[Ffmpeg-devel-irc] ffmpeg-devel.log.20140927

[burek]

[01:34] <cone-646> ffmpeg.git 03Luca Barbato 07master:d995f0a1550c: log: Unbreak no-tty support on 256color terminals
[01:34] <cone-646> ffmpeg.git 03Michael Niedermayer 07master:5a5d454dcce5: Merge commit 'd995f0a1550c9148140eab34c7d770e90925c004'
[01:40] <cone-646> ffmpeg.git 03Andreas Cadhalpun 07master:883e98bc2575: probe: Bump the score for mime type matching
[01:40] <cone-646> ffmpeg.git 03Michael Niedermayer 07master:b7a55388706b: Merge commit '883e98bc25755327565d8ae0f565a6c886f24816'
[01:48] <cone-646> ffmpeg.git 03James Almer 07master:26cd7b1e1ad5: x86/fmtconvert: add ff_int32_to_float_fmul_array8_{sse,sse2}
[02:12] <cone-646> ffmpeg.git 03Luca Barbato 07master:6cfbe1de5ac6: avconv: Use only audio and video to guess discontinuities
[02:12] <cone-646> ffmpeg.git 03Michael Niedermayer 07master:ac6b5bb834c0: Merge commit '6cfbe1de5ac6c57c41459626f7ac32841d63ace8'
[02:26] <cone-646> ffmpeg.git 03Luca Barbato 07master:9d2cee52d37c: configure: Assume a standard-compliant default libc
[02:26] <cone-646> ffmpeg.git 03Michael Niedermayer 07master:eb1ac04e59cb: Merge commit '9d2cee52d37c7340f85a5d41110282aac03e6855'
[02:58] <cone-646> ffmpeg.git 03Rémi Denis-Courmont 07master:1c80c9d7ef80: hwaccel: Call ->get_format again if hwaccel init fails
[02:59] <cone-646> ffmpeg.git 03Michael Niedermayer 07master:01831fd00490: Merge commit '1c80c9d7ef809180042257200c7b5f6b81d0b0e2'
[03:35] <cone-646> ffmpeg.git 03Rémi Denis-Courmont 07master:ce4231632114: h264: Always invoke the get_format() callback
[03:35] <cone-646> ffmpeg.git 03Michael Niedermayer 07master:28b1698a7ccc: Merge commit 'ce42316321145c435029dc288c2b5f1d48519dde'
[03:54] <cone-646> ffmpeg.git 03Rémi Denis-Courmont 07master:4e03b2097ca5: mpeg12: Always invoke the get_format() callback
[03:54] <cone-646> ffmpeg.git 03Michael Niedermayer 07master:be5dd8aa5558: Merge commit '4e03b2097ca59843ed65424a572b7acd5345f088'
[04:03] <cone-646> ffmpeg.git 03Michael Niedermayer 07master:2ec4a84dca60: ffplay: Separate AVFrame from AVSubtitle
[09:31] <cone-814> ffmpeg.git 03Paul B Mahol 07master:bfdf0f078a74: avfilter/signalstats: add more pixel formats
[12:35] <cone-385> ffmpeg.git 03Christophe Gisquet 07master:f183febcdafb: hevc: remove CodingTree
[13:07] <cone-385> ffmpeg.git 03Michael Niedermayer 07master:233d1acb7095: tests/fate/microsoft: add -flags +bitexact to ilaced_twomv
[13:57] <cone-385> ffmpeg.git 03lvqcl 07master:e58fc44649d0: avutil/x86/cpu: fix cpuid sub-leaf selection
[15:17] <ubitux> so... 2.4.1 in sid, nice
[15:19] <iive> in Sid?
[15:19] <JEEB> debian unstable
[15:20] <nevcairiel> indeed
[15:20] <nevcairiel> much better than experimental
[15:26] <iive> that's great. For some reason I thought that sid is already released version of debian. 
[15:27] <nevcairiel> no, sid is never a released version, its the permanent name for the unstable edition, it gets a new name when it gets promoted to testing, and the unstable branch retains the name sid
[15:29] <iive> aha, this explains where i've heard it.
[15:41] <wm4> now it just needs to make it to testing
[15:41] <wm4> then as I understand it'll be part of a release
[16:09] <nevcairiel> wm4: imo unlikely before the current testing is actually released
[16:10] <nevcairiel> but ubuntu pulls from sid kinda automatically, so it might get into 15.04
[16:10] <nevcairiel> too late for 14.10
[16:10] <iive> the plans for jessie release are for november, aren't they?
[16:17] <nevcairiel> dont think so, november has like the first freeze date, with further freezes for fixes in december and doc updates in january
[16:19] <iive> yes
[16:25] <Compn> libav doesnt like ffmpeg in debian as far as i can tell
[16:25] <Compn> or maybe some of the comments in the debian threads/bugs
[16:28] <wm4> of course they don't like it
[16:28] <wm4> just as ffmpeg didn't like it being kicked out of ffmpeg
[16:28] <wm4> this is 100% politics
[16:31] <ubitux> we're not actually kicking them out
[16:31] <ubitux> i'm extremly happy we can finally get back in debian though
[16:31] Action: wm4 too
[16:32] <wm4> finally I can drop some more ancient shitcode from my code
[16:32] <wm4> e.g. the old subtitle demuxer exists only because libav can't handle most subtitle formats
[16:32] <nevcairiel> so you're essentially dropping libav compat then?
[16:32] <wm4> I could port these ffmpeg changes back to libav myself, but jesus christ...
[16:33] <wm4> nevcairiel: no
[16:33] <wm4> it just means libav-mpv can handle not as many formats as ffmpeg-mpv
[16:33] <ubitux> not having subtitles support is gonna brings you some mad users :p
[16:33] <nevcairiel> you could've done that before
[16:34] <ubitux> you should probably wait until it's available in ubuntu
[16:34] <ubitux> not exactly sure when
[16:34] <nevcairiel> can only help to increase pressure on the distros to move
[16:36] <wm4> my userbase is tiny and most of these subtitle formats are rarely used, so that would be little pressure
[16:37] <ubitux> you still keep the srt and microdvd internal decoders?
[16:38] <wm4> currently they're still around
[16:39] <ubitux> i guess you plan to move away from them when i'll change the decoded state of subs?
[16:39] <wm4> probably
[16:39] <wm4> I should check what exactly "my" srt decoder seems to do better than the ffmpeg one
[16:39] <wm4> I recall there were some small issues
[16:40] <ubitux> i was worried about it when Aurelien wrote the one in ffmpeg after i did it in ffmpeg
[16:40] <ubitux> in mplayer*
[16:41] <ubitux> but i didn't look in details
[16:41] <ubitux> the srt decoder needs to move to the bprint first
[16:41] <ubitux> that's the only one that doesn't and i'm a bit uncomfortable at these snprintf with the fixed buffer
[16:42] <ubitux> not exactly a priority right now though
[16:43] <wm4> I think I prefer such fixed buffers over the bprint automagic, thougj
[16:43] <wm4> *though
[16:45] <ubitux> really?
[16:45] <ubitux> is it just a feeling or there is a technical ground?
[16:45] <wm4> with bprint it would be allocating and deallocating all the time
[16:46] <wm4> considering libav* does this all the time anyway, this is probably a weak concern
[16:46] <ubitux> not really
[16:46] <ubitux> bprint uses the stack for the first 1k iirc
[16:46] <ubitux> then it fallbacks on heap
[16:46] <wm4> I know
[16:46] <ubitux> also, it can be put in the context
[16:46] <ubitux> like i did with SAMI typically
[16:47] <ubitux> so re-used after each event
[17:34] <cone-844> ffmpeg.git 03Clément BSsch 07master:ff5f5cbb4653: avformat/assdec: output ASS packets
[17:34] <cone-844> ffmpeg.git 03Clément BSsch 07master:9ec52e5533cf: avformat/assdec: make sure "Comments:" events end up in the header
[17:34] <cone-844> ffmpeg.git 03Clément BSsch 07master:0aa5c5a89e2b: avformat/assdec: reindent after previous commit
[17:34] <cone-844> ffmpeg.git 03Clément BSsch 07master:3a6fa38fb1bd: avformat/assdec: add myself to Copyright
[17:34] <cone-844> ffmpeg.git 03Clément BSsch 07master:0e6bad1229d4: avformat/assenc: mux an event header if none are found in extradata
[17:53] <cone-844> ffmpeg.git 03Clément BSsch 07master:71f6275169ce: avformat/assdec: make sure to free allocated data in case of error
[17:56] <cone-844> ffmpeg.git 03Clément BSsch 07master:180d7f330bd9: avcodec/microdvddec: remove old compat code
[17:56] <cone-844> ffmpeg.git 03Clément BSsch 07master:5ec90c90c604: avcodec/subviewerdec: remove old compat code
[17:56] <ubitux> alright
[17:56] <ubitux> better working ground now
[18:56] <cone-844> ffmpeg.git 03Luca Barbato 07master:551746680462: configure: Provide a safe default for unknonw libcs
[18:56] <cone-844> ffmpeg.git 03Michael Niedermayer 07master:5672107ed696: Merge commit '551746680462538bb5ad7adf55fb1248190428ec'
[20:58] <cone-844> ffmpeg.git 03Michael Niedermayer 07master:d85ebea3f3b6: avcodec/ac3enc_template: fix out of array read
[00:00] --- Sun Sep 28 2014