[Ffmpeg-devel-irc] ffmpeg-devel.log.20120918

[burek]

[03:26] <burek> is there a diff between using -s and -vf scale ?
[03:34] <Compn> do both allow you to select scaling algorithm? e.g. bilinear lanczos ?
[03:35] <Daemon404> -sws_flags
[04:31] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * r66bdc58550 10ffmpeg/libavcodec/x86/cabac.h: 
[04:31] <CIA-55> ffmpeg: get_cabac_inline_x86: workaround clang bug with disabled optimizations
[04:31] <CIA-55> ffmpeg: gcc produces binary identical output relative to before this change
[04:31] <CIA-55> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[05:23] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * ra1fb2b935a 10ffmpeg/libavformat/nsvdec.c: 
[05:23] <CIA-55> ffmpeg: nsvdec: change 2 commented assert() to av_assert()
[05:23] <CIA-55> ffmpeg: they keep showing up if i grep for non av assert()
[05:23] <CIA-55> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[05:23] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * rdb436b17e9 10ffmpeg/libswscale/swscale.c: 
[05:23] <CIA-55> ffmpeg: swscale: change assert() to av_assert()
[05:24] <CIA-55> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[05:53] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * r70a9734544 10ffmpeg/libavcodec/dsputil.c: 
[05:53] <CIA-55> ffmpeg: dsputil: change assert(8) to av_assert2(8)
[05:53] <CIA-55> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[09:25] <CIA-55> ffmpeg: 03Bobby Bingham 07master * r6e03da7f35 10ffmpeg/doc/filter_design.txt: 
[09:25] <CIA-55> ffmpeg: doc: fix typos in filter documentation
[09:25] <CIA-55> ffmpeg: Signed-off-by: Bobby Bingham <uhmmmm at gmail.com>
[09:30] <ubitux> how the fuck is ffmpeg even able to play http://mindandvision.de/images/streams/sample.f4v
[09:30] <ubitux> it's insane
[09:30] <ubitux> the file is completely messed up
[10:10] <cbsrobot-> ubitux: that's what for ffmpeg is ...
[10:10] <ubitux> :)
[10:45] <CIA-55> ffmpeg: 03Bobby Bingham 07master * r8e59062597 10ffmpeg/libavcodec/targaenc.c: Fix typo in targa encoder comment
[10:45] <CIA-55> ffmpeg: 03Bobby Bingham 07master * r4aade44c4a 10ffmpeg/libavcodec/targaenc.c: targaenc: Use named TGA_RLE constant instead of magic number
[11:16] <burek> I'm not even shure what this guy is asking about, so if anyone knows, please help :) http://ffmpeg.gusari.org/viewtopic.php?f=16&t=673
[11:19] <av500> green screen
[11:19] <av500> replace green with other video content
[11:20] <burek> oh, I see
[11:20] <burek> a mask
[11:20] <burek> is that even possible with ffmpeg
[11:21] <av500> in a filter i guess
[11:23] <ubitux> burek: alphamerge filter maybe
[11:23] <ubitux> you might need a slightly complex filtergraph
[11:24] <burek> I see, thanks :) I'll suggest him to read the docs :D
[12:24] <cbsrobot-> ubitux: next week I can reference check the ebu128 filter
[12:50] <ubitux> cbsrobot-: oh really? would be great
[14:45] <ubitux> ./ffmpeg -f lavfi -i testsrc=d=10 -vf 'select=gt(n\,100)' -y test.mkv  OK
[14:45] <ubitux> ./ffmpeg -f lavfi -i testsrc=d=10 -vf 'select=gt(n\,100)' -y test.mp4  not OK
[14:45] <ubitux> looks like some ts weirdness again
[14:46] <ubitux> (playback wait 4 seconds before starting in the case of mp4)
[14:52] <CIA-55> ffmpeg: 03Mans Rullgard 07master * r1fce361d70 10ffmpeg/libavfilter/ (15 files): 
[14:52] <CIA-55> ffmpeg: lavfi: replace empty input/output lists with null pointers
[14:52] <CIA-55> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[14:52] <CIA-55> ffmpeg: 03Anton Khirnov 07master * r4e48aa8656 10ffmpeg/libavfilter/ (buffersrc.c buffersrc.h): buffersrc: add const to the AVFrame* argument of av_buffersrc_write_frame()
[14:52] <CIA-55> ffmpeg: 03Clément BSsch 07master * rffdd2e9144 10ffmpeg/libavformat/id3v2.c: 
[14:52] <CIA-55> ffmpeg: lavf/id3v2: do not export empty fields.
[14:52] <CIA-55> ffmpeg: This also avoids a memleak.
[14:52] <CIA-55> ffmpeg: Signed-off-by: Anton Khirnov <anton at khirnov.net>
[14:52] <CIA-55> ffmpeg: 03Clément BSsch 07master * r049ce4facb 10ffmpeg/libavformat/id3v2.c: 
[14:52] <CIA-55> ffmpeg: id3v2: strdup the genre name explicitly.
[14:52] <CIA-55> ffmpeg: It would have been done anyway in the av_dict_set() call.
[14:52] <CIA-55> ffmpeg: This simplifies the code and avoid a warning because of assigning a
[14:52] <CIA-55> ffmpeg: const string from ff_id3v1_genre_str to a non-const variable.
[14:52] <CIA-55> ffmpeg: Signed-off-by: Anton Khirnov <anton at khirnov.net>
[14:52] <CIA-55> ffmpeg: 03Anton Khirnov 07master * r9f64c8219a 10ffmpeg/libavcodec/utils.c: 
[14:52] <CIA-55> ffmpeg: lavc: add const to private codec class initialization.
[14:52] <CIA-55> ffmpeg: Silences a warning about discarding const.
[14:52] <CIA-55> ffmpeg: 03Anton Khirnov 07master * r466b39efaf 10ffmpeg/libavcodec/ (assenc.c avcodec.h dvbsub.c dvdsubenc.c utils.c xsubenc.c): 
[14:53] <CIA-55> ffmpeg: lavc: replace AVCodecContext.encode with subtitle-specific callback
[14:53] <CIA-55> ffmpeg: AVCodecContext.encode is currently used only for subtitles, encode2 is
[14:53] <CIA-55> ffmpeg: 03Anton Khirnov 07master * r34ebbbfeec 10ffmpeg/libavcodec/xsubenc.c: xsubenc: reindent
[14:53] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * r91af76099e 10ffmpeg/: (log message trimmed)
[14:53] <CIA-55> ffmpeg: Merge commit '23aae62c2cb4504a09ceb8cd0cabc1c8b260f521'
[14:53] <CIA-55> ffmpeg: * commit '23aae62c2cb4504a09ceb8cd0cabc1c8b260f521':
[14:53] <CIA-55> ffmpeg:  alsdec: Check k used for rice decoder.
[14:53] <CIA-55> ffmpeg:  avfiltergraph: silence an uninitialized variable warning
[14:53] <CIA-55> ffmpeg:  xsubenc: reindent
[14:53] <CIA-55> ffmpeg:  lavc: replace AVCodecContext.encode with subtitle-specific callback
[15:24] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * r9853e41aa0 10ffmpeg/libavcodec/alsdec.c: (log message trimmed)
[15:24] <CIA-55> ffmpeg: alsdec: check opt_order.
[15:24] <CIA-55> ffmpeg: Fixes out of array write in quant_cof.
[15:24] <CIA-55> ffmpeg: Also make sure no invalid opt_order stays in the context.
[15:24] <CIA-55> ffmpeg: Fixes CVE-2012-2775
[15:24] <CIA-55> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
[15:24] <CIA-55> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[15:24] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * r5b051ec3bd 10ffmpeg/libavcodec/alsdec.c: (log message trimmed)
[15:24] <CIA-55> ffmpeg: alsdec: Check that quantized parcor coeffs are within range.
[15:24] <CIA-55> ffmpeg: ALS spec:
[15:24] <CIA-55> ffmpeg:  11.6.3.1.1 Quantization and encoding of parcor coefficients
[15:24] <CIA-55> ffmpeg:  ...
[15:24] <CIA-55> ffmpeg:  In all cases the resulting quantized values ak are restricted to the range [-64,63].
[15:24] <CIA-55> ffmpeg: Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind
[15:25] <CIA-55> ffmpeg: 03Mans Rullgard 07master * r2a8a860ae6 10ffmpeg/configure: 
[15:25] <CIA-55> ffmpeg: configure: include flags in nm variable
[15:25] <CIA-55> ffmpeg: This simplifies ensuring proper flags are used when the default
[15:25] <CIA-55> ffmpeg: is overridden by the system or on the command line.
[15:25] <CIA-55> ffmpeg: Signed-off-by: Mans Rullgard <mans at mansr.com>
[15:25] <CIA-55> ffmpeg: 03Benjamin Larsson 07master * r6a08955c10 10ffmpeg/libavformat/flvenc.c: 
[15:25] <CIA-55> ffmpeg: flvenc: silence bogus warning
[15:25] <CIA-55> ffmpeg: The compiler fails to figure out that enc->codec_type can only
[15:25] <CIA-55> ffmpeg: have 3 different values.
[15:25] <CIA-55> ffmpeg: Thus when an if/else is encountered it triggers on the possibility
[15:25] <CIA-55> ffmpeg: of the else case has not initialized the flags variable.
[15:26] <CIA-55> ffmpeg:  flvenc: silence bogus warning
[15:26] <CIA-55> ffmpeg:  configure: include flags in nm variable
[15:26] <CIA-55> ffmpeg:  alsdec: fix misplaced parentheses.
[15:26] <CIA-55> (18 lines omitted)
[15:45] <kriegerod> git master HEAD is not compiling http://pastebin.com/azdLjjwt , could somebody look?
[15:48] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * ra44cbc1cbc 10ffmpeg/libavcodec/sonic.c: 
[15:48] <CIA-55> ffmpeg: sonic: update to encode2() API
[15:48] <CIA-55> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[15:48] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * r36e5c472c8 10ffmpeg/libavcodec/allcodecs.c: 
[15:48] <CIA-55> ffmpeg: allcodecs: re-enable sonic
[15:48] <CIA-55> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[15:48] <cbsrobot-> kriegerod: did you make clean ?
[15:54] <kriegerod> cbsrobot-: yes
[16:11] <michaelni> kriegerod, fixed locally, ill push it in a moment
[16:14] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * rc0dfc720ef 10ffmpeg/libavcodec/x86/mpegaudiodec.c: 
[16:14] <CIA-55> ffmpeg: mpegaudiodec: add prototypes for all functions that may need them
[16:14] <CIA-55> ffmpeg: fixes compile failure
[16:14] <CIA-55> ffmpeg: Found-by: kriegerod
[16:14] <CIA-55> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[17:54] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * r0b2ac679e2 10ffmpeg/libavcodec/sonic.c: 
[17:54] <CIA-55> ffmpeg: sonic: factorize tail_size
[17:54] <CIA-55> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[17:54] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * refac6f7472 10ffmpeg/libavcodec/x86/mpegaudiodec.c: 
[17:54] <CIA-55> ffmpeg: x86/mpegaudiodec: drop likely now unneeded #if HAVE_AVX_EXTERNAL
[17:54] <CIA-55> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[17:54] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * r512beea529 10ffmpeg/libavcodec/sonic.c: 
[17:54] <CIA-55> ffmpeg: sonicenc: limit quant so that golomb codes are less than 32 bits long
[17:54] <CIA-55> ffmpeg: before they could become 33 bits
[17:54] <CIA-55> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[18:39] <CIA-55> ffmpeg: 03Carl Eugen Hoyos 07master * r9e208ce6dd 10ffmpeg/configure: 
[18:39] <CIA-55> ffmpeg: Silence warning C4554 when compiling with msvc.
[18:39] <CIA-55> ffmpeg: warning C4554: '>>' : check operator precedence for possible error; use parentheses to clarify precedence
[18:39] <CIA-55> ffmpeg: Reviewed-by: Derek Buitenhuis
[18:43] <nevcairiel> eh
[18:43] <nevcairiel> silencing a whole warning because of one case where its annoying is rather stupid :p
[18:46] <merbanan> nevcairiel: it's a few more then just one
[18:47] <nevcairiel> sure it echos on every compiled file
[18:47] <nevcairiel> still
[18:47] <nevcairiel> fix the cause, not disable the warning
[18:48] <nevcairiel> warnings have a reason, just disabling them is bad style
[18:48] <merbanan> in this case I'm all for disabling that warning
[18:49] <CIA-55> ffmpeg: 03Clément BSsch 07master * rb89c16e73e 10ffmpeg/libavfilter/vf_select.c: 
[18:49] <CIA-55> ffmpeg: lavfi/select: make sure avctx is set before closing it.
[18:49] <CIA-55> ffmpeg: This avoid a crash when the filter fails before config_input(),
[18:49] <CIA-55> ffmpeg: typically with a syntax error in the next filter:
[18:49] <CIA-55> ffmpeg:  select=gt(scene\,.4),select=fail
[18:49] <ohsix> disabling the warning is what you do when you use cl/msvs :]
[18:49] <ohsix> and they've given you soooo much more to disable in the newer versions, it's great
[18:50] <nevcairiel> disabling the warning is what you do when you dont know any better, indedependent of compiler
[18:50] <ohsix> nuh uh
[18:51] <ohsix> if every warning has a number and you can use a pragma or the command line to inhibit them, that's what you're supposed to do! :]
[18:52] <ohsix> plus there's like no sanity in the warning levels with cl
[18:53] <ohsix> the only real downside to the pragmas and stuff is that they mask any warning level you set, there's no "if spamming every warning ever, inhibit", there's just inhibit
[18:54] <ohsix> the code gets festooned and crustified with pragmas for warnings that aren't even there any more, hiding new ones where it's actually important to know, it's really great
[18:57] <merbanan> nevcairiel: what is wrong with this ? ---> header_size = avctx->extradata_size-(ctx->off + 7) >> 3;
[18:57] <ohsix> in summary: suppressing them is ill advised in any circumstance
[18:57] <ohsix> nothing
[18:57] <ohsix> the warning is to help people that don't know, or are confused by operator precedence
[18:58] <merbanan> libavformat/latmenc.c(108) : warning C4554: '>>' : check operator precedence for possible error; use parentheses to clarify precedence
[18:58] <ohsix> yea, so add () to show what you meant, even if you full well knew precedence when you originally wrote it :]
[18:58] <nevcairiel> i personally agree with the compiler, an extra pair of bracers never hurt anyone
[18:59] <nevcairiel> but at least libav has the bracer police which wont let any bracers not required for syntax stand
[18:59] <ohsix> you will eventually be surprised
[18:59] <nevcairiel> which is stupid
[18:59] <nevcairiel> not every reader of the code may know the exact precedence rules 100%
[19:00] <ohsix> they should, but why not help them along if they don't
[19:00] <nevcairiel> its not like the extra set of bracers would make that particular line extra bad
[19:00] <ohsix> you expect them to know the precedence for adding and multiplication :> (which is from the maths, not the cprogramminglanguage)
[19:01] <nevcairiel> sure, but for clarity i usually still write it out :p
[19:01] <ohsix> nevcairiel: right, but like i said, you will eventually be surprised by it changing the result; which is the nature of the warning message about said thing
[19:24] <Daemon404> [12:59] < nevcairiel> but at least libav has the bracer police which wont let any bracers not required for syntax stand <-- that's it, i'm reporting you to the federal bureau of bracers
[19:25] <nevcairiel> bring it on!
[20:00] <Daemon404> can anyone explain wtf reimar is talking about to me
[20:01] <Daemon404> re: ffprobe
[20:03] <ubitux> Daemon404: what don't you understand?
[20:04] <ubitux> he wants to print everything disposition-related in one place
[20:04] <Daemon404> i responded.
[20:04] <ubitux> he explained why he wants such thing
[20:04] <Daemon404> his reasons are not valid
[20:05] <ubitux> if attached_picture happened to appear in another kind of stream (let's say subtitles), you would either duplicate, or move the field
[20:05] <ubitux> which could be problematic at some point
[20:05] <Daemon404> attached picture is teh flag for the picture itself
[20:06] <Daemon404> it will NEVER be anything btu vidoe
[20:06] <Daemon404> ever
[20:06] <Daemon404> period.
[20:06] <Daemon404> a 1 frame stream
[20:07] <ubitux> ok ok
[20:07] <ubitux> it's not like i'm trying to argue anything
[20:07] <Daemon404> i get the feelign reimar doesn't always even understand teh situation before he comments
[20:07] <ubitux> just trying to explain how i interpret the spirit of reimar :p
[20:07] <Daemon404> (like certain people form libav)
[20:07] Action: Daemon404 runs
[20:07] <ubitux> haha
[20:11] <Daemon404> hmmm
[20:11] <Daemon404> we have --host-cflags
[20:11] <Daemon404> but not --extra-host-cflags
[20:11] <Daemon404> slightly annoying
[21:35] Action: Daemon404 pings ubitux to weight in on patch
[21:35] <Daemon404> since you seem to do a lot of ffprove work
[21:35] <ubitux> to be honest i don't care much
[21:35] <ubitux> you're the author of the patch
[21:35] <ubitux> do what you think is best
[21:35] <ubitux> i don't really have any opinion in that particular case
[21:35] <Daemon404> i am ok with it moved up into one area
[21:36] <ubitux> ’ taste ’ your choice
[21:36] <Daemon404> i just need an oK
[21:36] <Daemon404> i am very much against using _ATTACHED_PIC as a generic metadata flag though
[21:36] <Daemon404> or even having one
[21:36] <ubitux> then keep it that way
[21:36] <ubitux> consider my opinion neutral on this one
[21:36] <ubitux> i'm ok with the patch
[21:36] <ubitux> maybe let saste a chance to comment though
[21:37] <Daemon404> well
[21:37] <Daemon404> :timezones:
[21:37] <ubitux> is the patch urgent?
[21:37] <Daemon404> yes
[21:37] <Daemon404> i could patch it locally
[21:37] <Daemon404> but i feel dirty
[21:39] <ubitux> did you change anything from the initial version?
[21:40] <ubitux> oh wait just a sec
[21:40] <Daemon404> ubitux, i noted everything changed in the chain\
[21:40] <Daemon404> s/is_//
[21:40] <Daemon404> and moving the check outside the switch
[21:41] <ubitux> i need to check somethign with your patch
[21:41] <ubitux> give me a few minutes
[21:41] <Daemon404> ok
[21:41] <Daemon404> its not a "RIGHT NOW" patch
[21:42] <Daemon404> as much as it is "before i leave work"
[21:42] <ubitux> i think it can be problematic with some writers
[21:42] <ubitux> with the optional fields printing stuff
[21:43] <Daemon404> i only tested with json
[21:43] <ubitux> basically, with some formats, you *need* to have the field printed
[21:43] <ubitux> typically with csv, where you have a number of columns
[21:43] <ubitux> see the print_str_opt() thing
[21:43] <Daemon404> i see
[21:44] <ubitux> and so if you add a if like you did, it's likely to get things broken
[21:44] <ubitux> also, another thing
[21:44] <ubitux> don't you need to add the type in the ffprobe.xsd for the XML output?
[21:44] <Daemon404> its not like any of this is documented
[21:44] <ubitux> it's a developer thing, so yeah sure
[21:45] <Daemon404> do you want me to fix it
[21:45] <Daemon404> or can you?
[21:45] <ubitux> saste is rewriting that stuff, and adding some tests
[21:45] <ubitux> well, i'd prefer if you do
[21:45] <Daemon404> ok
[21:45] <Skyler_> michaelni: check your email, have fun :3
[21:46] <Daemon404> money
[21:46] <Daemon404> ?
[21:46] <Skyler_> 14,500 euros or so.  michael should buy himself an upgrade for his pentium MMX.
[21:46] <Daemon404> x264 llc?
[21:47] <Skyler_> yesyes
[21:48] <Daemon404> ubitux, it isnt clear at all how to add a print_int_opt
[21:48] <Daemon404> with this stupid wanna-be-template code
[21:49] <ubitux> Daemon404: i don't remember seeing such conditionnal thing for int
[21:49] <Daemon404> there ist
[21:49] <Daemon404> isnt*
[21:49] <Daemon404> i was going to add it, since having if/else all the time is dumb.
[21:49] <ubitux> we have it for strings cause it's a bit better for "defaults"
[21:49] <ubitux> well
[21:49] <ubitux> just make it unconditionnal
[21:49] <ubitux> print the value all the time
[21:49] <Daemon404> it has to be an if/else
[21:50] <Daemon404> or well i could brping the result of the expr
[21:50] <Daemon404> but i think it's ugly
[21:50] <ubitux> "default" -> 0/1
[21:50] <Daemon404> print*
[21:50] <ubitux> what's the problem?
[21:51] <Daemon404> print_int("default", !!(stream->disposition & AV_DISPOSITION_ATTACHED_PIC));
[21:51] <Daemon404> is sort of ugly
[21:51] <ubitux> i hope you won't print the attached pic flag for "default" ;)
[21:51] <Daemon404> er
[21:51] <Daemon404> yeah, woops
[21:51] <Daemon404> but you see what i mean
[21:52] <ubitux> sure
[21:52] <ubitux> i'm sorry i can't change C limitations for you
[21:52] <Daemon404> it also means
[21:52] <Daemon404> that i must keep the attached_pic check
[21:52] <Daemon404> under teh video case
[21:52] <Daemon404> instead of having them all in teh same place
[21:53] <ubitux> ok, then let me propose you something else
[21:53] <ubitux> what about making a "disposition" with "+default +picture +..."?
[21:53] <Daemon404> thats horrible.
[21:53] <Daemon404> and parsing will be stupid
[21:53] <Daemon404> completely eliminates why i liek json output
[21:53] <Daemon404> if i have to parse strings
[21:54] <ubitux> true :)
[21:55] <Daemon404> if i had to choose
[21:55] <Daemon404> id do the !! method
[21:55] <ubitux> it's fine
[21:55] <Daemon404> and have attached_pic under the video case
[21:55] <ubitux> yeah, it's fine
[21:55] <Daemon404> ill send a new patch to the ML
[21:55] <ubitux> (and don't forget to update doc/ffprobe.xsd)
[21:56] <Daemon404> yes
[21:56] <Daemon404> may i add:
[21:56] <Daemon404> fuck xml
[21:56] <ubitux> of course
[21:56] <ubitux> :)
[21:56] <ubitux> when writing the json writer, i said to saste i hated xml
[21:57] <ubitux> he understood that as "i need to do it then"
[21:57] <Daemon404> lol
[21:57] <Daemon404> im using ffprobe from coffeescript (so hipster!)
[21:57] <Daemon404> and json is insanely convenient
[21:57] <ubitux> :)
[21:57] <ubitux> i kind of like the flat output
[21:58] <ubitux> i like how human readable it is :p
[21:58] <ubitux> even more than json actually
[21:58] <ubitux> (IMO)
[21:58] <michaelni> Skyler_, i already have a 6core i7 @ 4ghz, about the money, thx alot, this is quite a large amount of money for me ...
[21:59] <Daemon404> ubitux, the point of json isnt human readability :P
[21:59] <Skyler_> oh wow.  that is quite the upgrade
[21:59] <Daemon404> michaelni, gonna spend it on cameras?
[21:59] <ubitux> Daemon404: yeah but the indent kind of help :)
[21:59] <ohsix> xsl to and from json, problem solved
[22:00] <Skyler_> yay, michaelni has a faster computer than me now
[22:01] <michaelni> Daemon404, if i had the time to take pics i would spend maybe a small part of that on a better camera
[22:02] <Daemon404> maybe we'd get a canon raw decoder :P
[22:06] <Daemon404> ubitux, since im printing them unconditionally
[22:07] <Daemon404> use="required" in teh xsd?
[22:07] <ubitux> i admit i'm not much familiar with the xsd at all
[22:07] <ubitux> not much/at all... 
[22:07] <ubitux> :))
[22:08] <Daemon404> k
[22:08] <ubitux> sorry
[22:08] <Daemon404> it looks like i need to
[22:12] <Daemon404> ubitux, v2 sent
[22:14] <ubitux> Daemon404: any reason you only considered a limited set of dispositions?
[22:14] <Daemon404> some of them dont seem useful
[22:14] <Daemon404> or are ambiguous
[22:14] <ubitux> kay
[22:14] <Daemon404> like karaoke
[22:14] <Daemon404> and "original" is ambiguous
[22:14] <ubitux> you don't like karaoke?
[22:14] <ubitux> :))
[22:14] <Daemon404> lol
[22:14] <Daemon404> nothing but nut uses most of them
[22:15] <Daemon404> just fyi
[22:15] <Daemon404> ;)
[22:15] <ubitux> :D
[22:29] <burek> there is definitely some kind of bug which prevents compilation of ffmpeg+x264 on centos or I'm missing something obvious
[22:29] <burek> ffmpeg: error while loading shared libraries: libx264.so.125: cannot open shared object file: No such file or directory
[22:30] <burek> i did yum remove x264* and ffmpeg*
[22:30] <burek> compiled x264 from source (yasm also) and ffmpeg from latest tarball and got that error
[22:30] <Daemon404> you have not put libx264.so.* in your library path
[22:30] <Daemon404> ld.so cant find it
[22:30] <burek> i did ldconfig
[22:31] <Daemon404> burek, /usr/local/lib may not be checked
[22:31] <Daemon404> you may have to add it to ld.so.conf
[22:31] <llogan> don't build as shared
[22:31] <Daemon404> llogan, bad advice
[22:31] <burek> x264 had to be built as shared
[22:31] <burek> otherwise it doesn't create libx264.so
[22:32] <burek> but ffmpeg was built just with --enable-gpl --enable-libx264
[22:32] <burek> that's all
[22:32] <Daemon404> --enable-static
[22:32] <burek> i'll check ld.so.conf
[22:32] <Daemon404> will build libx264.a
[22:32] <llogan> Daemon404: why is that bad advice?
[22:32] <burek> Daemon404: include ld.so.conf.d/*.conf /usr/local/lib
[22:32] <Daemon404> llogan, building everything as static just isnt a good idea in many setups
[22:32] <Daemon404> e.g. where many things use libx264
[22:32] <Daemon404> or libavcodec
[22:32] <Daemon404> eyc
[22:32] <burek> that's the content of ld.so.conf
[22:32] <llogan> bah. nothing sane requires shared
[22:33] <Daemon404> yes it does
[22:33] <Daemon404> many things use libavcodec.so
[22:33] <Daemon404> some things that are essential
[22:33] <Daemon404> and cant be done with ffmpeg cli
[22:33] <JEEB> llogan, the point he was making that if you have multiple things using something, shared is better. In other cases, for custom installs of one thing, static is better
[22:33] <JEEB> (´
[22:35] <burek> sorry to interrupt, but, Daemon404, is that the correct content of ld.so.conf
[22:35] <Daemon404> latter should be on a newline
[22:35] <Daemon404> the dir
[22:35] <Daemon404> also you need to run ldconfig as root
[22:38] <burek> i am root
[22:38] <burek> im just confused with "include"
[22:39] <burek> yeah! :D
[22:39] <CIA-55> ffmpeg: 03Derek Buitenhuis 07master * r8e2a950b29 10ffmpeg/ (doc/ffprobe.xsd ffprobe.c): 
[22:39] <CIA-55> ffmpeg: ffprobe: Add a few common disposition cases
[22:39] <CIA-55> ffmpeg: This info is crucial in knowing which stream to pick in an
[22:39] <CIA-55> ffmpeg: automated setup.
[22:39] <CIA-55> ffmpeg: Signed-off-by: Derek Buitenhuis <derek.buitenhuis at gmail.com>
[22:39] <burek> it works :)
[22:39] <burek> I owe you a beer Daemon404 :D
[22:39] <Daemon404> lul
[22:40] <CIA-55> ffmpeg: 03Michael Niedermayer 07master * r8d07742cb1 10ffmpeg/libavformat/avidec.c: 
[22:40] <CIA-55> ffmpeg: avidec: workaround aac with broken block align
[22:40] <CIA-55> ffmpeg: Fixes Ticket1742
[22:40] <CIA-55> ffmpeg: Based on patch by: Nikola Vranic <nikola.vranic at rt-rk.com>
[22:40] <CIA-55> ffmpeg: Signed-off-by: Michael Niedermayer <michaelni at gmx.at>
[22:42] <burek> did you guys changed h264 to act as an encoder also (to use libx264) ?
[22:43] <burek> because I'm looking at ffmpeg -codecs | grep x264 and I can't see libx264 on a single line
[22:43] <burek> just this: DEV.LS h264                 H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10 (encoders: libx264 libx264rgb )
[22:43] <Daemon404> ?
[22:43] <Daemon404> it says right there that the h264 encoder is libx264
[22:44] <Daemon404> its just listing what can encode that codec id
[22:44] <llogan> i can see how it would be confusing given the previous behavior of -codecs
[22:44] <burek> yes, I see, I was used to libx264 being on a single line I guess
[22:50] <llogan> burek: there is also now -encoders and -decoders
[22:50] <burek> oh cool :)
[22:50] <burek> I might just add the suggestion to sort them
[22:58] <llogan> yeah. that would be nice.
[23:30] Action: cbsrobot- finally made a (more or less) copy-paste guide for OSX ^
[23:31] <ubitux> haha saste just got the same idea as me for the flags
[23:31] <ubitux> Daemon404 ^ :))
[23:32] <Daemon404> lol
[23:35] <llogan> cbsrobot-: thanks. perhaps you can use sed to change those names.
[23:36] <Daemon404> ubitux, and i told him the same thing i told you
[23:37] <burek> I cleaned spammers from the forum with a swing of my Return key :)
[23:37] <burek> god bless it :)
[23:37] <Daemon404> burek, i woulda hired little bobby drop tables
[23:38] <burek> well, sql injection wasn't a problem
[23:38] <burek> it's pure spam with bots..
[23:38] <Daemon404> you run what
[23:38] <Daemon404> smf?
[23:38] <burek> phpbb..
[23:38] <burek> should have installed vanilla
[23:38] <Daemon404> oh
[23:38] <Daemon404> you must be used to spam
[23:39] Action: Daemon404 runs
[23:39] <cbsrobot-> llogan: sure - next revision
[23:39] <burek> I found an ultimate sql command to wipe out all spam at once :>
[23:39] <burek> it took me an hour to craft it, but no more spam for us from now on :)
[23:39] <Daemon404> burek, me too
[23:39] <Daemon404> apt-get install mysql
[23:40] <Daemon404> s/install/remove/
[23:40] <burek> :D
[23:40] <burek> +1 :)
[23:40] <Daemon404> i recently started movng stuff to postgresql
[23:40] <Daemon404> since oracle shit the bed with mysql
[23:40] <Daemon404> postgresql isnt nearly as user friendly
[23:40] <JEEBsv> yeah, they really come from different positions
[23:41] <Daemon404> postgresql is the neckbeard sql server of choice
[23:41] <Daemon404> flatfiles are the hipster db of choice
[23:41] <Daemon404> etc
[23:41] Action: llogan supports all things neckbeard
[23:41] <burek> well, for flat files, did you use sqlite3 maube
[23:41] <burek> -u +y
[23:41] <Daemon404> no
[23:41] <Daemon404> stupid hipsters use text files
[23:42] <Daemon404> that store mardown
[23:42] <Daemon404> for their blogs
[23:42] <burek> well it's got its use case when you don't want a db server
[23:42] <llogan> i thought hipsters liked mongodb
[23:42] <JEEBsv> and you have a single thing accessing the db, yeah
[23:42] <Daemon404> llogan, not for blogs
[23:43] <Daemon404> the problem with all these flatfile blogs is that you cant actually edit shit in your browser
[23:43] <Daemon404> no wysiwyg
[23:43] <Daemon404> they want you to write in markdown
[23:43] <llogan> ah, blahgs.
[23:43] <Daemon404> and run soem ruby script after you upload
[23:43] <burek> yeah that's pretty stupid
[23:43] <burek> but the logic was all-in-one-file
[23:43] <Daemon404> no each post is a separate file
[23:43] <Daemon404> in a dir structure
[23:43] <burek> some people really have a problem copying/backing-up 3 files instead of 1
[23:44] <burek> oh
[23:44] <burek> well isn't that nice :)
[23:44] <JEEBsv> I think mplayer2 and then MPC-HC switched to something similar lately, just python-based
[23:44] Action: ubitux is a stupid hipster that use text files in markdown for his blog
[23:44] <Daemon404> hipsters/github coders/ruby coders
[23:44] <Daemon404> project pages are for losers
[23:44] <Daemon404> just a github page + README.md is enough
[23:44] <Daemon404> amirite?
[23:45] <burek> too often :)
[23:46] <llogan> the only "blog" i've updated recently are just html news entries.
[23:47] <llogan> burek: so how are you going to stop futrue spam?
[23:47] <llogan> *future.
[23:47] <burek> cron job which removes spam :D
[23:48] <llogan> how does it know what is spam?
[23:48] <ubitux> Daemon404: do you really care about wysiwyg? oO
[23:48] <Daemon404> ubitux, its nice to have
[23:49] <ubitux> well, i personally don't live in my browser like many ppl do so i can't really understand
[23:49] <llogan> maybe ubitux is the real neckbeard
[23:49] <ubitux> no, i'm just a stupid hipster that use text files in markdown for his blog
[23:49] <ubitux> :))
[23:50] <llogan> Is that the emote for a neckbeard?
[23:50] <ubitux> nope, i'm just very happy
[23:50] <Daemon404> editing markdown sucks
[23:51] <Daemon404> uploading files to make a post sucks
[23:51] <ubitux> uploading?
[23:51] <Daemon404> i actually like wp
[23:51] <ubitux> edit it directly
[23:51] <Daemon404> sshing in?
[23:51] <burek> llogan, well I taught it all it knows about spam :)
[23:51] <Daemon404> [17:50] <@ubitux> no, i'm just a stupid hipster that u
[23:51] <ubitux> sure, or vim ssh:// or whatever
[23:51] <Daemon404> wait
[23:51] <Daemon404> wrong paste
[23:51] <Daemon404> [17:50] <@llogan> maybe ubitux is the real neckbeard
[23:51] <Daemon404> usability etc
[23:51] <Daemon404> normal people dont wanna do that
[23:52] <ubitux> what's the problem with markdown btw?
[23:52] <JEEBsv> ubitux: you forgot git :3
[23:52] <Daemon404> i think it's ugly
[23:52] <Daemon404> but thats beside the point
[23:52] <JEEBsv> git yer blog to the internets!
[23:52] <ubitux> JEEBsv: i don't like it for blogging
[23:52] <JEEBsv> ah
[23:52] <ubitux> JEEBsv: i don't need any revert or branching or anything
[23:53] <ubitux> and i don't see the point in keeping history of all my typo errors
[23:53] <ubitux> or bad wordings
[23:53] <burek> blogs are stupid
[23:53] <burek> to cut the point down to the chase :)
[23:53] <Daemon404> you clearly havent read mike melansons
[23:53] <Daemon404> or kostyas
[23:54] <ubitux> ?
[23:54] <Daemon404> that was directed at burek 
[23:54] <ubitux> ah sorry
[23:54] <burek> no I didn't :)
[23:55] <burek> the last night's 3hrs presentation was boring enough
[23:55] <ubitux> anyway, the whole point of having a static blogging system was, for me, to avoid putting mysql/php/django or any huge unmaintainable monster 
[23:55] <burek> to think twice next time when I accept a link to watch again :)
[23:55] <ubitux> i believe text files in markdown are way more simpler, reliable and sustainable solution :p
[23:55] <burek> blog is just a fancy Changelog
[23:55] <Daemon404> ubitux, django...
[23:55] <Daemon404> seriously fuck django
[23:56] <Daemon404> i tried to set up patchwork for ffmpeg, which uses it
[23:56] <Daemon404> and i failed epically
[23:56] <JEEBsv> flask :3
[23:56] <ubitux> it's always the same story with web technos you know
[23:56] <Daemon404> ubitux, people dont run webservers anymore
[23:56] <Daemon404> they run 40 frameworks
[23:56] <Daemon404> and proxy them all with nginx
[23:56] <ubitux> :)
[23:57] <ubitux> and it sucks hard
[23:57] <Daemon404> and then they proxy nginx with nginx
[23:57] <Daemon404> to laod balance
[23:57] <Daemon404> (really)
[23:57] <ubitux> static blogging with markdown text files ftw
[23:58] <JEEBsv> Daemon404: I've set some stuf up with django, where did you fail?
[23:58] <JEEBsv> :3
[00:00] --- Wed Sep 19 2012