[Ffmpeg-devel-irc] ffmpeg-devel.log.20120320

[burek]

[00:00] <Compn> KSHawkEye : depends if you have static binaries with external libs? :P
[00:02] <KSHawkEye> Compn: Is only the ffmpeg code part licensed under GPL?
[00:03] <Compn> KSHawkEye : yes, are you asking if the external libs should be disabled if --enable-gpl is used ?
[00:03] <Compn> if they conflict with gpl
[00:03] <Compn> of course
[00:04] <KSHawkEye> I'm asking if I legally comply with the GPL is I only include one license file with my builds gpl-3.0.txt, and not each of the license files that come with each external library 
[00:06] <KSHawkEye> The current Windows builds include each license file that come with each external lib. I would like to only include the gpl-3.0.txt
[00:08] <durandal_1707> durandal_1707: actually ffv1 use avctx->context_model as index into array without checks...
[00:08] <durandal_1707> michaelni: ^
[00:10] <Compn> KSHawkEye : i like the license file that has each lib license :)
[00:11] <Compn> but i'm no license expert
[00:11] <Compn> theres not many license experts here
[00:14] <Daemon404> nobody truly understands the gpl
[00:14] <Daemon404> except rms
[00:16] <Compn> http://www.google-melange.com/gsoc/org/google/gsoc2012/libav
[00:16] <Compn> looks like libav got in
[00:16] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r72ec043af4 10ffmpeg/libavformat/omadec.c: (log message trimmed)
[00:16] <CIA-17> ffmpeg: oma: Fix out of array read.
[00:16] <CIA-17> ffmpeg: Input: 01-Untitled-partial.oma
[00:16] <CIA-17> ffmpeg: ZZUF params: zzuf[s=7157,r=0.001]
[00:16] <CIA-17> ffmpeg: Fixes Bugzilla #106
[00:16] <CIA-17> ffmpeg: Bug-found-by: darkshikari
[00:16] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[00:16] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * rc346f6304c 10ffmpeg/ (5 files in 2 dirs): adpcm: fix nb_samples rounding for adpcm_ima_dk3, and update reference.
[00:16] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r9604307dd8 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_ima_qt to bytestream2.
[00:16] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r39505f4240 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_ima_wav to bytestream2.
[00:16] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r689be85a9a 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_4xm to bytestream2.
[00:16] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r1da9587026 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_dk3 to bytestream2.
[00:16] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * rd570a96808 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_dk4 to bytestream2.
[00:16] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r9721901440 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_ima_iss to bytestream2.
[00:17] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r16b7a5e241 10ffmpeg/libavcodec/adpcm.c: adpcm: convert ima_ws to bytestream2.
[00:17] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r74d7ac95fb 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_ima_apc to bytestream2.
[00:17] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * redd9555537 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_ea_r1/2/3 to bytestream2.
[00:17] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r159831cc09 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_xa to bytestream2.
[00:17] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r8f2ed09289 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_ima_ea_eacs to bytestream2.
[00:17] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r22c48d399d 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_ea_xas to bytestream2.
[00:17] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r834c81f4c5 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_ct to bytestream2.
[00:17] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * rf184735a27 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_swf to bytestream2.
[00:17] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r72eda7786d 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_yamaha to bytestream2.
[00:17] <CIA-17> ffmpeg: 03Ronald S. Bultje 07master * r1744ab9e46 10ffmpeg/libavcodec/adpcm.c: adpcm: convert adpcm_thp to bytestream2.
[00:17] <CIA-17> ffmpeg: 03Martin Storsjö 07master * r27b7459b8d 10ffmpeg/libavutil/dict.h: 
[00:17] <CIA-17> ffmpeg: docs: Fix a stray reference to tags in the generic doxy on dicts
[00:17] <CIA-17> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[00:17] <CIA-17> ffmpeg: 03Martin Storsjö 07master * r57151f8674 10ffmpeg/libavformat/utils.c: (log message trimmed)
[00:17] <CIA-17> ffmpeg: libavformat: Use AVFormatContext.probesize in init_input
[00:18] <CIA-17> ffmpeg: This was forgotten in the transition from av_open_input_file to
[00:18] <CIA-17> ffmpeg: avformat_open_input, see 603b8bc2a1.
[00:18] <CIA-17> ffmpeg: This doesn't change anything for the default case where the
[00:18] <CIA-17> ffmpeg: option isn't set, since PROBE_BUF_MAX is 1048576 (which was
[00:18] <CIA-17> ffmpeg: used as max probe size earlier) while the default value for
[00:18] <CIA-17> ffmpeg: 03Nicolas George 07master * r01b0ade665 10ffmpeg/libavformat/url.h: 
[00:18] <CIA-17> (39 lines omitted)
[00:23] <michaelni> durandal_1707, how can i reproduce the issue ?
[00:27] <durandal_1707> michaelni: ffmpeg -i movie -vcodec ffv1 -context 4 -coder ac out.avi
[00:27] <michaelni> durandal_1707, bug reproduced
[00:27] <michaelni> -context 9999 here
[00:28] <durandal_1707> michaelni: behaviour is undefined if pointer points out of array
[00:39] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * re01f478dd2 10ffmpeg/libavcodec/ffv1.c: 
[00:39] <CIA-17> ffmpeg: ffv1enc: Check context_model
[00:39] <CIA-17> ffmpeg: Fixes crash
[00:39] <CIA-17> ffmpeg: Found-by: durandal_1707
[00:39] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[04:04] <Compn> Daemon404 : you can hardly blame mplayer filters for bad anime dvds ;P
[04:04] <Daemon404> Compn, pullup fails on anythign remotely bad
[04:04] <Daemon404> it's quite bad.
[04:04] <Compn> i dont use it ;)
[04:04] <Daemon404> it fails on all sorts of stuff, not just anime
[04:05] <Compn> luckily i dont run into such things, someone rips it properly first ;)
[04:05] <Daemon404> lol
[04:05] <Daemon404> yeah but im one of those weirdos that uses YATTA
[04:05] <Daemon404> to fine-tune IVTC
[04:05] <Daemon404> by hand
[04:06] <Compn> if you do write/port a filter, you might want to talk to dalias
[04:06] <Compn> he is deint master ;)
[04:06] <Compn> or ivtc whatevern
[04:06] <Daemon404> deint is entirely different from ivtc
[04:06] <Daemon404> and bcoudurier already ported a decent deint filt
[04:06] <Daemon404> :)
[04:07] <Daemon404> i dislike the idea of libavfilter anyway..
[04:07] <Daemon404> monolithic lib of filters ftl
[05:10] <michaelni> Daemon404, write a dynamic loader for the filters :)
[05:10] <michaelni> and a portable one :)
[05:12] <michaelni> or you could make the filters buildable as .so
[05:12] <michaelni> but it doesnt seem like a terribly good idea
[05:13] <Daemon404> for now ill just stick to avs filter
[05:13] <Daemon404> theyre generally the best available... and not so easy to port
[05:14] <Daemon404> i.e. triticals giant walls of msvc inline asm
[05:19] <CIA-17> ffmpeg: 03Derek Buitenhuis 07master * r41bd3519b0 10ffmpeg/tests/ (fate/lossless-video.mak ref/fate/zerocodec): 
[05:19] <CIA-17> ffmpeg: FATE: Add ZeroCodec test
[05:19] <CIA-17> ffmpeg: Signed-off-by: Derek Buitenhuis <derek.buitenhuis at gmail.com>
[05:19] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[05:19] <CIA-17> ffmpeg: 03Derek Buitenhuis 07master * r30f2d97afe 10ffmpeg/libavcodec/zerocodec.c: 
[05:19] <CIA-17> ffmpeg: zerocodec: Fix license
[05:19] <CIA-17> ffmpeg: ISC doesn't contain this line, so remove it to
[05:19] <CIA-17> ffmpeg: prevent confusion.
[05:19] <CIA-17> ffmpeg: Signed-off-by: Derek Buitenhuis <derek.buitenhuis at gmail.com>
[05:19] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[05:20] <michaelni> btw, Daemon404 ive checked trac but couldnt fine any excessive SQL qerry amounts
[05:21] <Daemon404> i dunno then
[05:21] <michaelni> ive also tuned a purge threshold that the net claimed to matter
[05:21] <michaelni> we will see if it does, not sure how long it takes 
[05:22] <michaelni> i think its only checked once a day if i didnt misread the code
[05:22] <michaelni> eventually we will figure out where the problem is ...
[05:23] <Daemon404> not sure if its faster
[05:23] <Daemon404> could be placebo
[05:23] <michaelni> ive also changed apache config, so actually quite a bit changed
[05:24] <Daemon404> hmm wtf happened with teh cosmetics patch
[05:24] <michaelni> dunno maybe the merge broke it
[05:24] <Daemon404> let's see.
[05:24] <michaelni> i merged some cosmetics from libav
[05:25] <Daemon404> the cosmetics i send to you were the ones they had me apply before they pushed
[05:25] <Daemon404> so they files will end up identical
[05:25] <Daemon404> s/they/the/
[05:26] <Daemon404> http://git.videolan.org/?p=ffmpeg.git;a=blob;f=libavcodec/zerocodec.c;h=1718818d925879b2a20a29ea42a939b34fd599c6;hb=HEAD
[05:27] <Daemon404> yeah looks liek you merged most of them
[05:27] <Daemon404> i do up a v2 patch later.
[08:00] <CIA-17> ffmpeg: 03Clément BSsch 07master * r9dd649c004 10ffmpeg/libavformat/ (flv.h flvdec.c): 
[08:00] <CIA-17> ffmpeg: flv: clarify use of video info/cmd frame.
[08:00] <CIA-17> ffmpeg: Also add generated key frame in the enum, and doxycomment the existing
[08:00] <CIA-17> ffmpeg: ones. Descriptions are directly taken from the public specifications.
[10:52] <ubitux> do we have a generic "get_index" helper or something?
[10:52] <ubitux> to loop over an array and check the value
[10:53] <ubitux> well, won't be easy to have it generic, but for int that would be nice
[10:53] <av500> to replace a 2 line for loop?
[10:55] <ubitux> generally we have a 5 lines function (body only)
[10:55] <ubitux> of course it doesn't matter much but well :)
[11:05] <Tjoppen> testing better fix for #1099
[11:37] <Tjoppen> seems to work fine. double-cecking that the proposed solution doesn't work
[13:01] <michaelni> Tjoppen, should i apply the patch ?
[13:01] <michaelni> or do you want to wait for someone reviewing it ...
[13:21] <Tjoppen> the guy seems happy about it, it fixes the problem and it's fairly easy to understand
[13:21] <Tjoppen> so yeah
[13:58] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * reeb792d862 10ffmpeg/libavcodec/pngdec.c: 
[13:58] <CIA-17> ffmpeg: pngdec: Print error messages for the various failure pathes.
[13:58] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[13:58] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r4ed47d3354 10ffmpeg/libavcodec/pngdec.c: 
[13:58] <CIA-17> ffmpeg: pngdec: dont discard incomplete images.
[13:58] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[13:58] <CIA-17> ffmpeg: 03Tomas Härdin 07master * r393b81f093 10ffmpeg/libavformat/mxfdec.c: 
[13:58] <CIA-17> ffmpeg: mxfdec: Only parse next partition pack if parsing forward
[13:58] <CIA-17> ffmpeg: This fixes ticket #1099.
[13:58] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[14:31] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r309d8ec19b 10ffmpeg/libavcodec/pngenc.c: 
[14:31] <CIA-17> ffmpeg: pngenc: allocate packets that have some chance of being large enough.
[14:31] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[14:32] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r8f394a6cf8 10ffmpeg/libavcodec/pngdec.c: 
[14:32] <CIA-17> ffmpeg: pngdec: print error message for truncated pngs even if we output them
[14:32] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[16:36] <TimNich> ping
[16:39] <ubitux> hi TimNich :)
[16:39] <ubitux> thanks for the feedback on the mov timecode patch, i'm going to look into it
[16:43] <TimNich> Hi ubitux, I wonder if it is related to the atom nesting that was unclear?
[16:43] <ubitux> no i don't think so
[16:43] <ubitux> i think i just messed up with the duration computation
[16:44] <TimNich> Uk oh. Do you have QT7?
[16:44] <ubitux> no i just use hexdump and vbindiff to compare :p
[16:45] <TimNich> Fine. I just happen to be currently running all my VM's on an OSX box as the memory upgrade was cheaper than anything else
[16:45] <TimNich> Means poiing into FCP and QT7 is easy :)
[16:46] <TimNich> s/poing/popping
[16:46] <spaam> :D 
[18:29] Action: Compn adds __gb__ 's proposal to ffmpeg soc page
[18:30] <Compn> in case anyone feels like working on it in the future, lots of good ideas :)
[20:51] <michaelni> Compn, ok if i add the project from "[FFmpeg-devel] FFmpeg-Based Projects" to the project page ? (didnt really look at what it is ...)
[21:14] <Daemon404> woah
[21:14] <Daemon404> michaelni, 
[21:14] <Daemon404> i fucked up
[21:14] <Daemon404> the permissions on zerocodec's fate test is bad
[21:14] <Daemon404> it definitely should NOT be 755
[21:25] <michaelni> Daemon404, fixed locally
[21:25] <Daemon404> thanks
[21:25] <Daemon404> that'll teach me to create files over CIFS
[21:25] <Daemon404> -_-
[21:35] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r15db6a9590 10ffmpeg/libavcodec/pngenc.c: 
[21:35] <CIA-17> ffmpeg: pngenc: Fix incorrect mask used for interlaced mode.
[21:35] <CIA-17> ffmpeg: Fixes Ticket1109
[21:35] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:35] <CIA-17> ffmpeg: 03Michael Niedermayer 07master * r745a33a443 10ffmpeg/tests/ref/fate/zerocodec: 
[21:35] <CIA-17> ffmpeg: fate/zerocodec: fix permissions
[21:35] <CIA-17> ffmpeg: Reported-by: Deamon404
[21:35] <CIA-17> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[21:37] <Daemon404> thanks
[22:11] <Daemon404> -bf <int> - use 'frames' B frames
[22:11] <Daemon404> ^ engrish if i ever saw it...
[22:12] <ohsix> that's racist
[22:12] <gnafu> He didn't say it was a particular race of Engrish.
[22:12] <gnafu> Engrish comes in all shapes and sizes.
[22:13] <ohsix> but most likely the kind that has no l sound
[22:13] <gnafu> But then it would say "flames".
[22:13] Action: gnafu ducks.
[22:44] <michaelni> noone around who wants to represent ffmpeg at linuxtag ?
[22:44] <michaelni> bcoudurier, ubitux, Compn maybe ?
[22:45] <bcoudurier> I'm far
[22:45] <michaelni> i know its kinda late to ask this 23 o clock on the last day of the deadline :(
[22:45] <ubitux> i'm in france sorry
[22:47] <michaelni> anyone else ?
[22:50] Action: Daemon404 pings j-b curiously (i do apologize for being gung-ho)
[23:43] <Compn> michaelni : i'm near detroit ;)
[23:44] <Compn> so kinda far from me
[23:48] <michaelni> Compn, ubitux bcoudurier if its just distance, finding someone to fund the back and forth travel should not be a major problem
[23:49] <michaelni> but i think the real problem is the little time left to the application deadline
[23:49] <michaelni> the actual linux tag will be in may if i understood correctly
[23:50] <michaelni> reimars mail points to 23th - 26 may
[23:53] <Compn> otherwise, i'm not interested , thanks for asking tho :)
[23:55] <Compn> i dunno, itd be fun to go to germany
[00:00] --- Wed Mar 21 2012