[Ffmpeg-devel-irc] ffmpeg-devel.log.20141116

[burek]

[01:04] <ubitux> https://www.youtube.com/watch?v=NAsM30MAHLg&index=1&list=PL0INsTTU1k2UYO9Mck-i5HNqGNW5AeEwq
[01:04] <ubitux> fun stuff fourier related
[01:06] <cone-398> ffmpeg.git 03Martin Storsjö 07master:c00365b46d46: aarch64: Make the function pointer tables position independent
[01:06] <cone-398> ffmpeg.git 03Michael Niedermayer 07master:f3cba01ccefe: Merge commit 'c00365b46d464ce47716315c1801818d811bdb9a'
[01:11] <cone-398> ffmpeg.git 03Michael Niedermayer 07master:08ee02deca81: avfilter/vf_tinterlace: remove unused variable
[01:13] <cehoyos> ubitux: I still believe dvdnav should be neither a protocol nor a demuxer but an in-device.
[01:13] <ubitux> yeah sure whatever, it's the same thing
[01:13] <cone-398> ffmpeg.git 03Lukasz Marek 07master:c727006616ef: lavc/options_table: set min to -1 for timecode_frame_start
[01:14] <cone-398> ffmpeg.git 03Lukasz Marek 07master:c544ffd2aeab: lavc/options_table: add pixel_format and video_size options
[01:14] <cone-398> ffmpeg.git 03Lukasz Marek 07master:4a30277a59ef: lavc/options: set timecode_frame_start to -1 as option default
[01:14] <cone-398> ffmpeg.git 03Lukasz Marek 07master:01974a58df40: lavc/options: initialize pkt_timebase
[01:14] <cone-398> ffmpeg.git 03Lukasz Marek 07master:eec693328a09: lavu/opt: introduce av_opt_is_set_to_default()
[01:14] <cone-398> ffmpeg.git 03Lukasz Marek 07master:bee5844ddd4e: lavu/opt: introduce av_opt_serialize()
[01:14] <cone-398> ffmpeg.git 03Lukasz Marek 07master:6690d4c3f53b: lavf/ffm: store/restore private codec context
[01:14] <cone-398> ffmpeg.git 03Lukasz Marek 07master:2657f00d3f99: ffmpeg_opt: use codec private context in ffserver streams
[01:14] <cone-398> ffmpeg.git 03Lukasz Marek 07master:a38e06c1aaac: ffserver_config: handle codec private options
[01:14] <cone-398> ffmpeg.git 03Lukasz Marek 07master:745730c9c208: lavf/ffm: use AVOption API to store/restore stream properties
[01:20] <jamrial> ubitux: nice find
[01:50] Action: ubitux wonders what's the point of that threshold filter given that they provide an equivalent with lut filter...
[02:03] <cone-398> ffmpeg.git 03Lukasz Marek 07master:ec012837527c: lavf/ffmdec: reident after last commit
[03:03] <cone-398> ffmpeg.git 03Michael Niedermayer 07master:ce80f9fee971: avformat/segment: export inner muxer timebase
[04:39] <cone-398> ffmpeg.git 03Michael Niedermayer 07master:530eb6acf8ee: avformat/hlsenc: Free context after hls_append_segment
[17:51] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:1de786777e0f: avcodec/dvdsubdec: Check all fseek()s return codes
[18:19] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:374c907fb35f: avcodec/vorbis_parser: Move vp check
[19:10] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:62eca2f827d4: avdevice/xcbgrab: Fix/remove unneeded NULL checks
[19:34] <ubitux> do we have any filter with slice threading and computation in each of them?
[19:34] <ubitux> like, typically each slice returns a score, and each of these needs to be summed at the end
[19:35] <ubitux> i'm wondering about allocating a tab of the number of jobs but i wonder if there is another way
[19:35] <ubitux> oh mmh there is a ret thing..
[19:37] <cone-86> ffmpeg.git 03Michael Niedermayer 07master:51ddaf65496b: avformat/mpeg: fix memleak of sub_name on error
[20:26] <ubitux> btw, stupid question but... what's exactly the point of the arg in threading execute function, given that filter have access to the private context anyway?
[20:28] <ubitux> everything in the thread data arg could be in the private context
[20:28] <ubitux> since AVFilterContext is passed
[22:33] <lukaszmluki> something happened with git server?
[22:35] <lukaszmluki> nvm, my dns is not working for some reason
[22:48] <iive> strange, i'm having dns problems too... 
[23:26] <ramiro> michaelni: didn't libswscale have floyd steinberg dithering at some point?
[23:28] <michaelni> some cases support error diffusion dither
[23:37] <ramiro> found it, thanks
[23:38] <kierank> ramiro: x264 has FS
[23:41] <ramiro> kierank: grep floyd -ir gives nothing. but i see Sierra-2-4A error diffusion on depth.c
[23:41] <kierank> ah ok
[00:00] --- Mon Nov 17 2014