[Ffmpeg-devel-irc] ffmpeg-devel.log.20121024

[burek]

[00:00] <llogan> lazy ass! that's isn't going to be good, is it?
[00:03] <Compn> heh
[00:10] <durandal_1707> saste: lol, that patch you pinged
[00:11] <saste> durandal_1707, ?
[00:13] <durandal_1707> you pinged patch which cannot be applied any more
[00:14] <saste> durandal_1707, no i just had to skip an hunk
[00:14] <saste> i was lazy to send an updated patch
[00:15] <durandal_1707> ok
[00:15] <saste> i was tempted to just push, but since that's not code which I "maintain" i preferred to get an explicit approval from someone else
[00:34] <cone-265> ffmpeg.git 03Stefano Sabatini 075ffadec30aaa: lavc/utils: propagate or return meaningful error codes in avcodec_open2()
[00:34] <cone-265> ffmpeg.git 03Stefano Sabatini 074a227a706d47: lavc/utils: fix a few grammar/style nits in log messages
[00:34] <cone-265> ffmpeg.git 03Stefano Sabatini 0733c4535db735: examples/decoding_encoding: add and use the decode_write_frame function
[00:51] <cone-265> ffmpeg.git 03Michael Niedermayer 071856162caa91: mpegvideo enc: choose VBV buffer size for mpeg1/2 when none is specified instead of failing
[00:51] <cone-265> ffmpeg.git 03Michael Niedermayer 073556e7ce737e: mpegvideo enc: choose VBV buffer size for mpeg4 when none is specified instead of failing
[01:33] <llogan> i don't see any commercial stuff on the projects page with my lazy search. are these also allowed to be listed? i don't see why not.
[01:33] Action: llogan is going through backlog of requests
[01:46] <llogan> and what's the consensus on old/abandoned/homeless stuff?
[01:47] <Compn> we generally dont link to commercial stuff much
[01:47] <Compn> abandoned stuff, update it to archive.org url
[01:49] <llogan> hhmm...i'm not going to be an old link farmer. i'd prefer to dump the shit.
[01:51] <llogan> if we don't list commercial stuff "much" the site should say so.
[01:56] <llogan> im wondering if the moldering list is even worth "maintaining" (not that it has much maintenance). how much value does it really have?
[01:57] <cone-265> ffmpeg.git 03Michael Niedermayer 076c87b2be7257: swscale_unscaled: remove unused variables
[01:57] <cone-265> ffmpeg.git 03Michael Niedermayer 07a494792961a0: yuv2rgb: fix integer overflow in fill_table
[01:57] <cone-265> ffmpeg.git 03Michael Niedermayer 07cccb45751e93: yuv2rgb: fix integer overflow in fill_gv_table()
[02:45] <Compn> llogan : well, the point is, who maintains a list of projects based on ffmpeg ?
[02:45] <Compn> its just us
[02:45] <Compn> theres no wikipedia list
[02:45] <Compn> maybe someone wants an ffmpeg based (whatever) , where should they look ?
[02:46] <Compn> i mean, if you dont want to maintain it, thats fine
[02:46] <Compn> how about maintaining a list of recommended projects? or what is your idea ?
[02:47] <iive> llogan: if the project have vanished, then dump the link. if the project is still accessible, but not changed for a long time, sink it down the list.
[02:48] <iive> linking to commercial stuff is sort of advertising, and this is not free.
[02:48] <Compn> its alpha betical iive
[02:48] Action: Compn wonders if iive seen ffmpeg project page
[02:48] Action: iive looks around. Is diego still here?
[02:49] <Compn> iive : i dont think he comes here anymore ?
[02:50] <iive> he likes to make things in alphabetical order.
[03:12] <llogan> Compn: no, maintaing a list of "recommended projects" wasn't my idea, AFAIK.
[03:15] <llogan> your points are valid, but i don't think the list is important enough to maintain.
[03:17] <llogan> and the little attention the site does get can be better used elsewhere.
[03:22] <Compn> i agree
[03:22] <Compn> but theres no need to delete it , or any links
[03:24] <Compn> oh lawd , 3rd party debate is getting a rocky start
[03:24] <Compn> oop wrong window
[13:22] <cone-609> ffmpeg.git 03Mans Rullgard 073f65eff408f7: Include sys/time.h before sys/resource.h
[13:22] <cone-609> ffmpeg.git 03Mans Rullgard 07da0d0ae9a429: network: add fallbacks for INADDR_LOOPBACK and INET_ADDRSTRLEN
[13:22] <cone-609> ffmpeg.git 03Mans Rullgard 07c3e73100af9b: network: use getservbyport() only if available
[13:22] <cone-609> ffmpeg.git 03Mans Rullgard 07fab0a8b2c6ca: libm: add fallbacks for various single-precision functions
[13:22] <cone-609> ffmpeg.git 03Mans Rullgard 07a805cefd8b81: swscale: avoid pointless use of compound literals
[13:22] <cone-609> ffmpeg.git 03Mans Rullgard 0780521c1997a2: build: allow targets to specify extra objects to link with executables
[13:22] <cone-609> ffmpeg.git 03Michael Niedermayer 072dbc93455c14: Merge commit '80521c1997a23e148edf89e11b939ab8646297ca'
[13:22] <cone-609> ffmpeg.git 03Michael Niedermayer 0742ee9f398197: build: fix detection of math functions with 2 parameters.
[13:46] <cone-609> ffmpeg.git 03Mans Rullgard 074ebc6a74108f: build: Plan 9 support
[13:46] <cone-609> ffmpeg.git 03Mans Rullgard 072a222a7255af: fate-lavfi: replace sed/grep/cut combos with awk
[13:46] <cone-609> ffmpeg.git 03Diego Biurrun 07c2dac8ac7a1f: rtmppkt: Avoid unescaped backslash in Doxygen comment
[13:46] <cone-609> ffmpeg.git 03Luca Barbato 070c03cc683864: mp3: exit on parsing error in mp_decode_frame
[13:47] <cone-609> ffmpeg.git 03Michael Niedermayer 07f3b8096bc0e7: Merge commit '0c03cc68386443f1e96ab6fb358220faf67cd5ff'
[14:08] <cone-609> ffmpeg.git 03Luca Barbato 079ab0874ea8b6: mp3: properly forward mp_decode_frame errors
[14:08] <cone-609> ffmpeg.git 03Diego Biurrun 07aa91fe80913f: dxtory: Replace AV_WN16A(AV_RN16A()) combination by AV_COPY16
[14:08] <cone-609> ffmpeg.git 03Diego Biurrun 07af6dd6de08d9: intreadwrite: Add AV_COPYxxU macros for copying to/from unaligned addresses
[14:08] <cone-609> ffmpeg.git 03Diego Biurrun 07330b864cda95: dsputil: Replace AV_WNxx(AV_RNxx()) combinations by AV_COPYxxU
[14:08] <cone-609> ffmpeg.git 03Reinhard Tartler 07a9f0dce283f8: prepare 9_beta2 release
[14:08] <cone-609> ffmpeg.git 03Diego Biurrun 07ceb754d041f5: lzo: Use AV_COPY*U macros where appropriate
[14:09] <cone-609> ffmpeg.git 03Michael Niedermayer 075555d2075a26: Merge commit 'ceb754d041f5f6327fd9195a5f43575af9516daa'
[14:14] <cone-609> ffmpeg.git 03Mans Rullgard 07d82f18850441: Remove need for padding in av_memcpy_backptr()
[14:14] <cone-609> ffmpeg.git 03Michael Niedermayer 075cf794e0c1f7: Merge commit 'd82f188504410fdfa446c5682c128c31bb5851a4'
[14:25] <cone-609> ffmpeg.git 03Diego Biurrun 07e831b3b852a2: av_memcpy_backptr: Drop no longer necessary malloc padding
[14:25] <cone-609> ffmpeg.git 03Mans Rullgard 07a153e45b953d: dfa: use av_memcpy_backptr() where previously impossible
[14:25] <cone-609> ffmpeg.git 03Diego Biurrun 07d7a39b334079: doxygen: Build Doxygen documentation in the doc/ subdirectory
[14:25] <cone-609> ffmpeg.git 03Diego Biurrun 075e28e974339e: Move Doxyfile into the doc/ subdirectory
[14:25] <cone-609> ffmpeg.git 03Justin Ruggles 07a5ef830b1226: lavc: use the correct API version guard macro for avcodec_encode_audio()
[14:25] <cone-609> ffmpeg.git 03Justin Ruggles 0744d854a518f9: atrac3: return an error if extradata_size is not a specific known size
[14:25] <cone-609> ffmpeg.git 03Justin Ruggles 07c68317ebbe49: lavc: fix documentation for AVCodecContext.delay
[14:25] <cone-609> ffmpeg.git 03Michael Niedermayer 07719fde47ca64: Merge commit 'c68317ebbe4915035df0b08c23eea7a0b80ab881'
[14:43] <cone-609> ffmpeg.git 03Anton Khirnov 071bc64c2814d4: lavc: initialize output AVFrame before decoding.
[14:43] <cone-609> ffmpeg.git 03Anton Khirnov 078e84f2055c41: avconv: remove now unneeded calls to avcodec_get_frame_defaults().
[14:43] <cone-609> ffmpeg.git 03Anton Khirnov 075a9567631a90: g.723.1: add missing CODEC_CAP_DR1
[14:43] <cone-609> ffmpeg.git 03Anton Khirnov 07292a08cbab33: asfdec: cosmetics, reformat ff_asf_parse_packet()
[14:43] <cone-609> ffmpeg.git 03Michael Niedermayer 07d0ab71ed11c5: Merge remote-tracking branch 'qatar/master'
[15:30] <cone-609> ffmpeg.git 03Michael Niedermayer 07e377b50b2df9: matroskaenc: avoid implicit truncation in avio_w*
[15:30] <cone-609> ffmpeg.git 03Michael Niedermayer 07a3fb3a670e8c: avio: avoid implicit truncation in avio_w*
[15:38] <durandal_1707> LOOOOOOL!!!! http://lists.libav.org/pipermail/libav-devel/2012-October/038308.html
[15:40] <Compn> durandal_1707 : careful, Daemon404 doesnt like us complaining about libav. :P
[15:41] <durandal_1707> well commit is somehow correct, but picked name is not
[15:45] <ubitux> i'm not sure that's a good idea to stimulate the flames from other entities with such flag
[15:45] <ubitux> but it made me smile :)
[15:46] <thegeek> perhaps have a flag for "recommended" codecs instead? :)
[15:46] <ubitux> the idea is to stem the generation of bad formats/codecs
[15:46] <ubitux> files
[15:46] <av500> then IDIOTIC is perfect
[15:47] <av500> also, when such codec is used the output filename should have "_made_by_an_idiot_" added
[15:47] <durandal_1707> av500: you forgot tags
[15:47] <av500> yes
[15:47] <av500> and hardcodec SUBS
[15:47] <ubitux> av500: yes, but what isn't IMO is that it's completely subjective
[15:47] <av500> coded*
[15:48] <av500> ubitux: of course it is
[15:48] <Compn> why are the tags AV_CODEC_BLAH_BLAH_LOSSY ? cant it just be tags: LOSSY ?
[15:48] <av500> how would you objectively define an idiot?
[15:48] <ubitux> av500: and so if you start doing that, you will likely end up adding the flag for every existing codec/format
[15:48] <Compn> av_our_function_names_are_going_to_get_longer_in_the_future_with_these_ideas
[15:48] <ubitux> because everyone has grudge about everything
[15:49] <av500> we can maybe spare PCM and YUV :)
[15:49] <Compn> they dont compress well
[15:49] <ubitux> it's idiot not to compress them
[15:49] <Compn> er, what ubitux said
[15:49] <ubitux> another problem is that it's not justified
[15:49] <durandal_1707> also, i think that IDIOTIC flag is more deserver for containers/formats and less for codecs
[15:49] <ubitux> it should at least have a comment with a few arguments
[15:49] <ubitux> for each one having the flag
[15:53] <durandal_1707> using AV_CODEC_PROP_BAD and AV_CODEC_PROP_UGLY is better
[15:54] <ubitux> AV_CODEC_PROP_HATERSGONNAHATE
[15:54] <thegeek> michaelni: the bug report for the asm crash I mentioned yesterday can be found here: https://ffmpeg.org/trac/ffmpeg/ticket/1844
[15:55] <michaelni> where do i add AV_FORK_PROPERTY_IDIOTIC ?
[15:56] <av500> michaelni: wherever you want, you have commit rights
[15:56] <ubitux> AV_PROPERTY_IDIOTIC_FORK?
[15:56] <thegeek> :P
[15:57] <durandal_1707> michaelni: please feel free to add it wherever it really is, so it can be fixed/rewrited
[15:57] <ubitux> i'm not against that AV_CODEC_PROP_IDIOTIC, but the message should be changed
[15:57] <ubitux> to something like "this format is generally not recommended" etc
[15:58] <ubitux> anyway, let's see how the bikeshed ends up on libav side
[15:58] <ubitux> unless they all really hates everything, it might takes a while to get pushed
[15:59] <Compn> huh i didnt know google donated a server to libav
[16:00] <av500> its only a small data center
[16:00] <ubitux> haha
[16:00] <Compn> hope that corporate influence does not warp their project
[16:01] <av500> mozilla got 300million, libav one server ....
[16:01] <Compn> is it me, or is it hard to turn off google 'protection' in firefox now ?
[16:02] <Compn> i didnt see an option for it in the prefs, had to go about:config
[16:02] <cone-609> ffmpeg.git 03Michael Niedermayer 078a34b85fa8de: avio: detect implicit truncation by assert in avio_w*
[16:02] <Compn> previous versions used to ask if you wanted to allow google to know every single website you visited
[16:06] <av500> why would you know allow it, they are not evil
[16:07] <Compn> just basic computing courtesy
[16:07] <Compn> 'dont do things behind the users back without asking'
[16:07] <Compn> maybe i'll go ask mozilla whats what
[16:08] <Tjoppen> >previous versions used to ask if you wanted to allow google to know every single website you visited
[16:08] <Tjoppen> wut?
[16:09] <Compn> old firefox used to popup a warning about safe browsing
[16:10] <Tjoppen> I see
[16:11] <Tjoppen> switching off the auto-google on mistyped URLs took some figuring out
[16:11] <Tjoppen> and all other stupid shit firefox does nowdays, like hiding http://
[16:12] <Tjoppen> but it still does when doing auto-complete. no idea how to make it not hide them at all
[16:12] <Compn> http://www.mozilla.org/en/firefox/phishing-protection/
[16:12] <Compn> that was the url to explain , back in firefox 2/3 days
[16:13] <Tjoppen> ah. I went and disabled it
[16:13] <Tjoppen> though it doesn't actually send any addresses to google
[16:14] <Tjoppen> time to reply to that idiotic patch
[16:14] <av500> Compn: its easy, dont google for Yahoo, just type yahoo.com
[16:14] <av500> solved!
[16:16] <Compn> you use yahoo as the search engine ?
[16:16] <av500> but yahoo has all the itnernet
[16:17] <kierank> no bing does
[16:17] <kierank> duhg
[16:18] <Compn> Tjoppen: heres what it used to do > http://code.google.com/p/google-safe-browsing/wiki/Protocolv2Spec
[16:19] <Compn> i turn off autocomplete, autocomplete is for jerks
[16:20] <Compn> Tjoppen : er nm, bad url
[16:20] <merbanan> I search for porn sometimes just to make google think I'm a pervert
[16:21] <av500> google knows
[16:21] <Tjoppen> I hear bing is the search engine to use for porn
[16:21] <Tjoppen> "google for work, bing for porn"
[16:22] <saste> what if you work in porn?
[16:22] <av500> then you bing google
[16:23] <Compn> i wasnt looking at porn at work, that was for research!
[16:23] <av500> Compn: I was only looking at encoding artifacts....
[16:25] <saste> ubitux: I found the culprit of weird pod \, => , conversion
[16:25] <ubitux> !
[16:25] <saste> but i think we should avoid "\," in single filter examples
[16:26] <saste> well not sure about this
[16:26] <saste> if we do foo=ARGS
[16:27] <ubitux> :)
[16:27] <saste> we're recalling the filtergraph syntax, implying that the snippet can be used literally in a filtergraph
[16:27] <saste> anyway it is currently inconsistent
[16:27] <ubitux> i have another "escaping" problem if you like them, but not in the documentation
[16:27] <ubitux> btw, in the documentation i think we have some '@' issues as well
[16:28] <saste> not in the documentation?
[16:29] <saste> and today i did cd doc; make clean
[16:29] <saste> and discovered it is broken since months
[16:33] <michaelni> saste, btw as you memtion docs, i think the example path & patterns in the Doxygen file where forgotten when they where moved to doc/
[16:36] <michaelni> saste, if you want i can look into fixing it or you can ?
[16:36] Action: michaelni wonders why doxygen is so slow
[16:36] <saste> michaelni, should be  trivial change in examples_path and pattern
[16:37] <ubitux> <@saste> and today i did cd doc; make clean // was it ever supported?
[16:37] <ubitux> we don't support this for libs, tests, etc
[16:37] <saste> i never use doxygen (yeah one reason is that it is sloooow to build)
[16:37] <saste> ubitux, maybe don't
[16:37] <ubitux> <@saste> not in the documentation? // yes, you can't put a '\n' in the option parameter
[16:38] <saste> i have no idea what is the "option parameter"
[16:38] <ubitux> -foobar $(printf 'hello\n')
[16:38] <ubitux> the \n is dropped
[16:39] <ubitux> -foobar "$(printf 'hello\n')"
[16:39] <ubitux> i don't know what part of the code is doing this
[16:39] <ubitux> but it's a problem for some options
[16:39] <ubitux> like -headers in the http protocol
[16:39] <ubitux> since it requires you to end the string with \r\n
[16:40] <ubitux> (try -headers "$(printf 'X-foobar: hello\r\n')")
[16:43] <durandal_1707> what subtitles are supported in mp4?
[16:43] <ubitux> mov_text
[16:44] <ubitux> (that's the name in ffmpeg)
[16:45] <durandal_1707> and nothing else?
[16:45] <ubitux> i don't think so, at least in ffmpeg
[16:45] <durandal_1707> by mp4 spec?
[16:48] <JEEB> oh boy
[16:48] <JEEB> "MP4"
[16:49] <saste> ubitux: texi2html is not escaping @@
[16:49] <ubitux> :(
[16:49] <JEEB> there's basically 3gpp subtitles that can be muxed into "MP4"
[16:49] <JEEB> you could poke MP4_maniac for details
[16:50] <MP4_maniac> j“^
[16:50] <MP4_maniac> oRúW_
[16:50] <JEEB> lol
[16:51] <JEEB> anyways, it's a mess, but officially'ishly the thing supported is something that is colloquially called "3gpp timed text" or whatever
[16:51] <durandal_1707> so subrip is not supported?
[16:51] <JEEB> no, not officially
[16:52] <JEEB> of course people have made their own proprietary additions to the "MP4" container
[16:53] <JEEB> ("MP4" being what is governed by the ISO Media Container and the MPEG-4/3GPP/SMPTE specs, mainly)
[16:53] <MP4_maniac> 3gpp timed text is for "3GPP", streaming text format (14496-17) is for "MP4"
[16:53] <saste> michaelni, are you already testing Doxygen, or should I do it myself?
[16:53] <MP4_maniac> streaming text format != 3gpp timed text
[16:53] <michaelni> saste, iam testing but it does not work
[16:53] <JEEB> ok
[16:54] <JEEB> so there were two types
[16:54] <ubitux> <@durandal_1707> so subrip is not supported? // i wonder if the mov text encoder strip them&
[16:55] <ubitux> mmh it should actually, since it's based on the ass markup
[16:58] <MP4_maniac> JEEB: streaming text format uses timed text formats optimized(?) for streaming. one of formats is 3gpp timed text. others are reserved/undefined yet
[17:02] <saste> another problem is that non-ASCII chars are not correctly rendered
[17:03] <saste> e.g. the special "x" symbol, both in HTML and MAN output
[17:06] <saste> doxygen is taking ages and heating my PC
[17:06] <saste> what's the recommended command to run: doxygen Doxyfile?
[17:08] <saste> also it seems in loop, always printing the same warnings again and again
[17:12] <michaelni> saste, you can set SEARCH_INLCUDES NO to make it run fast
[17:14] <saste> now it's not producing any output...
[17:15] <saste> enough, I'll post a patch and let's see if someone will be able to test it
[17:15] <cone-609> ffmpeg.git 03Michael Niedermayer 077d0ca5b7e436: doxygen: disable header/footer/stylesheet
[17:15] <saste> ubitux: texi2html is a monster 23K perl script
[17:16] <saste> i managed to fix the @@ escape issue, will send a patch if i'll find the motivation
[17:19] <ubitux> have no fear of the monster
[17:19] <ubitux> fear is the mind killer
[17:22] <cone-609> ffmpeg.git 03Stefano Sabatini 0751467ee86846: doc/filters: remove unnecessary and inconsistent double ',' escaping
[17:22] <cone-609> ffmpeg.git 03Stefano Sabatini 0755c77a0ca3ce: doc/filters: escape special Texinfo character @ in overlay example
[17:45] <cone-609> ffmpeg.git 03Tomas Härdin 07187630b2449f: mxfdec: Fix CID 732262
[17:45] <cone-609> ffmpeg.git 03Tomas Härdin 071d22d269f54c: mxfdec: Fix a potential DoS vector in mxf_read_pixel_layout()
[17:48] <frank_stl> Hi, We are developing one ffmpeg feature. So the update of ffmpeg is important to us. Just wondering does anybody know what is expected date of next ffmpeg version release?
[17:49] <saste> frank_stl, one release every 6 months more or less, which feature are you implementing?
[17:51] <frank_stl> We are trying to implement the GPU based filter acceleration. Target filters are deshake, scale, sharpening  and some other filters.
[17:53] <cone-609> ffmpeg.git 03Tomas Härdin 07fbb39f44f1d3: mxfdec: Fix potential leak in mxf_read_local_tags()
[17:53] <cone-609> ffmpeg.git 03Tomas Härdin 07e3ba817b95bb: mxfenc: Fix possible integer overflows
[17:54] <thegeek> frank_stl: gpl or lgpl?
[17:55] <michaelni> saste, iam trying to make a release every 3 month actually :)
[17:59] <frank_stl> saste, Thanks.
[18:46] <ubitux> frank_stl: deshake could get some nice efficiency improvements before you add gpu acceleration, or it might get meaningless
[18:46] <ubitux> (see for instance https://ffmpeg.org/trac/ffmpeg/ticket/1430)
[19:12] <michaelni> "<j-b>   [mpeg2video @ 0177ab80] a vbv buffer size is needed, for encoding with a maximum bitrate" // <--- should be fixed with latest ffmpeg 
[19:12] <j-b> michaelni: regression from 0.11.1 to 1.0.0 according to user
[19:13] <michaelni> with same vlc ? that would be odd
[19:14] <j-b> of FFmpeg
[19:14] <j-b> " Maybe it would be worth mentioning that downgrading ffmpeg from 1.0 to 0.11.1 solves the issue. "
[19:17] <michaelni> j-b, this cant be, the error message quoted does not exist in ffmpeg 1.0 nor in 0.11.1
[19:18] <j-b> michaelni: user moronic? :)
[19:18] <michaelni> it does exist in 0.10 though
[19:18] <Compn> is user using packaged ffmpeg ?
[19:19] <j-b> Compn: your guess is as good as mine
[19:20] <michaelni> j-b, the bug was real i think but the error messages dont match up, 1.0 should have shown "Either both buffer size and max rate or neither must be specified"
[19:21] <j-b> who|what introduced the bug?
[19:23] <michaelni> j-b, i do not know for sure, but possibly 1d981d16b744b000414cd5a2cc38db90d26d3903 in vlc, there may be other changes in vlc and or ffmpeg that affected it
[19:24] <j-b> is the VLC change wrong?
[19:24] <j-b> I can revert
[19:24] <michaelni> setting a max rate and no buffer size does not work with older ffmpeg nor libav AFAIK
[19:25] <michaelni> latest ffmpeg will fill in something reasonable for mpeg1/2/4 instead of failing
[19:26] <michaelni> if you dont set a max rate you also dont need to set a buffer size no matter if old or new ffmpeg
[19:29] <j-b> so a match up of ffmpeg before 1.0 and the 1d981d16b744b000414cd5a2cc38db90d26d3903 change to 0
[19:30] <michaelni> the fix is not in 1.0
[19:30] <michaelni> just master
[19:30] <j-b> but somehow the user says it is broken in 1.0 and not in 0.11.1
[19:31] <michaelni> did he say he was using the same vlc for both ?
[19:31] <j-b> yes
[19:31] <j-b> 2.0.4
[19:57] <cone-609> ffmpeg.git 03Paul B Mahol 07e12cfd044c2b: lavf: const correctness for avformat_new_stream
[20:15] <michaelni> j-b, after several tries, i manages to build 2.0.4 with 0.11.1 and with
[20:15] <michaelni> ./vlc -I dummy  --ffmpeg-hurry-up --no-loop --no-repeat --ffmpeg-skiploopfilter 4 --audio-language=any --start-time 0 --sout "#transcode{venc=ffmpeg,vcodec=mp2v,vb=7000,acodec=a52,ab=384,samplerate=48000,deinterlace}:std{access=file,mux=ts,dst="test.mpg"}" --sout-transcode-width 480 --sout-transcode-height 320 --sout-ffmpeg-keyint=3 "sample.avi"
[20:15] <michaelni> i get
[20:16] <michaelni> Either both buffer size and max rate or neither must be specified
[20:16] <michaelni> am i doing something wrong ?
[20:17] <michaelni> and i needed a few tries to get vlc link to 0.11.1, it prefers to link to the system installed libs it seems
[20:19] <michaelni> I think it would be quite usefull if vlc would print the libavcodec/libavformat versions it uses
[20:24] Action: michaelni is using: "strings ./modules/codec/avcodec/.libs/libavcodec_plugin.so | grep Lav.5" to find out which version actually gets used
[20:25] <michaelni> or ldd for shared ones ...
[20:35] <durandal_1707> pgmyuv should have IDIOTIC flag too
[20:38] <durandal_1707> or it should be changet to store real width and height
[20:45] <durandal_1707> ffprobe seems to ignore return value from decoder/ causing to report values which are not actually correct
[20:56] <durandal_1707> what happened to new coverage results?
[21:02] <cone-609> ffmpeg.git 03Paul B Mahol 076ddb03caf002: pgmyuvenc: do not create files which are not supported
[21:26] <durandal_1707> michaelni: how to get working seeking in voc?
[21:28] <llogan> durandal_1707: why say, "Backdoor.PHP.Rst.ap
[21:28] <llogan> damn it
[21:28] <llogan> This bug report is likely to be ignored.
[21:28] <llogan> doesn't give people much confidence to report bugs
[21:28] <llogan> (and no, i'm not installing any backdoors)
[21:30] <llogan> and i'm assuming you are richardpl
[21:35] <durandal_1707> what?
[21:36] <llogan> yeah, i obviously failed at communicating.
[21:36] <llogan> are you richardpl on trac?
[21:36] <durandal_1707> so?
[21:36] <llogan> don't tell people their bug reports are going to be ignored
[21:37] <durandal_1707> people use bug reports for patches - that is bad
[21:38] <nevcairiel> you should thank people for trying to contribute instead of telling them what they all did wrong
[21:39] <durandal_1707> so your solution is: post thank on bug report - result bug report is ignored
[21:39] <llogan> even better: you should tell that they are ignorant and shouldn't be wasting your precious time.
[21:41] <durandal_1707> exactly - they are wasting time sending patches to trac
[21:42] <nevcairiel> if any developer wants to fix the bug, he will find the patch. I don't see why it would be ignored just because tehre is a patch in it
[21:44] <nevcairiel> its like saying that any bug reports on trac are going to be ignored
[21:45] <av500> patches go to the ML, no?
[21:46] <av500> who is going to repost them from trac to the ML?
[21:47] <nevcairiel> noone is debating that posting a patch to trac reduces the likelyhood of it being applied, we're just pointing out that there are different ways to repond to people that post them on trac, and not all of them have to be all out hostile
[21:47] <Daemon404> ceheyos does it properly
[21:47] <Daemon404> explains the ysend to the ML
[21:47] <Daemon404> in a non-douchey way
[21:48] <llogan> michaelni: i recommend adding a note in the default ticket text (summary of the bug, etc) that patches should go to ffmpeg-devel, not trac
[21:48] <Daemon404> that sounds like a good idea
[21:48] <Daemon404> in giant red text, if possible
[21:48] <Daemon404> mayeb blinking
[21:49] <JEEB> lol <blink>
[21:49] <llogan> either i don't have permisison to change it or i'm too ignorant to see how.
[21:49] <llogan> either is highly likely
[21:49] <llogan> ...also a page counter with a monkey and some MIDI music would be nice.
[21:50] <Daemon404> diablo gif tiled background, you say?
[21:52] <llogan> http://wonder-tonic.com/geocitiesizer/content.php?theme=3&music=10&url=ffmpeg.org
[21:53] <nevcairiel> thinking back, you really have to wonder how anyone could ever make such pages without their eyes bleeding
[21:54] <Genia> i have a user question .. i'm trying to pass a single option into avformat_open_input, for a custom demux to read. in avformat_open_input, there is a bit of code that wants to see s->iformat->priv_class set, in order to pass those options to the demux. where/what should it be set to?
[21:54] <Daemon404> nevcairiel, by being 12
[21:54] <Daemon404> green text on black is cool, man.
[21:55] <nevcairiel> the only thing back from those days i did was add page counters and use the odd under construction gifs, never a full blown crazy-person site
[21:55] <Daemon404> you were clearly never 10-12 with a geocities account
[21:55] <JEEB> I remember using some editor with DHTML presets
[21:55] <JEEB> for fancy menus and stuff
[21:55] <JEEB> back in 1999-2000
[21:56] <llogan> i remember using AOLpress
[21:56] <JEEB> and no editor supported UTF-8
[21:56] <JEEB> so I had to either make a page with cyrillics, or one with öäå
[21:58] <llogan> my first page comprised of a starfield background and a photo of Inigo Montoya
[21:58] <michaelni> llogan, what where should be changed to what ?
[22:00] <llogan> michaelni: http://pastebin.com/HgVhVxTW
[22:03] <llogan> s/Note that p/P
[22:05] <michaelni> llogan, fixed
[22:06] <llogan> thanks
[22:10] <gnafu> "<@llogan> my first page comprised of a starfield background and a photo of Inigo Montoya" <-- I see nothing wrong with this.
[22:14] <cone-609> ffmpeg.git 03Clément BSsch 07c01d1d4ddf4d: http: add -content_type user option.
[22:15] <michaelni> nevcairiel, 953a3dcc4e28995072a1148e9d533e8d7b32248c breaks make tools/fourcc2pixfmt.o
[22:17] <cbsrobot> it's a fake !
[22:17] <cbsrobot> too much left aligned
[22:17] <ubitux> mmh i need to fix fate now
[22:18] <ubitux> with the new option saste added
[22:19] <michaelni> nevcairiel, simply adding "#undef printf" fixes it
[22:29] <cone-609> ffmpeg.git 03Michael Niedermayer 0795d07f33dc5c: libavfilter: include needed header for AVDictionary
[22:29] <cone-609> ffmpeg.git 03Michael Niedermayer 07f68ff2e0e79c: fourcc2pixfmt: fix compilation
[22:52] <ubitux> http://pastie.org/5111260
[22:53] <ubitux> nevcairiel: do you think this will make the substitution with msys?
[22:53] <ubitux> (see the printf)
[22:58] <ubitux> nevcairiel: if you have the possibility to try http://b.pkh.me/0001-WIP-use-the-lavfi-graph_file-option-to-workaround-pa.patch i'd be interested
[22:59] <ubitux> heh, i can even drop one escape
[23:00] <ubitux> mmh actually no
[23:03] <j-b> michaelni: http://git.videolan.org/?p=vlc.git;a=commit;h=50a0ceae0ffa535512c5fd8e0952f277b1732458 something like this?
[23:10] <michaelni> j-b, that should fix the failure but if the user intentionally sets min/max rate he might wonder why they have no effect
[23:10] <j-b> well, no magic solution
[23:18] <ubitux> ok patch updated, should be better
[23:18] <ubitux> i hope it works :(
[23:28] <ubitux> saste: i used your option
[23:29] <ubitux> i might have a patch to fix fate
[23:29] <saste> nice
[23:29] <saste> but?
[23:29] <ubitux> http://b.pkh.me/0001-fate-use-the-lavfi-graph_file-option-to-workaround-p.patch
[23:29] <ubitux> but i have no mean to test
[23:30] <saste> Daemon404, ^^
[23:30] <ubitux> i'm going to send the patch anyway
[23:37] <ubitux> saste: the issue is kind of fun actually
[23:37] <ubitux> funny puzzle.
[23:41] <cone-609> ffmpeg.git 03Stefano Sabatini 07ce875a613a16: doc/texi2pod.pl: drop unnecessary rule, which was converting "\," to ","
[23:41] <cone-609> ffmpeg.git 03Stefano Sabatini 07a161def1e4a5: Doxyfile: update examples path and patterns
[23:41] <cone-609> ffmpeg.git 03Stefano Sabatini 07a945607a78ce: ffprobe: add an array of section print buffers to the WriterContext
[23:45] <ubitux> ffprobe is getting pretty cute
[23:59] <ubitux> saste: thx for the escaping fix :)
[23:59] <saste> i'm the escaping man
[23:59] <ubitux> :D
[00:00] --- Thu Oct 25 2012