we don't need 30+ codecs, we really just need one |
| 19:10.28 | starseeker | right - so the question becomes how easy it is to strip libavcodec down to just what we need/want |
| 19:10.33 | brlcad | wrapping something in an ogg or avi or mpeg container is the easy bit |
| 19:10.48 | brlcad | or finding another small lib |
| 19:11.05 | starseeker | GPL seems to quite popular for video encoders |
| 19:12.28 | starseeker | hmm - Dirac may have a lossless mode |
| 19:13.49 | brlcad | http://freshmeat.net/projects/h264bitstream |
| 19:14.24 | starseeker | h264 is a patent minefield though |
| 19:20.21 | starseeker | huh, interesting: http://wiki.multimedia.cx/index.php?title=Apple_QuickTime_RLE |
| 19:21.35 | starseeker | wikipedia claims it works well for movies with "large areas of constant color" |
| 19:24.40 | brlcad | yeah, rle would be great |
| 19:24.59 | brlcad | might even be easy to just implement that format using libutahrle |
| 20:19.18 | ``Erik | 5~/cl |
| 23:17.54 | starseeker | ``Erik: heh, reading this Java article on slashdot is entertaining... |
| 23:18.30 | starseeker | from one of the comments: "Sounds like we need a new, and truly open, language and runtime for the 21st century." |
| 23:18.34 | starseeker | Lisp will rise again! |
| 23:33.27 | *** join/#brlcad louipc (~louipc@archlinux/trusteduser/louipc) |