[Ffmpeg-devel-irc] ffmpeg-devel.log.20120115

burek burek021 at gmail.com
Mon Jan 16 02:05:02 CET 2012


[00:16] <davidmitchell> If I'd like to submit a patch to ffmpeg, is it acceptable to submit a pull request to the GitHub mirror, or do I really need to subscribe to the mailing list and email it?
[00:20] <kierank> mailing list is better
[00:20] <durandal_1707> davidmitchell: http://ffmpeg.org/contact.html
[00:22] <davidmitchell> Thanks, durandal_1707, I read that and only found the link to the MPlayer patch guidelines, which are still geared towards svn
[00:22] <davidmitchell> did I miss a link?
[00:25] <durandal_1707> davidmitchell: http://libav.org/developer.html#Submitting-patches (our fork)
[00:26] <durandal_1707> replace every occurence of libav with ffmpeg
[00:34] <kierank> davidmitchell: http://ffmpeg.org/developer.html#Submitting-patches-1
[00:38] <davidmitchell> wow, that's buried pretty deeply
[00:42] <michaelni> suggestions, patches and pull requests to improve the docs are welcome!
[00:44] <michaelni> also you dont strictly have to subscribe but if you dont it may take a day or 2 until the mail shows up
[00:44] <michaelni> depends on how busy the ML people are
[00:57] <CIA-31> ffmpeg: 03Michael Niedermayer 07master * r1843e6c8aa 10ffmpeg/libavcodec/mjpegdec.c: 
[00:57] <CIA-31> ffmpeg: thp: Fix handling of restart interval
[00:57] <CIA-31> ffmpeg: Fixes Ticket903
[00:57] <CIA-31> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[00:57] <CIA-31> ffmpeg: 03Michael Niedermayer 07master * ra91f206665 10ffmpeg/libavcodec/mjpegdec.c: 
[00:57] <CIA-31> ffmpeg: ljpegdec: reset restart interval for each plane of a non interleaved file.
[00:57] <CIA-31> ffmpeg: Fixes Ticket907
[00:57] <CIA-31> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[02:18] <CIA-31> ffmpeg: 03Justin Ruggles 07master * re1a7af6f29 10ffmpeg/tests/fate/image.mak: FATE: fix targa tests on big-endian systems
[02:18] <CIA-31> ffmpeg: 03Justin Ruggles 07master * rfaaebcdf6b 10ffmpeg/libavcodec/targa.c: targa: add support for rgb555 palette
[02:18] <CIA-31> ffmpeg: 03Justin Ruggles 07master * r9c2f9b0e2d 10ffmpeg/tests/fate/image.mak: FATE: enable the 2 remaining targa conformance suite tests
[02:18] <CIA-31> ffmpeg: 03Janne Grunau 07master * r502bf3b462 10ffmpeg/tests/ref/fate/ (targa-conformance-CCM8 targa-conformance-UCM8): 
[02:18] <CIA-31> ffmpeg: fate: add missing reference files for targa tests in 9c2f9b0e2
[02:18] <CIA-31> ffmpeg: Fixes fate-targa-conformance-CCM8 and fate-targa-conformance-UCM8.
[02:18] <CIA-31> ffmpeg: 03Justin Ruggles 07master * r4c82c6d993 10ffmpeg/tests/ (3 files in 2 dirs): 
[02:18] <CIA-31> ffmpeg: FATE: convert output to rgba for the targa tests which currently output pal8
[02:18] <CIA-31> ffmpeg: fixes the tests on big-endian systems
[02:18] <CIA-31> ffmpeg: 03Ronald S. Bultje 07master * r3c172a4106 10ffmpeg/libswscale/x86/output.asm: swscale: change yuv2yuvX code to use cpuflag().
[02:18] <CIA-31> ffmpeg: 03Martin Storsjö 07master * r57facb73ab 10ffmpeg/libavcodec/libx264.c: (log message trimmed)
[02:18] <CIA-31> ffmpeg: libx264: Don't leave max_b_frames as -1 if the user didn't set it
[02:18] <CIA-31> ffmpeg: max_b_frames is initialized to -1 for libx264, to allow
[02:18] <CIA-31> ffmpeg: distinguishing between an explicit user set 0 and a default not
[02:18] <CIA-31> ffmpeg: touched 0 (see bb73cda2).
[02:18] <CIA-31> ffmpeg: If max_b_frames is left as -1, this affects dts generation (where
[02:18] <CIA-31> ffmpeg: expressions like max_b_frames != 0 are used), so make sure it is
[02:18] <CIA-31> ffmpeg: 03Diego Biurrun 07master * re44c11e9fa 10ffmpeg/ (4 files in 3 dirs): 
[02:18] <CIA-31> ffmpeg: cosmetics: Move static and inline attributes to more standard places.
[02:18] <CIA-31> ffmpeg: Fixes several "static is not at beginning of declaration" warnings.
[02:18] <CIA-31> ffmpeg: 03Stefano Sabatini 07master * rb5c2b5af6a 10ffmpeg/configure: 
[02:18] <CIA-31> ffmpeg: configure: provide libavfilter/version.h header to get_version()
[02:18] <CIA-31> ffmpeg: Fix libavfilter library version numbers generation, which was broken in
[02:18] <CIA-31> ffmpeg: 3167dc9515810bbdd86d99d773bcf84657d2e72a.
[02:18] <CIA-31> ffmpeg: Signed-off-by: Anton Khirnov <anton at khirnov.net>
[02:18] <CIA-31> ffmpeg: 03Aneesh Dogra 07master * r4cd0bdae9a 10ffmpeg/libavcodec/ (bytestream.h sgidec.c): 
[02:18] <CIA-31> ffmpeg: sgidec: Use bytestream2 functions to prevent buffer overreads.
[02:18] <CIA-31> ffmpeg: The patch also adds several bytestream macros to deal with native endian.
[02:18] <CIA-31> ffmpeg: Signed-off-by: Justin Ruggles <justin.ruggles at gmail.com>
[02:18] <CIA-31> ffmpeg: 03Michael Niedermayer 07master * r4640da7e58 10ffmpeg/: (log message trimmed)
[02:18] <CIA-31> ffmpeg: Merge remote-tracking branch 'qatar/master'
[02:18] <CIA-31> ffmpeg: * qatar/master:
[02:18] <CIA-31> (4 lines omitted)
[02:43] <MrNaz_YMAtv>  /j #libav
[02:43] <MrNaz_YMAtv> woops
[02:48] <Compn> dats a banning!
[02:48] <Compn> :p
[02:51] <Daemon404> irc protip: only /j from the server window
[02:53] <ohsix> or /channel add
[03:18] <Snaggle> I saw on the gitweb videolan page that libpostproc was being deprecated after a merge from libav.  However, a git pull doesn't show this happening.  What is the status for it in ffmpeg?
[03:19] <Compn> michaelni will know :)
[03:19] <Compn> whats libpostproc do again ? hasnt it been on depreciated list for years ?
[03:21] <Compn> of course sometimes michaelni merges things from libav that he shouldnt.. it happens when merging 1000s of commits
[03:23] <Snaggle> I'm just curious as to the discrepancy between gitweb and a git pull.  I don't have a preference as to which way to actually go
[03:23] <Compn> ah, sometimes merges show up in the commit but arent actually merged to git
[03:24] <Compn> it will list everything as merged, even things that are marked as not-to-be-merged
[03:24] <Compn> at least it did in the past on the list
[03:25] <Snaggle> ahh
[03:26] Action: Snaggle pokes gitweb for not being clear
[03:27] <michaelni> we wont deprecate libpostproc
[03:27] <michaelni> the commit was not merged but showed up, like compn said ...
[03:27] <ohsix> git log, there was some messages about it
[03:28] <ohsix> oicnm
[03:28] <michaelni> i guess i should have replied on the ML or added a note to the merge commit to clarify
[03:29] <Snaggle> michaelni: no biggie.  I was just curious
[03:29] <Compn> i find it wrong that libav removed sonic riff tags
[03:29] <Compn> i guess they dont care about decoding currently existing samples
[03:30] <Compn> wonder if anyone used it as a codec ever :)
[03:42] <Compn> michaelni : it might make sense to start a kind of running ffmpeg-libav changelog
[03:42] <Compn> well, thats probably a lot of work
[03:43] <Compn> and maybe no one even cares about the differences
[07:42] <CIA-31> ffmpeg: 03Michael Niedermayer 07master * r668a74e808 10ffmpeg/tests/Makefile: 
[07:42] <CIA-31> ffmpeg: buildsystem: add $EXESUF to the used tool for fate
[07:42] <CIA-31> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[07:42] <CIA-31> ffmpeg: 03Michael Niedermayer 07master * rd9816cd5df 10ffmpeg/libavcodec/imgconvert.c: 
[07:42] <CIA-31> ffmpeg: imgconvert: add PIX_FMT_GRAY8A to pix_fmt_info
[07:42] <CIA-31> ffmpeg: Fixes selecting gray8a for PAM
[07:42] <CIA-31> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[10:23] <CIA-31> ffmpeg: 03Nicolas George 07master * rbbf372e145 10ffmpeg/libavfilter/vf_overlay.c: overlay: do not leak x/y expressions.
[13:54] <CIA-31> ffmpeg: 03Reimar Döffinger 07master * r8af6b9711e 10ffmpeg/libavcodec/asv1.c: 
[13:54] <CIA-31> ffmpeg: Fix uninitialized reads for fate-vsynth1-asv2 test.
[13:54] <CIA-31> ffmpeg: This is not a real error and memsetting always even when the
[13:54] <CIA-31> ffmpeg: size did not change is overkill, but it still should be
[13:54] <CIA-31> ffmpeg: an acceptable trade-off.
[13:54] <CIA-31> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger at gmx.de>
[14:25] <CIA-31> ffmpeg: 03Michael Niedermayer 07master * r73da4fd520 10ffmpeg/tests/Makefile: 
[14:25] <CIA-31> ffmpeg: Revert "buildsystem: add $EXESUF to the used tool for fate"
[14:25] <CIA-31> ffmpeg: This reverts commit 668a74e808dad06b15549ed05aa3a879f76c17c3.
[14:25] <CIA-31> ffmpeg: 10l, this change was unneeded, target_exec can be used to achive the same.
[14:25] <CIA-31> ffmpeg: 03Peter Ross 07master * r69224582da 10ffmpeg/libavcodec/iff.c: 
[14:25] <CIA-31> ffmpeg: iff: support masking plane feature
[14:25] <CIA-31> ffmpeg: Fixes ticket 705.
[14:25] <CIA-31> ffmpeg: Reported-by: ami-stuff
[14:25] <CIA-31> ffmpeg: Signed-off-by: Peter Ross <pross at xvid.org>
[14:25] <CIA-31> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[14:53] <CIA-31> ffmpeg: 03Stefano Sabatini 07master * r4ee03fef53 10ffmpeg/libavfilter/vsrc_testsrc.c: lavfi/rgbtestsrc: make alpha plane opaque
[14:53] <CIA-31> ffmpeg: 03Stefano Sabatini 07master * rbb4c0abc7c 10ffmpeg/ffplay.c: 
[14:53] <CIA-31> ffmpeg: ffplay: mark i variable in toggle_full_screen() as unused
[14:53] <CIA-31> ffmpeg: Fix gcc warning:
[14:53] <CIA-31> ffmpeg: ffplay.c: In function toggle_full_screen:
[14:53] <CIA-31> ffmpeg: ffplay.c:2792:9: warning: unused variable i [-Wunused-variable]
[15:00] <KuroiTsuki> umm, someone from mplayer/mencoder team told me that someone from ffmpeg team don't want to fix issues when embedded subs of mkv are staying for longer than they are supposed to be in converted file. why ffmpeg team don't want to fix this issue? or i should ask this on mailing list?
[15:09] <saste> KuroiTsuki: ask on the mailing list, as for what i know we are happy to fix things (especially if there are patches for that)
[15:09] <saste> KuroiTsuki: or you may simply open a ticket on trac
[15:10] <KuroiTsuki> let me quote
[15:10] <KuroiTsuki> Yes, that is the issue that the maintainer (in FFmpeg) didn't want to have fixed.
[15:10] <KuroiTsuki> i guess  i go report on https://ffmpeg.org/trac/ffmpeg
[15:11] <KuroiTsuki> thanks for advice
[15:11] <saste> np
[15:11] <JEEB> just check if the issue itself isn't already there :P
[15:11] <JEEB> by your "quote" it would seem as if it is already there
[15:16] <KuroiTsuki> that's why i asked why ffmpeg team don't want to fix it :)
[15:24] <JEEB> you never pointed out what exact issue it is .-.
[15:24] <JEEB> you just pointed that it was supposedly an issue of type X
[15:24] <JEEB> but you didn't link the issue on the issue tracker or anywhere else
[15:27] <KuroiTsuki> when embedded subs of mkv are staying for longer than they are supposed to be in converted file
[15:28] <KuroiTsuki> pretty much explains issue :)
[15:28] <KuroiTsuki> and i didn't link to issue on tracker for the sole reason i've never see it there. i just quoted what person of mplayer team told me
[15:30] <JEEB> well, people usually officially comment on issues on trackers etc.
[15:30] <JEEB> anyways, get used to actually trying to find the issue that you're wishing to get fixed instead of just putting random rambling in .-.
[15:30] <KuroiTsuki> that's what i'm doing  atm :)
[15:31] <JEEB> not to mention that if there was some negative comment from the developers' side, that would be in there as well
[15:31] <JEEB> that is important
[15:41] <KuroiTsuki> i can't find it =/ shoild report it then?
[15:42] <JEEB> yes
[15:45] <KuroiTsuki> okay
[16:11] <durandal_1707> how to make audio filter ashowinfo to output something?
[16:23] <NapHtaKeRoSene> hi
[16:23] <NapHtaKeRoSene> will gcc optimize out the return overhead from my putpixel if i don't use the return value? (if i want it to return pixel offset)
[17:12] <saste> durandal_1707: you need to grep stderr
[17:12] <saste> better solutions are welcome
[17:19] <durandal_1707> saste: how? why it does not work like video variant?
[18:59] <NapHtaKeRoSene> what do you think will gcc optimize this for me?
[18:59] <NapHtaKeRoSene> (point2d_int_t){ ( ( (ofs) - (s) -> pixels ) % ( (s) -> pitch ) ) / (bpp), ( ( (ofs) - (s) -> pixels ) - ( ( (ofs) - (s) -> pixels ) % ( (s) -> pitch ) ) ) / ( (s) -> pitch ) }\
[19:54] <CIA-31> ffmpeg: 03Michael Niedermayer 07master * r719a576f04 10ffmpeg/tools/lavfi-showfiltfmts.c: 
[19:54] <CIA-31> ffmpeg: lavfi-showfiltfmts: fix typo that segfaulted haiku.
[19:54] <CIA-31> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[20:38] <CIA-31> ffmpeg: 03Reimar Döffinger 07master * rea3abcd58f 10ffmpeg/libavformat/riff.c: 
[20:38] <CIA-31> ffmpeg: Fix ff_get_guid for short reads or errors.
[20:38] <CIA-31> ffmpeg: Current code would just return uninitialized data with no way
[20:38] <CIA-31> ffmpeg: to detect this condition.
[20:38] <CIA-31> ffmpeg: Instead, fill the whole GUID with 0 in that case.
[20:38] <CIA-31> ffmpeg: Fixes valgrind uninitialized data errors in fate-seek-lavf_asf.
[20:38] <CIA-31> ffmpeg: Signed-off-by: Reimar Döffinger <Reimar.Doeffinger at gmx.de>
[22:15] <CIA-31> ffmpeg: 03Carl Eugen Hoyos 07master * red10ca4c76 10ffmpeg/libavcodec/pngenc.c: Support gray8a png encoding.
[22:15] <CIA-31> ffmpeg: 03Carl Eugen Hoyos 07master * rf94ce3bcab 10ffmpeg/libavcodec/pngenc.c: Support rgb48 and rgba64 png encoding.
[22:15] <CIA-31> ffmpeg: 03Carl Eugen Hoyos 07master * rfdc013da78 10ffmpeg/libavcodec/pngenc.c: Support gray16 png encoding.
[22:15] <CIA-31> ffmpeg: 03Carl Eugen Hoyos 07master * r107f55cb01 10ffmpeg/libavcodec/pngenc.c: Simplify RGB32 png encoding.
[00:00] --- Mon Jan 16 2012


More information about the Ffmpeg-devel-irc mailing list