[Ffmpeg-devel-irc] ffmpeg-devel.log.20120626

[burek]

[00:44] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * r8a0cd58729 10ffmpeg/libavcodec/sonic.c: 
[00:44] <CIA-119> ffmpeg: sonic: fix FPE
[00:44] <CIA-119> ffmpeg: Fixes Ticket1397
[00:44] <CIA-119> ffmpeg: Found-by: Piotr Bandurski <ami_stuff at o2.pl>
[00:44] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[00:54] <llogan> saste: any ideas for a new closed state for art tickets? I'm thinking "completed art".
[00:54] <saste> or just "completed"...
[00:55] <saste> but give me a night and I think at it when walking the dog
[00:56] <llogan> or "other" or "misc"
[00:56] <llogan> just spamming ideas
[00:57] <saste> i hate "other"... means all and nothing
[00:58] <saste> "archived" may be another...
[00:59] <llogan> i'm fine with "completed"
[01:01] <saste> for example a logo which is not completed, but to which the artist don't want to work no longer
[01:02] <saste> *doesn't want to work any longer
[01:02] <saste> completed (art), archived (art)
[01:05] <llogan> seems like we are overcomplicating it. i'd prefer just one closed state for art.
[01:07] <durandal_1707> i prefer if fffmpeg.org dont have images at all
[01:07] <CIA-119> ffmpeg: 03Paul B Mahol 07master * re3c2670539 10ffmpeg/libavcodec/mss1.c: 
[01:07] <CIA-119> ffmpeg: mss1: check number of free colours
[01:07] <CIA-119> ffmpeg: Prevents out of array write.
[01:07] <CIA-119> ffmpeg: Signed-off-by: Paul B Mahol <onemda at gmail.com>
[01:08] <burek> how about video :D
[01:08] <durandal_1707> what codec?
[01:09] <CIA-119> ffmpeg: 03yang 07master * r9b72041f80 10ffmpeg/libavutil/x86/intmath.h: 
[01:09] <CIA-119> ffmpeg: x86/intmath.h: Fix mull operand constraints
[01:09] <CIA-119> ffmpeg: Fixes Ticket1466
[01:09] <CIA-119> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[01:09] <Compn> ubitux : it was just a suggestion, if you think clone is better first, go for it :)
[01:30] <CIA-119> ffmpeg: 03Mans Rullgard 07master * r8725da49a2 10ffmpeg/libavcodec/x86/fft_mmx.asm: x86: fft: win64: fix stack alignment for memcpy() call
[01:30] <CIA-119> ffmpeg: 03Mans Rullgard 07master * r963cdf39b4 10ffmpeg/libavutil/x86/cpu.c: 
[01:30] <CIA-119> ffmpeg: x86: cpu: whitespace (mostly) cosmetics
[01:30] <CIA-119> ffmpeg: This adds whitespace around operators, aligns line continuation
[01:30] <CIA-119> ffmpeg: backslashes, and breaks long lines. Also fixes an ifdef halfway
[01:30] <CIA-119> ffmpeg: through a statement. The one line of duplication this saved is
[01:30] <CIA-119> ffmpeg: not worth the ugliness.
[01:30] <CIA-119> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[01:30] <CIA-119> ffmpeg: 03Ronald S. Bultje 07master * r246154a9af 10ffmpeg/libavutil/log.c: 
[01:30] <CIA-119> ffmpeg: log: Include io.h on windows
[01:30] <CIA-119> ffmpeg: This is required for isatty, which exists on MSVC and is found by
[01:30] <CIA-119> ffmpeg: configure, but is provided by io.h instead of unistd.h.
[01:30] <CIA-119> ffmpeg: Signed-off-by: Martin Storsjö <martin at martin.st>
[01:30] <CIA-119> ffmpeg: 03Justin Ruggles 07master * r14a34d90ad 10ffmpeg/libavresample/x86/audio_convert_init.c: lavr: x86: merge some branches
[01:30] <CIA-119> ffmpeg: 03Michael Niedermayer 07master * r3b0ad040b3 10ffmpeg/: (log message trimmed)
[01:30] <CIA-119> ffmpeg: * qatar/master:
[01:30] <CIA-119> ffmpeg:  log: Include io.h on windows
[01:30] <CIA-119> ffmpeg:  lavr: x86: merge some branches
[01:30] <CIA-119> ffmpeg:  x86: cpu: whitespace (mostly) cosmetics
[01:30] <CIA-119> ffmpeg:  x86: fft: win64: fix stack alignment for memcpy() call
[04:03] <llogan> how many more stupid mistakes in this patch will i make today?
[04:03] Action: llogan blames the heat. wool was a bad idea.
[04:09] <Compn> wool is always a bad idea
[04:09] <Compn> so itchy
[05:51] <Zeranoe> I'm a little confused by this patch: http://git.videolan.org/?p=ffmpeg.git;a=commit;h=9ec5e956a2a4e359f21d48cc700527ca6ae9e32e "ffmpeg: disable threading on mingw, it doesnt work due to dependance on internal code." is this only for win32thread?
[09:31] <CIA-119> ffmpeg: 03Carl Eugen Hoyos 07master * rc77bcbbb32 10ffmpeg/libavcodec/mss1.c: 
[09:31] <CIA-119> ffmpeg: Signal MSS1 palette change.
[09:31] <CIA-119> ffmpeg: Reviewed-by: Paul B Mahol
[09:55] <CIA-119> ffmpeg: 03Carl Eugen Hoyos 07master * r6eff277284 10ffmpeg/libavcodec/sanm.c: 
[09:55] <CIA-119> ffmpeg: Make LucasArts Smush SANM palette opaque.
[09:55] <CIA-119> ffmpeg: Reviewed-by: Paul B Mahol
[11:45] <CIA-119> ffmpeg: 03Stefano Sabatini 07master * rf054dbee6c 10ffmpeg/tools/graph2dot.c: tools/graph2dot: make dot graph representation a bit more compact
[11:45] <CIA-119> ffmpeg: 03Stefano Sabatini 07master * r8dd0e87d7b 10ffmpeg/ (5 files in 3 dirs): 
[11:45] <CIA-119> ffmpeg: lavfi: remove old video sink API
[11:45] <CIA-119> ffmpeg: It was deprecated since a long time and removed after the 2->3 major
[11:45] <CIA-119> ffmpeg: bump.
[11:45] <CIA-119> ffmpeg: 03Stefano Sabatini 07master * r720ec62012 10ffmpeg/libavfilter/ (avfilter.h buffersink.h defaults.c formats.c version.h): 
[11:45] <CIA-119> ffmpeg: lavfi: drop deprecated and disabled packing API
[11:45] <CIA-119> ffmpeg: It was deprecated and removed after the recent 2->3 major bump.
[11:45] <CIA-119> ffmpeg: 03Stefano Sabatini 07master * r79a7451d06 10ffmpeg/ffplay.c: 
[11:45] <CIA-119> ffmpeg: ffplay: add configure_filtergraph() helper
[11:45] <CIA-119> ffmpeg: Will help factorization with the pending -af patch, and add some checks
[11:45] <CIA-119> ffmpeg: missing in the original code.
[11:53] <burek> is anyone here interested in improving the documentation by making changes to allow easy editing for both developers and regular users?
[11:53] <burek> I've left the link yesterday to DocBook: http://en.wikipedia.org/wiki/DocBook
[11:54] <burek> but I didn't see anyone bothered to reply to it..
[11:54] <ubitux> i thought we agreed with the wiki trac :)
[11:54] <ubitux> burek: btw i sent a patchset to add the wiki and stuff
[11:55] <ubitux> in a "community"-like section
[11:55] <burek> ubitux, hi :)
[11:55] <burek> well, what's the point of the trac wiki if the documentation is still left split?
[11:55] <ubitux> the main documentation will stay in the repository
[11:56] <burek> by that, do you mean "editable only by developers" or "tracked for version changes" ?
[11:56] <saste> burek: technical documentation -> git, user level documentation (guides, walkthrough, overview) -> wiki
[11:56] <ubitux> that is IMO the best way to get sure developers will update it when there is a new features in
[11:56] <ubitux> exactly what saste said
[11:57] <burek> isn't that putting all the work for tech docs to developers only?
[11:57] <burek> i.e. it will stay messy like it is now, for that exact reason..
[11:57] <cbsrobot> ah please not xml
[11:58] <ubitux> :D
[11:58] <burek> cbsrobot, it's not xml
[11:58] <cbsrobot> maybe http://johnmacfarlane.net/pandoc/ could be used
[11:58] <burek> read more thoroughly
[11:58] <ubitux> burek: it won't be better with a wiki
[11:58] <ubitux> for technical doc, it can't
[11:58] <burek> it's saved as xml, but it offers wide variety of presentations, even man pages format
[11:58] Action: cbsrobot is meber of the noxml party
[11:58] <burek> you can edit it even in html if you like
[11:58] Action: ubitux is a member of the killxml parthy
[11:58] <ubitux> -h
[11:59] <ubitux> html is like xml
[11:59] <cbsrobot> lol
[11:59] <saste> burek: in order to contribute effectively to technical documentation, you need to be a developer anyway
[11:59] <burek> saste, to correct syntax errors too?
[11:59] <ubitux> burek: just raise them on irc, we will fix them
[12:00] <saste> so you already knows how to use git, patches, etc, using a wiki won't help, but will make harder to maintain the thing (wiki<->docs synch, versions, etc.)
[12:00] <burek> ubitux, it's not a problem to leave things as they are.. but isn't it obvious so far that it's a double work
[12:00] <ubitux> burek: we have reviews to avoid most of the syntax errors
[12:00] <burek> maybe I'm wrong about all this..
[12:00] <ubitux> i think you are
[12:00] <cbsrobot> burek: I like the intention, but the implementation is the problem
[12:00] <ubitux> i can explain why again
[12:00] <ubitux> but you will need to read me this time
[12:01] <cbsrobot> hehe
[12:01] <cbsrobot> read more guess less
[12:01] <burek> fine
[12:01] <ubitux> burek: to me there is just no user contrib, because the trac has not much content, so there is no double work
[12:01] <ubitux> and as saste said, user documentation is a different documentation from user
[12:02] <ubitux> from dev*
[12:03] <saste> an idea could be i heard is implemented maybe by libboost, an user can edit the web page, the change is converted to a patch and sent to the ML
[12:03] <saste> this may work but i guess it's a lot of work to setup
[12:03] <ubitux> that's a bit what you can do with github
[12:03] <ubitux> since you can edit files directly from the web ui
[12:03] <saste> and most patches created in this way will need to be reviewed more or less, so it's again moving the work on the shoulder of the developers
[12:05] <ubitux> or you can just raise the typo on irc and we will fix them asap
[12:21] <saste> ubitux: ideally we should be able to set metadata on filtered frames, and use select accordingly (or a selectmetadata variant)
[12:21] <saste> rather than adding more and more features to select
[12:21] <saste> indeed i don't like the lavc dependency
[12:21] <saste> not that it is an issue, since now lavfi depends on lavc anyhow
[12:41] <saste> michaelni: what's your plan with the opaque field in filters init?
[12:41] <saste> I'd prefer to keep it, it is used in the sink and i can see some scenarios when it is actually useful
[13:07] <CIA-119> ffmpeg: 03Nicolas George 07master * rf767658414 10ffmpeg/libavcodec/x86/fft_mmx.asm: 
[13:07] <CIA-119> ffmpeg: Revert "x86: fft: win64: fix stack alignment for memcpy() call"
[13:07] <CIA-119> ffmpeg: This reverts commit 8725da49a2090de05b4b2d05e33727f45cb9d970.
[13:07] <CIA-119> ffmpeg: Necerrary to revert 82992604706144910f4a2f875d48cfc66c1b70d7.
[13:07] <CIA-119> ffmpeg: 03Nicolas George 07master * rfd91a3ec44 10ffmpeg/libavcodec/x86/ (Makefile fft_mmx.asm fft_sse.c): 
[13:07] <CIA-119> ffmpeg: Revert "x86: fft: convert sse inline asm to yasm"
[13:07] <CIA-119> ffmpeg: This reverts commit 82992604706144910f4a2f875d48cfc66c1b70d7.
[13:07] <CIA-119> ffmpeg: It breaks shared builds on x86_64.
[13:15] <CIA-119> ffmpeg: 03Stefano Sabatini 07master * re6674e46ec 10ffmpeg/ (7 files in 3 dirs): (log message trimmed)
[13:15] <CIA-119> ffmpeg: lavu/imgutils: create misc functions for dealing with buffers
[13:15] <CIA-119> ffmpeg: Move the lavc/imgconvert functions and rename them as follows:
[13:15] <CIA-119> ffmpeg: avpicture_get_size -> av_image_get_buffer_size()
[13:15] <CIA-119> ffmpeg: avpicture_fill -> av_image_fill_arrays()
[13:15] <CIA-119> ffmpeg: avpicture_layout -> av_image_copy_to_buffer()
[13:15] <CIA-119> ffmpeg: The new functions have an align parameter, which allows to define the
[13:36] <CIA-119> ffmpeg: 03Nicolas George 07master * ra2bd8a9384 10ffmpeg/libavfilter/buffersink.h: buffersink: group libav API functions.
[13:36] <CIA-119> ffmpeg: 03Nicolas George 07master * r784675ca91 10ffmpeg/libavfilter/sink_buffer.c: sink_buffer: make opaque argument optional.
[14:01] <ubitux> saste: ah yeah maybe
[14:01] <ubitux> saste: didn't you say you wanted to add ifdef?
[14:01] <ubitux> (in the select/scene)
[14:01] <saste> yes
[14:02] <ubitux> i can do it, maybe tonigh, otherwise
[14:02] <saste> but i don't want the filter to do things which is not required to
[14:02] <ubitux> sure
[14:02] <saste> and decrease performance in case a certain "feature" has not to be selected
[14:03] <ubitux> it doesn't run scene detect if you don't use it
[14:03] <saste> but the more general approach is to decouple selecting logic from feature detection
[14:03] <ubitux> it could be nice to access the metadata when requesting a string indeed...
[14:03] <saste> that is you detect the feature and put the data in frame metadata, and use a select filter matching a metadata pattern
[14:04] <saste> with no need to further extend the select code more and more
[14:21] <ubitux> feel free to design something :)
[16:18] <CIA-119> ffmpeg: 03Mans Rullgard 07master * r37c3864ef7 10ffmpeg/libavcodec/x86/fft_mmx.asm: 
[16:18] <CIA-119> ffmpeg: x86: fft: elf64: fix PIC build
[16:18] <CIA-119> ffmpeg: In a 64-bit PIC build, external functions must be called
[16:18] <CIA-119> ffmpeg: through the PLT.
[16:18] <CIA-119> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[16:18] <CIA-119> ffmpeg: 03Nicolas George 07master * r91765594dd 10ffmpeg/libavcodec/x86/ (Makefile fft_mmx.asm fft_sse.c): 
[16:18] <CIA-119> ffmpeg: Revert "Revert "x86: fft: convert sse inline asm to yasm""
[16:18] <CIA-119> ffmpeg: This reverts commit fd91a3ec44de38251b2c15e03e26d14e983c4e44.
[16:18] <CIA-119> ffmpeg: The bug it introduced has been fixed.
[16:19] <CIA-119> ffmpeg: 03Nicolas George 07master * rd4c45b8adf 10ffmpeg/libavcodec/x86/fft_mmx.asm: 
[16:19] <CIA-119> ffmpeg: Revert "Revert "x86: fft: win64: fix stack alignment for memcpy() call""
[16:19] <CIA-119> ffmpeg: This reverts commit f767658414fc85dea4006cb82969b6a925fdd380.
[16:19] <CIA-119> ffmpeg: The bug it introduces has been fixed.
[17:00] <jesk> what is a simple/cheap way to create free atom space in mp4
[17:00] <jesk> don't want to code that :-)
[17:15] <ubitux> jesk: use a "free" or "wide" atom?
[17:27] <jesk> ubitux: any tool for doing that?
[17:28] <ubitux> no
[17:28] <ubitux> you can do it with your shell
[17:28] <ubitux> as long as you're appending it at the end
[17:33] <jesk> ubitux: i wouldn't have any problem with an initial rewrite, have to do that anyway to get the moov to the beginning
[17:36] <jesk> how would you insert it to the end, and how would you rewrite so that moov and/or free atom is in the the beginning?
[17:36] <jesk> sorry if those questions seem to be stupid, but google isn't that helpful with that.. :-(
[17:36] <ubitux> i don't know what you want to do
[17:37] <ubitux> are you willing to add a faststart to the mov muxer?
[17:37] <jesk> that's something i'am not quite sure yet, when writing with quicktime or mp4box it has always a faststart
[17:38] <jesk> with ffmpeg only with qtfaststart
[17:38] <jesk> i'am troubleshooting yet what the more accepted way with my devices is
[17:39] <jesk> but anyway, without free atom there is no cheap way of modifying metadata when the data will increase a bit
[17:40] <jesk> and that's unacceptable
[17:40] <jesk> so I would like to try out what the better practice is
[17:40] <ubitux> 17:36:50 <@ubitux> i don't know what you want to do
[17:41] <jesk> would like to try with free atom in the beginning, either with procedure which focus *only* on placing free atom or with procedure with moving moov and freeing some atoms
[17:42] <jesk> creating free at the end is something I would also like to test
[17:42] <jesk> but yet I have no clue at all :D
[17:42] <av500> what point would a free at the end have?
[17:42] <jesk> *no clue how to create free atoms for placing metadata in both ways
[17:42] <av500> you can add that any time
[17:43] <jesk> av500, true
[17:43] <jesk> very true
[17:44] <jesk> so lets focus with free in beginning moov :-)
[17:45] <av500> well, so change the muxer to add that
[17:45] <jesk> what would you suggest, which muxer can do that natively?
[17:45] <jesk> does ffmpeg has option for that?
[17:47] <ubitux> ffmpeg doesn't support faststarting yet
[17:48] <ubitux> but it shouldn't be that hard to do
[17:48] <av500> change ffmpeg then
[17:50] <jesk> ok, nothing easier than that ;-)
[17:51] <av500> you are in #ffmpeg-devel after all
[17:53] <jesk> take it as feedback from user and developer but ffmpeg noob 
[17:53] <jesk> :)
[17:53] <jesk> usecase for many apple fanbox
[17:53] <jesk> s,fanbox,fanboys,
[17:53] <av500> but what is the use case?
[17:53] <av500> what do you need free for?
[17:54] <jesk> creating mp4, compatible with iDevice and ATV, and ability to modify/add iTunes/Movie metadata without the need to wait hours for rewrite
[17:55] <ubitux> if the header is at the end, it doesn't matter, you can just add the udta/meta crap at the end (and resize parent atoms)
[17:55] <ubitux> if it's on the top you need to adjust the offsets
[17:56] <jesk> header at the end is more or less incompatible with Playstation and Apple stuff (imho)
[17:56] <av500> then you need to faststart it after encoding
[17:56] <ubitux> then look how qt-faststart does
[17:56] <jesk> ffmpeg is different in this regard I believe
[17:56] <jesk> qt-faststart can create free?
[17:57] <jesk> not that I know of, but will have a look again
[17:57] <ubitux> well adapt the code
[17:57] <av500> why do you need to create free?
[17:57] <ubitux> av500: to make space to add metadata later
[17:57] <av500> ah, ok for metadata
[17:57] <av500> well, then modify faststart to add a free atom
[17:57] <av500> no big deal
[17:57] <jesk> yet it is for me, but I will investigate
[17:58] <ubitux> iirc qt-faststart is far from perfect
[17:58] <ubitux> especially while looking for the offsets table
[17:58] <ubitux> (just a memsearch)
[17:58] <ubitux> i wonder if it doesn't trigger false-positive often
[17:59] <ubitux> anyway,
[17:59] Action: ubitux &
[18:00] <jesk> ffmpeg/qtfaststart has a lot of problems creating mp4s working on Apples
[18:01] <av500> and you reported them?
[18:01] <jesk> not yet, still in the debugging/understanding phase
[18:01] <jesk> working with mp4's since a week now
[18:05] <av500> k
[18:09] <cbsrobot> last time I saw a file some adobe enc added xml at the end of a mov and qtfaststart just failed with that ... so if you rewrite it keep that in mind
[18:25] <Compn> jesk : we'd love patches :)
[18:25] <Compn> when you are ready
[18:26] <jesk> cbsrobot: not only Adobe, Apple too
[18:27] <jesk> Compn: ok :-)
[19:05] <CIA-119> ffmpeg: 03Gavin Kinsey 07master * r37b5959d96 10ffmpeg/doc/examples/ (filtering_audio.c filtering_video.c): 
[19:05] <CIA-119> ffmpeg: examples/filtering: fix packet memleak
[19:05] <CIA-119> ffmpeg: Free packets unconditionally after demuxing, and not only when the
[19:05] <CIA-119> ffmpeg: packets belong to a given stream.
[19:05] <CIA-119> ffmpeg: Signed-off-by: Stefano Sabatini <stefasab at gmail.com>
[19:05] <CIA-119> ffmpeg: 03Gavin Kinsey 07master * r9ebed95db6 10ffmpeg/doc/examples/filtering_video.c: 
[19:05] <CIA-119> ffmpeg: examples/filtering_video: update to the new API
[19:05] <CIA-119> ffmpeg: Update the video filtering example program based on the audio one.
[19:05] <CIA-119> ffmpeg: Signed-off-by: Stefano Sabatini <stefasab at gmail.com>
[19:38] <ubitux> the re-align wasn't required
[19:38] <ubitux> and now it's wrong :)
[19:42] <TXH350> Provide me ffmpeg PGP public key ID for release, plz. I can't find it.
[19:47] <ubitux> http://ffmpeg.org/download.html#releases ?
[19:49] <TXH350> there are only signature files
[19:58] <Compn> TXH350 : ask michaelni for it 
[19:58] Action: Compn afk
[21:42] <ubitux> burek: you should update your forum favicon
[21:44] <llogan> do you think it's too qataresque?
[21:44] <ubitux> yes
[21:44] <ubitux> and i think it might get us some trouble
[21:46] <llogan> what happened to the forum.ffmpeg.org proposal? (or am i confusing myself?)
[21:50] <burek> ubitux, is it ok now?
[21:53] <llogan> looks fine to me
[21:55] <burek> ok
[22:20] <Compn> llogan : what was the proposal ?
[22:20] <Compn> iirc someone came up with a list of ffmpeg forums
[22:20] <Compn> so its kinda weird for us to chose one to link to 
[22:22] <ubitux> burek: looks like it's still in the cache for me :p
[22:25] <burek> ubitux, open in your browser this page http://ffmpeg.gusari.org/favicon.png and press ctrl+F5
[22:25] <burek> and then try the home page again
[22:26] <burek> Compn, the idea was to merge forums into one, but the process has taken too long (more than 3-4 months) and finally it was abandoned (at least by me)
[22:26] <llogan> what's the other forum?
[22:27] <burek> not to blame anyone, just I don't think it should take that much time, so I didn't want to participate any more..
[22:27] <burek> well, there is a list of forums at ffmpeg's contact page I think
[22:27] <ubitux> burek: i did it, it worked, but it didn't update the one in cache when browsing the site :p
[22:27] <ubitux> creepy firefox
[22:27] <ubitux> but i believe you ;)
[22:27] <ubitux> thx :)
[22:28] <burek> but there were 3 of them iirc.. what happened with the 3rd one?
[22:28] <ubitux> i added yours
[22:28] <burek> ubitux, ok
[22:28] <ubitux> it should stress a bit the process
[22:32] <llogan> burek's is the one linked to in #ffmpeg /topic. i don't see why forum.ffmpeg.org shouldn't point to it if he's willing to admin it.
[22:34] <burek> well I am, I just didn't want to diminish the value of other existing forums
[22:34] <burek> so I suggested that we merge all of them into one official
[22:45] <ubitux> reminds me the story of the wiki&
[22:45] <ubitux> ;)
[22:57] <burek> well, all I'm trying to do is to improve something a little bit and make it easy for development of the project to get more help from other people
[22:57] <burek> either by reporting bugs (forums) or by improving docs (wiki), etc
[22:58] <burek> it's not something egoistic, it's just a wish to make things better
[22:59] <ubitux> i know :)
[22:59] <ubitux> and thank you for your contributions
[22:59] <burek> :beer: :)
[22:59] <ubitux> :)
[23:00] <burek> :juice: :D
[23:00] <ubitux> yeah much better
[23:08] <burek> is there somewhere a table, a list, something.. of supported video codecs inside existing formats..
[23:08] <burek> like for example, flv can contain following video codecs: h264, flv, bla bla
[23:09] <burek> and following audio: aac, mp3, bla bla
[23:09] <burek> I'd like to compile a table that can have it all in one place
[23:11] <durandal_1707> i know only two tables riff one and isom one
[23:11] <burek> is there any url for them, please? :)
[23:11] <JEEB> someone tell burek not to ignore me please
[23:11] <JEEB> http://pastebin.com/pFt55P3v
[23:12] <burek> I'm not ignoring you :)
[23:12] <burek> I just choose not to talk to you :)
[23:12] <JEEB> great
[23:12] <JEEB> flv/f4v specs are also available @ http://www.adobe.com/devnet/f4v.html
[23:13] <burek> thanks for that :)
[23:14] <JEEB> I have no idea why ffmpeg let that dude remux mpeg-4 part 2 into flv (technically it's possible naturally, the format is just not spec'd for it)
[23:28] <ePirat> anyone has an idea why static linking of lav* libraries seems to fail while shared works? http://pastebin.com/zqtu17Xt maybe I am doing a noob mistake there& 
[23:30] <nevcairiel> it seems odd that you provide a path to avformat and avutil, but use -lavcodec
[23:31] <ePirat> oh& huh
[23:33] <ePirat> hm now i Undefined symbols for architecture http://pastebin.com/MBwK1DXT 
[23:33] <nevcairiel> sounds like its expecting some more libraries to be linked
[23:34] <ePirat> why does it worked when using shared libraries then? 
[23:35] <nevcairiel> probably because its also loading the shared versions of all the other things? I dunno, i prefer shared anyway
[23:35] <ePirat> hm but i need static& else i have to recompile on every machine&
[23:40] <ePirat> seems maybe there is something wrong with my static libraries& I'll recompile everything and see if it works then&
[00:00] --- Wed Jun 27 2012