[Ffmpeg-devel-irc] ffmpeg-devel.log.20121010

burek burek021 at gmail.com
Thu Oct 11 02:05:02 CEST 2012


[02:53] Action: Daemon404 builds clang ioc
[03:35] <Daemon404> make: *** No rule to make target `doc/../tests/fate_config.sh.template', needed by `doc/fate.html'.  Stop.
[03:35] <Daemon404> make: *** Waiting for unfinished jobs....
[03:35] <Daemon404> ...
[03:35] <Daemon404> i know ive seen this mentioned somewhere before
[04:08] <Daemon404> man this thing needs a ton of ram
[04:33] <burek> are we supposed to handle these kind of cases where some project X is based on ffmpeg and when an issue pops up in that project, they redirect users to our support: http://ffmpeg.gusari.org/viewtopic.php?f=12&t=697&p=1043
[04:33] <burek> what is the usual practice in such cases anyway?
[04:34] <Compn> if its our bug
[04:34] <Compn> they should report a sample to us and follow proper bug report guidelines
[04:34] <burek> "I posted on the Serviio forum, and it was suggested that I post on the ffmpeg forum, hence why I did!"
[04:34] <burek> now, his ffmpeg log is clear
[04:35] <Compn> oh
[04:35] <Compn> pebcak :P
[04:35] <Compn> he needs to use ps3mediaserver or something
[04:35] <Compn> :P
[04:35] <Daemon404> the known issue is that we simply do not support dts-hd
[04:36] <burek> oh yes, durandal11707 was speaking about that recently..
[04:36] <Compn> hes trying to reinvent ps3mediaserver i think?
[04:36] <Compn> just point him in right direction and hope he steers that way :)
[04:46] <Compn> oh 
[04:46] <Compn> is servillo the thing like ps3mediaplayer ?
[04:46] <Compn> er ps3mediaserver
[04:46] <Compn> maybe i misunderstood his report
[04:47] <Compn> looks like he has a mkv which stops after 3 minutes unless he uses -ac 2
[04:47] <Compn> maybe we need his sample to reproduce
[04:50] <burek> well, yes, he said it partially works
[04:50] <burek> I redirected him to the bug tracker
[04:56] <cone-919> ffmpeg.git 03Michael Niedermayer 0725f44b041c42: x86/dsputil_mmx: change assert() to av_assert() * 03http://tinyurl.com/9xbujvh03
[07:25] <ubitux> <@Daemon404> make: *** No rule to make target `doc/../tests/fate_config.sh.template', needed by `doc/fate.html'.  Stop. // the file was moved, 'must be a .d lying
[07:25] <ubitux> distclean should solve the problem :p
[07:42] <Daemon404> ubitux, right.
[09:29] <ubitux> meh
[09:29] <ubitux> i don't remember the trick to specify protocols options...
[09:36] <ubitux> oh my bad i think i get it
[09:54] <cone-946> ffmpeg.git 03Paul B Mahol 07b4e516e30e70: pcx: use meaningful return values * 03http://tinyurl.com/9m2rtaz03
[10:10] <durandal_1707> j-b: does libdca handle xxch?
[10:10] <j-b> of course not.
[10:10] <j-b> libdca handles core and core
[10:11] <j-b> and core
[10:11] <j-b> did I mention core-only?
[10:11] <durandal_1707> looks like libav one have more and ffmpeg one even more
[10:13] <j-b> yes
[10:13] <j-b> this is not really news
[10:13] <durandal_1707> so why vlc still use libdca?
[10:14] <j-b> I thought FFMpeg has X96 and both have XCh, or so
[10:14] <j-b> durandal_1707: SPDIF
[10:18] <merbanan> libdca fails to decode the lfe channel also
[10:18] <merbanan> it is only silence
[11:09] <nevcairiel> ffmpeg also has XBR and XXCH now
[11:10] <nevcairiel> full support for HD HRA :)
[11:11] <j-b> nevcairiel: does it?
[11:12] <j-b> nevcairiel: never new that
[11:12] <j-b> knew
[11:12] <nevcairiel> only thing missing is XLL for lossless mode (also known as HD MA)
[11:13] <durandal_1707> and lbr and x96
[11:13] <j-b> lbr is not really important, IMVHO
[11:13] <nevcairiel> oh right it doesnt do X96 yet
[11:13] <nevcairiel> lbr is only used on secondary audio on blu-rays, nice to have, but not crucial, imho
[11:26] <j-b> nevcairiel: weren't there 2 different ways to store some extensions in the stream, btw?
[11:27] <nevcairiel> yes
[11:27] <nevcairiel> you can have a core+extensions, usually called dts-es
[11:27] <nevcairiel> or is DTS HD, which has a HD header, and extensions after that
[11:27] <nevcairiel> but core can only have x96 or xch, really
[11:28] <j-b> nevcairiel: but XCH and X96 can be in dts-es, no?
[11:29] <nevcairiel> yeah
[11:29] <nevcairiel> so you can have 6.1 DTS-ES
[11:29] <nevcairiel> or even 96Khz
[11:32] <j-b> nevcairiel: and FFmpeg supports XCH in both core and ext?
[11:32] <nevcairiel> yes
[11:33] <nevcairiel> and lately, even XXCH, which usually only appears in HD extensions
[11:33] <nevcairiel> have a 7.1 DTS HD HRA file which plays fine now :)
[11:34] <j-b> nevcairiel: cool
[12:22] <burek> this is absurd
[12:23] <burek> cehoyos first flames me in one email about me not motivating people to submit tickets+samples
[12:23] <burek> so that developers can investigate more on users' issues
[12:23] <burek> instead of just showing them ways to workaround stuff so that it just works for them
[12:23] <burek> (losing the opportunity to send user to bug trac to upload samples of media)
[12:23] <burek> and when I did start to redirect them
[12:24] <burek> then his attitute (again and again) drives people crazy
[12:24] <burek> somebody should really tell him not to close any bug reports on the first reply
[12:24] <burek> because it is really rude to cut off people like that
[12:25] <burek> who have spent their time to register and create a new ticket, which takes time most of people don't have
[12:25] <nevcairiel> yeah he can be quite annoying on the bug tracker
[12:25] <burek> so at least that time and will should be respected and not just ignored by simply closing tickets
[12:27] <burek> I mean, after all, ffmpeg team needs those reports, people dont care.. they will either fix things, pay someone to that for them or wait for sime time until the bug is fixed
[12:27] <burek> so no big deal for them if they dont report bugs
[12:27] <burek> but its a huge deal for ffmpeg if we loose those reports
[12:40] <pettter> Apparently, MPEG streams can be enclosed in MOV files
[12:41] <av500> yes
[12:41] <pettter> codec tag " m1s"..
[12:51] <divVerent>     .long_name         = NULL_IF_CONFIG_SMALL("FFM (FFserver live feed)"),
[12:51] <divVerent> oops, sorry
[13:05] <cone-946> ffmpeg.git 03Duncan Salerno 07a6363e3d89e7: url: Don't treat slashes in query parameters as directory separators * 03http://tinyurl.com/8k5wh7s03
[13:05] <cone-946> ffmpeg.git 03Duncan Salerno 07eea003814cc5: url: Handle relative urls being just a new query string * 03http://tinyurl.com/9e683wd03
[13:05] <cone-946> ffmpeg.git 03Duncan Salerno 0733893e6abcdc: url: Handle relative urls starting with two slashes * 03http://tinyurl.com/8oya5u703
[13:05] <cone-946> ffmpeg.git 03Martin Storsjö 077bc433b36dbb: fate: Add tests of the ff_make_absolute_url function * 03http://tinyurl.com/95mvl3q03
[13:05] <cone-946> ffmpeg.git 03Janne Grunau 073fbda309e553: avcodec: free extended_data instead address of it * 03http://tinyurl.com/9rwjtyx03
[13:05] <cone-946> ffmpeg.git 03Janne Grunau 078dd0650fe673: rtpdec_mpeg4: au_headers is a single array, simple av_free is enough * 03http://tinyurl.com/9h8uk4t03
[13:05] <cone-946> ffmpeg.git 03Diego Biurrun 07ac56ff9cc9d4: build: non-x86: Only compile mpegvideo optimizations when necessary * 03http://tinyurl.com/94dh2tb03
[13:05] <cone-946> ffmpeg.git 03Justin Ruggles 0723d53c547391: atrac1: use planar sample format * 03http://tinyurl.com/8wymev903
[13:05] <cone-946> ffmpeg.git 03Justin Ruggles 07c1a9cfd1feb2: mace: use planar sample format * 03http://tinyurl.com/99pbnrd03
[13:05] <cone-946> ffmpeg.git 03Justin Ruggles 077e5f04503901: pcmdec: use planar sample format for pcm_lxf * 03http://tinyurl.com/9av8zho03
[13:05] <cone-946> ffmpeg.git 03Justin Ruggles 07176db0b8928a: adpcmdec: use planar sample format for adpcm_ima_qt * 03http://tinyurl.com/92qfd3h03
[13:05] <cone-946> ffmpeg.git 03Justin Ruggles 071b9ac7290868: adpcmdec: use planar sample format for adpcm_ima_wav * 03http://tinyurl.com/9xxgskc03
[13:05] <cone-946> ffmpeg.git 03Justin Ruggles 074356d66d784c: adpcmdec: use planar sample format for adpcm_4xm * 03http://tinyurl.com/8osnn6503
[13:05] <cone-946> ffmpeg.git 03Justin Ruggles 078b854283c346: adpcmdec: use planar sample format for adpcm_ima_ws for vqa version 3 * 03http://tinyurl.com/8cxa5uk03
[13:05] <cone-946> ffmpeg.git 03Justin Ruggles 07d2b6ae02aa4d: adpcmdec: use planar sample format for adpcm_xa * 03http://tinyurl.com/9pa62yw03
[13:05] <cone-946> ffmpeg.git 03Justin Ruggles 07da9620e8e593: adpcmdec: use planar sample format for adpcm_ea_r1/r2/r3 * 03http://tinyurl.com/9ck2k2k03
[13:05] <cone-946> ffmpeg.git 03Justin Ruggles 07327cdb04e354: adpcmdec: use planar sample format for adpcm_ea_xas * 03http://tinyurl.com/9ptkwdl03
[13:05] <cone-946> ffmpeg.git 03Justin Ruggles 074ebd74cec75c: adpcmdec: use planar sample format for adpcm_thp * 03http://tinyurl.com/8wah7oj03
[13:05] <cone-946> ffmpeg.git 03Justin Ruggles 07cbcd497f384f: adxdec: use planar sample format * 03http://tinyurl.com/9vtmtfe03
[13:05] <cone-946> ffmpeg.git 03Michael Niedermayer 07eadba3e94daa: Merge commit 'cbcd497f384f0f8ef3f76f85b29b644b900d6b9f' * 03http://tinyurl.com/8d7wcdx03
[13:06] <j-b> planar!
[13:06] <durandal_1707> vlc borken
[13:07] <j-b> yep
[13:12] <cptspiff> i've said that for years.
[13:13] <j-b> cptspiff: unlike your software...
[13:13] <durandal_1707> xbmc supports planar sample formats?
[13:15] <cptspiff> lol. take the lame joke for what it is
[13:16] <nevcairiel> i still use rather crude small C snippets to interleave the audio again, i should add native support to my own audio buffers and let avresample do it :d
[13:16] <cptspiff> vlc is the best rat oriented video player there is. i just don't like the rat, so if anything, i'm broken.
[13:24] <cone-946> ffmpeg.git 03Mans Rullgard 070fb3b24adac1: build: link test programs only against static libs * 03http://tinyurl.com/8gvj3mz03
[13:24] <cone-946> ffmpeg.git 03Janne Grunau 0753e122dd4afb: swfenc: error out for more than 1 audio or video stream * 03http://tinyurl.com/93f8r3k03
[13:24] <cone-946> ffmpeg.git 03Janne Grunau 074ffbe3f3a5d9: matroskaenc: check cue point validity before reallocation * 03http://tinyurl.com/8ejpe8t03
[13:24] <cone-946> ffmpeg.git 03Janne Grunau 0718ff4d20201a: avconv: simplify memory allocation in copy_chapters * 03http://tinyurl.com/9fy7q4n03
[13:24] <cone-946> ffmpeg.git 03Janne Grunau 071afd7a118fd7: af_channelmap: free old extended_data on reallocation * 03http://tinyurl.com/95pa4p703
[13:24] <cone-946> ffmpeg.git 03Janne Grunau 07714f5ab59780: vc1dec: prevent memory leak on av_realloc error * 03http://tinyurl.com/9rkqhjv03
[13:24] <cone-946> ffmpeg.git 03Janne Grunau 076f8ef5320f4d: vc1dec: prevent memory leak in error path * 03http://tinyurl.com/9bwefs803
[13:24] <cone-946> ffmpeg.git 03Janne Grunau 078501c098687b: af_amix: prevent memory leak on error path * 03http://tinyurl.com/93dgfbz03
[13:24] <cone-946> ffmpeg.git 03Janne Grunau 07ac9a89562adc: af_resample: unref out_buf when avresample_convert returns 0 * 03http://tinyurl.com/948q5gh03
[13:24] <cone-946> ffmpeg.git 03Janne Grunau 07b94e4acb4874: cmdutils_read_file: increment *size after writing the trailing \0 * 03http://tinyurl.com/9odn5ya03
[13:24] <cone-946> ffmpeg.git 03Michael Niedermayer 0750b5477616c8: Merge commit 'b94e4acb4874843e914fd3cb8e089aff0756bb4a' * 03http://tinyurl.com/95xz54c03
[13:26] <burek> durandal11707, have you been working something on dts-hd lately?
[13:26] <durandal11707> why?
[13:26] <burek> if so, can you please take a look at this topic http://ffmpeg.gusari.org/viewtopic.php?f=12&t=697
[13:27] <nevcairiel> i havent had a DTS-HD file yet which didnt play properly
[13:27] <nevcairiel> of course, in most cases it only plays the core
[13:27] <nevcairiel> but it usually works just fine
[13:27] <durandal11707> .dtshd files
[13:28] <cone-946> ffmpeg.git 03Paul B Mahol 071470c8a9a3a7: takdec: remove redundant/wrong avio_tell() * 03http://tinyurl.com/9tlq4xs03
[13:28] <cone-946> ffmpeg.git 03Paul B Mahol 07eb71f027f058: dtshd: remove redundant/wrong avio_tell() * 03http://tinyurl.com/8rgobs303
[13:28] <nevcairiel> that person is playing a mkv :p
[13:28] <burek> that's what he says
[13:36] <durandal11707> burek: DTS-HD have limited support, OP in this case have DTS-HD-MA file which is not supported but core should still work, in other words all 6 chans should play without problems
[13:39] <durandal11707> burek: i see nowhere that OP claim that issue is in ffmpeg, so clarification is needed...
[13:40] <burek> durandal11707, ok, I was just asking if that's what you were working on, if he creates the ticket, so that you know what is it related to
[13:40] <pettter> are there provisions for restarting parsing etc. with a different format?
[13:43] <durandal11707> burek: DTS-HD-MA is not for streaming...
[13:45] <burek> durandal11707, what do you suggest (if ffmpeg can actually decode that dts-hd stream), which format should he use? aac?
[13:47] <durandal11707> burek: he is not transcoding (which works) but he wants to stream it as is
[13:48] <durandal11707> and nothing in topic points that issue is in ffmpeg
[13:49] <durandal11707> burek: and just to be sure - ffmpeg can decode dts-hd
[13:49] <burek> ok, I'll try to get some more log outputs from him, to make sure what exactly is the issue
[13:49] <burek> oh, I should correct my reply then
[13:50] <burek> somebody did suggest that ffmpeg doesn't
[13:50] <burek> [04:35] <Daemon404> the known issue is that we simply do not support dts-hd
[13:50] <burek> :)
[13:50] <j-b> who is gusari.org, btw?
[13:50] <ubitux> burek
[13:51] <ubitux> eeeh regression in webm encode!
[13:51] <ubitux> (libvpx or libvorbis)
[13:52] Action: ubitux bisects
[13:53] <burek> j-b, gusari.org is just used as a domain hoster, to avoid registering new domains just for ffmpeg related stuff
[13:53] <durandal11707> burek: that just not true, there are extensions which are part of dts-hd which are supported and there are some which are not
[13:54] <burek> when we (if) decide to create an official ffmpeg forum at forum.ffmpeg.org, I'll shut that subdomain down
[13:54] <burek> durandal11707, I've added a reply about that, asking him for more logs
[13:54] <burek> I just wanted to be sure if I understand what the actual problem was
[13:56] <cone-946> ffmpeg.git 03Anton Khirnov 07fb722a900fc5: avconv: remove -same_quant * 03http://tinyurl.com/8lp6ozf03
[13:56] <cone-946> ffmpeg.git 03Michael Niedermayer 07b7ebb49d0352: Merge commit 'fb722a900fc5cc9e003b9fef25b27ed7fc5547a2' * 03http://tinyurl.com/9v6owxf03
[14:10] <cone-946> ffmpeg.git 03Anton Khirnov 07233a5a807e76: lavc: split asv12 encoder/decoder * 03http://tinyurl.com/8u74zxz03
[14:10] <cone-946> ffmpeg.git 03Janne Grunau 0779e6e8eba203: avfilter: correct memcpy size avfilter_copy_buf_props() * 03http://tinyurl.com/8mul2oa03
[14:10] <cone-946> ffmpeg.git 03Janne Grunau 074a7c0c455547: http: use av_strlcpy instead of strcpy() without size checks * 03http://tinyurl.com/8npupt903
[14:10] <cone-946> ffmpeg.git 03Janne Grunau 07f1de23faaa61: g722enc: fix size argument in memset * 03http://tinyurl.com/9n5yxzt03
[14:10] <cone-946> ffmpeg.git 03Janne Grunau 07d5ef9354ce01: rtspdec: use av_strlcpy for writing into fixed size buffer * 03http://tinyurl.com/9fshz9l03
[14:10] <cone-946> ffmpeg.git 03Diego Biurrun 076d0beefbf6ee: swscale: Do not make ff_ symbols globally visible. * 03http://tinyurl.com/8m9dqkj03
[14:10] <cone-946> ffmpeg.git 03Michael Niedermayer 07cd6f5c4895c4: Merge commit '6d0beefbf6ee6dbf8efb522a9307e54c6ed5f702' * 03http://tinyurl.com/8c56f3403
[14:11] <ubitux> michaelni: ping
[14:12] <ubitux> michaelni: 2c34367b4a17856584b3e8b64cefa1900342ebcd broke webm encoding (output unplayable with firefox for instance)
[14:15] <ubitux> to reproduce: ./ffmpeg -f lavfi -i 'testsrc[out0]; aevalsrc=sin(440*2*PI*t)[out1]' -t 5 -y out.webm
[14:15] <michaelni> ubitux, does -avoid_negative_ts 1 help ?
[14:15] <ubitux> michaelni: yes
[14:16] <michaelni> then add it to the write header of webm (and probably mkv) (could be under if(avoid_negative_ts<0))
[14:18] <cone-946> ffmpeg.git 03Diego Biurrun 07096a5d76a547: fate: Refactor setting of environment variables for groups of tests * 03http://tinyurl.com/8r93uhn03
[14:18] <cone-946> ffmpeg.git 03Mans Rullgard 074b895cb294d6: build: sanitize linking of tools and test programs * 03http://tinyurl.com/9o2bypd03
[14:18] <cone-946> ffmpeg.git 03Michael Niedermayer 07c6d39fb3c5a7: Merge remote-tracking branch 'qatar/master' * 03http://tinyurl.com/8ug7a3b03
[14:19] <ubitux> michaelni: ok will try, thx
[14:28] <ubitux> michaelni: ffplays it fine btw
[14:28] <ubitux> -s
[14:28] <ubitux> wow.
[14:28] <ubitux> ffplay plays* it fine
[14:30] <divVerent> I wonder how avealsrc=tan(440*2*PI*t) would sound, if using floating point samples ;)
[14:31] <pettter> is there any comprehensive list of defined ISOM/FourCC tags, apart from the one in isom.c?
[14:32] <nevcairiel> for mp4 specifically? there probably is an official list somewhere
[14:32] <pettter> well, I was hoping for more general .mov
[14:32] <nevcairiel> well ISOM is mp4, not mov :d
[14:33] <pettter> true enough, my bad
[14:34] <ubitux> divVerent: ./ffplay -f lavfi -i 'aevalsrc=tan(440*2*PI*t),aformat=sample_fmts=flt'
[14:34] <ubitux> like this
[14:34] <divVerent> my hardware can't do this ;)
[14:34] <divVerent> seriously, I am just calculating its fourier coefficients
[14:36] <divVerent> apparently, its fourier series nowhere converges ;)
[14:36] <MP4_maniac> pettter: http://www.mp4ra.org/codecs.html
[14:37] <pettter> MP4_maniac: thanks, but the tag I'm looking for isn't there..
[14:37] <durandal11707> what tag you are looking for?
[14:38] <pettter> " m1s"
[14:38] <pettter> I'm currently checking if I can share the sample
[14:38] <nevcairiel> what codec is that supposed to be?
[14:39] <MP4_maniac> there are many unofficial fourcc in the world :P
[14:39] <pettter> it looks like it's not a codec per se, but a complete MPEG stream (haven't checked if it's ES/PS/TS yet)
[14:39] <nevcairiel> ew, i know of those files
[14:39] <nevcairiel> they should be burned
[14:39] <nevcairiel> with a passion
[14:39] <av500> there is also MPEG TS in asf
[14:40] <av500> TS can go anywhere :)
[14:40] <nevcairiel> you can get lucky if you just open it wit ha mpeg-ts demuxer
[14:40] <MP4_maniac> i cant find m1s in headers in xcode and quicktime
[14:40] <nevcairiel> because they are meant to re-sync when the sync bytes get lost
[14:41] <ubitux> michaelni: the tests change, i'm not sure if it makes sense: http://b.pkh.me/0001-lavf-mkv-avoid-negative-ts-by-default.patch
[14:41] Action: MP4_maniac guesses " m1s" is one of unofficial fourcc
[14:41] <pettter> MP4_maniac: really? I'm told the file opens in both QT and WMP
[14:44] <pettter> it's actually a PS stream
[14:47] <Compn> pettter : i made a list of isom for mp4 ...
[14:48] <Compn> from quicktime
[14:48] <Compn> but it was a few years ago
[14:48] <Compn> pettter : does it play with mplayer -vc +ffmpeg12 yourfile ?
[14:48] Action: Compn digs up list
[14:49] <pettter> heh, I don't actually have mplayer installed
[14:49] <Compn> http://wiki.multimedia.cx/index.php?title=QuickTime_container
[14:49] <Compn> doh
[14:50] <Compn> this site lists m1s as a isom
[14:50] <Compn> http://wiki.awkwardtv.org/wiki/QT_Codec_Information
[14:50] <Compn> M1S MPEG component
[14:50] <Compn> google for "m1s" imdc to get good results
[14:50] <Compn> imdc is the name of the quicktime decoder
[14:52] <pettter> Compn: thanks!
[14:52] <Compn> pettter : is it mpeg1 or mpeg2 ? 
[14:53] <Compn> i can just add it to ffmpeg now...
[14:53] <Compn> if you cant share sample
[14:53] <Compn> altho we love having samples :)
[14:53] <pettter> mpeg1 video
[14:53] <MP4_maniac> so mpeg-1 video in mpeg-2 ps?
[14:53] <pettter> format_long_name=MPEG-PS format
[14:54] <nevcairiel> its not mpeg video, its supposedly a mpeg program stream
[14:54] <nevcairiel> you cant justh ook that up to the mpeg decoder
[14:54] <Compn> is it .mov or ?
[14:54] <pettter> it's a .mov originally yes
[14:54] <Compn> m1s in .mov
[14:54] <Compn> heh
[14:55] <pettter> stripping the MOV header (990 bytes) gives me a parseable MPEG-PS stream
[14:55] <pettter> with mpeg1video and mp2 audio
[14:55] <Compn> does ffplay -f mpeg1video yourfile work ?
[14:55] <nevcairiel> ffmpeg -f mpeg -i file.mov? :p
[14:55] <michaelni> ubitux, probably makes sense ... 
[14:56] <pettter> nevcairiel: indeed
[14:57] <pettter> so it's a matter of recognising the tag and restarting the parsing with the proper format?
[14:57] <nevcairiel> i dont think anything like that is supported at this point
[14:57] <Compn> ffmpeg is setup so that you have to add .mov tags to libavformat/isom.c
[14:58] <pettter> yeah, that much I know, but as it's a format and not a codec..
[14:58] <Compn> but i'm not sure if adding it there will fix it
[14:58] <Compn> yeah
[14:58] <Compn> sounds like dv-in-avi stuff
[14:58] <Compn> they kind of jammed a format in another format
[14:58] <pettter> myeah
[15:00] <ubitux> michaelni: ok, submitted, i'll wait a bit for more comments
[15:00] <ubitux> thx :)
[15:11] <cone-946> ffmpeg.git 03Paul B Mahol 0703cc52a07d51: takdec: fix seeking * 03http://tinyurl.com/8osje4k03
[15:17] <michaelni> ubitux, if there are no comments then the patch should be ok (/me drops it from my que of still needing a review patches)
[15:17] <ubitux> michaelni: i keep that in mind
[15:17] <ubitux> i'll push tomorrow or after tomorrow then
[15:18] <ubitux> unless you want to fix quick
[15:18] <michaelni> no hurry
[15:51] <pettter> Compn, nevcairiel: http://titan.codemill.se/~peteri/scarecrowdance2.mov
[15:51] <pettter> that's the sample
[16:00] <Compn> ehe, mplayer plays it fine
[16:01] <pettter> huh, fancy that
[16:01] <Compn> it does report BUG! Invalid demuxer type in new_demuxer(), big troubles ahead.
[16:01] <Compn> :)
[16:01] <pettter> :)
[16:01] <pettter> well, ffplay/ffmpeg doesn't
[16:01] <Compn> yes, its a sample that needs some fixing
[16:03] <pettter> and there's no Easy(tm) fix, is there?
[16:14] <MP4_maniac> pettter: apparently, that sample is indeed mpeg-ps. one sample is mapped to one mpeg-ps stream. MPC MPEG splitter can split it lol
[16:15] <pettter> MP4_maniac: lovely
[16:16] <pettter> MP4_maniac: should I make a ticket?
[16:19] <MP4_maniac> pettter: better than nothing
[16:20] <pettter> right-o
[16:26] <MP4_maniac> that mov sample has only one track http://up-cat.net/p/0b5aed3f . so the track has two separate stream (audio and video). this is similar with dv-in-avi i think
[16:29] <pettter> sounds reasonable
[16:30] <pettter> oh dear, and that is implemented by keeping a DVDemuxContext* inside the AVIContext?
[16:49] <cone-946> ffmpeg.git 03Matthieu Bouron 07f94edfe484e9: vf_idet: remove some unnecessary statement and unused variable * 03http://tinyurl.com/9f54ows03
[16:49] <cone-946> ffmpeg.git 03Michael Niedermayer 07229e33a2b6de: avidec: override sample size of 1024 for VBR AAC * 03http://tinyurl.com/8dx2wr303
[17:02] <ubitux> michaelni: av_log(s, AV_LOG_DEBUG, "overriding sample_size\n", ast->dshow_block_align); looks wrong
[17:03] <maker> how am i going to use ld for using the development version of ffmpeg? I have a foo.c file using libavsample, and after setting the correct PKG_CONFIG_PATH I get this error: /usr/bin/ld: cannot find -lavresample
[17:03] <maker> I tried to remap LD_PATH, but probably I should do also $ ldconfig. Is it going to break something?
[17:04] <ubitux> yay animated gif patch \o/
[17:04] <pettter> LD_LIBRARY_PATH maybe?
[17:04] <ubitux> is avresample built? (--enable-avresample)
[17:05] <durandal_1707> not and should not
[17:05] <maker> same error, note that I have not done anything with $ ldconfig. 
[17:06] <maker> ubitux: yay
[17:06] <ubitux> or you could just use libswresample.
[17:07] <maker> ubitux: probably not, I was looking at the trac for some issues, and found one notifying a memoryleak there. 
[17:09] <ubitux> is Vitaliy Sugrobov on IRC?
[17:09] <cone-946> ffmpeg.git 03Michael Niedermayer 0773ad355d23ab: avidec: fix 10l typo * 03http://tinyurl.com/8n2m2ws03
[17:53] <cone-946> ffmpeg.git 03Paul B Mahol 07ee109c6bc2cf: dtshd: fix seeking * 03http://tinyurl.com/8msoeun03
[18:27] <Daemon404> durandal_1707, do not push things with useless commit messages while blatantly ignoring replies.
[18:36] <Daemon404> michaelni, just fyi, ffmpeg passes clang-ioc
[18:36] <Daemon404> ill set up an ioc vm once my ram arrives this week
[18:36] <Daemon404> and once baptiste gets my key approved...
[18:42] <michaelni> Daemon404, great, thx
[18:50] <cone-946> ffmpeg.git 03Paul B Mahol 072c5b1a2a6abb: txd: remove duplicated header inclusion * 03http://tinyurl.com/9c4rwu303
[18:51] <cone-946> ffmpeg.git 03Michael Niedermayer 0792c3173c84a9: qt-faststart: fix printf argument type * 03http://tinyurl.com/8nkrugp03
[18:51] <cone-946> ffmpeg.git 03Michael Niedermayer 079586db6ae504: lavc: docs: the field_order is set by the demuxer and not libavcodec for decoding * 03http://tinyurl.com/8wkuq2v03
[19:14] <burek> can someone help me with some h264 raw video from logitech c920 camera
[19:14] <burek> a guy wrote a tool to capture the h264 stream from the web cam
[19:14] <burek> https://github.com/csete/bonecam/blob/master/capture/capture.c#L498
[19:14] <burek> im just not sure how to pipe it to ffmpeg
[19:15] <burek> something like ./capture -o | ffmpeg -f ? -r ? -i - ...
[19:17] <relaxed> burek: -o - 2>/dev/null| ffmpeg -f h264 -vcodec h264 ?
[19:17] <burek> let me try
[19:18] <relaxed> you may also need framerate, framesize and pix_fmt
[19:18] <burek> http://pastebin.com/R2BnEbYC
[19:19] <burek> ill update ffmpeg just to be sure
[19:19] <relaxed> add -r and -s
[19:20] <burek> i dont know where to see -r
[19:20] <burek> -s would be 640x480 i guess
[19:22] <relaxed> output a sample from ./capture (without piping) and analyze the stream
[19:22] <burek> I have a sample video.raw if it will help
[19:22] <burek> :)
[19:26] <burek> this is with the latest ffmpeg: http://pastebin.com/QGSAZC99
[19:41] <relaxed> -s $framesize -pix_fmt yuv420p
[19:42] <Daemon404> you should not need -s for an annexb strea,
[19:43] <burek> Option video_size not found.
[19:43] <burek> it does not allow -s before -i
[19:43] <Daemon404> i do not think its feasible to pipe h264 like that
[19:43] <burek> ./ffmpeg -f h264 -s 640x480 -pix_fmt yuv420p -vcodec h264 -i video.raw -c copy -f null -
[19:43] <Daemon404> it uses v4l2
[19:43] <burek> well c920 has its own internal h264 encoder (in the webcam)
[19:44] <Daemon404> the proper solution is to add h264 support to our v42l code
[19:44] <Daemon404> if it is not there yet
[19:44] <burek> so im just trying to pipe/mux that with ffmpeg and thats it
[19:44] <burek> yes, there was some talk about v4l2 not yet supporting h264 pix format
[19:44] <burek> something like that
[19:44] <burek> but i think its implemented in their latest head
[19:45] <burek> I can upload a sample
[19:45] <burek> just a sec
[19:45] <Daemon404> support needs to be added to out code too
[19:46] <burek> http://ffmpeg.gusari.org/uploads/sample-raw-h264-video-from-logitech-c920-webcam.raw
[19:46] <burek> first 10 MB of data
[20:03] <michaelni> burek, the raw can be played with ./ffplay -f rawvideo -pixel_format yuyv422 -s 640x480 -i ...
[20:04] <burek> oh great :)
[20:04] <burek> thanks :)
[20:12] <burek> michaelni, is that sample a rawvideo or h264 video or both :D
[20:13] <burek> i would expect less size with h264 for 200 frames
[20:14] <burek> it's raw..
[20:14] <burek> i forgot -f
[20:14] <Daemon404> lol
[20:19] <ubitux> is it normal to have an aweful quality with libvpx and -quality best?
[20:20] <ubitux> (and it's slow to encode)
[20:23] <ubitux> it looks like something fishy: http://lucy.pkh.me/libvpx/orig.mkvhttp://lucy.pkh.me/libvpx/encoded.webm
[20:24] <durandal_1707> i64?
[20:24] <ubitux> (encoded with ffmpeg -i orig.mkv -avoid_negative_ts 1 -quality best -q:a 7 encoded.webm)
[20:25] <ubitux> -quality best being a libvpxenc specific option
[20:25] <ubitux> durandal_1707 ?
[20:25] <ubitux> yes it's in a i64?
[20:27] <durandal_1707> ubitux: bisect
[20:27] <ubitux> yeah i guess so :(
[20:27] <durandal_1707> ubitux: you do not set bitrate?
[20:28] <ubitux> i don't, isn't it facultative with a "quality" setting?
[20:29] <durandal_1707> dunno, maybe it is merge bug
[20:32] <burek> michaelni, this was supposed to be the raw h264 output from the webcam: http://ffmpeg.gusari.org/uploads/sample-h264-video-from-logitech-c920-webcam.raw
[20:32] <burek> sorry :)
[20:33] <burek> when I run it like this: # ./capture -f -o | ffmpeg -f h264 -i - -f null -
[20:34] <burek> it doesn't start at all, but after I press ctrl+c, it says this http://pastebin.com/JKbp1bMH
[20:34] <ubitux> same in 1.0... same with -crf... :(
[20:35] <durandal_1707> ubitux: bisect from last time it worked, do you know when it was?
[20:35] <ubitux> i don't know if it ever worked
[20:36] <ubitux> i'm wondering if it's not a bad usage or something
[20:38] <michaelni> burek, -f h264
[20:39] <ubitux> i don't get it, the encode really is a lot slower with quality best
[20:39] <ubitux> but the video is still awful
[20:39] <Daemon404> ubitux, -crf works fine
[20:39] <burek> michaelni, it works like: ffmpeg -f h264 -i sample.raw ...
[20:39] <Daemon404> vimeo uses it
[20:39] <ubitux> Daemon404: what setting did you use?
[20:39] <burek> but it doesn't work like this: ./capture -f -o | ffmpeg -f h264 -i - -f null -
[20:40] <Daemon404> ubitux, -crf <crf val>
[20:40] <Daemon404> as seen by x264
[20:40] <Daemon404> generally 18-20
[20:40] <ubitux> Daemon404: yes right, but what value?
[20:40] <burek> -o just redirects output to stdout instead of file
[20:40] <ubitux> i've tried 15-18
[20:40] <ubitux> it's still awful
[20:40] <Daemon404> from what input
[20:40] <Daemon404> and preset
[20:40] <ubitux> i didn't set any preset
[20:40] <Daemon404> medium then
[20:40] <ubitux> 20:23:59 <@ubitux> it looks like something fishy: http://lucy.pkh.me/libvpx/orig.mkvhttp://lucy.pkh.me/libvpx/encoded.webm
[20:40] <ubitux> this input
[20:40] <Daemon404> >webm
[20:40] <Daemon404> why are you using h264 with webm
[20:41] <ubitux> ffmpeg -i orig.mkv -crf 18 -q:a 7 encoded.webm ?
[20:41] <Daemon404> thats vp8
[20:41] <Daemon404> nto x264
[20:41] <Daemon404> -_-
[20:41] <Daemon404> crf is x264.
[20:41] <ubitux> where did i said it was h264?
[20:41] <Daemon404> >crf
[20:41] <Daemon404> is an x264 option
[20:41] <ubitux> (libvpx also has a crf private option)
[20:41] <Daemon404> ubitux, well its abusing it
[20:41] <Daemon404> because avaik libvpx has no concept ofr 'crf'
[20:41] <Daemon404> of*
[20:41] <ubitux> {"crf", "Select the quality for constant quality mode", offsetof(VP8Context, crf), AV_OPT_TYPE_INT, {.dbl = 0}, 0, 63, VE},
[20:42] <ubitux> whatever, i tried to play with quality and crf
[20:42] <ubitux> and it doesn't help
[20:42] <Daemon404> thats not what crf is in x264
[20:42] <Daemon404> constant rate factor != constant quality
[20:42] <Daemon404> anyway, i cant help with vp8
[20:43] <Daemon404> because... lolvp8 who uses that
[20:47] <burek> wholly cannoli it works :)))
[20:47] <burek> thanks a lot michaelni :)))
[20:51] <burek> :beer: to relaxed and Daemon404 too, of course :)))
[20:51] <burek> this is awesome... cpu on my machine is 0% :)
[20:51] <burek> i can now use even raspbery :)
[21:47] <ubitux> using the bitrate seems to work
[21:47] <ubitux> but well..
[21:48] <cone-946> ffmpeg.git 03Michael Niedermayer 0794f5470a20d4: lavf: add a AVPROBE_SCORE_RETRY instead of using hardcoded values. * 03http://tinyurl.com/8gbjnua03
[21:48] <cone-946> ffmpeg.git 03Michael Niedermayer 07b47396b6143c: img2dec: detect .raw files only with a low score as img2 * 03http://tinyurl.com/9lur9mv03
[21:48] <cone-946> ffmpeg.git 03Michael Niedermayer 07317505b56691: lavf: do not prematurely accept a format with low score in init_input() * 03http://tinyurl.com/8uvudqb03
[22:53] <llogan> fuck. people are STILL thinking FFmpeg is dead due to that damned "This program is not developed anymore" message in ubuntu.
[22:54] Action: llogan leaves to pick up ppc machine for saste to monkey with.
[23:08] <ubitux> llogan: i proposed burek to write a wiki page about this 
[23:08] <ubitux> but he's not confortable with it
[23:08] <ubitux> comfortable*
[23:08] <ubitux> so feel free to write one
[23:08] <Daemon404> if they didnt care to do anything but read that line that it output
[23:09] <Daemon404> what makes you think theyd find said wiki page
[23:09] <Daemon404> also why do you care what some lazy person thinks
[23:09] <ubitux> what are they supposed to do?
[23:09] <ubitux> "oh ok ffmpeg is renaming the tool okay"
[23:10] <ubitux> now you're not on #ffmpeg, and regularly ppl are coming there "hey i have a problem with avconv" etc
[23:10] <ubitux> it would be easier to paste them the link to the wiki page
[23:10] <ubitux> with a brief summary
[23:10] <ubitux> instead of wasting time derping about ffmpeg, libav, debian etc
[23:11] <ubitux> like it happens half of the time
[23:12] <ubitux> Daemon404: i don't think you realize how much users are abused by that message :p
[23:13] <Daemon404> itll go away with 0.9
[23:13] <Daemon404> at least
[23:13] <Daemon404> (libav)
[23:13] <ubitux> lts and stuff will likely ship it for a while
[23:14] <ubitux> and it's just a fucking patch to apply to drop/reword that message
[23:14] <ubitux> they just refuse to change it... for obvious reasons.
[23:14] <iive> btw, there is debian bugzilla entry, and somebody (you) should post a patch to change this message with a better one.
[23:14] <ubitux> it's already done
[23:14] <ubitux> i'm not a debian user
[23:14] <Daemon404> i agree the message is overly vague
[23:14] <iive> siretart is really just dragging his feet on the formality that there is no patch.
[23:15] <iive> is it just proposal, or real patch?
[23:15] <ubitux> no it's not vague Daemon404, it's a pretty obvious lie :)
[23:15] <Daemon404> no it isnt
[23:15] <Daemon404> it's true
[23:15] <Daemon404> for libav.
[23:15] <Daemon404> which is what they package.
[23:15] <Daemon404> it's just misleadingly vague
[23:15] <ubitux> yes but it's packaged under what name? "ffmpeg"
[23:16] <iive> it is misleading at best and outright lie.
[23:16] <ubitux> and the sentence doesn't say it's only true in the small libav mind scope
[23:16] <Daemon404> thats for backwards compat with debian
[23:16] <Daemon404> and it is being changed
[23:16] <Daemon404> quite soon i hear
[23:16] <iive> ubitux: this too. you should ask Debian packagers to stop abusing your trademark.
[23:16] <Daemon404> ffmpeg package is beign dropped entirely.
[23:16] <ubitux> but in that situation, it's a lie Daemon404.
[23:16] <Daemon404> also
[23:16] <Daemon404> lol @ you all
[23:16] <Daemon404> so butthurt
[23:16] <ubitux> it's not the problem about "it's gonna change"
[23:16] <iive> Daemon404: it have been 2 years.
[23:16] <ubitux> it's been 2yr like this
[23:16] <ubitux> and it will stay for at least 2-3 yrs
[23:16] <Daemon404> talking to siretart, he's wanted to change it for a long time
[23:16] <ubitux> (lts & stuff)
[23:17] <Daemon404> but been held up by otehr package maintainers
[23:17] <ubitux> yeah
[23:17] <ubitux> lol
[23:17] <ubitux> please.
[23:17] <llogan> i didn't get that impression
[23:17] <iive> Daemon404: ffmpeg have been dropped from debian 2 years ago. without any discussion on the wim of a single developers.
[23:17] <Daemon404> i met and talked with him irl
[23:17] <Daemon404> so im pretty sure i got teh right impression
[23:17] <ubitux> Daemon404: he is the sole maintainer...
[23:17] <Daemon404> ubitux, OTHER package maintainers
[23:17] <Daemon404> who use the fucking package
[23:17] <Daemon404> asa dep
[23:18] <ubitux> why do they care about the message?
[23:18] <Daemon404> i was talkign about teh "ffmpeg" package name
[23:18] <Daemon404> which offends you so
[23:18] <iive> Daemon404: this is why you don't drop the original project when a fork appears.
[23:18] <ubitux> Daemon404: in that case, if it has to be kept as "ffmpeg" the message must be changed
[23:18] <Daemon404> i agree the message should be less vague and misleading
[23:18] <ubitux> because it's willingly misleading
[23:19] <Daemon404> i think you should stop peppering everythign with conspiratorial words
[23:19] <ubitux> it's been 2yr man
[23:19] <ubitux> and you're not doing user support
[23:19] <Daemon404> .. you are nto familiar with debian i see
[23:19] <iive> Daemon404: are you trying to say that we are over reacting?
[23:19] <Daemon404> yes
[23:20] <Daemon404> yes i am
[23:20] <iive> this is kind of insulting.
[23:20] <Daemon404> stop being butthurt
[23:20] <Daemon404> youre reminded me of mans and diego
[23:20] <ubitux> you don't understand Daemon404, this is a real problem
[23:20] <Daemon404> with your ffmepg tribalism
[23:20] <ubitux> we have to do avconv support on #ffmpeg most of the time
[23:20] <ubitux> and explain users
[23:20] <ubitux> it's a fucking problem for 2 years
[23:20] Action: llogan wouldn't say "most".
[23:20] <iive> Daemon404: then tell me, why is there no ffmpeg package in debian.
[23:21] <ubitux> it's not about butthurting
[23:21] <Daemon404> it really is
[23:21] <ubitux> it's a real pain in practice
[23:21] <Daemon404> every time libav makes a commit oyu dont like
[23:21] <Daemon404> there's this bitchign session
[23:21] <ubitux> that is another story
[23:21] <Daemon404> its the same thing.
[23:21] <llogan> all one all one all one, all one or none!
[23:21] <Daemon404> and the answer is really simple
[23:21] <Daemon404> "debian switched to libav. go to #libav."
[23:22] <Daemon404> end. of. story.
[23:22] <iive> it is the wrong approach altogether.
[23:22] <ubitux> this is not helping the users, and doesn't help us either
[23:22] <Daemon404> it is the correct approach
[23:22] <llogan> Daemon404: did you have a bad morning or something?
[23:22] <Daemon404> send them to teh channel of the people who maintain the fucking package
[23:22] <ubitux> (especially when only ffmpeg is solving their problem)
[23:22] <Daemon404> its simple logic
[23:22] <Daemon404> a 2 year old coudl do it
[23:22] <Daemon404> yes
[23:23] <ubitux> we care about our abused users :)
[23:23] <iive> no, it is not users fault that the most spread distribution have switched versions.
[23:23] <Daemon404> remember what i said about tribalism
[23:23] <Daemon404> and butthurt
[23:23] <nevcairiel> if they want ffmpeg, send them to a distro that has it available :P
[23:23] <Daemon404> ubitux, youre doign them a disservice.
[23:23] <ubitux> how?
[23:23] <ubitux> what am i doing wrong?
[23:23] <Daemon404> [17:22] <@Daemon404> send them to teh channel of the people who maintain the fucking package
[23:23] <ubitux> i believe they are happy about the services i provide
[23:23] <Daemon404> either you do this
[23:24] <Daemon404> or you stop bitching
[23:24] <Daemon404> because its your own damn fault
[23:24] <iive> Daemon404: actually it is the opposite of tribalism. We are helping their users too, because we are nice guys. And for this they may decide that they do want ffmpeg.
[23:24] <ubitux> now it's my fault...
[23:24] <ubitux> Daemon404: users installed "ffmpeg"
[23:24] <Daemon404> "i need help with this package"
[23:24] Action: llogan now blames ubitux for the dumb libav message
[23:24] <ubitux> i believe they wanted ffmpeg.
[23:24] <ubitux> llogan :DD
[23:24] <Daemon404> "we dont maintain this package but let em entertain you with our fork backstory before we try and help you with a different packages"
[23:25] <ubitux> i don't do that.
[23:25] <ubitux> i make them try the static build first
[23:25] <Daemon404> thats altogether wrong
[23:25] <ubitux> and.or help them with the avconv cmd line
[23:25] <Daemon404> if youre having them install or build ffmpeg
[23:25] <Daemon404> youre doing it wrong
[23:26] <ubitux> i'm not
[23:26] <ubitux> 23:25:29 <@ubitux> i make them try the static build first
[23:26] <iive> Daemon404: if you were been sold iPhone6 and go into apple store, and instead of kicking you, people there give you a real iphone5, would you be happy?
[23:26] <Daemon404> ^ tribalism
[23:26] <Daemon404> "ffmpeg is better so its ok"
[23:26] <ubitux> isn't it?
[23:26] <ubitux> :)
[23:27] <ubitux> if we're working on ffmpeg it's because we believe the project is better, same for libav...
[23:27] <iive> Daemon404: the whole point is, when you buy iPhone, you expect that it is Apple product. This is why you go to apple Store, not Samsung or Nokia
[23:27] <Daemon404> yes but users dont give a fuck
[23:27] <ubitux> exactly
[23:27] <Daemon404> so dotn waste their time tryign to convert them
[23:27] <Daemon404> send them to the proper place for THEIR package
[23:27] <ubitux> so i'm helping the users instead of making them go in the wrong direction
[23:27] <ubitux> aka: not to the project they didn't selected
[23:27] <Daemon404> ... you people
[23:27] <Daemon404> serious
[23:27] <Daemon404> for real?
[23:28] <ubitux> and yes i often point out the libav support
[23:28] <Daemon404> sending them to teh support channel for the teh fuckign package theyre using is wrong?
[23:28] <Daemon404> because the distro chose for them?
[23:28] <ubitux> if they insist on using the non-working avconv
[23:28] <Daemon404> what sort of fuckign convluted logic is this?
[23:28] <ubitux> ppl come to #ffmpeg to get help on the ffmpeg package they installed
[23:28] <ubitux> they ask why it's deprecated often
[23:28] <nevcairiel> but its not a ffmpeg package :)
[23:28] <Daemon404> ^
[23:28] <ubitux> or just use avconv
[23:28] <ubitux> nevcairiel: yes, because it's a lie
[23:29] <Daemon404> ugh
[23:29] <nevcairiel> make them go complain at debian or ubuntu why their ffmpeg package installs libav
[23:29] <Daemon404> sanity has no place in libav or ffmpeg
[23:29] Action: Daemon404 goes off for a whi;e
[23:29] <ubitux> nevcairiel: that's generally in the panel of options i provide
[23:30] <ubitux> "try this cmd line" -> "try static build" -> if it works: build/install it (!wiki), and/or complain to your distro + long explanations
[23:30] <ubitux> this is why a wiki page with a summary of these steps would be welcome.
[23:31] <ubitux> if it works with the avconv, well then be it, i don't want to enter in long explanations, unless they ask
[23:31] Action: michaelni upgraded debian and read a message that debian is deprecated and not maintained anymore and that i should use gentoo instead, after asking on the debian lists iam beint told that my package is now maintained by gentoo and i should ask for support there  aha
[23:31] <ubitux> :D
[23:31] <iive> this may be a good 1'st april joke
[23:32] <ubitux> oh again a deprecated message user who doesn't understand on the ml
[23:32] <ubitux> Daemon404: i don't think you understand how much pain it is to deal with this
[23:32] <j-b> use Windows :)
[23:33] <iive> ubitux: i think he understands, and his position is that we shouldn't deal with it at all.
[23:34] <llogan> back to the message...the "original" said: "This program is not developed anymore and is only provided for compatibility. Use avconv instead (see Changelog for the list of incompatible changes)"
[23:34] <iive> "5B G5;>25:, =5B ?@>1;5<0" as Stalin likes to say.
[23:34] <llogan> the new says: "*** THIS PROGRAM IS DEPRECATED ***\n "This program is only provided for compatibility and will be removed in a future release. Please use avconv instead."
[23:34] <ubitux> haha
[23:34] <ubitux> much better!...
[23:34] <ubitux> caps!
[23:34] <llogan> anyone know which ubuntu versions, if any, still use the old one?
[23:35] <iive> indeed. much better. at least now it means that this program may be still developed.
[23:35] <ubitux> the message should be "This program is deprecated in the Libav project, an alternative to the FFmpeg project"
[23:36] <llogan> i tried to get some wording like that, but it was not accepted. i'll try again.
[23:37] <iive> honestly, we should first force debian to stop package libav as ffmpeg
[23:37] <ubitux> i can understand they have some technical issues with the name
[23:37] <iive> they are quite sensitive to IP violations and we can make a huge fuss about it.
[23:37] <iive> this is not our problem.
[23:37] <llogan> iive: i don't know if that will cause dependent programs to hate us.
[23:37] <ubitux> but refusing to change the message is a blatant hostile move
[23:38] <iive> the easiest solution would be to package ffmpeg.
[23:38] <iive> i mean, the real ffmpeg.
[23:38] <ubitux> ffmpeg is deprecated iive !
[23:38] <iive> trademarks exists to prevent THIS exact situation.
[23:39] <iive> that is, somebody smearing your good name.
[23:39] <burek> you should stop wasting your time on libav.. just find a decent lawyer and sue them for copyright infringement and take some money from that
[23:39] <llogan> i'm too poor to need money.
[23:39] <burek> they don't care about changing that message, since they came up with the same, for a good reason
[23:40] <iive> what copyright infringement?
[23:40] <ubitux> it's easier in the debian policy to just add a patch changing the message
[23:40] <burek> but since they are a fork, they are not allowed (legally) to use your brandings, names, copyrighted stuff and trade marks (like FFmpeg or ffmpeg)
[23:40] <ubitux> "This program is deprecated in the Libav project, an alternative to the FFmpeg project" really sounds perfect for that...
[23:40] <Compn> we should just ask debian to stop packaging ffmpeg
[23:40] <burek> especially not in a context "Copyright by Libav developers team"
[23:40] <nevcairiel> libav doesnt even have a ffmpeg anymore, debian is either shipping a rather old version or they re-added the app for compat reasons :p
[23:40] <Compn> :P
[23:41] <Compn> nevcairiel : they cant just axe ffmpeg binary, too many programs _require_ it
[23:41] <ubitux> nevcairiel: can you say that for stable/unstable/testing and the various ubuntu releases?
[23:41] <burek> the point is libav shouldn't (and mustn't) use the name ffmpeg anywhere in their fork
[23:41] <iive> they can always make a symlink named ffmpeg.
[23:41] <llogan> ok. my next step is to see what ubuntu versions use the old really crappy message, if any, and re-visit the bug report to get it re-worded.
[23:41] <burek> no matter if it breaks the whole damn debian
[23:42] <Compn> burek : mail fabrice
[23:42] <nevcairiel> looks like it was released after the last release, libavs last major release is quite old already
[23:42] <ubitux> the problem doesn't lie in libav anymore burek 
[23:42] <nevcairiel> er
[23:42] <nevcairiel> removed after the last release
[23:42] <ubitux> burek: they don't have ffmpeg for a long time now
[23:42] <ubitux> it's just in the debian packaging
[23:42] <ubitux> for retro compat issues and stuff
[23:42] <ubitux> that is EXACTLY why we are pushing to fix that message
[23:42] <burek> Compn, you guys are way longer in this whole business to suggest me to send an e-mail to Fabrice :)
[23:43] <ubitux> i believe Fabrice is aware and don't want to mess with that
[23:43] <burek> well, then why are we even discussing it anymore :)
[23:43] <burek> if he doesn't care, why should we?
[23:43] <ubitux> because we do user support
[23:43] <Compn> no one wants to mess with it
[23:43] <Compn> :P
[23:44] <ubitux> arh i'm tired :(
[23:44] <burek> I would mess with it a lot, if it was my project
[23:44] <Compn> so email fabrice then
[23:44] <Compn> dont matter if you are new or old
[23:44] <burek> those guys showed who they are by the way they were acting
[23:44] <Compn> fabrice is nice guy
[23:44] <ubitux> burek: remember, libav can't do much now to change that
[23:44] <Compn> i dont think i've ever actually talked to him...
[23:44] <burek> well, it's a kinda insult that a newbee cares more about the project, then its developers who are far longer in all this..
[23:44] <ubitux> burek: the problem is now totally on debian-like distro side
[23:45] <burek> so, I'm gonna skip sending that email
[23:45] <ubitux> burek: and that mean the packager(s?) are just refusing to fix the message
[23:45] <nevcairiel> the debian package maintainer for that package is a libav dev, so there is that
[23:45] <j-b> Fabrice is aware
[23:45] <ubitux> is he really a "dev"?
[23:45] <nevcairiel> well some sort of dev anyway
[23:45] <burek> you should've sue debian community for changing the official ffmpeg source code with their lie
[23:45] <ubitux> but anyway, there is nothing to change on libav side
[23:45] <ubitux> except the maintainer(s)
[23:46] <llogan> i'll revisit the "message" issue, so hopefully i can resolve that, and then i'll try to get ffmpeg package to actually be FFmpeg in Ubuntu+. or at least see what they have to say.
[23:46] <iive> fabrice holds the ffmpeg.org domain too, so the restoration of the project under this name happened with his blessing.
[23:46] <ubitux> llogan: would you mind proposing my message so they just have to do a copy/paste? :)
[23:46] <llogan> yes. a copy/paste lazyproof is in mind.
[23:46] <iive> He is not active ffmpeg developer at the moment, so there is no reason for him to engage into legal conflicts on his own.
[23:48] <iive> but there is that thing with the trademarks that if you don't protect them, you can loose them.
[23:48] <iive> So, writing a page with explanation, then publishing a news on the homepage would be a good start.
[23:49] <Compn> burek : also is good faith to give fork team some time to migrate
[23:49] <burek> Compn, they gave you a fair fight for you to give them anything? :)
[23:49] <Compn> burek : thats the point, we are nice people
[23:49] <burek> I would just ddos them for a month to return the favor of taking over the servers like a bunch of kids
[23:50] <iive> 1 year should be enough for such elite coders like them.
[23:51] <iive> burek: ddos-ing service provider is quite lame and may not be entirely legal. We should not drop to their level.
[23:51] <burek> anyway, I gave up on that topic, because I'm too fresh and newbee to even discuss those things.. you guys do whatever you feel like it's right
[23:51] <llogan> ok, i should actually do some meatspace work today...
[23:51] <burek> iive I agree, I'm just comparing things
[23:52] <iive> we have given them enough time to sort out the problems.
[23:52] <iive> I want to know if everybody agree that we cannot tolerate the situation in its current state.
[23:53] Action: ubitux just wants to fix the message
[23:54] <ubitux> there is not a single technical problem to do that
[23:54] <ubitux> (in comparison to renaming the package, or trying to distribute ffmpeg in addition to ffmpeg or other solutions)
[23:54] <iive> there is one, it should be done by the person who wrote the original message.
[23:55] <ubitux> mmh that message comes from libav come, and in the libav scope it's "true"
[23:55] <ubitux> and it got droppped from their repo so who care
[23:55] <ubitux> again, the problem is in the packaging policy
[23:56] <ubitux> anyway, unrelated: no one to comment/review the metadata patchset?
[00:00] --- Thu Oct 11 2012


More information about the Ffmpeg-devel-irc mailing list