[Ffmpeg-devel-irc] ffmpeg-devel.log.20180613

[burek]

[00:05:33 CEST] <atomnuker> yeah, I read that
[00:21:51 CEST] <cone-633> ffmpeg 03Marton Balint 07master:643123b29d40: avdevice/decklink_dec: use a custom memory allocator
[00:21:51 CEST] <cone-633> ffmpeg 03Marton Balint 07master:ba8a1d161820: avdevice/decklink_dec: use std::atomic for decklink_input_callback refcounting
[00:21:51 CEST] <cone-633> ffmpeg 03Jon Morley 07master:fb480a1f1e0c: avdevice/decklink_common: Move DECKLINK_* string functions into header
[00:21:51 CEST] <cone-633> ffmpeg 03Jon Morley 07master:0946c0ec177d: avdevice/decklink_dec: capture timecode to metadata when requested
[07:02:07 CEST] <cone-412> ffmpeg 03Gyan Doshi 07master:67747c89ad4d: avformat: deprecate unused MP4A_LATM flag
[12:01:26 CEST] <cone-448> ffmpeg 03Carl Eugen Hoyos 07master:9461e7d3a598: configure: The eac3_core bitstream filter needs the ac3 parser.
[16:47:41 CEST] <cone-386> ffmpeg 03Jacob Trimble 07master:4aba45ca1fec: avformat/mov: Add check for per-sample IV size.
[16:47:42 CEST] <cone-386> ffmpeg 03Timo Teräs 07master:12205d2c896b: avformat/movenc: properly handle cover image codecs
[16:47:43 CEST] <cone-386> ffmpeg 03Timo Teräs 07master:2223811b0159: avformat/movenc: fix recognization of cover image streams
[20:43:57 CEST] <January> atomnuker: https://github.com/MathieuDuponchelle/FFmpeg.git using meson, might be of interest
[20:46:05 CEST] <January> atomnuker: only missing fate tests which require samples (and maybe some other smaller things), I'd say it's actually a reality. The guy who wrote it says it works on Windows and Linux, and it works on macOS for me
[20:47:14 CEST] <January> JEEB: ^ also maybe
[21:00:39 CEST] <Mathieu_Du> only missing fate tests and half a thousand details*
[21:00:46 CEST] Action: Mathieu_Du disclaiming :)
[21:03:33 CEST] <Mathieu_Du> btw this requires meson master
[21:32:44 CEST] <cone-386> ffmpeg 03Lou Logan 07master:30af29a712b8: doc/devices: order alphabetically and add missing v4l2 outdev
[22:09:30 CEST] <akravchenko188> jkqxz: hi. Some while ago I sent patch with hwcontext_amf implementation and amdenc refactoring to use it. You decided to postpone applying patch until one more usage of hwcontext_amf is implemented. Now I have implemented vf_scaler_amf. Should I sent new patch containing all hwcontext_amf + amdenc + vf_scaler_amf for initial review or to split it to two patches? 
[22:12:08 CEST] <January> akravchenko188: better to split it
[22:13:27 CEST] <January> akravchenko188: you'd probably want it in 3 patches if I understand correctly: add new hwdec, refractor amdenc, add filter
[22:17:39 CEST] <akravchenko188> I think it should be 2 patches. 1) movement some amf common code from amfenc to hwcontext_amf; 2) vf_scaler_amf just new filter
[22:18:34 CEST] <cone-386> ffmpeg 03Michael Niedermayer 07master:2fc108f60f98: avcodec/mpeg4videodec: Clear bits_per_raw_sample if it has originated from a previous instance
[22:18:35 CEST] <cone-386> ffmpeg 03erankor 07master:ee09ffbfd2f7: qt-faststart: add validation on ftyp atom size
[22:18:36 CEST] <cone-386> ffmpeg 03erankor 07master:dc16c9dd9059: qt-faststart - stco offset bug fix
[22:18:37 CEST] <cone-386> ffmpeg 03erankor 07master:c5e6c0b5f680: qt-faststart: add fate test for stco overflow
[22:21:10 CEST] <January> akravchenko188: yes, that'd make sense (note I didnt actually look at the previous patches you sent so I just guessed :)
[23:20:31 CEST] <jkqxz> akravchenko188:  Definitely split the filter.  Mild preference for splitting the other one (so it's one change to libavutil and then a separate one to libavcodec), but I don't feel very strongly about it.
[23:24:48 CEST] <akravchenko188> ok, thanks
[23:59:06 CEST] <klaxa> ok i got first tests with hls working, but found that the matroska serving is broken for other files :/
[00:00:00 CEST] --- Thu Jun 14 2018