[Ffmpeg-devel-irc] ffmpeg-devel.log.20140721

[burek]

[00:01] <cone-390> ffmpeg.git 03Andreas Cadhalpun 07master:e59ce54412f7: doc: fix two spelling errors
[01:26] <thardin> any ffmpeg devs going to europython?
[02:01] <cone-390> ffmpeg.git 03Diego Biurrun 07master:4de8b60684ce: idct: Move arm-specific declarations to a header in the arm directory
[02:01] <cone-390> ffmpeg.git 03Michael Niedermayer 07master:110420aac01d: Merge commit '4de8b60684ce13dff3e3d372dae4f49b9e53f755'
[02:12] <cone-390> ffmpeg.git 03Carl Eugen Hoyos 07master:b31d76e45fc3: tiff: support reading gray+alpha at 8 bits
[02:12] <cone-390> ffmpeg.git 03Michael Niedermayer 07master:08b597b1f8f2: Merge commit 'b31d76e45fc3c6529dd7109e721676f3ec376d00'
[02:22] <cone-390> ffmpeg.git 03Ronald S. Bultje 07master:c39059bea3ad: h264: Fix direct temporal mvs for bottom-field-first poc order
[02:22] <cone-390> ffmpeg.git 03Michael Niedermayer 07master:dc021a0218e1: Merge commit 'c39059bea3adebcd888571d1181db215eee54495'
[02:34] <cone-390> ffmpeg.git 03Vittorio Giovara 07master:819bb3e81c72: fate: add h264-direct-bff test
[02:34] <cone-390> ffmpeg.git 03Michael Niedermayer 07master:ec59f027ac74: Merge commit '819bb3e81c72939ae1c1a4e5fc3dfdf3e98df074'
[03:25] <cone-390> ffmpeg.git 03Michael Niedermayer 07master:437da3e349c2: avfilter/vf_owdenoise: use av_malloc_array()
[03:25] <cone-390> ffmpeg.git 03Michael Niedermayer 07master:71e515c180a5: avformat/sapenc: Use av_mallocz_array()
[04:22] <cone-390> ffmpeg.git 03Muhammad Faiz 07master:ce385c869c0d: avfilter/biquads: avoid clipping when using floating type
[07:49] <anshul> in get_bits function why are we doing av_assert2(n>0 && n<=25); 
[07:49] <anshul> why not only av_assert2(n<=25);
[07:55] <nevcairiel> Because <= 0 is bad?
[09:22] <anshul> ok i was looking for condition when this <= 0 arise, anyway leave it will find out  myself
[09:27] <nevcairiel> assert2 is usually disabled anyway
[12:55] <michaelni> db0company, as you are maintaining our webpage, you should have op here
[13:18] <cone-601> ffmpeg.git 03Janne Grunau 07master:a9f3f5fadb57: Revert "tiff: support reading gray+alpha at 8 bits"
[13:18] <cone-601> ffmpeg.git 03Michael Niedermayer 07master:5b998dae597a: Merge commit 'a9f3f5fadb57bae3f3ff0be69e56b2c6014f2513'
[13:51] <cone-601> ffmpeg.git 03Diego Biurrun 07master:81b9bf319226: dct-test: Move arch-specific bits into arch-specific subdirectories
[13:51] <cone-601> ffmpeg.git 03Michael Niedermayer 07master:d986c414dea1: Merge commit '81b9bf319226fe03436c80aaa8a2c91767cab7ce'
[13:57] <Compn> harumph
[13:57] <Compn> new website looks great 
[13:57] <Compn> i will make some small tweaks and send patches.
[13:58] <Compn> e.g. putting packages above source on download page
[13:58] <Compn> putting OS icons on the download buttons
[13:58] <Compn> btw ffmpeg website looks like mplayer website colorscheme now :D
[14:00] <cone-601> ffmpeg.git 03Nidhi Makhijani 07master:c3ec963d0d0f: mjpeg: return proper error code
[14:00] <cone-601> ffmpeg.git 03Michael Niedermayer 07master:c05dbb8744f5: Merge commit 'c3ec963d0d0fac2d817f7fbce2e89ed46eeba550'
[14:01] <Compn> oh, the os icons show up with javascript
[14:01] <Compn> but the icons dont show up if js is disabled 
[14:01] <Compn> i see.
[14:02] <Compn> did we ever get the copyright on the website design ?
[14:02] <Compn> db0company : is your design under CC or gpl license ?
[14:07] <cone-601> ffmpeg.git 03Michael Niedermayer 07master:40541716661e: avcodec/dct-test: remove fixed array sizes
[14:17] <ubitux> still no tweet about 2.3 and the new website?
[14:20] <db0company> I just tweeted it o/
[14:20] <ubitux> db0company: we have a twitter/ffmpeg
[14:22] <db0company> I know, I mentioned it @FFmpeg
[14:32] <kierank> new site looks nice
[14:39] <wm4> typical hipster design
[14:39] <wm4> with a download button covering half of the screen, literally
[14:40] <wm4> but I guess the only effect that needs to be achieved is looking modern
[14:40] <db0company> I like hipster design :)
[14:41] <db0company> I also tried to make it look accessible for newbies even though I haven't been asked to do so
[14:47] <Compn> i like giant button for newbies
[14:47] <Compn> :P
[14:47] <Compn> ubitux : i think logan is twitter guy, and he said he is without internets
[14:48] <ubitux> yes it's llogan
[14:48] <Compn> i'm also happy its not-libav design anymore :P
[14:48] <Compn> db0company : what license is the site ? i forgot...
[14:54] <db0company> there were no license when I started editing it and I'm not sure if I can just add one myself, I don't own most of the code
[14:55] <db0company> I usually use apache2 for some reason but I don't think it would fit here
[14:57] <Compn> oh ok
[14:57] <Compn> so its based on the old website code ?
[14:58] <Compn> thats ok :)
[14:58] <Compn> the old author gave ffmpeg permission to use that :)
[14:58] <Compn> i just mean your parts are under some kind of open source, thats fine with me
[15:21] <cone-601> ffmpeg.git 03Muhammad Faiz 07master:4927c0a9880b: avutil/opt.c: fix bug when parsing option
[16:31] <ubitux> j-b: is it possible to get/print the exact libav* version vlc is built against?
[16:31] <ubitux> mmh, and the one used at runtime
[17:01] <cone-601> ffmpeg.git 03Michael Niedermayer 07master:2c5c7f108926: avutil/opt: remove dead code from set_string_number()
[17:01] <cone-601> ffmpeg.git 03Michael Niedermayer 07master:88efc013b10c: avutil: move QP2LAMBDA constant to eval
[17:01] <cone-601> ffmpeg.git 03Michael Niedermayer 07master:265dadbe7de5: avutil/opt: Support general expressions involving AVOption constants in set_string_number()
[17:04] <db0company> Compn: but I think it's a good idea to add a license
[17:44] <Compn> db0company : yes, we've had issues with copyrights in the past ... :\
[18:08] <db0company> do you think that'd be a problem if I add one? considering i'm not the only contributor and I can't really reach all of them one by one to ask
[18:19] <j-b> ubitux: depends, if yo provide the function or not
[18:19] <Compn> db0company: yeah, i realize that. maybe just a notice that any of your work is licensed under whichever license you choose
[18:20] <Compn> i put all my stuff in public domain. i never really wrote it down anywhere except a mail or two on the mailing list
[19:51] <kurosu> BBB: there's a benefit to using x/y reporting
[19:51] <kurosu> but that's actually less trivial that one may have thought
[19:51] <kurosu> https://github.com/kurosu/ffmpeg/commits/hevc_mt
[19:51] <kurosu> around 2% on my dualcore (seems to run 3 or 4 threads?)
[19:52] <kurosu> more than 5% on the samples I tested with on a 6-cores
[19:55] <kurosu> maybe not the best design, but the raster_scan / increment is required
[19:55] <kurosu> first because of the way dbf and sao are ran, second because of wpp/slices
[21:51] <cone-227> ffmpeg.git 03Diego Biurrun 07master:bd499d9af668: build: Conditionally build and test iirfilter
[21:51] <cone-227> ffmpeg.git 03Michael Niedermayer 07master:94c4ad27dc63: doc/APIchanges: mention FF_INPUT_BUFFER_PADDING_SIZE chage
[21:51] <cone-227> ffmpeg.git 03Michael Niedermayer 07master:79df91bfc403: Merge commit 'bd499d9af668aef979ec9f3f3215b8dd508c7ec1'
[22:21] <cone-227> ffmpeg.git 03Diego Biurrun 07master:16b7328058fa: build: Conditionally build and run DCT test program
[22:22] <cone-227> ffmpeg.git 03Michael Niedermayer 07master:9a436ae30f10: Merge commit '16b7328058fa600d5158c84d9cc621a134eb88bc'
[23:05] <cone-227> ffmpeg.git 03Diego Biurrun 07master:2ce53880bfeb: build: Conditionally compile all avcodec test programs
[23:05] <cone-227> ffmpeg.git 03Michael Niedermayer 07master:16c66f3151cf: Merge commit '2ce53880bfeb431decdd2fc82298c0aff823ab47'
[23:23] <cone-227> ffmpeg.git 03Katerina Barone-Adesi 07master:4d2d0595cf9d: fate: Split audio- and video-related RealMedia targets
[23:23] <cone-227> ffmpeg.git 03Michael Niedermayer 07master:c75c594e5a73: Merge commit '4d2d0595cf9dd199c35aefef450952d209a56573'
[00:00] --- Tue Jul 22 2014