[Ffmpeg-devel-irc] ffmpeg-devel.log.20150530

[burek]

[02:47:33 CEST] <cone-365> ffmpeg 03Martin Storsjö 07master:b2f0f37d242f: rtmpdh: Generate the whole private exponent using av_get_random_seed() with nettle/gmp
[02:47:33 CEST] <cone-365> ffmpeg 03Michael Niedermayer 07master:2e15f07cfddd: Merge commit 'b2f0f37d242f1194fe1f886557cf6cefdf98caf6'
[04:10:48 CEST] <cone-365> ffmpeg 03Michael Niedermayer 07master:dae89802a6ee: avformat/h264dec: Fix detection of invalid h264 with 0x100 startcodes
[04:10:49 CEST] <cone-365> ffmpeg 03Michael Niedermayer 07master:aa9c6b6986ea: avformat/h264dec: Add ff_tlog() with trace information about h264 probing
[04:39:05 CEST] <cone-365> ffmpeg 03banastasov 07master:debf4d6e67df: avcodec/dvbsubdec: Fix buf_size check in dvbsub_parse_display_definition_segment()
[09:08:21 CEST] <rcombs> wm4: hey, cehoyos just poked me regarding securetransport; apparently both of us forgot to add a copyright notice in the license header
[09:08:48 CEST] <rcombs> wm4: do you care if you're listed, or should I just put myself down?
[09:09:37 CEST] <wm4> don't care
[09:11:30 CEST] <rcombs> alright
[09:11:49 CEST] <rcombs> <insert rant about how copyright law is dumb and this shouldn't have to matter>
[09:24:05 CEST] <chrisjunkie> Looking to patch FFMPEG with some code to use the last previous (full, non-corrupt) frame until we can finally decode a non-corrupt frame with h.264 
[09:24:30 CEST] <chrisjunkie> We're seeing massive packet loss on a customer's CCTV network and as much as we'd like to help them fix it, for certain reasons we cant
[09:24:40 CEST] <chrisjunkie> However when we stream from their cameras, we get LOTS of error concealment
[09:25:18 CEST] <chrisjunkie> Unlike normal video, it doesn't matter if we replicate the previous frame
[09:25:30 CEST] <chrisjunkie> Jumpy CCTV video is better than concealed video as its distracting
[12:12:05 CEST] <cone-766> ffmpeg 03Rodger Combs 07master:4dd22531fd25: lavf/tls_securetransport: add missing copyright notice
[16:17:36 CEST] <gr1sha> anyone have an experience with adding another thread before the actual display of the YUV overlay?
[16:17:48 CEST] <gr1sha> I'm having a big problem when using this thread
[16:46:48 CEST] <cone-766> ffmpeg 03Michael Niedermayer 07master:1607f774949a: avformat/mov: Append place name instead of overwriting for loci
[16:46:49 CEST] <cone-766> ffmpeg 03Michael Niedermayer 07master:08b7d19deff9: avformat/mov: Also export loci altitude
[16:46:50 CEST] <cone-766> ffmpeg 03Michael Niedermayer 07master:a3ec1d6abddb: avformat/movenc: Add loction/xyz support to allow copying data between mov files
[19:32:48 CEST] <cone-766> ffmpeg 03Michael Niedermayer 07master:161d0ac7219d: doc/ffmpeg: Document xerror
[20:02:38 CEST] <kierank> So, does ffmpeg need this NUMA thing that x265 has?
[20:59:33 CEST] <cone-766> ffmpeg 03Andreas Cadhalpun 07master:3ef5702926c4: libopenjpegdec: check existence of image component data
[21:28:21 CEST] <cone-766> ffmpeg 03Luca Barbato 07master:5ecabd3c54b7: msrle: Use FFABS to determine the frame size in msrle_decode_pal4
[21:28:22 CEST] <cone-766> ffmpeg 03Michael Niedermayer 07master:7d1cfbbc3e84: Merge commit '5ecabd3c54b7c802522dc338838c9a4c2dc42948'
[22:03:44 CEST] <cone-766> ffmpeg 03Luca Barbato 07master:0181ae9af2de: h264: Make sure reinit failures mark the context as not initialized
[22:03:45 CEST] <cone-766> ffmpeg 03Michael Niedermayer 07master:e9d646f44c6a: Merge commit '0181ae9af2de1526464d23209b82e6674d362f5d'
[22:48:23 CEST] <cone-766> ffmpeg 03wm4 07master:c41e0aedf5fc: network: prevent SIGPIPE on OSX
[00:00:00 CEST] --- Sun May 31 2015