[Ffmpeg-devel-irc] ffmpeg-devel.log.20130320

[burek]

[00:00] <kangstarr> thanks guys
[00:01] <kangstarr> i created a new enhancement ticket and put as much info in it as i could
[00:01] <kangstarr> https://ffmpeg.org/trac/ffmpeg/ticket/2382
[00:01] <kangstarr> ubitux: thanks again!
[00:03] <saste> more tickets are open than we can close
[00:05] <ubitux> saste: oh, i'm stupid, i was still using the comma separator
[00:06] <ubitux> ok, now it works indeed
[00:06] <ubitux> saste: want me to fix that in the fate patch?
[00:07] <saste> ubitux, why not
[00:16] <cone-134> ffmpeg.git 03Clément BSsch 07master:17e1b3cd14eb: fate: make metadata fate tests only output ts and metadata.
[00:50] <llogan> so we go ahead and put up a link to the non-free FFmpeg Basics Book and then the author ignores my requests to rename his twitter account from FFmpeg
[00:50] Action: llogan waits another week
[00:57] <ubitux> all this stuff sounds like some stupid scam to me :P
[00:57] <michaelni> llogan, did you CC me ? if so i could reply to him maybe that would help
[00:57] <ubitux> did anyone have a look to the book?
[01:05] <llogan> ubitux: i think i suggested we must take a look first, but i can't remember.
[01:05] <llogan> michaelni: i'll forward it to you.
[01:06] <llogan> lol. i misspelled my name in the message.
[01:11] <cone-134> ffmpeg.git 03Michael Niedermayer 07master:d244a615f63e: er: set error_occured on missing slices
[01:18] Action: michaelni looks as llogan leaves 
[01:18] <michaelni> just wanted to say i have a different email address from the ffmpeg book guy
[01:19] <michaelni> maybe he doesnt read that one he used for ffmpeg-devel subscription ....
[01:19] <michaelni> whatever, reply sent to both addresses
[03:24] <Compn> Chrome/Chrome OS via a Linux kernel bug, config file error and a video parsing flaw."
[03:24] <Compn> wonder if the video parsing bug in chrome was ffmpeg fault ?
[04:05] <highgod> libavutil/atomic.c:101:2: error: #error "Threading is enabled, but there is no implementation of atomic operations available"
[04:05] <highgod> does anyone has this problem when compile?
[04:06] <Daemon404> what compiler
[04:06] <highgod> mingw gcc gcc version 4.6.2 (GCC)
[04:08] <Daemon404> im not sure what version of mingw introduced atomics
[04:08] <Daemon404> it should be ussing gcc atomics
[04:11] <highgod> the latest master branch, I think it is my git configure problem. but the pervious version I get from git is correct
[04:12] <Daemon404> make distclean?
[04:14] <highgod> I don't konw. how can I update the version base on my old version using git? what's the command?
[04:14] <highgod> I want to try this way, thanks very much
[04:17] <cone-134> ffmpeg.git 03Sebastian Sandberg 07master:db8403d04afc: vc1dec: interlaced B-frames
[04:32] <highgod> the same error
[06:13] <ubitux> oh fun the ivtc seems to work
[06:49] <ubitux> WTF @interlaced sprite overlay in EYESHIELD_ENDING.VOB o_o
[08:27] <nevcairiel> its a rather common thing in broadcast too, a telecined video which has a interlaced overlay like a ticker or something
[09:03] <ubitux> nevcairiel: the thing is, the overlay is only interlaced
[09:03] <ubitux> i mean, it's like you have one static interlaced picture
[09:03] <ubitux> which is moved around
[09:04] <ubitux> wtff
[09:13] <JEEB> ubitux, sounds like most of those anti-piracy things they add on TV :P
[09:13] <JEEB> which move at 60 fields per sec
[09:14] <JEEB> and then everything else is telecined
[09:14] <JEEB> which is how we got that one script by firesledge to deint and motion estimate such stuff prettily into 24/1.001
[09:50] <pr0digy1> hello
[09:51] <pr0digy1> i have a question about ffmpeg api's, is anyone available to try and help me out?
[09:53] <av500> ask, dont ask to ask
[11:02] <Bor0> could anyone explain me this piece of code (h264.c) http://pastebin.com/MwwAM6U4. in what cases will the if() clause fail? what's the meaning of those variables?
[11:30] <onto> Hi! I am trying to write a simple program that can capture audio + video but it cannot open the video device. Here's my code so far #include <stdio.h>
[11:30] <onto> #include <libavformat/avformat.h>
[11:30] <onto> #include <libavdevice/avdevice.h>
[11:30] <onto> int main(int argc, char **argv){
[11:30] <onto> av500: _register_all();
[11:30] <onto> av500: device_register_all();
[11:30] <onto> irock: f(!(pFormat = av_find_input_format("v4l2"))){
[11:30] <cone-581> ffmpeg.git 03Anton Khirnov 07master:3cd93cc7b898: Revert "asfenc: return error on negative timestamp"
[11:30] <cone-581> ffmpeg.git 03Anton Khirnov 07master:e5c32d6da783: avplay: remove the -debug option.
[11:30] <cone-581> ffmpeg.git 03Anton Khirnov 07master:2c328a907978: pixdesc: add a function for counting planes in a pixel format.
[11:30] <cone-581> ffmpeg.git 03Michael Niedermayer 07master:17ebef2fc809: Merge commit '2c328a907978b61949fd20f7c991803174337855'
[11:32] <onto> sorry for the flood, I accidentally pasted the code instead of the pastebin URL (http://pastebin.ca/2336458)
[11:33] <durandal_1707> onto: you are on wrong channel
[11:33] <onto> I am trying to create a simple program to capture video + audio
[11:33] <onto> durandal_1707: oh, how come?
[11:34] <durandal_1707> there is channel info
[11:34] <onto> ah sorry
[11:35] <Bor0> not sure if you received my question, my connection resetted, so here goes: durandal_1707: is there any official channel for libavcodec/h264 development, or is this the right place for it?
[11:36] <durandal_1707> what libavcodec/h264 development means?
[11:36] <Bor0> the h264.c file within libavcodec/
[11:36] <durandal_1707> this channel is for development of FFmpeg(libavcodec/libavformat/ etc) itself
[11:36] <durandal_1707> and not for development of code that use such libs
[11:37] <Bor0> is h264.c a wrapper for libx264?
[11:37] <av500> is it?
[11:37] <av500> look inside
[11:37] <Bor0> I am, from what I can see it isn't
[11:37] <cone-581> ffmpeg.git 03Anton Khirnov 07master:c977039e585b: lavc, lavfi: fix counting number of planes in AVBufferRef wrappers
[11:38] <cone-581> ffmpeg.git 03Anton Khirnov 07master:6599b087de62: buffersrc: fix a typo.
[11:38] <cone-581> ffmpeg.git 03Michael Niedermayer 07master:df804041bd42: Merge commit '6599b087de62a5f9f2a8d61a1952d777d1bff804'
[11:38] <av500> notice there is also libx264.c
[11:38] <Bor0> ok, my issue is that I'm getting "no picture" error, mainly because pics > s->avctx->has_b_frames conditional fails. could anyone explain to me what it means?
[11:42] <michaelni> Bor0, that means you have fewer pictures than there is reordering
[11:44] <Bor0> is the video badly transcoded? ffplay plays it correctly
[11:50] <cone-581> ffmpeg.git 03Anton Khirnov 07master:1516bf7af1e8: lavc: update the doxy for avcodec_decode_{video,audio} wtih refcounting.
[11:50] <cone-581> ffmpeg.git 03Hendrik Leppkes 07master:4e70d66ded53: mpegvideo: fix allocation of the hwaccel_picture_private data
[11:50] <cone-581> ffmpeg.git 03Michael Niedermayer 07master:54b0bef390ed: Merge commit '4e70d66ded537cadd32dbd02a38c3d86a203c812'
[11:57] <cone-581> ffmpeg.git 03Hendrik Leppkes 07master:c71c80f53b6a: pthread: unref already decoded frames when flushing the decoder
[11:57] <cone-581> ffmpeg.git 03Michael Niedermayer 07master:084510d12ab6: Merge commit 'c71c80f53b6a63fd6360d9de5a262f9e7c85681f'
[11:58] <michaelni> Bor0, why do you think the video is badly transcoded ?
[11:59] <durandal_1707> ubitux: what fate test is relevant for lavu/opt?
[12:04] <cone-581> ffmpeg.git 03Hendrik Leppkes 07master:9c9ede44f37e: h264: allocate hwaccel privdata after the frame buffer
[12:04] <cone-581> ffmpeg.git 03Hendrik Leppkes 07master:c3ebfcd6e132: mpegvideo: allocate hwaccel privdata after the frame buffer
[12:04] <cone-581> ffmpeg.git 03Michael Niedermayer 07master:f7fee8f700c2: Merge commit 'c3ebfcd6e1327ca7bbcaee822e593c2da6cfd352'
[12:06] <Bor0> michaelni: I don't know how to handle that error. I'm working on a browser plugin and it works good for small vids but the browser crashes when I try to play a bit larger vids, namely the "no picture" error occurs
[12:07] <michaelni> what error ?
[12:07] <michaelni> "no picture" is a debug message no error
[12:08] <Bor0> hmm
[12:08] <Bor0> but that debug message does not show when I play small vids
[12:09] <cone-581> ffmpeg.git 03Martin Storsjö 07master:a5e6080a8d85: rtmp: Pass the parameters to do_adobe_auth in the right order
[12:09] <cone-581> ffmpeg.git 03Michael Niedermayer 07master:e7dd0365f293: Merge remote-tracking branch 'qatar/master'
[12:10] <Bor0> so I assumed that could be the root of the problem, which I could be wrong?
[12:16] <michaelni> Bor0, maybe the application doesnt handle "no picture" correctly but thats just a guess. If it crashes you can look at where it crashes ...
[12:27] <Bor0> michaelni: strangely, it crashes at while (is->pictq_size >= VIDEO_PICTURE_QUEUE_SIZE && !is->videoq.abort_request); @ queue_picture(), on the line pthread_cond_wait(&is->pictq_cond, &is->pictq_mutex);
[12:27] <Bor0> it doesn't make any sense to me.
[12:31] <Bor0> I just now noticed that the crash are occuring at random lines, so maybe you were right about not handling "no picture", I'll look into that, thanks!
[12:40] <cone-581> ffmpeg.git 03Bojan Zivkovic 07master:26f3924d782b: mips: Optimization of AAC coefficients encoder functions
[12:47] <cone-581> ffmpeg.git 03Bojan Zivkovic 07master:7d29c6eebc42: mips: Optimization of IIR filter function
[13:33] <cone-581> ffmpeg.git 03Paul B Mahol 07master:7aa9af51db84: lavu/opt: add AV_OPT_VIDEO_RATE option
[13:33] <cone-581> ffmpeg.git 03Paul B Mahol 07master:ac44e52f508f: lavfi/showwaves: make use of AV_OPT_TYPE_VIDEO_RATE
[13:50] <cone-581> ffmpeg.git 03Michael Niedermayer 07master:eaa5882e714e: avutil: fix compilation
[13:50] <cone-581> ffmpeg.git 03Michael Niedermayer 07master:b274703493fb: avutil/atomic: use av_assert0()
[13:54] <cone-581> ffmpeg.git 03Nedeljko Babic 07master:271f95940b0f: libavcodec/mips: fix for mips optimized fp fft with hard coded tables disabled
[14:03] <cone-581> ffmpeg.git 03Nicolas George 07master:5eb273b2e767: lavfi/buffersink: check av_frame_ref() failure.
[14:03] <cone-581> ffmpeg.git 03Nicolas George 07master:de1568a452d8: lavf/concatdec: fix possible leak in case of malloc failure.
[14:03] <cone-581> ffmpeg.git 03Nicolas George 07master:0d09aa9d8ec4: lavf: avformat_seek_file(): validate stream_index.
[14:03] <cone-581> ffmpeg.git 03Nicolas George 07master:a964d6a8a26d: lavu/frame: use channels rather than channel_layout.
[14:03] <cone-581> ffmpeg.git 03Michael Niedermayer 07master:9ea6c14820da: Merge remote-tracking branch 'cigaes/master'
[15:32] <JEEB> can someone try with current ffprobe or something if current ffmpeg's libavformat still tries to load this whole file into memory? http://x264.fushizen.eu/random/otonajoshi_avformat_tries_to_load_whole.ts
[15:33] <JEEB> I think Daemon404 replicated it like two days ago on ffprobe
[15:33] <JEEB> (he got the full 4GB file while it was still available at its original URL)
[16:05] <nevcairiel> three day old FFmpeg still dos
[17:30] <Tjoppen> michaelni: that mxf bug on the list is related to #1916
[17:37] <Daemon404> JEEB, the problem is there's a "fix"
[17:37] <Daemon404> but do we want that "fix" or a fix
[17:38] <Daemon404> "fix" has potential to cause regressions
[17:39] <JEEB> yeah... that's the problem always :V
[17:40] <nevcairiel> i'm not sure there is a clear and risk-free fix
[17:40] <nevcairiel> not without a lot of complexity at least
[17:41] <Daemon404> nevcairiel, yes exactly
[17:41] <Daemon404> but saying "only try N times" seems hacky at best
[17:41] <Daemon404> whats a good value of N? etc
[17:41] <nevcairiel> easy fix: try to figure out what the stream its not managing to probe actually is, and teach it to probe it :D
[17:41] <Daemon404> its probably mangled
[17:41] <nevcairiel> result: file works
[17:42] <nevcairiel> ;)
[17:42] <Daemon404> i.e. not a valid stream
[17:42] <nevcairiel> there are already settings for max probe size and stuff
[17:42] <Daemon404> i peeked at some vars, and i sear i saw it thought there were 119 streams
[17:42] <nevcairiel> it could obey that
[17:42] <Daemon404> i could be misremembering
[17:42] <JEEB> Daemon404, PIDs are over 100 at many places
[17:42] <JEEB> or did it really find many streams?
[17:42] <JEEB> and not just PID number something?
[17:43] <Daemon404> i saw i=119, and i starts at 0
[17:43] <Daemon404> so...
[17:43] <j-b> m
[17:43] <j-b> oops, sorry
[17:43] <Daemon404> i thought it was a PID too
[17:43] <JEEB> nevcairiel, yeah -- exactly. What's the probe size for if it's not used for that stuff?
[17:43] <nevcairiel> its only checked whenever it managed to finish reading a packet :P
[17:43] <JEEB> :D
[17:51] <michaelni> nevcairiel, i think i know how to fix it
[17:58] <cone-134> ffmpeg.git 03Michael Niedermayer 07master:06a43baecd2c: av_find_stream_info: change read_size to 64bit
[17:58] <cone-134> ffmpeg.git 03Michael Niedermayer 07master:710cd0fddf39: avformat: Fix infinite probing that reads the whole file
[17:58] <michaelni> nevcairiel, JEEB should be fixed
[17:58] <JEEB> wohoo
[17:59] Action: nevcairiel is not sure how that fix works
[18:01] <JEEB> if remaining size in buffer is zero or less, probe codec... not sure as I haven't read through that code at all
[18:03] <nevcairiel> ah i see
[18:03] <nevcairiel> it has a defined size of how large its supposed to grow
[18:03] <nevcairiel> and if thats exceeded, just abort probing
[18:04] <JEEB> ah
[18:06] <Daemon404> you know
[18:07] <Daemon404> why didnt i enable ffmpeg's native mingw fate
[18:07] Action: Daemon404 does that now
[18:07] <Daemon404> it's caught errors tht wine hasntt a bunch of times
[18:11] <BBB-work> I know why
[18:11] <BBB-work> it's because nobody wants to voluntarily maintain a windows machine that's connected to a network
[18:12] <BBB-work> it's like a nightmare-come-true
[18:12] <Daemon404> BBB-work, i have a native mingw fate instance on libav
[18:12] <Daemon404> since like october
[18:12] <JEEB> more like setting all the rsync stuff up is a bitch
[18:12] <JEEB> and such
[18:12] <JEEB> :V
[18:12] <JEEB> on msys/mingw
[18:12] <Daemon404> its just a matter of adding one line to a bat file
[18:12] <JEEB> I don't even know how you'd be doing the cron'ing there
[18:12] <BBB-work> I have a windows laptop and I make sure it never connects to a network anywhere in the world
[18:12] <Daemon404> JEEB, windows has a scheduler
[18:12] <Daemon404> its part of the OS
[18:13] <BBB-work> only way to make sure it doesn't die of "natural" causes within a few weeks
[18:13] <JEEB> Daemon404, OK so you wrote a batch file for it :)
[18:13] <JEEB> gotcha
[18:13] <Daemon404> worse
[18:13] <nevcairiel> i have native windows MSVC and mingw fate boxes
[18:13] <nevcairiel> all fine here :p
[18:13] <Daemon404> a batch file that launches sh.exe script.sh
[18:13] <JEEB> that's better :P
[18:13] <JEEB> not worse
[18:13] <Daemon404> an source /fate/builtenv
[18:13] <JEEB> the less batch scripting you have to read/write
[18:13] <JEEB> the better
[18:14] <JEEB> for sanity and otherwise
[18:14] <Daemon404> the script is literally two lines
[18:14] <Daemon404> set PATH=<msys>;%PATH
[18:14] <Daemon404> sh C:\...\script.sh
[18:14] <JEEB> ok, that's how I'd do it too :D
[18:17] <wm4> And now, it seems that Anton, once again, decided that the wheel needed more
[18:17] <wm4> reinventing and more squareness. Shall I push this patch series or shall I
[18:17] <wm4> wait for the merge to hit the fan (sorry, the pun was too tempting)?
[18:17] <wm4> ^ nice mail
[18:18] <nevcairiel> the way i see it, since lavfi is much more matured over here, just let them do their thing :p
[18:18] <nevcairiel> and ignore it
[18:18] <BBB-work> I like daemon404's assessment of lavfi
[18:19] <BBB-work> if one is 0.000000000000% mature and the other is 0.00000006174% mature
[18:19] <BBB-work> is any of them significantly mature?
[18:19] <iive> infinitely more
[18:49] <wm4> so, uh, what's the point merging lavfi changes anyway
[18:50] <wm4> most development happens in ffmpeg's lavfi
[19:04] <michaelni> wm4, the point is keeping future merges simpler, keeping the API similar (nicer for user apps  and filter authors)
[19:17] <iive> michaelni: you get the idea that wm4 is one of the user apps maintainers?
[19:18] <michaelni> iive yes
[19:18] <iive> imho, it may make more sense to wait for all anton's changes and merge them in one go, rather than redoing all filters after every variable rename.
[19:18] <ubitux> he will likely push all the changes in one go anyway ;)
[19:19] <wm4> you can't even support both ffmpeg and libav lavfi without ifdeffery (though that one was ffmpeg's fault)
[19:20] <michaelni> wm4, where is a ifdef needed ?
[19:20] <iive> i don't think he is over with lavfi
[19:20] <wm4> michaelni: avfilter_graph_parse() has incompatible arguments
[19:23] <ubitux> michaelni: saste changed avfilter_graph_parse() prototype at some point (likely when the api was unstable), and then later libav added avfilter_graph_parse2() with our prototype (or something similar but different)
[19:23] <michaelni> we should fix this when major is bumped the next time
[19:24] <michaelni> or better said we should fix this and put it under appropriate #if
[21:03] <cone-134> ffmpeg.git 03Alexander Kojevnikov 07release/1.0:7fc39f51cd00: mp3dec: Fix VBR bit rate parsing
[21:03] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/1.0:10d59d29bc93: msrledec: move output pointer test up
[21:03] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/1.0:cc0662daefba: msrledec: move loop into switch
[21:03] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/1.0:bbefdb393861: msrledec: merge switches
[21:03] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/1.0:f84ddb0c0fbb: msrledec: fix output_end checks
[21:03] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/1.0:31f337c1e44e: iff: fix integer overflow
[21:03] <cone-134> ffmpeg.git 03David Favor 07release/1.0:2440424f1dcf: Slight bug building ffmpeg-1.1.3 on OSX + patch to fix
[21:03] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/1.0:4bee3eba7535: doc/ffmpeg: remove non ascii char
[21:03] <cone-134> ffmpeg.git 03Nicolas George 07release/1.0:dbe57f7e9f08: lavfi/avf_concat: fix invalid exclusive test.
[21:03] <cone-134> ffmpeg.git 03Andrea3000 07release/1.0:2f00d28aa013: matroska: fix missing ,
[21:03] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/1.0:94576cec549e: update for 1.0.6
[21:40] <cone-134> ffmpeg.git 03Frank Galligan 07master:b853103fe0e5: matroskadec: Add support for parsing Matroska ContentEncKeyID
[21:43] <Daemon404> it's scary that someone actually is looking into using matroska's DRM features
[21:43] <Daemon404> wave of the future!
[21:43] <nevcairiel> i think its more scary that a format like matroska has drm features
[21:44] <nevcairiel> although iirc it was just added and ffmpeg is probably the only thing that can read it now
[21:44] <Daemon404> yea
[21:44] <wm4> can ffmpeg actually decode such encrypted files?
[21:44] <Daemon404> no
[21:45] <Daemon404> teh encryption is arbitrary iirc
[21:45] <nevcairiel> it only exports the key in metadata
[21:45] <Daemon404> yea
[21:45] <nevcairiel> usually encryption also needs some second part of the key, otherwise its a pointless encryption
[21:46] <wm4> so why does lavf export something useless and obscure as encryption keys, and not, say, stuff needed to get ordered chapters to work?
[21:47] <nevcairiel> because noone send a patch for that?
[21:47] <nevcairiel> you always sound like someone is planning out these features, stuff gets implemented as people are implementing it
[21:48] <wm4> don't know, I remember uau saying something about this some years ago
[21:48] <BBB-work> uau is exactly what nevcariel just said
[21:48] <BBB-work> complains a lot
[21:48] <wm4> and there were patches to implement ordered chapters etc.
[21:48] <BBB-work> refuses to send patches
[21:48] <wm4> (none which got accepted)
[21:48] <BBB-work> then amend them until they got accepted
[21:48] <BBB-work> hey man, I got msvc support in; it prooves that any patch can make it in here
[21:49] <BBB-work> just gotta massage it until it fits
[21:51] <uau> wm4: i haven't sent any patches for libavformat ordered chapters
[21:52] <uau> nor have i requested that feature on libavformat side IIRC
[21:52] <Daemon404> elenril sent one years ago
[21:52] <Daemon404> but it was FLAME FLAME FLAME
[21:52] <Daemon404> i think "worst patch i have ever seen" was the reply
[21:52] <uau> yes elenril did a patch
[21:53] <uau> the main issue as that the libavformat api is pretty limited - it tries to handle all formats through the same API, and there's no clear framework for format-specific features
[21:54] <Daemon404> uau, its the same reason why we cant have multiple edit list support for mp4/mov
[21:54] <Daemon404> exact same reason.
[21:54] <uau> and doing ordered chapters properly would require features beyond generic-container ones
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:de1a71595de7: lavf/concatdec: support seeking.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:68fb7e260bcd: lavfi/concat: fix silence duration computation.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:8d928a9d9967: lavu/opt: add AV_OPT_TYPE_DURATION.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:21b092de7c1e: lavfi: add sine audio source.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:6a90f6d7891b: ffmpeg: fix freeing of sub2video frame.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:cdd78a65f470: lavf/vobsub: free index pseudo-packet.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:befbcc374d8b: lavfi: add common code to handle options parsing.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:8c6c811b2137: lavfi/avf_concat: use standard options parsing.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:e9f45a833d49: lavfi/vf_tile: use standard options parsing.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:e9ca55aeddc4: lavfi/af_volume: use standard options parsing.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:6202cf5dd809: lavfi/vf_crop: use standard options parsing.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:40ecd44f00dc: lavfi/vf_decimate: use standard options parsing.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:37438d0a7230: lavfi/vf_delogo: use standard options parsing.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:80580b405aea: lavi/vf_drawbox: use standard options parsing.
[21:55] <cone-134> ffmpeg.git 03Nicolas George 07master:4e2bcec95f81: lavfi/vf_fade: use standard options parsing.
[21:56] <cone-134> ffmpeg.git 03Nicolas George 07master:b201c167d096: lavfi/vf_fps: use standard options parsing.
[21:56] <cone-134> ffmpeg.git 03Nicolas George 07master:392ec7ec1cf2: lavfi/vf_pad: use standard options parsing.
[21:56] <cone-134> ffmpeg.git 03Nicolas George 07master:769cc300723c: lavfi/vf_transpose: use standard options parsing.
[21:56] <cone-134> ffmpeg.git 03Nicolas George 07master:8d4d11cd89ec: lavfi/vf_unsharp: use standard options parsing.
[21:56] <cone-134> ffmpeg.git 03Nicolas George 07master:2753d4ebf067: lavfi/vf_yadif: use standard options parsing.
[21:56] <cone-134> ffmpeg.git 03Michael Niedermayer 07master:426ebdf923f9: Merge remote-tracking branch 'cigaes/master'
[22:11] <iive> elenril is the Anton who redesigned avformat api, isn't he? ... so maybe now it can handle ordered chapters?
[22:11] <nevcairiel> avformat api hasn't really changed much for a long time
[22:12] <wm4> I still don't see where the problem is, you'd just have to export the list of chapters and their associated metadata
[22:13] <wm4> there's per-chapter metadata, so that could be used to store chapter-specific data like segment ID
[22:14] <nevcairiel> its really rather pointless if you have to do all the virtual timeline things yourself, and that only covers ordered chapters in the same segment
[22:15] <wm4> I'd find it dirty if libavformat loaded other files itself
[22:16] <nevcairiel> if you only export that metadata, thats still plenty manual labor before it works properly
[22:17] <iive> i think that actual design decision was taken ages ago. too bad it haven't been implemented.
[22:51] <cone-134> ffmpeg.git 03James Almer 07release/0.11:c095137d1b02: lavc/bink: Chech for malloc failure
[22:51] <cone-134> ffmpeg.git 03Xi Wang 07release/0.11:a31be9dd065b: rtpenc: fix overflow checking in avc_mp4_find_startcode()
[22:51] <cone-134> ffmpeg.git 03Xi Wang 07release/0.11:dd6462602283: rtmp: fix multiple broken overflow checks
[22:51] <cone-134> ffmpeg.git 03Xi Wang 07release/0.11:3fb1a6c70253: rtmp: fix buffer overflows in ff_amf_tag_contents()
[22:51] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/0.11:562aa82d2a22: huffyuvdec: Check init_vlc() return codes.
[22:51] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/0.11:864a7e73b960: huffyuvdec: Skip len==0 cases
[22:51] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/0.11:b5469fec6a21: movenc: hotfix, dont store fiel for h264 / mpeg4-asp / dnxhd
[22:51] <cone-134> ffmpeg.git 03Diego Biurrun 07release/0.11:8e2c8aaba4ec: configure: Make warnings from -Wreturn-type fatal errors
[22:51] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/0.11:ad352157505f: doc/APIchanges: fix odd .01 versions
[22:51] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/0.11:c3a5e7be84c8: apichanges: fix date
[22:51] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/0.11:753c6377d3b2: apichanges: Use , instead of / to seperate multiple hashes
[22:51] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/0.11:78317ad8b927: apichanges: fix 2 wrong hashes
[22:51] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/0.11:9faf323d7ed1: doc/APIchanges: List merge commit hashes and version numbers
[22:51] <cone-134> ffmpeg.git 03Alexander Kojevnikov 07release/0.11:7cb24ecd4c41: mp3dec: Fix VBR bit rate parsing
[22:51] <cone-134> ffmpeg.git 03Michael Niedermayer 07release/0.11:fff5f6554006: Merge remote-tracking branch 'jamrial/release/0.11' into release/0.11
[23:37] <saste> any feedback on the doc stuff?
[23:38] <beastd> saste: what doc stuff? is it on -devel ml?
[23:38] <saste> beastd, sure
[23:38] <saste> --enable-monolithic-tools-doc
[23:39] <beastd> uh ok, will try to have a look tomorrow
[23:40] <saste> i didn't post all the patches, waiting for comments on the overall design
[23:42] <beastd> i see. will try to comment.
[23:42] <ubitux> saste: the patch is ok by itself, but i'd like ppl wanting the feature to comment instead
[23:43] <ubitux> (hi Daemon404)
[00:00] --- Thu Mar 21 2013