[Ffmpeg-devel-irc] ffmpeg-devel.log.20111208

[burek]

[00:31] <CIA-41> ffmpeg: 03Diego Biurrun 07master * ree41963f19 10ffmpeg/ (6 files in 2 dirs): cosmetics: drop some completely pointless parentheses
[00:31] <CIA-41> ffmpeg: 03Luca Barbato 07master * r73585620b8 10ffmpeg/libavfilter/vf_drawtext.c: 
[00:31] <CIA-41> ffmpeg: drawtext: introduce explicit draw/hide variable
[00:31] <CIA-41> ffmpeg: Signed-off-by: Luca Barbato <lu_zero at gentoo.org>
[00:31] <CIA-41> ffmpeg: 03Martin Storsjö 07master * r1eef08f98c 10ffmpeg/libavformat/rtmpproto.c: 
[00:31] <CIA-41> ffmpeg: rtmp: Use nb_invokes for all invoke commands
[00:31] <CIA-41> ffmpeg: 704af3e29c3ddbc22ac5c8f40e5a0f860d53ac4c broke publishing
[00:31] <CIA-41> ffmpeg: of rtmp streams, at least publishing to Wowza servers.
[00:31] <CIA-41> ffmpeg: This changes all invoke commands to use nb_invokes.
[00:31] <CIA-41> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[00:31] <CIA-41> ffmpeg: 03Paul B. Mahol 07master * rfc9489f6ad 10ffmpeg/libavcodec/cljr.c: 
[00:32] <CIA-41> ffmpeg:  cljr: simplify CLJRContext
[00:32] <CIA-41> ffmpeg: There is no need to have delta, offset and gb in CLJRContext.
[00:32] <CIA-41> ffmpeg: Signed-off-by: Janne Grunau <janne-libav at jannau.net>
[00:32] <CIA-41> ffmpeg: 03Luca Barbato 07master * rf5edfc9e92 10ffmpeg/libavfilter/vf_drawtext.c: 
[00:32] <CIA-41> ffmpeg: drawtext: introduce rand(min, max)
[00:32] <CIA-41> ffmpeg: 03Diego Biurrun 07master * rff159e7816 10ffmpeg/ (4 files in 3 dirs): doxygen: Replace '\' by '@' in Doxygen markup tags.
[00:32] <CIA-41> ffmpeg: 03Ronald S. Bultje 07master * r8b03d7fe95 10ffmpeg/libavcodec/w32pthreads.h: w32thread: port fixes to pthread_cond_broadcast() from x264.
[00:32] <CIA-41> ffmpeg: 03Hendrik Leppkes 07master * rbc057f7592 10ffmpeg/libavcodec/pcm-mpeg.c: 
[00:32] <CIA-41> ffmpeg: pcm-mpeg: implement new audio decoding api
[00:32] <CIA-41> ffmpeg: Signed-off-by: Justin Ruggles <justin.ruggles at gmail.com>
[00:32] <CIA-41> ffmpeg: 03Luca Barbato 07master * r78212cefe1 10ffmpeg/libavfilter/vf_drawtext.c: 
[00:32] <CIA-41> ffmpeg: drawtext: remove typo
[00:32] <CIA-41> ffmpeg: It slipped through the last two iterations.
[00:32] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r8e2bab5d4b 10ffmpeg/: (log message trimmed)
[00:32] <CIA-41> ffmpeg: Merge remote-tracking branch 'qatar/master'
[00:32] <CIA-41> ffmpeg: * qatar/master:
[00:32] <CIA-41> ffmpeg:  drawtext: remove typo
[00:32] <CIA-41> ffmpeg:  pcm-mpeg: implement new audio decoding api
[00:32] <CIA-41> ffmpeg:  w32thread: port fixes to pthread_cond_broadcast() from x264.
[00:32] <CIA-41> ffmpeg:  doc: add editor configuration section with Vim and Emacs settings
[00:45] <CIA-41> ffmpeg: 03Mashiat Sarker Shakkhar 07master * rc5cd231151 10ffmpeg/libavcodec/wmalosslessdec.c: Partially fix CDLMS prediction
[00:45] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * rec8a3e25d8 10ffmpeg/: 
[00:45] <CIA-41> ffmpeg: Merge remote-tracking branch 'shariman/wmall'
[00:45] <CIA-41> ffmpeg: * shariman/wmall:
[00:45] <CIA-41> ffmpeg:  Partially fix CDLMS prediction
[00:45] <CIA-41> ffmpeg: Conflicts:
[00:45] <CIA-41> ffmpeg:  libavcodec/wmalosslessdec.c
[00:45] <CIA-41> ffmpeg: Merged-by: Michael Niedermayer <michaelni at gmx.at>
[01:17] <CIA-41> ffmpeg: 03Laurent Aimar 07release/0.8 * r494cfacdb9 10ffmpeg/libavcodec/vmdav.c: 
[01:17] <CIA-41> ffmpeg: vmd: fix segfaults on corruped streams
[01:17] <CIA-41> ffmpeg: Signed-off-by: Janne Grunau <janne-libav at jannau.net>
[01:17] <CIA-41> ffmpeg: Signed-off-by: Reinhard Tartler <siretart at tauware.de>
[01:17] <CIA-41> ffmpeg: 03Laurent Aimar 07release/0.8 * ra31ccacb1a 10ffmpeg/libavcodec/qdm2.c: 
[01:17] <CIA-41> ffmpeg: Check for out of bound writes in the QDM2 decoder.
[01:17] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:17] <CIA-41> ffmpeg: Signed-off-by: Justin Ruggles <justin.ruggles at gmail.com>
[01:17] <CIA-41> ffmpeg: (cherry picked from commit 291d74a46d32183653db07818c7b3407fd50a288)
[01:17] <CIA-41> ffmpeg: Signed-off-by: Reinhard Tartler <siretart at tauware.de>
[01:17] <CIA-41> ffmpeg: 03Laurent Aimar 07release/0.8 * r0d93d5c461 10ffmpeg/libavcodec/qdm2.c: 
[01:17] <CIA-41> ffmpeg: Fix out of bound reads in the QDM2 decoder.
[01:17] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:17] <CIA-41> ffmpeg: Signed-off-by: Justin Ruggles <justin.ruggles at gmail.com>
[01:17] <CIA-41> ffmpeg: imgutils: Fix illegal read.
[01:17] <CIA-41> ffmpeg: Found with address sanitizer.
[01:17] <CIA-41> ffmpeg: Signed-off-by: Alex Converse <alex.converse at gmail.com>
[01:17] <CIA-41> ffmpeg: (cherry picked from commit c693aa6f71b4f539cf9df67ba42f4b1932981687)
[01:18] <CIA-41> ffmpeg: Signed-off-by: Reinhard Tartler <siretart at tauware.de>
[01:18] <CIA-41> ffmpeg: 03Michael Niedermayer 07release/0.8 * r8acf9905a1 10ffmpeg/: (log message trimmed)
[01:18] <CIA-41> ffmpeg: Merge remote-tracking branch 'qatar/release/0.7' into release/0.8
[01:18] <CIA-41> ffmpeg: Note, all these commits where already in our release, this merge thus
[01:18] <CIA-41> ffmpeg: changes nothing, its just for metadata
[01:18] <CIA-41> ffmpeg: * qatar/release/0.7:
[01:18] <CIA-41> ffmpeg:  vp6: Fix illegal read.
[01:18] Last message repeated 1 time(s).
[01:18] <CIA-41> ffmpeg: 03Alex Converse 07release/0.8 * r67a7ed623b 10ffmpeg/libavcodec/vp6.c: 
[01:18] <CIA-41> ffmpeg: vp6: Fix illegal read.
[01:18] <CIA-41> ffmpeg: (cherry picked from commit 2a6eb06254df79e96b3d791b6b89b2534ced3119)
[01:18] <CIA-41> ffmpeg: Signed-off-by: Reinhard Tartler <siretart at tauware.de>
[01:18] <CIA-41> (22 lines omitted)
[02:34] <CIA-41> ffmpeg: 03Anatoliy Wasserman 07master * r6778f9d9f2 10ffmpeg/libavcodec/ (libopenjpegdec.c pthread.c): 
[02:34] <CIA-41> ffmpeg: pthread: fixed thread initialization issues related to thread-safe buffer allocator
[02:34] <CIA-41> ffmpeg: Signed-off-by: Anatoliy Wasserman <anatoliy.wasserman at yandex.ru>
[02:34] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[02:50] <CIA-41> ffmpeg: 03Carl Eugen Hoyos 07master * r51d7cd3596 10ffmpeg/libavcodec/ (mjpegdec.c version.h): 
[02:50] <CIA-41> ffmpeg: Decode RGB ljpeg to RGB24.
[02:50] <CIA-41> ffmpeg: No alpha channel is decoded, the output used to be completely transparent.
[02:50] <CIA-41> ffmpeg: 03Carl Eugen Hoyos 07master * r9608e3a182 10ffmpeg/libavcodec/bmp.c: 
[02:50] <CIA-41> ffmpeg: Support decoding 56 byte BMP Bitmap Information Header.
[02:50] <CIA-41> ffmpeg: Fixes ticket #719.
[02:50] <CIA-41> ffmpeg: 03Carl Eugen Hoyos 07master * r0cc9949ead 10ffmpeg/libavcodec/ (gif.c version.h): 
[02:50] <CIA-41> ffmpeg: Mark one colour as transparent in the gif image encoder
[02:50] <CIA-41> ffmpeg: if the input palette contains at least one colour
[02:50] <CIA-41> ffmpeg: that is more than 50% transparent.
[02:50] <CIA-41> ffmpeg: Fixes ticket #715.
[06:02] <burek> btw, pasteeater, I've just realized it would maybe be good to put the article about Eclipse on the forum, so that people can ask questions while troubleshooting that process on their machines
[06:03] <burek> also, michaelni has the forum been created/published so far?
[06:03] <CIA-41> ffmpeg: 03Paul B Mahol 07master * r1e8f77324e 10ffmpeg/libavcodec/escape130.c: escape130: remove trailing whitespace
[06:03] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r8ff3d1ccec 10ffmpeg/libavcodec/escape130.c: 
[06:03] <CIA-41> ffmpeg: escape130: make sure cliping is done on signed values.
[06:03] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[06:03] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r06798c6e3b 10ffmpeg/libavcodec/escape130.c: 
[06:03] <CIA-41> ffmpeg: escape130: Fix y_base handling, this fixes some artifacts, vissible in most videos.
[06:03] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[06:03] <CIA-41> ffmpeg: 03Paul B Mahol 07master * r7bd8b70f19 10ffmpeg/libavcodec/escape130.c: escape130: fix colors
[06:03] <CIA-41> ffmpeg: 03Eli Friedman 07master * r7d64391441 10ffmpeg/libavcodec/ (Makefile allcodecs.c avcodec.h escape130.c): Escape 130 (RPL) decoder
[06:03] <CIA-41> ffmpeg: 03Paul B Mahol 07master * r06029fbb7f 10ffmpeg/libavcodec/escape130.c: escape130: minimal effort to make it compile without warnings
[06:03] <CIA-41> ffmpeg: 03Paul B Mahol 07master * r30961a8dd5 10ffmpeg/libavformat/rpl.c: rpl: enable escape130 codec
[06:03] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r4f79c7dd08 10ffmpeg/: (log message trimmed)
[06:03] <CIA-41> ffmpeg: Merge remote-tracking branch 'richardpl/escape130'
[06:03] <CIA-41> ffmpeg: * richardpl/escape130:
[06:03] <CIA-41> ffmpeg:  escape130: fix colors
[06:03] <CIA-41> ffmpeg:  escape130: remove trailing whitespace
[06:03] <CIA-41> ffmpeg:  rpl: enable escape130 codec
[06:03] <CIA-41> ffmpeg:  escape130: minimal effort to make it compile without warnings
[06:05] <michaelni> burek, ask zeranoe when he logs on next time, i know nothing
[06:06] <burek> ok
[08:22] <GordonFreeman> hellouuuu
[10:17] <Tjoppen> right, FATE should be run with samples for proper results
[10:28] <Tjoppen> there we go
[12:31] <Tjoppen> there we go.. computing proper packet offsets in mxfdec. now I should only need to do what mov.c does and just demux based on AVIndex
[12:31] <Tjoppen> this way turned out easier to understand, and hopefully prettier that the patch on the ML
[12:35] <Compn> j-b : did vlc ever get back in the apple app store ?
[12:36] <Compn> or is the apple app store still anti-gpl ?
[12:36] <j-b> Compn: we never got back
[12:36] <j-b> and it is so-so anti-gpl
[12:37] <Compn> theres a bunch of apps using ffmpeg in there :P
[12:47] <av500> Compn: sue them!!!
[12:50] <Compn> i'd rather infect apple with gpl, make them a party of 'induced copyright infringement' and take all of apple down! :P
[12:56] <av500> that'll be hard
[12:56] <av500> they have 70b$ to pay lawyers
[12:57] <cbsrobot> Compn: It's rather ironic - you can only add apps after apple inspected them, because it could destroy your iWhatever - well if they inspected it they are aware of the libraries used in it.
[12:57] <av500> cbsrobot: you think they do code analysis?
[12:57] <Compn> cbsrobot : you know they only inspect apps for homosexuality, abortion, and other such nonsense
[12:58] <cbsrobot> no idea - but why do they claim "it could destroy yout iThing" ?
[12:58] <Compn> thats to keep you locked into their store
[12:58] <Compn> not to use 3rd party stores, they are unclean!
[12:58] <cbsrobot> in my opinion if this is their argument, they need to analyze the code aswell
[12:59] <Compn> i think they arent touching vlc due to the complaint, not really due to the gpl
[12:59] <av500> afaik you cannot analyse code due to the halting problem
[12:59] <Compn> i mean the gpl is the complaint, but in a roundabout way...
[13:02] <cbsrobot> yeah it might be - I'm sure in thir eula they excluded any reliability towards apps violating a license
[13:02] <cbsrobot> or devs violating a license
[13:03] <av500> or course
[13:03] <av500> of
[13:13] <CIA-41> ffmpeg: 03Nicolas George 07master * re91c088055 10ffmpeg/libavformat/utils.c: 
[13:13] <CIA-41> ffmpeg: lavf: lower the log level of "parser not found".
[13:13] <CIA-41> ffmpeg: The information is relevant, but under normal circumstances
[13:13] <CIA-41> ffmpeg: it raises far too many false alarms.
[13:13] <CIA-41> ffmpeg: Signed-off-by: Nicolas George <nicolas.george at normalesup.org>
[13:13] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[13:16] <Compn> j-b : well i didnt want to pour salt in the wound. just curious... :)
[13:16] <Compn> when people say they cant get mplayer to do something, i point them at vlc, because i know vlc has more support 
[13:25] <iive> Compn: your plan to take apple store down is really cunning. We'll just have to wait for the SOPA/PrIP to be signed as law.
[13:26] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r31c55c572a 10ffmpeg/libavcodec/escape130.c: 
[13:26] <CIA-41> ffmpeg: escape130: 10l for myself for not initializing y_base.
[13:26] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[13:32] <cbsrobot> michaelni: with the commit of "pthread: fixed thread initialization with default buffer allocator" does ffmpeg ever need to call ff_thread_finish_setup ?
[16:05] <ubitux> nice idea from nicolas the -report
[16:21] <Compn> iive : because apple is distributing it, they must also distribute source , right? or is it hosted on peoples own servers ?
[16:25] <iive> Compn: store license forbids RE and modifications... that's not very gpl compatible. I remember that the issue was raised in relation to GPL3, but ... yeh, I guess asking for copyright notice and offer to the source could do it too.
[16:49] <Tjoppen> what do the columns in the fate demux tests mean? a few of them are off
[16:50] <Tjoppen> like:
[16:50] <Tjoppen>  0, 172800, 3688, 0x1db45852
[16:50] <Tjoppen> -0, 176400, 38412, 0x2ee26a63
[16:50] <Tjoppen> +0, 176400, 166412, 0x7a4fd591
[16:50] <Tjoppen>  0, 180000, 8385, 0x0bc20a27
[16:50] <Tjoppen> -1, 180000, 32000, 0x8f7e5009
[16:50] <Tjoppen>  0, 183600, 3733, 0xa3e2a9a0
[17:16] <CIA-41> ffmpeg: 03Carl Eugen Hoyos 07master * r8ac6469575 10ffmpeg/libavcodec/escape130.c: 
[17:16] <CIA-41> ffmpeg: Silence a warning when compiling Escape 130 decoder.
[17:16] <CIA-41> ffmpeg: Fixes:
[17:16] <CIA-41> ffmpeg: libavcodec/escape130.c:177:17: warning: ISO C90 forbids mixed declarations and code
[17:16] <CIA-41> ffmpeg: Reviewed-by: Paul B Mahol
[17:25] <michaelni> Tjoppen: pkt->stream_index, pkt->dts, pkt->size, crc
[17:25] <ubitux> framecrc format
[17:27] <Tjoppen> the parser must be changing the size then. hm
[17:29] <Tjoppen> the FATE sample is really ugly
[17:30] <Tjoppen> it violates the spec in various ways
[18:18] <Mista_D> While encoding to 4 output files (1st of 2 passes) only 2.5 CPU cores are utilized. I tried decreasing the number of outputs, `-threads` or `-thread_type` are, the utilization stays at ~2.5 cores of 16 available. Any way to increase CPU utilization. Pass 2 uses up to 9 cores. http://pastebin.com/m5ctKziU
[18:44] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r59d3656f6e 10ffmpeg/libavcodec/escape130.c: 
[18:44] <CIA-41> ffmpeg: escape130: Check dimensions to be a multiple of the block size.
[18:44] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[18:44] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r40d6296095 10ffmpeg/libavcodec/escape130.c: 
[18:44] <CIA-41> ffmpeg: escape130: The minimum read in and after decode_skip_count() is 4 bits.
[18:44] <CIA-41> ffmpeg: Thus we can check for 4 being available.
[18:44] <CIA-41> ffmpeg: If the next block is skiped we need 4 bits to encode the skip. If the
[18:44] <CIA-41> ffmpeg: next block is not skiped then we need 1 bit for the skip code and
[18:44] <CIA-41> ffmpeg: at least 3 bits for the block.
[18:44] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[18:44] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * rf7da12350e 10ffmpeg/libavcodec/escape130.c: 
[18:44] <CIA-41> ffmpeg: escape130: give all av_log() a context
[18:44] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[18:44] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * rbef8dfa056 10ffmpeg/libavcodec/escape130.c: 
[18:44] <CIA-41> ffmpeg: escape130: replace can_safely_read() by get_bits_left()
[18:44] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[18:44] <CIA-41> ffmpeg: 03Paul B Mahol 07master * r118ca0c79a 10ffmpeg/doc/general.texi: 
[18:44] <CIA-41> ffmpeg: Add Escape 130 to the documentation
[18:44] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[18:44] <CIA-41> ffmpeg: 03jb at kdenlive.org 07master * re2baaa21aa 10ffmpeg/libavformat/mpegtsenc.c: 
[18:44] <CIA-41> ffmpeg: mpegtsenc: recognize .mts as MPEG Transport Stream (encoding)
[18:44] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[18:47] <CIA-41> ffmpeg: 03Mashiat Sarker Shakkhar 07master * r6d4deecfdf 10ffmpeg/libavcodec/wmalosslessdec.c: Pass pointer to integer buffer instead of integer to memcpy
[18:47] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r15c481614b 10ffmpeg/: 
[18:47] <CIA-41> ffmpeg: Merge remote-tracking branch 'shariman/wmall'
[18:47] <CIA-41> ffmpeg: * shariman/wmall:
[18:47] <CIA-41> ffmpeg:  Pass pointer to integer buffer instead of integer to memcpy
[18:47] <CIA-41> ffmpeg: Merged-by: Michael Niedermayer <michaelni at gmx.at>
[18:50] <Tjoppen> aha - set stream_index etc. *after* av_get_packet(), not before :)
[18:53] <CIA-41> ffmpeg: 03Nicolas George 07master * r26c6fec9d9 10ffmpeg/ (4 files in 2 dirs): lavu: introduce av_log_format_line.
[19:02] <CIA-41> ffmpeg: 03Carl Eugen Hoyos 07master * r935ac228e4 10ffmpeg/libavformat/wav.c: Fix typo: Correctly parse 64 byte umid tags in bext metadata.
[19:34] <Tjoppen>  23177c8a72f34e243e9ffc4f6c70d3c7 *./tests/data/lavf/lavf.mxf_d10
[19:34] <Tjoppen>  5330989 ./tests/data/lavf/lavf.mxf_d10
[19:34] <Tjoppen> -./tests/data/lavf/lavf.mxf_d10 CRC=0x81602ff1
[19:34] <Tjoppen> does this mean the CRC changed or..?
[19:35] <Tjoppen> oh, wait.. I need to port over the AES stuff too
[20:15] <Tjoppen> bah, punting for D-10 for now
[20:16] <Tjoppen> it's.. special
[20:21] <Compn> d-10 ?
[20:21] Action: Compn forgets
[20:21] <Compn> camera ?
[20:22] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r7c41995e00 10ffmpeg/ (avconv.c ffmpeg.c): 
[20:22] <CIA-41> ffmpeg: ffmpeg: more exact comment about libavformat.
[20:22] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[20:22] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * rfab97adb66 10ffmpeg/doc/ (developer.texi faq.texi): 
[20:22] <CIA-41> ffmpeg: docs: Minor project name cleanup
[20:22] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[20:22] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * ra4bf093e7b 10ffmpeg/doc/indevs.texi: 
[20:22] <CIA-41> ffmpeg: indevs.texi: fix application name to match the code
[20:22] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[20:23] <Tjoppen> yeah, IMX50 basically
[20:23] <Tjoppen> it's CBR MPEG-2 + AES3 audio, all in their own little packet
[20:24] <Tjoppen> make fate w/ fate-suite passes \o/
[20:35] <Tjoppen> https://github.com/Tjoppen/FFmpeg/tree/opatom_demuxing_and_seeking
[20:37] <Tjoppen> time to head home. that tree works with the opatom files I linked on the ML, the fate sample and the regression files
[20:37] <Tjoppen> plus a few more
[20:38] <Tjoppen> and the demuxing code doesn't piss me off
[21:11] <bcoudurier> Tjoppen, great
[21:16] <michaelni> Tjoppen, is that branch ready to be merged ?
[21:22] <Tjoppen> I rebased it this morning. it might need a touch more work, mainly with d-10 (though it isn't worse than before)
[21:23] <Tjoppen> I'll be hacking away at it tomorrow too. I need to decode PTSes from the index tables (hence why I parse TemporalOffset)
[21:24] <Tjoppen> but yeah, should merge fine. then we can add atoms to fate
[21:27] <Tjoppen> PTS decoding isn't actually that hard. it mostly involes a bucket sort
[21:29] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * ra502939d64 10ffmpeg/libavfilter/ (avfilter.h buffersink.h sink_buffer.c): 
[21:29] <CIA-41> ffmpeg: buffersink: Implement a poll function.
[21:29] <CIA-41> ffmpeg: With this the application can find out how many frames are available.
[21:29] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:29] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * rec9d92700e 10ffmpeg/ffmpeg.c: 
[21:29] <CIA-41> ffmpeg: ffmpeg: Use the buffer sinks poll instead of its input.
[21:29] <CIA-41> ffmpeg: This fixes the case that the buffer sinks fifo contains frames which
[21:29] <CIA-41> ffmpeg: would not show up on the poll from its input
[21:29] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:29] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r6f7f239604 10ffmpeg/libavformat/libnut.c: 
[21:30] <CIA-41> ffmpeg: libnut: check for av_malloc failure.
[21:30] <CIA-41> ffmpeg: No, ive no testcase, i just spoted these when looking ...
[21:30] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:30] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * re21f8a07ca 10ffmpeg/libavformat/libnut.c: 
[21:30] <CIA-41> ffmpeg: libnut: Check nut_demuxer_init() return value.
[21:30] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:30] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r74436ad51e 10ffmpeg/libavformat/libnut.c: 
[21:30] <CIA-41> ffmpeg: libnut: NULL priv->nut after nut_demuxer_uninit() as a saftey precaution.
[21:30] <CIA-41> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:31] <michaelni> Tjoppen, ok, ill merge it in a moment, just double checking make fate
[21:34] <ubitux> michaelni: are you ok with http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/2011-December/117401.html ?
[21:35] <ubitux> basically i keep the ff_ for consistency at first, and then remove all of then in a row
[21:36] <CIA-41> ffmpeg: 03Tomas Härdin 07master * r1170749b39 10ffmpeg/libavformat/mxfdec.c: mxfdec: Add EssenceContainer UL found in 0001GL00.MXF.A1.mxf_opatom.mxf
[21:36] <CIA-41> ffmpeg: 03Tomas Härdin 07master * re1914b5a2f 10ffmpeg/libavformat/mxfdec.c: 
[21:36] <CIA-41> ffmpeg: mxfdec: Make sure DataDefinition is consistent between material track and source track
[21:36] <CIA-41> ffmpeg: This fixes 0001GL.MXF.V1.mxf_opatom.mxf and 0001GL00.MXF.A1.mxf_opatom.mxf
[21:36] <CIA-41> ffmpeg: getting two streams each due to both using the same SourcePackageID.
[21:36] <CIA-41> ffmpeg: 03Tomas Härdin 07master * re5f9c8927b 10ffmpeg/libavformat/mxfdec.c: mxfdec: Parse ThisPartition
[21:36] <CIA-41> ffmpeg: 03Tomas Härdin 07master * r336246238e 10ffmpeg/libavformat/mxfdec.c: mxfdec: Compute essence container offsets and lengths into mxf->partitions
[21:36] <CIA-41> ffmpeg: 03Tomas Härdin 07master * r623128d782 10ffmpeg/libavformat/mxfdec.c: 
[21:36] <CIA-41> ffmpeg: mxfdec: av_dlog():ify 'no corresponding source package found'
[21:36] <CIA-41> ffmpeg: This isn't an error. It's also expected behavior for OPAtom files.
[21:36] <CIA-41> ffmpeg: 03Tomas Härdin 07master * r7560c26495 10ffmpeg/libavformat/mxfdec.c: mxfdec: Parse TemporalOffsets
[21:36] <CIA-41> ffmpeg: 03Tomas Härdin 07master * rfd34dbea58 10ffmpeg/libavformat/mxfdec.c: 
[21:36] <CIA-41> ffmpeg: mxfdec: Parse more values in PartitionPack
[21:36] <CIA-41> ffmpeg: These values include KAGSize, HeaderByteCount and IndexByteCount.
[21:36] <CIA-41> ffmpeg: The length of the pack itself is also stored, and KAGSize is sanity checked.
[21:36] <CIA-41> ffmpeg: The FATE sample has KAGSize == 0, which is adjusted to 512.
[21:36] <CIA-41> ffmpeg: Other bad KAGSizes are set to 1.
[21:36] <CIA-41> ffmpeg: 03Tomas Härdin 07master * r5fb800f49a 10ffmpeg/libavformat/mxfdec.c: mxfdec: Use MaterialPackage - Track - TrackID instead of the system_item hack
[21:36] <CIA-41> ffmpeg: 03Michael Niedermayer 07master * r4ab495b4ee 10ffmpeg/: (log message trimmed)
[21:36] <CIA-41> ffmpeg: Merge remote-tracking branch 'tjoppen/opatom_demuxing_and_seeking'
[21:36] <CIA-41> ffmpeg: * tjoppen/opatom_demuxing_and_seeking:
[21:36] <CIA-41> ffmpeg:  mxfdec: Index table driven demuxing and seeking
[21:37] <CIA-41> ffmpeg:  mxfdec: Compute packet offsets properly
[21:37] <CIA-41> ffmpeg:  mxfdec: Use MaterialPackage - Track - TrackID instead of the system_item hack
[21:37] <CIA-41> ffmpeg:  mxfdec: Parse more values in PartitionPack
[21:37] <CIA-41> ffmpeg: 03Tomas Härdin 07master * rddcf3e0535 10ffmpeg/ (libavformat/mxfdec.c tests/ref/seek/lavf_mxf): 
[21:37] <CIA-41> ffmpeg: mxfdec: Index table driven demuxing and seeking
[21:37] <CIA-41> ffmpeg: This adds OPAtom support and proper seeking.
[21:37] <CIA-41> ffmpeg: D-10 and non-seekable OP1a streams still use the old demuxing/seeking code.
[21:37] <CIA-41> ffmpeg: 03Tomas Härdin 07master * rdcd30b83b4 10ffmpeg/libavformat/mxfdec.c: (log message trimmed)
[21:37] <CIA-41> ffmpeg: mxfdec: Speed up metadata and index parsing
[21:37] <CIA-41> ffmpeg: Specifically, this means parsing as before until we run into essence.
[21:37] <michaelni> ubitux, i fear this renaming breaks ABI
[21:38] <ubitux> ah, ok
[21:38] <CIA-41> ffmpeg: At that point we seek to the footer and parse until EOF. After that we start
[21:38] <CIA-41> (9 lines omitted)
[21:38] <michaelni> i mean rename in libavcodec and then old libavformat would fail to find it
[21:39] <ubitux> and with a bump?
[21:40] <ubitux> removing functions means major bump though...
[21:40] <michaelni> yes ...
[21:41] <michaelni> you can rename and provide wrapers under the old name if you want
[21:41] <michaelni> and put the wrapers under #if VERSION blah
[21:41] <michaelni> a lot of boring work ...
[21:41] <ubitux> sounds overkill too :p
[21:43] <nevcairiel> since these functions are private, do they really break ABI ? (unless someone starts mixing different avformat with avcodec versions, if thats supposed to be supported)
[21:45] <michaelni> it depends, on how distros package things and how they define dependancies
[21:47] <michaelni> btw, Tjoppen, you might want to add yourself as mxf maintainer
[21:51] <Tjoppen> ah, right
[21:58] <ubitux> haha stefano sending another toy filter :)
[22:15] <Tjoppen> come to think of it, I didn't try atoms with multiple EditUnitByteCounts
[22:56] <CIA-41> ffmpeg: 03Stefano Sabatini 07master * rb229485f1a 10ffmpeg/doc/APIchanges: APIchanges: add entry after av_buffersink_poll_frame() addition
[00:00] --- Fri Dec  9 2011