[Ffmpeg-devel-irc] ffmpeg-devel.log.20130707

[burek]

[00:00] <cone-826> ffmpeg.git 03Michael Niedermayer 07master:9a6e814be7c0: oggparseskeleton: avoid header parsing failure
[01:54] <cone-826> ffmpeg.git 03Michael Niedermayer 07master:abce6dfd9ece: avcodec/x86/vp3dsp_init: move mmx functions under HAVE_MMX_INLINE
[01:54] <cone-826> ffmpeg.git 03Michael Niedermayer 07master:707b2135fda3: libavcodec/x86/mpegvideo: Move mmx functions under HAVE_MMX_INLINE
[01:54] <cone-826> ffmpeg.git 03Michael Niedermayer 07master:104dfaadaa0b: dct_unquantize_h263_intra_c: fix assert condition
[02:03] <llogan> what's the status/plan of ffmpeg-devel at mplayerhq.hu? i noticed carl still sends to it, so it still works.
[02:06] <Compn> is it the original mail addy ?
[02:06] <Compn> i dont remember ...
[02:06] <llogan> i believe so
[02:06] <Compn> well then we leave it working in case someone wants to respond to a super old mail :P
[02:07] Action: Compn has no idea really
[02:07] <Compn> is there any reason to get rid of it ?
[02:08] <llogan> i can't think of any good enough reasons.
[02:08] Action: llogan is also ignorant
[04:17] <cone-826> ffmpeg.git 03Michael Niedermayer 07master:dc2a13aa802f: avformat/mov: Fix duration of fragmented mov
[10:59] <cone-757> ffmpeg.git 03Luca Barbato 07master:8f689770548c: kmvc: use fixed sized arrays in the context
[10:59] <cone-757> ffmpeg.git 03Michael Niedermayer 07master:1aeddc3cbfe7: Merge commit '8f689770548c86151071ef976cf9b6998ba21c2a'
[11:20] <cone-757> ffmpeg.git 03Luca Barbato 07master:4e7f0b082d8c: kmvc: Clip pixel position to valid range
[11:20] <cone-757> ffmpeg.git 03Michael Niedermayer 07master:2adf8a9c09c9: Merge commit '4e7f0b082d8c4b360312216b9241bec65ff63b35'
[11:35] <cone-757> ffmpeg.git 03Luca Barbato 07master:73142e753381: fifo: K&R formatting cosmetics
[11:35] <cone-757> ffmpeg.git 03Diego Biurrun 07master:6516632967da: tests: Only run noproxy test if networking is enabled
[11:35] <cone-757> ffmpeg.git 03Michael Niedermayer 07master:7bd417c22a90: Merge commit '6516632967da5e6bd7d6136e8678f826669ed26e'
[11:42] <cone-757> ffmpeg.git 03Diego Biurrun 07master:cd7b6deeaaf1: tools: Wording and formatting cosmetics
[11:42] <cone-757> ffmpeg.git 03Luca Barbato 07master:a4baf96abd64: avprobe: rename the prettyprint structures
[11:42] <cone-757> ffmpeg.git 03Michael Niedermayer 07master:b7ce51000327: Merge commit 'a4baf96abd64cda68e3e4692b182bdf60745a14d'
[11:53] <cone-757> ffmpeg.git 03Luca Barbato 07master:f5e646a00ac2: wtv: Mark attachment with a negative stream id
[11:53] <cone-757> ffmpeg.git 03Michael Niedermayer 07master:b9d55a07f3cd: Merge remote-tracking branch 'qatar/master'
[12:01] <cone-757> ffmpeg.git 03Michael Niedermayer 07master:a71832f826bb: avformat/Makefile: remove stray tab
[12:27] <superware> michaelni: hi
[12:27] <michaelni> superware, hi
[12:30] <superware> Kyle (zeranoe) is asking "Can you explain to me what aviocat does?"
[12:54] <superware> michaelni: I know you're busy, but let me ask when will you have a few minutes to receive the udp stream itself?
[12:55] <michaelni> superware, sure but whats the point if i dont have hours time to debug it ?
[12:57] <michaelni> aviocat: "a tool that uses avio to read and write, doing a plain copy of data" (from log)
[12:57] <superware> yeah, I just told him
[13:00] <superware> Since the TS itself is playing normally, and since the stream source is sending quite large datagrams, the bug is in the way ffmpeg is handling incoming data/buffer. I don't think hours will be required, I just think it worth a shoot if you might be able to spot the problem in a few minutes.
[13:20] <superware> michaelni: what are you suggesting?
[13:21] <superware> by the way, I really appreciate your help so far.
[13:22] <cone-757> ffmpeg.git 03Michael Niedermayer 07master:4cc32e5cad98: avformat/mp3dec: read xing toc independant of usetoc, only skip filling index if requested
[13:22] <cone-757> ffmpeg.git 03Michael Niedermayer 07master:19d416333114: avformat/mp3dec: read TOC even if not all needed information is available
[13:24] <michaelni> superware, like i said i dont really have time but if you want to send that udp stream i dont mind
[13:25] <superware> great, can I please /msg you?
[13:25] <michaelni> sure
[16:25] <cone-757> ffmpeg.git 03Michael Niedermayer 07master:c7cfd4cce98a: avformat_find_stream_info: use fps_dts as last resort to determine duration
[16:25] <michaelni> superware, should be fixed
[16:35] <superware> michaelni: you found it?!!
[16:45] <michaelni> superware, yes
[16:45] <superware> thanks, well done! where's the problem in code?
[16:51] <superware> do you still need the stream? :)
[16:55] <michaelni> superware, no you can stop the stream
[16:56] <superware> ok. if you'll require it further to test the patch, please tell me. can you direct me to the problematic code
[16:56] <superware> ?
[17:05] <michaelni> superware, see last commit
[17:24] <superware> ok, thanks
[17:29] <durandal_1707> what? ffmpeg doesn't suppport 16bit PCM DVD audio?
[18:59] <cone-757> ffmpeg.git 03Michael Niedermayer 07master:be25614177ef: mmsh: reimplement seeking
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/0.10:8514e3e08e23: rmdec: Pass AVIOContext to rm_read_metadata()
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/0.10:c25c89a53095: mpegts: only reopen pmt_cb filter if its different from the previous.
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/0.10:e445dc9237ae: avformat/mov: Fix duration of fragmented mov
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/0.11:139005452d0f: rmdec: Pass AVIOContext to rm_read_metadata()
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/0.11:cbbfdcf9ee1c: mpegts: only reopen pmt_cb filter if its different from the previous.
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/0.11:ec882942ce45: avformat/mov: Fix duration of fragmented mov
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/0.7:dc61d4484269: mpegts: only reopen pmt_cb filter if its different from the previous.
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/0.8:c997dcd38be7: mpegts: only reopen pmt_cb filter if its different from the previous.
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/0.9:0ca658a71659: rmdec: Pass AVIOContext to rm_read_metadata()
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/0.9:d6641c0bab21: mpegts: only reopen pmt_cb filter if its different from the previous.
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.0:5fa940d9c71e: rmdec: Pass AVIOContext to rm_read_metadata()
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.0:ee829bacdeb5: mpegts: only reopen pmt_cb filter if its different from the previous.
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.0:8c5c4adebbf6: libavcodec/x86/mpegvideo: Move mmx functions under HAVE_MMX_INLINE
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.0:e68b62a014b6: avformat/mov: Fix duration of fragmented mov
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.1:93fc80f8bfb3: avcodec/x86/dsputil_init: only use xvid idct for lowres=0
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.1:af95e174c5d9: rmdec: Pass AVIOContext to rm_read_metadata()
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.1:b186a5d08cde: mpegts: only reopen pmt_cb filter if its different from the previous.
[19:13] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.1:25ed0f05fdc3: libavcodec/x86/mpegvideo: Move mmx functions under HAVE_MMX_INLINE
[19:14] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.1:ba8d68462228: avformat/mov: Fix duration of fragmented mov
[19:14] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.2:99036565ca8b: avcodec/x86/dsputil_init: only use xvid idct for lowres=0
[19:14] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.2:b88bea1d6c8a: rmdec: Pass AVIOContext to rm_read_metadata()
[19:14] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.2:cb340ecd7d96: mpegts: only reopen pmt_cb filter if its different from the previous.
[19:14] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.2:5effd65312dd: avcodec/x86/vp3dsp_init: move mmx functions under HAVE_MMX_INLINE
[19:14] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.2:5ce57e0248ec: libavcodec/x86/mpegvideo: Move mmx functions under HAVE_MMX_INLINE
[19:14] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.2:627772d9886f: avformat/mov: Fix duration of fragmented mov
[19:57] <Daemon404> sure is commit spam 
[19:58] <durandal11707> 8 concurent releases
[20:06] <durandal11707> you must had lobotomy
[20:36] <Daemon404> durandal_1707, +1
[20:36] <Daemon404> ive been complaining about it for ages
[21:09] <durandal11707> wtf: https://trac.ffmpeg.org/attachment/ticket/2255/%C3%89N1.jpg
[21:11] <durandal11707> michaelni: can you remove this "art"
[21:11] <durandal11707> and ban that uploader with lack sense of humor
[21:11] <Daemon404> whats wrong with it?
[21:11] <durandal11707> its not ffmpeg art
[21:11] <Daemon404> i guess i lack context
[21:12] <durandal11707> look at it....
[21:12] <Daemon404> yes but an image along withotu context means nothing
[21:12] <Daemon404> s/along/alone/
[21:12] <Daemon404> the ticket provides context
[21:13] <Daemon404> which i have now read
[21:13] <durandal11707> Daemon404: do you want pose for ffmpeg banner?
[21:15] <Daemon404> im flattered
[21:15] <Daemon404> how much does it pay?
[21:18] <durandal11707> hmm, gonna ask saste
[21:21] <durandal11707> michaelni: can same psnr code be used for audio?
[21:22] <michaelni> in principle yes
[21:22] <durandal11707> i'm thinking this filter could be improved a lot, added simd and made slice threading support, so it could be used in fate
[21:22] <Daemon404> i dislike that
[21:22] <Daemon404> it creates a hard dep on lavfi for FATE
[21:23] <durandal11707> lavfi everywhere!
[21:49] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/0.11:e452e4ca5cdb: mmsh: dont close context on seeking failure
[21:49] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.0:40c521aec0ed: mmsh: dont close context on seeking failure
[21:49] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.1:ce74b92c0981: mmsh: dont close context on seeking failure
[21:49] <cone-757> ffmpeg.git 03Michael Niedermayer 07release/1.2:430ef8a7166b: mmsh: dont close context on seeking failure
[23:39] <BBB> Daemon404: I think some people consider lavfi the best thing since sliced bread
[23:40] <Daemon404> indeed
[23:40] Action: Daemon404 continue his 4000-commit bisect into ffmpeg
[23:41] <BBB> I bet you the resulting diff will be a massive merge commit
[23:41] <BBB> gl figuring that out
[23:41] <Daemon404> wouldnt be the first time
[23:41] <Daemon404> im using a work server w/ 32 cores and 100 gb of ram... speeds things up a bit
[23:42] <Daemon404> but configure is still dog slow
[23:42] <Daemon404> whole lib builds instancely, except for dsputils
[23:42] <Daemon404> which sits and crunches
[23:42] <Daemon404> instantly*
[23:55] <Daemon404> 1e8b9738fa70e20967ddb542d2f9d5552fc51ec6 is the first bad commit
[23:55] <Daemon404> commit 1e8b9738fa70e20967ddb542d2f9d5552fc51ec6
[23:55] <Daemon404> Author: Hendrik Leppkes <h.leppkes at gmail.com>
[23:55] <Daemon404> Date:   Fri Mar 22 15:13:42 2013 +0100
[23:55] <Daemon404> what the fuck
[23:55] <Daemon404> this doesnt make sense
[23:55] <Daemon404> this is from libav
[23:55] <Daemon404> and ths p roblem is ffmpeg-specific
[23:55] <Daemon404> goddamn merge shit
[23:56] <wm4> lol
[23:57] <Daemon404> >_<
[23:58] <wm4> I'm amazed that it still goes so well
[23:58] <wm4> such a big codebase, maintained separately for such a long time
[00:00] --- Mon Jul  8 2013