[Ffmpeg-devel-irc] ffmpeg-devel.log.20121003

[burek]

[01:34] <Daemon404> the mplayer dev list is quite the source of hilarity
[01:34] <Daemon404> re: mp3lib
[01:50] <KGB> [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/x8jRyg
[01:50] <KGB> [FFmpeg/master] 8svx: copy start value in output samples. - Nicolas George
[01:54] <Compn> Daemon404 : you just wish you could fork and maintain a dead library for at least 5 years until someone comes along to maintain it again
[01:54] <Daemon404> lol
[01:54] <Daemon404> also
[01:54] <Daemon404> loling at "performance concernss"
[01:54] <Daemon404> for fucking mp3
[01:54] <Daemon404> lolz
[01:54] <Compn> 10% cpu on a 150mhz is the difference between video and a slideshow
[01:54] <Compn> you know this
[01:55] <Daemon404> >150 mhz
[01:55] <Daemon404> i cat hear you back there in 1993
[01:55] <Daemon404> cant*
[01:56] <Compn> there are lots of low powered arm devices too 
[01:56] <KGB> [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/FXGMTA
[01:56] <KGB> [FFmpeg/master] 8svx: fix delta_decode cliping limits - Michael Niedermayer
[01:56] <Daemon404> Compn, not that low powered
[01:56] <Daemon404> maybe a first gen ipod
[01:56] <Daemon404> from 2000
[01:56] <Daemon404> or w/e
[01:56] <Compn> pretty much every device mplayer runs on , but that mplayer does not have accelerated output
[01:57] <Compn> e.g. 700mhz arm tablet
[01:57] <Compn> 700mhz is barely fast enough for most videos now
[01:57] <Compn> unless you have mpeg2 .
[01:57] <Compn> ah you know this already
[01:57] <Daemon404> im reminded of all the bikeshedding over removing dead 3dnow! code
[01:57] <Daemon404> that was only ever useful on a single cpu
[01:57] <Compn> diego is the one with the k3...
[01:58] <Compn> we cant all have i7s 
[01:58] <Compn> :P
[01:58] <Daemon404> bro
[01:59] <Daemon404> i can go to the garbage dump
[01:59] <Daemon404> and find abetter computer
[01:59] <Daemon404> for free
[02:00] <cbsrobot_> Daemon404: you as a bsd lover should know that everybody lot of people have a net4501 at home &.
[02:00] <Daemon404> o u
[02:00] <cbsrobot_> s/everybody//
[02:00] <Daemon404> i just lol at the neckbeards who cling to ancient hw 
[02:01] <Daemon404> like the slackware main servers running on a p3 until 2012
[02:01] <cbsrobot_> firewall and mp3 enocer in one peace
[02:01] <Daemon404> cbsrobot_, because ffmpeg on a firewar is surely no security risk
[02:01] <cbsrobot_> maybe he pipes the spamd log to mp3 - who knows
[02:01] <Daemon404> firewall even
[02:01] Action: Daemon404 also lols at 'pros' who run irssi on their router
[02:01] <Daemon404> also very safe!
[02:02] <cbsrobot_> apple has even a name for such thing &.
[02:04] <Compn> Daemon404 : anyway, this is curse of mplayer , trying to be the 'fastest' :P
[02:05] <Compn> so we support ancient hacks, and optimizations
[02:05] <Daemon404> mplayer itself is an ancient hack
[02:05] <Daemon404> <_<
[02:05] <Compn> am curious what player you use
[02:05] <Daemon404> depends what os im on.
[02:05] <Compn> ugh
[02:05] <Compn> well what players do you use then
[02:06] <Daemon404> i use mpc-hc with lav filters (see: nevcairiel) on windows
[02:06] <Daemon404> and vlc anywhere else
[02:06] <Compn> i use mplayer cross-platform
[02:06] <Compn> :P
[02:06] <Daemon404> i like having a gui
[02:06] <Daemon404> with a seekbar
[02:06] <Daemon404> and such.
[02:06] <Compn> i just bind keys to certain points in the video and never leave my keyboard
[02:06] <Compn> the mouse sits there alone
[02:07] <Daemon404> fine for you
[02:07] <Daemon404> not for most humans
[02:07] Action: Compn cant believe anyone uses mplayer over vlc
[02:07] Action: Compn keeps converting users to vlc
[02:08] <Compn> new strategy, send all users who complain to other software
[02:08] <Compn> enjoy current software with less complaints :)
[02:08] <Daemon404> oh yeah? you dont like libavcodec?! well why dont you go and use --- nvm
[02:08] <Compn> for a second i thought you were going to say you used quicktime on mac :P
[02:09] <Daemon404> you asked what player i used
[02:09] <Daemon404> quicktime is now a player
[02:09] <Daemon404> not
[02:09] <Daemon404> *
[02:09] <Compn> does quicktime still make users pay to get fullscreen ? :P
[02:09] <Compn> ala 'quicktime pro'
[02:10] <Daemon404> dunno
[02:10] <cbsrobot_> quicktime x has no pro
[02:10] <Compn> i havent been on a mac in a while
[02:10] <cbsrobot_> and lost almost all features qt 1 - 7 had
[02:11] <Compn> apple finally got tired of supporting its own dumb features like putting flash in mov
[02:11] <Compn> and i mean swf flash, not flash video
[02:11] <Daemon404> cbsrobot_, so it's like fcp x
[02:11] <Daemon404> killed all usefulness
[02:11] <cbsrobot_> yeah - almost as bad
[02:12] <cbsrobot_> you still can watch movies though
[02:12] <Compn> so whos ready for apple stock to start lowering when the next iphone starts to suck ?
[02:12] <Daemon404> no theyll just release a new one
[02:12] <Daemon404> that sucks 4% less
[02:14] <Compn> any big flamewars i miss this week ?
[02:15] <Daemon404> no
[02:16] <iive> I just started one bakeshed.
[02:16] <Daemon404> im thinking of creating a com[elx versioning system for ffmpeg
[02:16] <Daemon404> based on solar cycles
[02:16] <Daemon404> and the mayan calender
[02:17] <Daemon404> calendar*
[02:53] <KGB> [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/58-TAw
[02:53] <KGB> [FFmpeg/master] 8svx: avoid custom clip, avoid +128 for compressed data. - Michael Niedermayer
[03:03] <iive> Daemon404: actually solar cycles is fine. I think that the sun had a big activity cycle that lasts about 10-11 years, the sun core rotates with about 1 month cycle, so we can use it for regular releases. The best thing... it seems we already follow that scheme :P
[03:09] <Daemon404> lol
[03:16] <ohsix> dont' do one based on the mayan calendar, it's almost over
[03:52] <Compn> Welcome to http://www.multimedia.cx/
[03:52] <Compn> This domain been suspended, please use the Registration URL link to contact your registrar
[03:52] <Compn> hmm
[03:52] <Daemon404> old new
[03:53] <Compn> where was the news posted ?
[03:53] Action: Compn is out of loop
[03:54] <Daemon404> cant remember
[03:54] <Daemon404> mike was notified
[03:54] <Daemon404> and he paid
[03:54] <Daemon404> waitning now
[03:58] <Compn> the ole domain hostage situation
[03:58] <Compn> i've seen it before
[03:59] <Daemon404> i dunno how people miss paying though
[04:00] <Daemon404> i get 10 metric tons of spam saying "YOUR DOMAIN WILL EXPIRE"
[04:00] <Daemon404> if it's coming up anywhere close to soon
[04:02] <KGB> [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/FfTxAA
[04:02] <KGB> [FFmpeg/master] 8svx: Fixing header size, move decoding to per call instead of the first call. - Justin Ruggles
[05:48] <KGB> [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/Hm8MGw
[05:48] <KGB> [FFmpeg/master] ffv1dec: print bps for pict debug too - Michael Niedermayer
[09:36] <ubitux> mmh is the date= thing broken in ffserver?
[09:39] <ubitux> it seems to seek way ahead the requested pts
[09:39] <ubitux> or well, the requested pts is outbounded (>)
[09:39] <ubitux> so it seeks at the end
[09:39] <ubitux> and thus has no effect whatsoever
[09:46] <ubitux> burek: do you remember some users complaining about that?
[09:47] <ubitux> maybe i'm using it wrong though
[09:48] <ubitux> with this configuration: http://pastie.org/4901196, i'm trying:
[09:48] <ubitux> ./ffserver -f ./ffserver.conf
[09:48] <ubitux> ./ffmpeg -re -f lavfi -i testsrc 'http://localhost:8090/video.ffm'
[09:49] <ubitux> ./ffplay 'http://localhost:8090/video.ts?date=09:47:50'
[09:49] <ubitux> with date being a few seconds after starting the push
[09:50] <ubitux> and it's leading to something like wanted_pts=1349250470000000 with pts_min=-40000 (pos=4096) pts_max=16720000 (pos=24854528)
[09:50] <ubitux> wanted_pts is compared against pts_min/pts_max
[09:50] <ubitux> it either a rescale problem or a relative/absolute ts problem
[09:51] <ubitux> (imo)
[09:51] <ubitux> i thought it was related to 16b9156b7 but doesn't look like to :p
[14:12] <Tjoppen> is there a nice way to get decodec video ptses?
[14:13] <Tjoppen> *decoded
[14:13] <Tjoppen> ffprobe -show_frames perhaps :)
[14:15] <Compn> in the debug code somewheres
[14:15] <ubitux> -debug_ts in ffmpeg too
[14:15] <ubitux> showinfo filter too maybe
[14:16] <ubitux> -show_frames/-show_packets
[14:16] <ubitux> infinite possibilities.
[14:16] <Tjoppen> I did -show_frames | grep -A5 video | grep pkt_pts= | head
[14:17] <Tjoppen> I have a file which is behaving a bit strangely. trying to figure out why
[14:18] <ubitux> -of flat might be easier to grep :p
[14:19] <Tjoppen> I just want the numbers, so.. :)
[15:40] <burek> ubitux, i don't remember anything related to that.. :S
[15:40] <ubitux> okay
[15:41] <ubitux> did you see some users having with the date= thing?
[15:43] <burek> I've seen ?buffer=X and Preroll
[15:43] <burek> but not date
[15:44] <ubitux> ok
[16:42] <KGB> [FFmpeg] michaelni pushed 7 new commits to master: http://git.io/L3icoA
[16:42] <KGB> [FFmpeg/master] ARM: set Tag_ABI_align_preserved in all asm files - Mans Rullgard
[16:42] <KGB> [FFmpeg/master] ARM: bswap: drop armcc version of av_bswap16() - Mans Rullgard
[16:42] <KGB> [FFmpeg/master] segment: Pass the interrupt callback on to the chained AVFormatContext, too - Martin Storsjö
[16:51] Action: Daemon404 wonders why he only found out about app verifier today
[16:56] <Compn> Daemon404 : you like ass subs right ?
[16:57] <Compn> do you like ass in srt? http://pastebin.com/mCYwKkmY
[16:57] <Compn> ehe
[16:57] <Daemon404> sadly, vsfilter would likely actually render that
[16:58] <KGB> [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/90xAjw
[16:58] <KGB> [FFmpeg/master] framecrcenc: print flags and side data elements - Michael Niedermayer
[16:58] <Compn> user reports it working in windows with mpc
[16:59] <j-b> of course
[17:00] <j-b> Italians...
[17:05] <Daemon404> Compn, because vsfilter is gabest code
[17:05] <Compn> racissssssssst
[17:06] <Daemon404> i am gabestist
[17:06] <Compn> ?
[17:06] <Compn> ah
[17:06] <Compn> so you support this hybrid new thing ?
[17:06] <Daemon404> opposite
[17:06] <JEEB> I think the only reason it works with GabestCode is because he probably was lazy to limit ass syntax to ass
[17:07] <JEEB> and I think everyone agrees to the fact that it's an abomination no matter how you look at it
[17:07] <j-b> well, the biggest issue is that the first line is not escaped in {\ stuff
[17:09] <ubitux> Compn: just remove the { } escaping in subassconvert if you want to support that crap
[17:09] <ubitux> i explicitely added the escape because i want users to be able to print { } in srt
[17:10] <j-b> ubitux: but wouldn't you print the rest of the first line anyway ?
[17:10] <j-b> since they are outside of { } ?
[17:10] <ubitux> « J'ai fait {Epitech.} »
[17:11] <ubitux> https://encrypted-tbn3.gstatic.com/images?q=tbn:ANd9GcRYyQqRG1-lt3ZeKGnfWLqz6_tLOSQ_4MoE8bap4J3EMZ5RXwSz
[17:11] <j-b> et?
[17:11] <ubitux> what happens when you want to print that?
[17:11] <ubitux> ’ comments
[17:11] <ubitux> the more important word disappear
[17:11] <j-b> then, stop escaping { }
[17:11] <j-b> and do as we do
[17:11] <j-b> {\ } escape
[17:12] <ubitux> i don't understand
[17:12] <j-b> but still, this is {\something} something else
[17:12] <Daemon404> um
[17:13] <j-b> how do you not print the "something else" part?
[17:13] <Daemon404> i would just tell the users to stop using insanely unsupported files
[17:13] <Compn> ubitux : only hide {\ tags with slash}
[17:13] <Daemon404> the fact that vsfilter even works is due to bugs
[17:13] <Daemon404> and it not caring what format it parses
[17:13] <Compn> j-b : user said it didnt work with vlc either :P
[17:14] <j-b> Compn: of course not
[17:14] <j-b> Compn: see my question
[17:14] <Compn> ah
[17:14] <ubitux> <@j-b> how do you not print the "something else" part? // why would i not print it?
[17:14] <j-b> Compn: VLC would hide the first part {\ xxxx }
[17:14] <ubitux> i'm sorry i'm a bit lost
[17:14] <ubitux> you want to escape only with "{\" and not "{"?
[17:15] <j-b> ubitux: because the something else part is gibberish numbers ?
[17:15] Action: Daemon404 cant believe you people are actually talkign about supporting this abomination
[17:16] <Compn> ubitux : sorry, i havent a clue what needs to be done. i just find bug reports ;P
[17:16] <Daemon404> its not a bug
[17:16] <ubitux> it's always the same things imo
[17:16] <Compn> i didnt say 'valid bug reports'
[17:16] <Daemon404> lol
[17:16] <ubitux> either you escape ass tag, or you don't
[17:17] <ubitux> mplayer escape by default
[17:17] <ubitux> so libass renders the subs verbatim
[17:17] <Daemon404> as it should
[17:17] <j-b> Daemon404: for VLC we escape all {\xxx} in srt
[17:17] <Daemon404> in this case, escape = it gets printed verbatim on screen?
[17:17] <j-b> no
[17:17] <Daemon404> oh
[17:17] <j-b> we hide those
[17:18] <Daemon404> then that is wrong.
[17:18] <ubitux> :)
[17:18] <Daemon404> and not teh srt format
[17:18] <j-b> lol
[17:18] <j-b> there is no srt format
[17:18] <Daemon404> there must be! webvtt is based off it!
[17:18] <Daemon404> :)
[17:18] <ubitux> :D
[17:18] <j-b> anyway, on this abomination, this is worse
[17:18] <j-b> because of {\an9\fscx73\1a&HC0&\3a&HFF&\4a&HFF&\p5}
[17:18] <j-b> m 49 99 s 32 131 21 161
[17:18] Action: Compn lulz
[17:18] <ubitux> i'd personnally propose -sub-hack-no-escape
[17:18] <j-b> how do you decide to hide the rest ?
[17:18] Action: Compn rolls d12
[17:19] <ubitux> so ass can interpret it directly.
[17:19] <ubitux> it was proposed several times
[17:19] <Daemon404> j-b, you dont and you tell the person who made it to stop making hilariously wrong files
[17:19] <ubitux> not sure what we do in ffmpeg btw...
[17:19] <j-b> Daemon404: well, here, sure.
[17:19] <Compn> Daemon404 : whip up some ass drama... keeps everyone on their toes :D
[17:20] <Daemon404> is like if i started complaining my perl interpreter wont run nodejs code
[17:20] <j-b> people using ass are morons anyway...
[17:20] <Daemon404> j-b, theres Nothing Better for styled subs
[17:21] <Daemon404> i dont think using it for typesetting is very nice though.
[17:21] <Compn> Daemon404 : you just dont know the power of RealText
[17:21] <ubitux> Daemon404: when will the anime scene switch to photoshopped bitmap subs?
[17:21] <Compn> or vrhtml something -scoobydoo
[17:22] <Daemon404> ubitux, never?
[17:22] <ubitux> :)
[17:23] <Daemon404> also, re: time text
[17:23] <Daemon404> it seems everyone and their mom is making their own 'standard' now
[17:24] <Daemon404> upon the many existing 'standards'
[17:24] <Daemon404> fun times ahead
[17:25] <Daemon404> (and come on... embedded css in subtitles? whats not to love? vlc can start linking webkit!)
[17:25] Action: ubitux needs to start the new subtitles api
[17:25] <ubitux> i need to rewrite almost every text sub codec :(
[17:25] <ubitux> (but the encoders will be way more simpler)
[17:26] <Compn> Daemon404 : you ready to include xml parser ?
[17:27] <Compn> everyone needs an xml parser!
[17:27] <Compn> webkit probably does it ?
[17:28] <ubitux> no you can't use a xml parser Compn 
[17:28] <ubitux> i tried.
[17:28] <ubitux> it's impossible to parse sami and stuff with it
[17:28] <j-b> Daemon404: WebVTT! TTML 4 different specs!
[17:28] <ubitux> \p/
[17:28] <ubitux> \o/
[17:28] <Daemon404> \o/
[17:28] <j-b> Daemon404: http://blog.gingertech.net/2012/09/18/what-is-interoperable-ttml/
[17:28] <Daemon404> yeah i read it
[17:30] <Compn> why you guys dont just patch vsfilter to break these files
[17:31] <Compn> obviously its vsfilter fault
[17:31] <Compn> break _his_ legs!
[17:31] <ubitux> oh.
[18:17] <Compn> Daemon404 : freakin firefox switched to msvc2010 so the builds wont work on win2k either
[18:17] <Compn> OPERA still works :D
[18:17] <Compn> maybe i could find someone with msvc2005 to make me win2k builds
[18:17] <Compn> of firefox 
[18:17] <Compn> :P
[18:18] <JEEB> I think they partially updated their compilers because the old ones would run out of memory during compilation or linking? :D
[18:19] <Compn> jerks.
[18:19] <Compn> at least i know what the problem is now
[18:19] <Compn> next time zeranoe is back in here
[18:19] <Compn> well i guess he still uses mingw :\
[18:19] <Compn> not related
[18:20] <iive> Compn: you should try linux some time. ;)
[18:21] <Compn> ubuntu wouldnt give me root , but wanted me to login to root whenever i made a change
[18:22] <Compn> kind of upset me :P
[18:22] <Compn> couldnt remember my root pw, had to chroot and such
[18:24] <iive> you mean, sudo. try slackware, they got 14.0 few days ago. It is nothing like ubuntu, for good or bad :)
[18:31] <KGB> [FFmpeg] michaelni pushed 1 new commit to master: http://git.io/w6v4KQ
[18:31] <KGB> [FFmpeg/master] tiffenc: remove unused variable - Michael Niedermayer
[18:44] <j-b> Daemon404: http://www.html5rocks.com/en/tutorials/track/basics/ hmmm JavaScript inside subtitles!!
[18:45] <Daemon404> >_>
[18:45] Action: Daemon404 vomits a bit
[18:47] Action: j-b pushes Daemon404 
[18:48] <Daemon404> why so confrontational
[18:49] <Daemon404> bro
[18:49] <j-b> jsut to laugh :)
[19:36] <ubitux> Daemon404: gix seems to have some trouble with the MSVC thing, the doc/plateform.texi might not be accurate
[19:36] <ubitux> maybe you can help him
[19:36] <ubitux> < gix> that's what i did. either you guys use a different version of the c99 wrapper or something is missing. it states to make sure that msvc link.exe is used, but the wrapper calls it with -o instead of -out. hardcoding LD_O in the makefile changes that, but then link.exe cannot cope with /tmp-paths
[19:37] <nevcairiel> all you need to do these days is specify --toolchain=msvc and it should work
[19:37] <nevcairiel> wonder whats in that doc
[19:37] Action: nevcairiel reads
[19:38] <gix> it says the same
[19:39] <Daemon404> did you uild your own wrapper?
[19:39] <Daemon404> or get a binary from somewhere
[19:39] <gix> i compiled the current master of the github repo
[19:40] <gix> (with current llvm/clang master)
[19:41] <Daemon404> humm
[19:41] <Daemon404> can you revert the latest commit to the c99 converter repo
[19:41] <Daemon404> and retyr
[19:41] <Daemon404> i have a hunch...
[19:42] <nevcairiel> the cosmetics patch?
[19:43] <Daemon404> yeah
[19:43] <Daemon404> inb4 i mangled something
[19:48] <gix> Daemon404: nah, doesn't seem like it. same error running configure --toolchain=msvc: http://codepad.org/8His5uDM
[19:48] <Daemon404> which link
[19:48] <Daemon404> inb4 /bin/link.exe
[19:48] <Daemon404> because msvc's link sure doesnt have --help
[19:49] <Daemon404> please note i explicitly mentioned this in the deocs
[19:49] <gix> there's only one link in the path
[19:49] <gix> the one from msvc11
[19:50] <Daemon404> that is wrong
[19:50] <Daemon404> so please show me teh output of which link
[19:50] <gix> $ which link
[19:50] <Daemon404> Try `link --help' for more information.
[19:50] <Daemon404> ^ wiat
[19:50] <Daemon404> is this part of our configure
[19:50] <gix> /c/Program Files (x86)/Microsoft Visual Studio 11.0/VC/BIN/link.exe
[19:50] <Daemon404> -_-
[19:50] <Daemon404> yeah
[19:51] <Daemon404> this is and should be handled by c99wrap.exe
[19:51] <Daemon404> it's bizzare
[19:52] <Daemon404> ive never build compilewrap.c with msvc11 though
[19:52] <Daemon404> only 10
[19:52] <gix> must the wrapper be built with msvc?
[19:53] <Daemon404> i dont think so
[19:53] <Daemon404> ive not tested it with mingw though
[19:53] <gix> i compiled it with clang
[19:53] <Daemon404> oh.
[19:53] <Daemon404> did you also compile teh converter with clang?
[19:54] <Daemon404> because that wont work
[19:54] <Daemon404> libclang and the converter should all be built with msvc
[19:54] <nevcairiel> i compiled it with gcc, and it worked fine
[19:55] <Daemon404> because of some annoyances with libclang
[19:55] <nevcairiel> at least c99wrap, not the converter
[19:55] <Daemon404> they hardcode compiler-specific things
[19:55] <nevcairiel> but i guess thats what causing issues here
[19:56] Action: Daemon404 out for foo
[19:56] <Daemon404> d
[19:57] <gix> well, gonna build c99conv with msvc then. though i don't get why that should matter
[20:08] <ubitux> is c99conv buildable with msvc?
[20:08] <ubitux> i thought it was written in c99
[20:08] <ubitux> can it convert itself btw?
[20:22] <gix> ubitux: just built it with msvc without problems (result is the same though)
[20:23] <Daemon404> ubitux, its not c99
[20:23] <Daemon404> that sounds wrong
[20:23] <Daemon404> many people have been uisng and building it with msvc (msvc10)
[20:23] Action: Daemon404 starts to worry about msvc11
[20:24] <Daemon404> tho nevcairiel said it works with gcc too
[20:24] <crtmpserver> hello to all
[20:24] <crtmpserver> I see a lot of packets being dropped on RTP over UDP with h264 content
[20:24] <crtmpserver> all the sources I have tried (my server, an IP camera, some internet sources), all have the same problem
[20:25] <crtmpserver>  RTP: missed 3 packets
[20:27] <gix> Daemon404: but shouldn't configure call the linker with -out when the msvc toolchain is set? that's at least what it looks like with LD_O etc.
[20:27] <Daemon404> c99wrap converts it
[20:27] <Daemon404> thats the point
[20:29] <Daemon404> btw is this git master?
[20:31] <Daemon404> er, my bad
[20:31] <Daemon404> it does indeed take -out, but ffmpeg's configure DOES set it correctly
[20:31] <Daemon404> (in git master)
[20:36] <gix> Daemon404: yes, 05e5a24f7 to be exact
[20:37] <Daemon404> gix, this is rather strange
[20:37] <Daemon404> can you post your whole config.log
[20:38] <Daemon404> im betting it's environmental and/or something we dotn account for
[20:38] <durandal_1707> crtmpserver: you should report bug, this channel is not for reporting bugs
[20:38] <crtmpserver> tx, I will
[20:38] <crtmpserver> sorry for the noise
[20:41] <gix> Daemon404: http://codepad.org/nfhzb09v
[20:42] <Daemon404> uh
[20:42] <Daemon404> looks like it got cut off
[20:42] <Daemon404> that doesnt contain a single test log
[20:42] <gix> indeed
[20:43] <gix> Daemon404: http://gix.myftp.org/config.log
[20:51] <Daemon404> gix, gonna try something
[20:51] <Daemon404> sec
[20:57] <Daemon404> gix, x86 or x64?
[20:58] <gix> x86
[20:58] <Daemon404> k
[21:02] <ubitux> okay.
[21:02] <ubitux> so it's not the cabac tables!
[21:02] <ubitux> :)
[21:04] <Daemon404> gix, hrm
[21:04] <Daemon404> works here
[21:06] <durandal_1707> michaelni: can we get bot that reports commits in real-time?
[21:07] <microchip_> what happened to CIA?
[21:07] <ubitux> microchip_: you missed all the drama..
[21:07] <microchip_> ?
[21:08] <durandal_1707> DCMA takedown
[21:08] <microchip_> Oo
[21:09] <gix> Daemon404: with -o passed to "c99wrap link"?
[21:09] <ubitux> microchip_: http://pastebin.com/9RBBniM1
[21:09] <microchip_> you mean DMCA?
[21:09] <Daemon404> gix, no
[21:09] <Daemon404> configure detects my compiler correctly
[21:09] <Daemon404> it isnt detecting yours correctly
[21:09] <Daemon404> and is using the defaults
[21:09] <Daemon404> run 'c99wrap cl' and see what it outputs
[21:10] <microchip_> ubitux: ic
[21:10] <gix> the usual output from cl.exe
[21:10] <microchip_> drama
[21:13] <gix> Daemon404: what should "c99wrap link" show?
[21:13] <Daemon404> the usual output
[21:13] <Daemon404> the ms logo
[21:13] <Daemon404> and options
[21:13] <burek> who is cehoyos
[21:13] <burek> is he online here 
[21:13] <Daemon404> carl doesnt see the merits of irc
[21:13] <Daemon404> unfortunately
[21:17] <gix> Daemon404: i see. that's where it went wrong. and you were right and wrong about link.exe. my wrappers picked up an old link.exe from coreutils in the path the wrappers are stored.
[21:17] <Daemon404> ah
[21:17] <Daemon404> isnt this so fun
[21:18] <gix> the wrappers probably shouldn't do that ;)
[21:20] <gix> Daemon404: anyway, want a makefile and some compat stuff (inttypes, snprintf, oldnames) to compile the wrapper with msvc?
[21:25] <Daemon404> no
[21:25] <Daemon404> i have a msvs solution
[21:25] <Daemon404> i just havent pushed it yet
[21:25] <Daemon404> also a README
[21:25] <Daemon404> for how to build a static clang
[21:25] <Daemon404> if one is so inclined
[21:32] <michaelni> j-b, whats the status of a CIA replacement for ffmpeg git ?
[21:39] <j-b> michaelni: almost ready
[21:50] <Compn> burek : carl reads irc logs tho, and you can mail him :)
[21:52] Action: Compn waves at carl
[21:54] <Compn> poor mike
[21:54] <Compn> hope he gets multimedia.cx back
[22:18] <KGB> [FFmpeg] michaelni pushed 2 new commits to master: http://git.io/-aUwhg
[22:18] <KGB> [FFmpeg/master] mux/nut: factorize ff_choose_timebase() out of nut - Michael Niedermayer
[22:18] <KGB> [FFmpeg/master] ff_choose_timebase: only try factors upto 14 - Michael Niedermayer
[23:32] <Skyler_> michaelni: I think this code in the h264 prober is problematic
[23:32] <Skyler_>                 if(p->buf[i+2]&0x0F)
[23:32] <Skyler_>                     return 0;
[23:32] <Skyler_> because those bits are no longer reserved; constraint_set4 and 5 exist now.
[23:41] <michaelni> Skyler_, i think i fixed this last year: http://git.videolan.org/?p=ffmpeg.git;a=commitdiff;h=5deedf3552819e5bea3327a450155bb57643a999
[23:42] <Skyler_> oh, it's not in the latest version, I see
[23:42] Action: Skyler_ goes to slap libav people
[23:44] <j-b> michaelni: the bot for commit is ready. It will be tested on #videolan and pushed here when ready.
[23:44] <ubitux> has it colors? *_*
[23:45] <JEEBsv> fabulous colors needed
[23:46] <iive> wow double rainbow.
[23:46] <j-b> ubitux: yes
[23:46] <ubitux> \o/
[23:48] <ubitux> wbs: i see you're doing some segmenter work; did you see the work done in ffmpeg on it?
[23:54] <wbs> ubitux: I didn't, and I'm not really doing work on it for the sake of adding functionality, I'm just trying to clean up whatever code we have there already
[23:55] <Daemon404> j-b, what is it based off of?
[23:55] <Daemon404> the bot
[23:55] <j-b> Daemon404: you have objections against a language?
[23:55] <Daemon404> no?
[23:55] <Daemon404> im curious
[23:55] <Daemon404> i had an idea for my own setup
[23:55] <Daemon404> which may or may not suck
[23:55] <Daemon404> <-- idea stealer
[23:56] <j-b> Daemon404: ask thresh
[23:59] <ubitux> wbs: well there is some kind of interesting features like HLS; also you might be able to find some fixes
[23:59] <ubitux> anyway, you obviously do as you wish :p
[23:59] <wbs> ubitux: as said, that was not my current target
[00:00] --- Thu Oct  4 2012