[FFmpeg-devel-irc] IRC log for 2010-02-15

[irc at mansr.com]

[01:03:54] <astrange> two days of ffmpeg-mt debugging and i think it's going to result in 2 one-line commits
[01:03:57] <astrange> oh well
[01:04:24] <Dark_Shikari> it happens.
[01:04:56] * Kovensky pets astrange
[01:05:07] <mru> one line per day, not bad
[01:05:16] <Dark_Shikari> the most annoying deadlock I ever dealt with was a one line fix
[01:05:55] <mru> those are always the most annoying
[01:06:02] <Dark_Shikari> >Last Name is invalid. Name cannot contain numeric or special characters (i.e. %,$,#).Last name is required
[01:06:04] <mru> especially when it's the hardware that deadlocks
[01:06:06] <Dark_Shikari> Fuck you, I have a hypenated name
[01:06:40] * mru hands Dark_Shikari a dehyphenator
[01:06:54] <astrange> i signed up for a bioware account using astrange+bw at gmail, then found you couldn't type + in the xbox keyboard...
[01:06:55] <mru> now you have a camelcased name
[01:07:20] <Kovensky> Dark_Shikari: http://www.cracked.com/article_17575_7-true-stories-that-prove-airlines-hate-you_p1.html <-- #4
[01:08:06] <Dark_Shikari> mru: yup
[02:04:00] <saintdev> Dark_Shikari: that deadlock that we spent a few days debugging =p
[02:04:05] <Dark_Shikari> yup
[03:42:17] <astrange> ==91450== Source and destination overlap in memcpy_chk(0x11bd6ab08, 0x11bd6ab08, 40)
[03:42:29] <astrange> is memcpying something to itself forbidden?
[03:42:42] <astrange> the warning is technically wrong anyway, darwin memcpy is memmove
[03:45:21] <astrange> oh, it is undefined. that's a pain, i have to add boilerplate to handle it
[04:06:03] <astrange> hm, it'd be easier to debug mt seeking if regular seeking didn't already deadlock
[04:15:35] <elenril> Honoome: ?
[05:38:17] <elenril> lol at xkcd
[05:41:35] <astrange> Kovensky: try theora seeking now (mpeg1 was broken too, i half-fixed it but just turned it off for now anyway)
[05:46:20] <saintd3v> moose and squirrel =D
[06:47:59] <pJok> mornings jai :)
[06:50:49] <jai> good morning pJok :)
[06:51:58] <benoit-> Good morning
[07:38:06] <jai> http://fate.multimedia.cx/index.php?stderr=181631
[07:38:07] <jai> heh
[07:41:34] <siretart`> good morning
[07:42:05] <av500> gm
[07:44:51] <KotH> salut
[07:53:38] <av500> KotH: about the player, I think I have to pass
[07:54:00] <av500> 64g I can only offer in a hdd based pmp
[07:54:26] <av500> hmm, I could offer 60gb in an a5 though, 1.8" hdd
[08:29:47] <siretart`> FYI, I've prepared test packages for ubuntu lucid with ffmpeg 0.5+libx264 API version 84 in my PPA. testers/comments welcome
[08:29:49] <siretart`> Dark_Shikari: ^^
[08:33:40] * Dark_Shikari doesn't use ubuntu
[08:33:42] <Dark_Shikari> but nice.
[08:38:53] <siretart`> Dark_Shikari: I didn't try yet (I think), but what behavior do you expect if the updated presets are used with an libx264 api version 67? just work or weird behavior?
[08:39:47] <Dark_Shikari> it will work fine
[08:39:55] <Dark_Shikari> it _won't_ work with an old _ffmpeg_
[08:40:34] <superdump> people shouldn't mix presets between ffmpeg packages anyway...
[08:40:47] <siretart`> Dark_Shikari: that's interesting to know, because I have to declare that dependency manually (and my test packages don't have them)
[08:40:58] <Dark_Shikari> why do you have to?
[08:41:01] <Dark_Shikari> they're part of the package
[08:41:35] <siretart`> actually not
[08:41:57] <siretart`> hm. I could probably move the presets from the 'ffmpeg' package to 'libavcodec'
[08:42:03] <Dark_Shikari> agreed.
[08:42:06] <Dark_Shikari> they don't make sense with anything else
[08:42:18] <superdump> where's the parsing code?
[08:42:44] <siretart`> that'd be a bit of waste for users that don't install 'ffmpeg', but only install libavcodec via some dependency
[08:42:58] <siretart`> the presets are only useful for the /usr/bin/ffmpeg binary
[08:43:05] <siretart`> that's why I've placed them there in the first place
[08:43:51] <Dark_Shikari> a bit as in 10 kilobytes
[08:43:53] <Dark_Shikari> big whoop
[08:45:20] <siretart`> I'm just stating my rationale for why I did so that time
[08:45:36] <siretart`> I don't think it's a big deal to change that
[08:45:44] <Dark_Shikari> ah k
[08:47:06] <siretart`> Dark_Shikari: I've posted your patch to ffmpeg-devel. maybe you want to comment on that :-)
[08:49:56] <KotH> av500: hmm.. i'm looking rather for something like an a2
[08:50:18] <av500> 8gb only
[08:50:20] <KotH> av500: so i guess, i have to wait another year or so until flash sizes double
[08:50:29] <KotH> .o0(quatruple)
[08:50:48] <av500> wait 18mo for 2x (simple moore
[08:50:51] <av500> wait 18mo for 2x (simple moore)
[09:04:25] <KotH> currently it's a bit faster, as prices for flash chips are dropping rapidly due to larger volumes
[09:04:55] <av500> yes
[09:05:05] <av500> still I doubt "simple" mp3 players will be 64GB soon
[09:05:13] <av500> that will be pmp with video etc..
[09:05:22] <KotH> yeah..
[09:05:41] <KotH> though, there are "simple" players on the market with taht size
[09:05:46] <Dark_Shikari> you can already get a 2GB mp3 player for $10
[09:05:49] <KotH> like the iaudio x5 and the ipod
[09:05:56] <Dark_Shikari> you can get a $30 mp3 player with video support and a screen
[09:05:58] <superdump> i saw a toshiba development where they're using short range radio EM and stacking chips with controllers to make a 1TB SSD with a footprint the size of a stamp
[09:06:11] <superdump> i guess they're a bit chunkier as the chips are stacked though
[09:06:31] <KotH> superdump: nope, stacked chips are not much bigger than normal chips
[09:06:32] <superdump> data cube
[09:06:38] <superdump> oh
[09:06:48] <superdump> even if it's 8 chips stacked?
[09:06:49] <KotH> the little bit of additional height is neglicible in most applications
[09:06:59] <Dark_Shikari> remember the silicon itself is thin
[09:07:09] <twnqx> a fully cased IC can be less than 1mm
[09:07:15] <KotH> a die is usualy 0.5mm-1mm
[09:07:17] <superdump> nifty
[09:07:23] <KotH> you can thin it down to 0.1mm
[09:07:38] <superdump> a 1TB stamp
[09:07:40] <superdump> :)
[09:07:43] <twnqx> yeah, the once used for wirebonding to PCBs are like that :P
[09:07:45] <twnqx> ones*
[09:07:50] <Dark_Shikari> datacube ftw
[09:07:55] <superdump> lol
[09:09:44] <superdump> they said it'd be 2012 before they roll out a proof of concept though
[09:09:47] <av500> KotH: ok, I stand corrected
[09:10:35] <superdump> a computer stuff retailer in the uk had an offer on a 32GB SSD and 1TB HDD the other week
[09:10:50] <superdump> i think that's being encouraged as the way to go for the time being
[09:11:24] <superdump> i also saw an in-one-package SSD/HDD combination that used some kind of 'smart' (yeah... right) caching
[09:13:42] <KotH> av500: ?
[09:14:18] <av500> [10:05] <KotH> though, there are "simple" players on the market with taht size
[09:14:19] <av500> :)
[09:14:30] <KotH> superdump: that has been done on bigger nfs servers for years
[09:14:45] <KotH> av500: :-)
[09:15:00] <KotH> av500: did you find any with flash storage? or just the x5 and the ipod?
[09:15:46] <superdump> it would be nice to eliminate all audible noise from electrical devices
[09:16:01] <av500> KotH: 60gb in the x5...
[09:16:10] <av500> thats flash
[09:16:22] <av500> or a drive?
[09:16:34] <KotH> av500: 1.8" drive
[09:16:40] <KotH> av500: i have one :)
[09:16:40] <av500> ah
[09:16:57] <av500> so the same drive we use in the a5/60gb
[09:17:16] <superdump> av500: you work for cowon?
[09:17:32] <KotH> unfortunately, the battery needs to be replaced and the software has problems with more than 10k files
[09:17:47] <av500> KotH: 65536 ftw! :)
[09:17:55] <av500> superdump: nope :)
[09:18:04] <Bagder> KotH: rockbox doesn't!
[09:18:14] <KotH> Bagder: yeah.. i know, i should switch somewhen
[09:18:28] <KotH> Bagder: and i will if i cannot find another mp3 player within a reasonable time
[09:18:32] <av500> Bagder: I remember a "max # of files per folder" setting, no?
[09:18:45] <Bagder> yes but that's not quite like the X5's problem
[09:19:08] <KotH> but that means i have to replace the battery, which means i have to find a PL553450 compatible single cell lion pack with electronics
[09:19:49] <av500> with?
[09:19:50] <KotH> and finding single cell lion packs with electronics is difficult if you need just one
[09:19:59] <KotH> av500: schutzelektronik
[09:20:07] <av500> feigling
[09:20:14] <KotH> ^^'
[09:20:40] <av500> superdump: I work for a competitor ... :)
[09:20:46] <superdump> aha
[09:20:50] <KotH> av500: the current cell has one, hence i'm not going to replace it with one that hasn't
[09:20:59] <av500> reuse the one you have
[09:21:24] <KotH> av500: have you worked with single cell lion packs?
[09:21:51] <av500> yup
[09:22:08] <av500> never with such tiny one though :)
[09:22:09] <KotH> then you know how easy to solder they are ^^'
[09:22:37] <av500> no idea, we have MMe Lam for that
[09:24:01] <KotH> ^^'
[09:33:04] <Dark_Shikari> siretart`: new x264 is up.
[09:33:12] <Dark_Shikari> give it a few days of testing and hopefully we'll be good for beta release
[09:33:34] <siretart`> Dark_Shikari: you mean you have a candidate for lucid?
[09:35:05] <Dark_Shikari> I was thinking of just making the latest the candidate for now, plus any bugfixes after
[09:35:28] <siretart`> sounds reasonable to me
[09:35:43] <Dark_Shikari> 26 commits at once ftw
[09:35:58] <siretart`> Dark_Shikari: btw, perhaps you can suggest a better version string than '1:0.svn20100213+gitfcf70c' (that's what I'm currently using)
[09:36:23] <siretart`> how about a proper release number?
[09:36:45] <Dark_Shikari> why not use the output of version.sh?
[09:36:50] <Dark_Shikari> i.e. what x264 itself uses?
[09:37:05] <siretart`> that would be ""
[09:37:16] <Dark_Shikari> erm...
[09:37:20] <Dark_Shikari> look a tad more carefully
[09:37:46] <siretart`> my packaging branch is not based on 'your' x264 branch. I'm importing tarballs from yours
[09:37:53] <siretart`> for consistency reasons
[09:37:55] <Dark_Shikari> what difference does that make?
[09:37:59] <Dark_Shikari> and we don't have branches
[09:38:16] <Dark_Shikari> as long as you have .git, you can get a valid version number
[09:38:17] <siretart`> branch as in 'a series of commits'
[09:38:20] <Dark_Shikari> #define X264_VERSION " r1442 781d300"
[09:38:21] <Dark_Shikari> #define X264_POINTVER "0.85.1442 781d300"
[09:40:31] <siretart`> hm. even in the upstream branch ./version.sh gives me ""
[09:40:33] <siretart`> not helpful
[09:40:47] <Dark_Shikari> it writes it to config.h
[09:41:12] <siretart`> so I need to build it to get a version number? that's clumsy
[09:41:18] <Dark_Shikari> no
[09:41:25] <Dark_Shikari> you could just run version.sh and read config.h
[09:41:25] <Dark_Shikari> lol
[09:41:34] <siretart`> can't you just use 'git tag' and I'm basing the package version on that?
[09:41:43] <Dark_Shikari> we don't do tags or branches
[09:41:51] <siretart`> that'd be helpful
[09:41:55] <Dark_Shikari> if it would be useful, maybe we could tag every single revision with its number
[09:42:00] <Dark_Shikari> scripts welcome
[09:42:22] <siretart`> tagging each and every revision is not helpful
[09:42:30] <siretart`> tagging revision you expect distros to use would be
[09:42:37] <Dark_Shikari> sure it's helpful
[09:42:37] <siretart`> that's not scriptable
[09:42:42] <Dark_Shikari> it lets you match revision numbers with revisions
[09:43:20] <siretart`> I guess I miss what you mean with 'revision'
[09:43:55] <Dark_Shikari> I'm just saying people have complained about that
[09:43:57] <twnqx> x264 revisions are git revisions, there's nothing else
[09:44:02] <Dark_Shikari> hard to convert "r1283" to a git rev number
[09:44:05] <Dark_Shikari> er, git rev hash
[09:44:09] <jez9999> Are there any known problems when piping packets through Windows ICS that cause video decoding problems in ffmpeg?  I'm having a problem where, when the same camera is routed through a Linux box or router, its data decodes OK, but when through Windows ICS, avcodec_decode_video2 fails.
[09:44:11] <Dark_Shikari> twnqx: we use rev numbers though
[09:44:30] <twnqx> which are derived from the number of git commits, if i read the script correctly :)
[09:44:35] <jez9999> it's UDP (RTP), by the way
[09:44:42] <Dark_Shikari> twnqx: yes, but they're based off the svn numbers
[09:45:26] <siretart`> #define X264_VERSION " r40+9 4e3eff0" and #define X264_POINTVER "0.84.40+9 4e3eff0" is what ./version.h produces in 'my' packaging branch. that's nonsense
[09:45:43] <Dark_Shikari> lol
[09:45:44] <twnqx> no
[09:45:46] <Dark_Shikari> wtf did you do to it
[09:46:27] <siretart`> Dark_Shikari: i'm cloning from http://git.debian.org/?p=pkg-multimedia/x264.git;a=summary and use the 'ubuntu' branch
[09:46:45] <siretart`> as said, ./version.h is pretty useless to me in its current form
[09:46:58] <siretart`> oh, I didn't push my revision there yet, btw
[09:47:06] <Dark_Shikari> um
[09:47:08] <Dark_Shikari> you could like
[09:47:08] <Dark_Shikari> run it
[09:47:14] <Dark_Shikari> and copy the output by hand (omg!)
[09:47:20] <Dark_Shikari> once every 6 months when you update the package (what!)
[09:47:26] <Dark_Shikari> that's what I originally suggested
[09:47:34] <Dark_Shikari> it's 20 extra seconds for each update
[09:47:36] <Dark_Shikari> 20 seconds every 6 months
[09:47:54] <siretart`> yes, but you have sucessfully distracted from my original question
[09:48:03] <Dark_Shikari> no I haven't
[09:48:05] <Dark_Shikari> that's the format I recommend
[09:48:32] <siretart`> please state an example. what should the version number look like instead of '0.svn20100213+gitfcf70c'
[09:48:35] <twnqx> 0.79.1330M af612d7... hm, i should update, too
[09:49:11] <Dark_Shikari> siretart`: what does ubuntu use for ffmpeg for example?
[09:49:41] <siretart`> currently lucid is at: 4:0.5+svn20090706-5ubuntu4
[09:50:02] <siretart`> to indicate that the 0.5 branch is tracked and last updated at that date
[09:50:17] <siretart`> 5 revisions in debian, +4 in ubuntu
[09:50:35] <Dark_Shikari> something like 0.85.r1442 git781d300, maybe with a date stuck in there somewhere
[09:50:42] <elenril> why do you use tarballs btw?
[09:50:44] <siretart`> spaces are not allowed
[09:50:54] <siretart`> elenril: I need to compile a source package, which requires a tarball
[09:50:55] <Dark_Shikari> siretart`: whatever, use a -
[09:50:58] <Dark_Shikari> I didn't mean literally that
[09:51:05] <twnqx> wait, source packages require tarballs?
[09:51:17] <twnqx> oh well.
[09:51:39] <elenril> can't you just git archive or whatever that's called
[09:52:16] <twnqx> what the hell did my 264 update just do
[09:52:28] <siretart`> Dark_Shikari: 1:0.85.r1442+git781d300 would indeed be an option.
[09:52:28] <Dark_Shikari> basically siretart`
[09:52:41] <Dark_Shikari> would it be too unreasonable to find a way to get a datestamp there too?
[09:52:50] <Dark_Shikari> though that's already freaking long
[09:52:50] <siretart`> though '1:0.85.r1442' might do as well
[09:52:57] <Dark_Shikari> people would complain about r1442
[09:53:01] <Dark_Shikari> no immediate match to git
[09:53:20] <Dark_Shikari> unless we implemented autotagging/etc
[09:53:23] <siretart`> you can look that up from /usr/share/doc/libx264/changelog.Debian.gz
[09:53:38] <siretart`> elenril: that's what I'm using
[09:54:11] <twnqx> my live ebuild is broken :( it build the new x264 executable against the old lib...
[09:55:06] <twnqx> Dark_Shikari: 0.85.1442M <- wouldn't the M indicate local changes? :S
[09:55:13] <Dark_Shikari> Yes
[09:55:55] <elenril> siretart`: i mean why isn't git://git.debian.org/pkg-multimedia/x264.git a fork of official repo
[09:56:33] <elenril> iirc for ffmpeg you said it was because of svn externals
[09:56:43] <elenril> but that doesn't apply here
[09:56:52] <siretart`> elenril: for consistency reasons. x264 is team maintained and having different repo styles on team git archives is not helpful
[10:01:32] <elenril> and that is worth losing history/making it unmergeable with official repo? :/
[10:03:57] <siretart`> elenril: have you ever worked with packaging branches?
[10:05:48] <elenril> no
[10:06:01] <siretart`> I imagined so
[10:06:17] <elenril> well explain it to me then (if you have time)
[10:07:20] <siretart`> in short: yes, it is worth, we use quilt to manage upstream changes.
[10:07:55] <siretart`> sorry, no time for more elaborated explanation
[10:10:17] <elenril> ok, maybe later
[10:10:30] <elenril> i was just wondering why some repos track upstream and some don't
[11:15:55] <Compn> cmt                 FlixEngineLinux_8.0.15.3 (www.on2.com)
[11:16:06] <Compn> hah, i found a flix-encoded-mp4 file :D
[11:16:20] <Compn> too                 Lavf52.16.0
[11:16:46] <jez9999> Are there any known problems when piping packets through Windows ICS that cause video decoding problems in ffmpeg?  I'm having a problem where, when the same camera is routed through a Linux box or router, its data decodes OK, but when through Windows ICS, avcodec_decode_video2 fails.
[11:16:49] <jez9999> it's UDP (RTP), by the way
[12:00:15] <pross-au> wb kshishkov
[12:01:13] <kshishkov> howdy cobber
[13:35:49] <mru> Dark_Shikari: why 3 mails per commit on x264-devel?
[13:36:20] <kshishkov> mru: one for devs, one for public and one for archival purposes
[13:42:54] <merbzt> how do I define a vararg function pointer ?
[13:43:05] <mru> ...
[13:43:12] <kshishkov> void (*func)(int arg0,...)
[13:43:28] <Honoome> mru: yeah that's the right answer in this case ;)
[13:45:12] <elenril> Honoome: you wanted something about metadata?
[13:45:44] <Honoome> elenril: ah yeah… the ffmpeg snapshot in gentoo has been updated a couple of days ago, and now my flac→alac recording drops the Artist field
[13:46:11] <Honoome> before I go knocking on the maintainer's door, can you tell me whether something changed/was broken recently? :)
[13:46:34] <elenril> yeah, you should probably blame me
[13:46:56] <elenril> what container?
[13:47:17] <merbzt> so I need at least one argument and the the ...?
[13:47:20] <Honoome> flac→m4a[alac] -- can alac be encoded in anything else? :P
[13:47:37] <merbzt> hmm int (*callback)(, ...); seemed to work
[13:47:39] * elenril has no idea wtf alac is
[13:47:50] <Honoome> -f ipod fwiw as that's the only way to get it to be seen by an ipod to begin with (and my only reason to use alac in the first place)
[13:47:54] <merbzt> apple lossless
[13:48:23] <merbzt> mov/mp4
[13:48:39] <elenril> yeah, somebody forgot to change author->artist in the mov muxer
[13:48:48] * elenril wonders who could that be :)
[13:50:21] * Honoome gets ready to ask for a new snapshot/backport of FFmpeg for Gentoo :D
[13:50:46] <Kovensky> someone should ask for a new snapshot of mplayer on fbsd
[13:50:51] <Kovensky> they still have rc2 on ports
[13:50:52] <Kovensky> lol
[13:50:53] <Honoome> I wouldn't have noticed at all if it wasn't that I'm ransacking Magnatune after getting a subscription ;)
[13:51:06] <Kovensky> (they somehow managed to beat debian stale)
[13:51:13] <Honoome> Kovensky: I'm afraid until the next 7.x release is out they have frozen ports
[13:51:22] <Kovensky> I'm on 8.0
[13:51:48] <Kovensky> well, I'm actually on win7, since it's the only OS that SiS still supports <_<
[13:52:11] <Kovensky> I thought about putting fbsd as main but they don't have sis191 drivers
[13:52:24] <Kovensky> so I'm with it on a vm ._.
[13:53:06] <elenril> write one for linux ;)
[13:56:47] <mru> merbzt: variadic functions need at least one fixed argument
[13:56:52] <Kovensky> linux already has one =p
[13:56:56] <Kovensky> just not a video one
[13:59:49] <elenril> Kovensky: you had some autoaway on detach script, right?
[14:00:09] <Kovensky> it's some script on scripts.irssi.org
[14:01:18] <merbzt> mru: msvc didn't choke on int (*callback)(, ...);
[14:01:33] <merbzt> does that mean the first argument is void ?
[14:04:33] <kshishkov> merbzt: IIRC, first arg is needed for varargs to access the rest of them
[14:04:56] <mru> merbzt: that's not valid C
[14:05:01] <mru> never trust msvc
[14:05:02] <merbzt> :)
[14:05:15] <mru> there is no such thing as a void arg
[14:05:34] <kshishkov> mru: but I trust MSVC to not compile vanilla FFmpeg
[14:06:59] <Honoome> kshishkov: something along the lines of “if it builds with MSVC there's something fundamentally wrong with it”?
[14:08:29] <kshishkov> Honoome: no, not necessary. Helloworld should work fine with it, for example
[14:08:45] <Honoome> ah true
[14:08:49] <KotH> is MSVC finally c99 compatible?
[14:08:55] <mru> no
[14:09:00] <KotH> or are they still stuck with c89+extensions?
[14:09:03] * Honoome has had some nefarious experience with MSVC and C++…
[14:09:12] <Honoome> last time I had to develop under Windows I went for Borland^W CodeGear
[14:09:26] <mru> it's still c89+extension-amputations
[14:11:08] * KotH knows hwy he doesnt develop apps for windows
[14:11:34] <Kovensky> they will never get c99 compatible
[14:11:54] <Kovensky> they're a Visual C compiler, and that's C++ with extensions
[14:12:03] <Kovensky> c89 just happens to compile on it ._.
[14:14:17] <KotH> so, c will move on and they will be stuck with their old cruft?
[14:14:26] <KotH> .o0(why does that sound so familliar?)
[14:14:50] <kshishkov> KotH: because it's usual lifecycle?
[14:15:49] <Kovensky> well, the only way they'll be c99 compatible is if c++0xA allows it
[14:15:54] <KotH> depends, most systems get rid of such stuff quite fast
[14:15:56] <Kovensky> or is it 0xB? =p
[14:16:30] <KotH> unlike windows, which could run 16bit os/2 apps until vista
[14:17:15] <kshishkov> Kovensky: the funniest thing is MSVS _Embedded_ - it compiles only C++ (and refuses to take .c completely) but does not allow using any of standard C++ headers or libraries
[14:17:38] <Kovensky> lol?
[14:17:46] <Kovensky> how the hell are you supposed to do anything on it
[14:18:06] <kshishkov> use <stdio.h> instead of <fstream.h> in C++, of course
[14:18:48] <Kovensky> =p
[14:19:23] <kshishkov> I found that gcc ported to WinCE served me better
[14:19:43] <KotH> eh.. same here
[14:19:47] <kshishkov> (and it's hard to type on that small onscreen keyboard)
[14:19:51] <KotH> we use lcc for windows apps :)
[14:20:40] <Kovensky> it
[14:20:44] <Kovensky> hurf
[14:22:42] <kshishkov> offtopic - if SMART reports error reading sector, does that mean HDD has badblock?
[14:23:41] <mru> it can mean that
[14:23:48] <Kovensky> probably, specially if the error repeats itself
[14:23:53] <av500> kshishkov: replace hdd
[14:24:01] <merbzt> most likely the HD shouldn't be used for vital info
[14:24:23] <av500> some people need anime to live, no? :)
[14:24:25] <kshishkov> and if I hear click, system freezes for a second and reports read error?
[14:24:45] <kshishkov> and that can be reliably reproduced?
[14:24:47] <Kovensky> definitely something bad is going on
[14:24:47] <mru> I always replace drives on the first sign of error
[14:24:56] * Kovensky used to get that with maxtor drives
[14:25:08] <mru> unless I know I accidentally nudged a cable or something
[14:25:11] <Kovensky> I had all 4 of my maxtors die on me on the same way
[14:25:25] <mru> always the same story with maxtor
[14:25:29] <kshishkov> I gave that netbook to warranty service center. Thay haven't found a thing and touchpad stopped working.
[14:25:48] * kshishkov had Fujitsu MPG
[14:26:28] <elenril> how do i check why the regtests are dying?
[14:26:53] <elenril> is ffmpeg's stdout/stderr logged somewhere?
[14:28:06] <kshishkov> yes
[14:28:23] <mru> elenril: make test V=2
[14:28:42] <elenril> thanks
[14:28:54] * thresh doesnt run any build commands without |& tee logfile
[14:30:46] * Kovensky detects bashism
[14:31:00] <thresh> nope, a zshism
[14:31:09] * KotH detects people who do not believe in being bourne again
[14:31:27] * mru bashes shell extensions
[14:32:06] <av500> dash ftw!
[14:32:27] <KotH> mru: are you a solaris 6 user? ;)
[14:33:09] * Kovensky is reminded to leech a solaris iso
[14:33:17] * Kovensky wants to experiment with more OSes
[14:33:21] <KotH> kshishkov: dont!
[14:33:31] <KotH> kshishkov: do not follow the path to darknes!
[14:33:45] <av500> Kovensky: I have signed minix3 cd, willing to sell it :)
[14:33:46] * thresh somehow feels KotH's IRC client sucks
[14:33:47] * Kovensky pets kshishkov
[14:33:56] <KotH> av500: lol
[14:34:10] <KotH> er..
[14:34:12] <KotH> well...
[14:34:23] <KotH> tab complete, you know ^^'
[14:34:26] <Kovensky> I have toyed with DOS, win3.11, win9x, winnt, several flavors of linux2.6, now fbsd
[14:34:31] <Kovensky> next on the list are solaris and osx
[14:34:59] <KotH> Kovensky: os/2 missing!
[14:35:08] <KotH> beos too!
[14:35:23] * mru has messed with OSE
[14:39:04] <elenril> lawl, asfdec can't read files written by asfenc
[14:39:15] <elenril> if there's any metadata in them
[14:39:30] <av500> M$ does not want anybody to convert from ASF, so that makes sense :)
[14:39:32] * elenril wonders if he broke it himself
[14:42:59] <elenril> it seems asfenc is used really often :)
[14:43:14] <elenril> mplayer plays them fine though
[14:58:33] <elenril> anyone knows what asf version does asfenc write?
[15:01:14] <kshishkov> do you expect anyone to remember the whole version GUID?
[15:01:49] <elenril> no, i expect 1.0 or 2.0
[15:01:57] <kshishkov> 1.0 then
[15:02:02] <elenril> you sure?
[15:02:09] <kshishkov> "version 2 is patented and nobody uses it"
[15:02:11] <elenril> it has no docs
[15:03:20] <BBB_> version 2 is documented by MS and unused
[15:03:25] <BBB_> version 1 is undocumented and used
[15:03:35] <BBB_> so presumably we write 1.0
[15:04:00] <elenril> actually 2.0 describe what the muxer is doing quite well
[15:05:51] <elenril> pecifies an array of Unicode characters that contains
[15:06:11] <elenril> pure w00tnes, didn't they hear there are multiple representations of unicode?
[15:06:27] <kshishkov> yes, M$ way and right way
[15:06:37] <Kovensky> no, to microsoft there is only one way
[15:06:39] <Kovensky> that is UTF-16LE
[15:06:40] <kshishkov> IIRC they always used UCS-16 for Unicode
[15:06:48] <elenril> then we're DoingItWrong
[15:06:54] <elenril> not that anybody cares
[15:06:55] <Kovensky> kshishkov: was UCS2 until win2k, when they changed to UTF16LE IIRC
[15:07:09] <Kovensky> they're almost identical, except UTF16 supports surrogate pairs
[15:07:17] <kshishkov> Kovensky: oh yeah, what about bug-for-bug compatibility?
[15:07:24] * justlooking if Kovensky likes obscure OS's to try then he should perhaps look at  #aros irc://irc.freenode.net/aros.dev http://aros.sourceforge.net/
[15:12:50] <elenril> hmm, so asfdec fails at reading the extended content header
[15:13:01] * elenril wonders why is he wasting time at this
[15:14:53] <Honoome> self-hatred?
[15:15:33] <elenril> probably
[15:16:14] <Honoome> that's my answer when I ask myself why I keep on working on Ruby
[15:16:28] <elenril> lol
[15:32:06] <BBB> grin @ last line of peloverde message on ffmpeg-devel@
[15:32:43] <peloverde> BBB that actually is/was an issue
[15:33:53] <bilboed-pi> peloverde, alex converse ?
[15:34:17] <bilboed-pi> peloverde, saying some software is bad because you don't understand it doesn't really help
[15:35:24] <bilboed-pi> (unless you meant demuxers from gst-plugins-bad :) )
[15:35:49] <KotH> .o0(round one! fight!)
[15:36:02] <peloverde> bilboed-pi, I believe you said the same thing
[15:36:15] <peloverde> then I explained to you exactly in what cases you need to set needs parsing for aac
[15:36:23] <Honoome> KotH: should we provide weapons or leave them at using punches?
[15:36:23] <peloverde> and then... nothing happened
[15:36:31] <bilboed-pi> peloverde, maybe we've got other things to do ?
[15:36:37] <bilboed-pi> peloverde, you can blame me for not reactin
[15:36:57] <peloverde> bilboed-pi, comments in your code still says that my code is broken
[15:37:03] <bilboed-pi> peloverde, anyway, having OOB fixes is always a good thing imho
[15:37:11] <peloverde> true
[15:37:32] <KotH> Honoome: we want them to stay alive, so no weapons
[15:37:58] <bilboed-pi> Honoome, I choose fencing then !
[15:38:23] * elenril agrees with KotH, peloverde needs to stay alive until sbr is committed
[15:38:33] <Honoome> bilboed-pi: you can use the unlootable dagger given at character creation then :P
[15:38:34] <Kovensky> then he'll be disposable?
[15:38:52] <siretart`> peloverde: do you think that and/or similar patches should go to the 0.5 branch as well?
[15:39:15] <peloverde> siretart`: probably makes sense
[15:40:35] <merbzt> is Baptiste on holiday or something ?
[15:41:40] <peloverde> I've had that patch kicking around since around when 0.5 was released
[15:43:58] <superdump> peloverde: so what does need fixing in gst demuxers with respect to aac?
[15:45:49] <bilboed-pi> superdump, the fixing's been done in demuxers
[15:45:53] <peloverde> superdump, http://bugzilla.gnome.org/show_bug.cgi?id=566250
[15:46:49] <bilboed-pi> the demuxers now properly include in the caps how the aac data is packetized (or not)
[15:47:06] <bilboed-pi> so basically we only want to use the AVParser IFF it's raw ADTS
[15:48:47] <superdump> bilboed-pi: yup
[15:52:42] <peloverde> mru, which const is misplaced?
[15:52:52] <mru> one of them
[15:53:04] <mru> const char const * should be const char *const
[15:53:10] <mru> but it should all be const char foo[]
[16:02:20] <peloverde> superdump, Can I tag decode_pce() as av_cold? We only call it once now
[16:02:33] <superdump> sure
[16:02:35] <mru> once per stream?
[16:02:38] <mru> then it should be av_cold
[16:02:55] <peloverde> mru, yes, unless there is garbage at the beging of the stream than it is until we get a valid frame
[16:03:03] <peloverde> but i'm not worried about optimizing for broken streams
[16:03:05] <mru> same thing
[16:03:47] <Honoome> hmm it has been quite a while since I last ran cowstats over ffmpeg
[16:04:12] <peloverde> Decreases aac.o size by ~3.1%
[16:04:27] <kshishkov> Honoome: try also sheepstats and pigstats
[16:05:00] <Honoome> kshishkov: is that Position Indipendence Gruesomness statistics? :P
[16:05:18] <KotH> pigs?
[16:05:32] <KotH> isnt' that a short for certain countries with money problems
[16:05:34] * kshishkov values pigs over cows
[16:05:34] <KotH> ?
[16:05:41] <Kovensky> http://multimedia.cx/eggs/call-for-samples/ <-- are SSA/ASS samples that hard to find? :V
[16:05:53] <kshishkov> KotH: you are thinking of financial black holes
[16:06:10] <Honoome> Kovensky: Mike doesn't watch anime, I guess
[16:06:45] <kshishkov> he stated it in his blog too
[16:07:12] <kshishkov> that's the reason he's not a MPlayer dev for a long time
[16:07:26] <mru> mike was never an mplayer dev iirc
[16:08:02] <Honoome> he was xine ;)
[16:08:07] <mru> yep
[16:08:08] <BBB> that's what I thought also
[16:08:20] <kshishkov> mru: he was
[16:08:25] <BBB> but mike is corporate now :)
[16:08:40] <mru> that's irrelevant
[16:13:13] <KotH> kshishkov: nope, mike never worked on mplayer
[16:13:47] <kshishkov> I may be wrong but I think he mentioned one or two patches for MPlayer
[16:13:57] <KotH> oops.. he was
[16:14:11] <KotH> at lest he's been mentioned in AUTHORS
[16:14:28] <Honoome> well, if it's one or two patches, I think I have one or two patches in VLC :P on the other hand I'm pretty sure I can only be classed as a xine dev :P
[16:14:50] * kshishkov is MPlayer dev of Compn-class
[16:14:50] * KotH wonders why he doenst remember mike working on mplayer
[16:14:54] * mru has patches in all kinds of scary places
[16:15:07] <KotH> mru: nope
[16:15:15] <KotH> mru: you're lacking the mplayer trophy
[16:15:29] <mru> are you sure?
[16:16:15] <thresh> having patches in mplayer is like having a black label if you're a pirate
[16:16:50] <thresh> or is it called a black spot?
[16:17:16] <kshishkov> black mark or black subpoena ;)
[16:17:58] <kshishkov> at least the source calls it "black spot"
[16:18:53] * elenril headdesks
[16:18:54] <mru> KotH: there are entire files in mplayer with my (c)
[16:19:31] <KotH> mru: what have you done?!?
[16:19:36] <kshishkov> elenril: try Russian traditional way of facepalming - killing yourself with a wall
[16:19:59] <kshishkov> KotH: from external repos like libavcodec ;)
[16:20:36] <CIA-34> ffmpeg: alexc * r21833 /trunk/libavcodec/aac.c: AAC: Mark functions that are only called when the output configuration is not locked as av_cold.
[16:20:39] <mru> KotH: some 3dlabs drivers
[16:21:12] <KotH> mru: ...you're tainted....
[16:21:25] <elenril> kshishkov: i'd rather kill the person who wrote this stuff
[16:21:27] <kshishkov> KotH: says matrox driver man
[16:21:56] <kshishkov> elenril: what stuff? ASF?
[16:22:16] <elenril> kshishkov: yeah
[16:22:49] <Honoome> elenril: if you can get the guys who wrote RTSP as well, I'd be happy to lend a hand
[16:22:49] <kshishkov> throw a chair at them
[16:23:14] * thresh repairs a 19T partition
[16:23:18] * kshishkov counts: Adobe, M$, Buffering Inc., what else
[16:23:24] <elenril> why can't they use utf8 like all sane people do
[16:23:43] <Honoome> because it makes soooo much sense to re-use HTTP, which has the protocol identifier at the end of the request line…
[16:23:57] <Honoome> so you cannot validate it until you reach the end, and then you start back over…
[16:24:32] <mru> you don't need to parse the whole line to extract the last word
[16:24:44] <KotH> kshishkov: you're just jealous that you dont have so many matrox cards yourself
[16:24:59] * mru has a matrox card that KotH doesn't have
[16:25:02] <Honoome> mru: I still need to backtrack though
[16:25:13] <mru> Honoome: strrchr(req, " ")
[16:25:17] <kshishkov> KotH: but I had Trident and S3 Virge!
[16:25:29] * mru has a dual-vga g400 on agp
[16:25:31] <mru> rare stuff
[16:25:56] <Honoome> mru: yeah but since I need to parse the first two tokens anyway (but differently depending on the third, last one)…
[16:26:08] <Honoome> plus I'd need to have the lines split already, to use strrchr
[16:26:34] <Honoome> so the best method I can come up with is to “validate” the first two tokens, parse the third, then backtrack and parse (with the correct parser) the first two again
[16:26:35] <mru> I agree it's annoying, but not *that* annoying
[16:26:54] <Honoome> not especially bad, but it would have made much more sense to have the protocol identifier at the *beginning* of the line
[16:30:35] <CIA-34> ffmpeg: alexc * r21834 /trunk/libavcodec/aac.c:
[16:30:35] <CIA-34> ffmpeg: AAC: Mark che_configure() as av_cold.
[16:30:35] <CIA-34> ffmpeg: It is also only called when the output configuration is not locked.
[16:32:02] <KotH> mru: right, you didnt bring it to fosdem
[16:33:43] <mru> if you really wanted it you'd come and collect it here
[16:35:28] <av500> mru: I think I have 3 of those here, want any? :)
[16:36:06] <mru> one's plenty
[16:36:28] <mru> I have only 2 agp slots in the house
[16:36:41] <av500> kitchen and loo? :)
[16:36:50] <mru> and the mga only works in one of them
[16:37:01] <mru> and that one will probably never be rebooted
[16:37:22] <mru> it's running now, but when it goes down I probably won't bother fixing it
[16:37:23] <av500> and does not need dualhead anyway? :)
[16:38:02] <CIA-34> ffmpeg: stefang * r21835 /trunk/libavcodec/cavs.c: avoid using DECLARE_ALIGNED on stack variable as suggested by Reimar
[16:44:35] <CIA-34> ffmpeg: stefang * r21836 /trunk/libavcodec/ (cavs.h cavsdec.c):
[16:44:35] <CIA-34> ffmpeg: add heuristic to discern the old sample clips from streams encoded
[16:44:35] <CIA-34> ffmpeg: with rm52j encoder, a marker_bit has been added in the I-Frame syntax
[16:47:43] <KotH> huh? you dont have a video wall in your toilet?
[16:48:04] <av500> now that got me thinkine
[16:48:05] <av500> now that got me thinking
[16:48:18] <av500> prolly the only place my wife lets me put it :)
[16:48:30] <superdump> she might get suspicious
[16:48:31] <KotH> hehe
[16:49:14] <benoit-> and anyway, that's often a (too?) small wall
[16:50:59] <KotH> you dont know the german toilets!
[16:51:25] <av500> I park my S600 in there too :)
[16:58:19] <KotH> no BMW?
[16:58:36] <KotH> you disappoint me!
[16:59:53] <av500> BMW, in the loo? are you crazy?
[17:00:42] <KotH> well, would you leave it outside instead?
[17:00:47] <mru> crap cars, bmw...
[17:00:49] <KotH> in germany, none the less
[17:01:01] <mru> KotH: nonetheless is one word
[17:01:07] <KotH> thanks
[17:01:23] <av500> KotH: it's in the shed (color open for discussion)
[17:01:33] <KotH> lapisblau!
[17:01:41] <kshishkov> mru: do you prefer English cars?
[17:01:56] <Honoome> are there English cars? o-O
[17:02:01] <mru> there used to be some good english cards
[17:02:05] <mru> cars
[17:02:14] <kierank> rolls royce
[17:02:20] * kshishkov knows the only true (bike)shed is Faluröda
[17:02:20] <mru> bentley
[17:02:26] <kshishkov> the same thing
[17:02:36] <av500> mru: did I mention I know somebody who collects bentley....sigh
[17:02:38] <kshishkov> what about Vauxhall? Morris?
[17:02:48] <KotH> now, that i've painted the shed, it's time for me to leave the battlefield
[17:03:05] <KotH> readya tomorrow
[17:16:50] <Kovensky> http://esr.ibiblio.org/?p=1705 <-- the story of GCC
[17:17:03] <Kovensky> it's fun how he recommends building with -O0 though; does ffmpeg even build with -O0? IIRC it runs out of registers...
[17:17:25] <kshishkov> it does so with -O3 as well
[17:18:15] * elenril wonders if anybody maintains asfenc
[17:18:25] <av500> you? :)
[17:18:39] <kshishkov> elenril: we have _default_ maintainer
[17:18:52] * elenril sighs
[17:18:52] <kierank> [17:02] <@kshishkov> what about Vauxhall? Morris? --> I guess by .ua standards they are godlike cars
[17:18:56] <Honoome> Kovensky: it doesn't fail because of the registers
[17:19:04] <elenril> why do all my patches have to go through mn
[17:19:07] <Honoome> Kovensky: it fails earlier because of the missing DCE
[17:19:24] <kshishkov> kierank: no, what was designed here (FIAT ripoffs) cannot be considered cars at all
[17:19:26] <Kovensky> DCE?
[17:19:38] <Honoome> Dead Code Elimination phase
[17:19:56] <kshishkov> av500: just wondering, do ICE trains have power outlets and Internet access?
[17:20:08] <Kovensky> oic
[17:20:12] <av500> yes
[17:20:13] <Honoome> if ( 0 ) call_undefined_function(); → no problem if DCE is done, but it's a problem at -O0
[17:20:20] <av500> internet access not on all line
[17:20:21] <av500> internet access not on all lines
[17:20:30] <av500> but I prefer my 3g anyway
[17:20:41] <av500> power in all 1st class for sure, 2nd at some places I think
[17:20:54] <av500> kshishkov: you come and visit me?
[17:20:55] <kshishkov> our trains don't even have standard cell coverage for even major lines :P
[17:21:03] <kshishkov> av500: unlikely
[17:22:20] <BBB> Honoome, that's not true, afaik
[17:22:33] <Honoome> BBB: hm? what do you mean?
[17:22:34] <BBB> Honoome, I've used if (0) and -O0 and afaics the code is being removed
[17:22:40] <Kovensky> gl even finding trains around here ._.
[17:22:41] <BBB> (by gcc)
[17:22:53] <Honoome> BBB: uhm okay maybe I cannot guarantee about proper if ( 0 )
[17:23:09] <Kovensky> and I remember I once was on a car trip to a neighboring state; we only had cellphone signal on two or so hub towns
[17:23:12] <Honoome> it definitely is the case that it's not dropped for if ( something_always_false() )
[17:23:19] <Kovensky> everywhere else was dead phone ._.
[17:23:48] <kshishkov> Kovensky: http://en.wikipedia.org/wiki/Railway_stations_in_Brazil
[17:24:20] <Kovensky> lulz
[17:24:24] <Kovensky> yes, that's very incomplete
[17:25:09] <kshishkov> only "Planning" section is incomplete
[17:25:12] <Kovensky> but the only railway in this entire region is Vale do Rio Doce's Carajás road, which brings iron from the carajás mine to the port in my city
[17:25:43] <Kovensky> the only reason it transports passengers is because there's a law that mandates that private railroads must transport at least a certain amount of passengers
[17:25:53] <BBB> Honoome, might be true, never tried :)
[17:26:11] <Dark_Shikari> mru: videolan git server was broke
[17:26:32] <Kovensky> the 1st class has AC, somewhat comfy seats and a tv passing boring documentaries
[17:26:36] <Honoome> BBB: last I checked, FFmpeg failed at -O0 because of similar constructs in the initialisation of codecs
[17:26:36] <Kovensky> 3rd class doesn't even have seats
[17:27:03] <mru> Kovensky: in sweden 3rd class doesn't even have trains
[17:27:11] <Kovensky> mru: lol
[17:27:11] <kshishkov> same here
[17:27:24] <kierank> why is there a contest to see who's country is the crappiest?
[17:27:25] <Kovensky> well '3rd class' because there's no 2nd class
[17:27:35] <kshishkov> looks like we have only 64th, 63th and 62th classes
[17:27:41] <Kovensky> we only count '1st class' and '3rd class' (don't ask me where the 2 has gone to)
[17:28:08] <Kovensky> kierank: dunno
[17:28:15] <Kovensky> kierank: though our contries ARE crappy :P
[17:28:22] <mru> kierank: http://www.youtube.com/watch?v=Xe1a1wHxTyo
[17:28:31] <Kovensky> well, not mru's
[17:28:44] <Kovensky> since both me and kshishkov want to escape to there
[17:28:44] <kshishkov> the only place where we have "1st" and "2nd" classes are "expresses" to capital. 1st class is actually much worse
[17:29:01] <kshishkov> Kovensky: mru lives in GB
[17:29:28] <kierank> mru: yes, I know ;)
[17:29:51] <BBB> Honoome, syntax checking, including code references, come before DCE
[17:29:56] <BBB> so I'd expect it to warn about that
[17:29:58] <Kovensky> kshishkov: o rly
[17:30:01] * Honoome would love to escape to GB
[17:30:04] <BBB> anyway
[17:30:09] <Kovensky> kshishkov: I thought he lived in .se
[17:30:10] <Honoome> BBB: note undefined and not undeclared :)
[17:30:27] <Honoome> BBB: symbol definition is resolved at link time, not compile time
[17:31:39] <Kovensky> kshishkov: http://www.mangafox.com/manga/mudazumo_naki_kaikaku/v02/c011/22.html
[17:33:31] <kshishkov> Kovensky: quite predictable. For the second candidate to appear in manga you need Photoshop from another 20 years in future
[17:33:46] <Kovensky> lol
[17:34:07] <Kovensky> kshishkov: read a few pages
[17:35:53] <kshishkov> Kovensky: I think of her as of female version of Hugo Chavez, dunno if you heard about that guy
[17:37:16] <Kovensky> oh yes, a lot
[17:37:27] <Kovensky> lolvenezuela
[17:37:35] <Kovensky> kshishkov: did you see page 24?
[17:37:51] <kshishkov> yes
[17:38:28] <Kovensky> heh
[17:38:32] <CIA-34> ffmpeg: cehoyos * r21837 /trunk/libavcodec/ivi_common.h:
[17:38:32] <CIA-34> ffmpeg: Remove outdated comment.
[17:38:32] <CIA-34> ffmpeg: Patch by Maxim, max_pole gmx de
[17:38:46] <Kovensky> I started reading that manga a few days ago; it's pretty awesome
[17:39:12] <Kovensky> will get an OVA later this year
[17:39:43] <astrange> i always thought she looked like saber
[17:41:10] * Kovensky often gets pretty lost on the mahjong sections
[17:41:21] <Kovensky> I have studied a bit of mahjong but I don't get the riichi rules at all
[17:41:54] * kshishkov used it to play solitaires long before he got computer
[17:42:00] <Dark_Shikari> Kovensky:
[17:42:03] <Dark_Shikari> 1) watch Akagi
[17:42:07] <Dark_Shikari> 2) play Touhou Mahjong
[17:42:08] <Dark_Shikari> 3) ???
[17:42:11] <Dark_Shikari> 4) Play on Tenhou!
[17:42:34] <Kovensky> Dark_Shikari: lol
[17:42:41] <Kovensky> Dark_Shikari: if you don't already, read that manga :D
[17:42:46] <Dark_Shikari> touhou mahjong is awesome though
[17:43:07] <Dark_Shikari> surprisingly good music.
[17:57:16] * BBB thinks dark_shikari should spend more time optimizing ffmpeg's h264 decoder
[17:59:00] <BBB> Dark_Shikari, would you optimize ff's h264 decoder for money? (and how much?)
[18:03:31] <j-b> BBB: hello :)
[18:16:35] <kshishkov> j-b: BBB - ping timeout
[18:23:44] <BBB> I'm still here
[18:23:46] <BBB> hi j-b :)
[18:23:58] <BBB> I think jason is timeout'ing
[18:24:10] <iive> no, he is still counting.
[19:01:20] <Compn> i think its more that ass samples are buired within 400mb mkv files
[19:01:38] <Compn> errrr
[19:02:41] <kshishkov> huh? ass buried in itself?
[19:02:50] <av500> ass deep
[19:03:57] * kshishkov has another synonym for 'arse' - Ukraine
[19:05:43] <elenril> fun fact: mplayer -demuxer lavf doesn't work with ass
[19:06:15] <Compn> i remember some ffmpeg devs didnt like ass or the way it was stored
[19:06:15] <kshishkov> yes, FFmpeg is not good with subtitles (yet)
[19:06:22] <Compn> i think theres a flame about ass in -devel
[19:06:28] <Kovensky> Compn: aurel
[19:06:32] <elenril> yes, aurel wanted his own format
[19:06:39] <elenril> which nobody else likes/uses
[19:06:47] * Compn wont point any fingers in a logged chat ;p
[19:06:47] <Kovensky> everybody called him on it but michael
[19:06:57] <Kovensky> michael just told him to do wtf he wanted to since he was the maintainer
[19:07:14] <elenril> btw anyone wants to apply a few patches for me?
[19:07:15] <Kovensky> Compn: :P
[19:08:15] <Kovensky> http://blogs.msdn.com/oldnewthing/archive/2010/02/15/9963387.aspx#9963721
[19:09:39] <elenril> hmm, using subversion to overthrow goverment...
[19:10:03] <elenril> maybe this is the reason ffmpeg still hasn't switched to git
[19:10:44] <Kovensky> are there even merikan devs, other than loren and dark?
[19:11:04] <Kovensky> or compn
[19:11:14] <Kovensky> [[AndZoidberg Compn]]
[19:11:49] <kshishkov> Mike
[19:11:58] <Compn> a few of us around , surely
[19:12:00] <elenril> Kovensky: http://tvtropes.org/pmwiki/pmwiki.php/Main/TvTropesWillRuinYourVocabulary
[19:12:10] <kshishkov> Ronald Bultje
[19:12:11] <Kovensky> elenril: inorite
[19:13:02] <kshishkov> Justin "flac" Ruggles
[19:14:10] <Kovensky> <@Watarase_Jun> http://saber.kawaii-shoujo.net/~jun-chan/MSDOS6.22.rar
[19:14:10] <Kovensky> <@Watarase_Jun> for whoever wants it
[19:14:17] * elenril wonders if kshishkov is grepping some secret KGB db
[19:14:26] <Kovensky> < twnqx> -rw-r--r--  1 charlie users 123148 1996-01-14 09:46 WIN101-1.ZIP
[19:14:26] <Kovensky> < twnqx> -rw-r--r--  1 charlie users 181434 1996-01-14 09:46 WIN101-2.ZIP
[19:14:26] <Kovensky> < twnqx> -rw-r--r--  1 charlie users 156209 1996-01-14 09:47 WIN101-3.ZIP
[19:14:26] <Kovensky> < twnqx> -rw-r--r--  1 charlie users 168145 1996-01-14 09:47 WIN101-4.ZIP
[19:14:27] <Kovensky> < twnqx> -rw-r--r--  1 charlie users 101806 1996-01-14 09:48 WIN101-5.ZIP
[19:14:29] <Kovensky> < twnqx> :>
[19:14:32] <Kovensky> lol
[19:14:43] * twnqx slaps Kovensky 
[19:14:55] <twnqx> ZIS CHANNEL IS BUGGED
[19:15:01] <kshishkov> elenril: nope, just memory. I'd like to keep blackmailing DB on others
[19:15:03] <Compn> get your warez out of here
[19:15:20] <twnqx> Compn: make ffmpeg work on that.
[19:15:26] <Kovensky> twnqx: indeed
[19:15:36] <Kovensky> just found it somehow appropriate since we were discussing weird OSes earlier today
[19:15:40] <Kovensky> lol
[19:15:43] <Compn> sherpya has built mplayer on some kind of extended dos
[19:15:48] <twnqx> cool
[19:15:51] <Compn> and someone runs ffmpeg on dos fate
[19:15:55] <twnqx> likely freedos, though
[19:16:09] <Compn> check fate page for dos ffmpeg status ;p
[19:16:40] <Dark_Shikari> BBB: hmm.  maybe put out a $1000 per % bounty for speedups (with some particular test sample used for testing)
[19:17:17] <Compn> someone should enter michael's h264 speedups into roundup
[19:17:21] <Compn> and add whos working on what
[19:17:33] <Compn> i mean, the suggested speedups
[19:18:50] <Kovensky> Dark_Shikari: $_$
[19:19:18] <Dark_Shikari> that means it'll only cost $50k to be as fast as coreavc.
[19:19:33] <Dark_Shikari> oh, and like a $20k bounty for multithreading
[19:19:52] <twnqx> >_>
[19:20:08] <Honoome> Dark_Shikari: and penalties if somebody get the speed down? :P
[19:20:15] <elenril> lol
[19:20:21] <Dark_Shikari> lol
[19:20:24] <Dark_Shikari> awesome
[19:20:24] <Kovensky> €_€
[19:21:03] <kshishkov> in your case it's Kr_Kr or something
[19:21:12] <twnqx> more like
[19:21:14] <twnqx> 0_0
[19:21:15] <kshishkov> (or Cr)
[19:21:32] <mru> £_£
[19:21:39] <CIA-34> ffmpeg: michael * r21838 /trunk/libavcodec/h264_cabac.c:
[19:21:39] <CIA-34> ffmpeg: Merge decode_cabac_mb_type_b() into calling code.
[19:21:39] <CIA-34> ffmpeg: This avoids a conditional branch and is about 3 cpu cyclues faster.
[19:21:46] <twnqx> lol
[19:21:48] <Honoome> I wonder… if it's M$, should we have Adob€ then? :P
[19:21:53] <twnqx> speaking of h264 speedups :D
[19:22:07] <elenril> so how much money was that commit worth
[19:22:09] <mru> and Goog£e
[19:22:16] <Honoome> lol
[19:22:32] <Kovensky> kshishkov: â‚¢_â‚¢ ?
[19:22:46] <Kovensky> kshishkov: that stopped being used in the 60s
[19:22:50] <Kovensky> current currency is R$
[19:23:02] <CIA-34> ffmpeg: michael * r21839 /trunk/libavcodec/h264_cabac.c:
[19:23:02] <CIA-34> ffmpeg: Simplify decode_cabac_mb_intra4x4_pred_mode().
[19:23:02] <CIA-34> ffmpeg: same speed
[19:23:03] <Kovensky> which started being used in the 90s
[19:23:11] <kshishkov> el dollaro real?
[19:23:21] <Kovensky> Real / Reais
[19:23:22] <kierank> any mxf experts around?
[19:23:35] <Kovensky> http://pt.wikipedia.org/wiki/Cruzeiro_real <-- lol, this was used for less than a year
[19:24:04] <kshishkov> Kovensky: I haven't heard even of Cruzado
[19:24:15] <Kovensky> lol
[19:24:33] <Kovensky> brazil used like o9k currencies between the 40s and the 90s
[19:25:42] <astrange> http://github.com/astrange/ffmpeg/commit/cc9a11d728aa0e8a0dd0ee8d34793c4f174b680c i think this will be 2-3% faster if rewritten to not be incredibly ugly
[19:26:06] <astrange> i need to stop rebasing that branch, i had to cherry-pick that commit out of the detached gt objects
[19:26:10] <astrange> i
[19:26:12] <kshishkov> so now it's 6% faster?
[19:26:23] <Kovensky> kshishkov: perharps even more weirdly, my keyboard has a â‚¢ key
[19:26:45] <Kovensky> kshishkov: apparently they thought it was more useful for you to have a key for typing a 40-years-old outdated currency than a key for typing a slash or question mark
[19:26:53] <Kovensky> I need to use altgr to type those if I use br-abnt2
[19:27:01] <Kovensky> thankfully I use US layout
[19:27:03] <kshishkov> altbr key
[19:27:14] <Dark_Shikari> astrange: oh wow, that messy patch
[19:27:20] <Dark_Shikari> it doesn't help on normal builds that aren't cpu-specific though :/.
[19:27:44] <astrange> well, most arches will have an equivalent, but x86-32 generic isn't one of them unfortunately
[19:28:10] <Dark_Shikari> easy, default to i686 when on x86-32 ;)
[19:28:14] <Dark_Shikari> like we do for x264
[19:28:16] <astrange> it could use AV_COPY128, but i wrote that using explicit asm to avoid having to benchmark lots of compilers
[19:28:20] <astrange> so it always reloads the input
[19:28:23] * Kovensky makes uau a pull request in astrange's stead
[19:28:35] <mru> looks like fill_rectangle needs to be broken out into lavc/$arc/foo.h
[19:28:42] <mru> arch
[19:28:44] <astrange> yes, that'd be the first step
[19:29:03] <kshishkov> mru: lavu/intrectwrite.h :P
[19:29:51] <Kovensky> talking about intwhateverwrite
[19:30:00] <Kovensky> there's a public header that includes intreadwrite.h
[19:30:11] <mru> there's an ongoing bikeshed about that
[19:30:19] <Dark_Shikari> lol
[19:30:20] <Dark_Shikari> indeed
[19:30:21] <mru> expect it to be fixed in 6-8 months
[19:30:25] <Kovensky> lol
[19:30:39] <Kovensky> meanwhile, every time I compile ffms2 I get warnings about a deprecated function :(
[19:31:08] <Kovensky> it was moved to libavutil but the avutil header that defined it includes intreadwrite.h
[19:31:12] <Dark_Shikari> mru: and by "fixed" we mean "solidified" not "solved"
[19:31:20] <Kovensky> lol
[19:31:28] <mru> paint dried in can
[20:00:22] <BBB> Dark_Shikari, holy shit, $1000 per %?
[20:00:38] <BBB> buying a license for avc is cheaper then :-p
[20:00:40] <Kovensky> BBB: ¥_¥
[20:01:10] <Kovensky> \_\ <-- I swear those are two yen signs
[20:01:19] <Dark_Shikari> BBB: making something as good as coreavc isn't worth $50k? =p
[20:01:21] <BBB> I saw then yens in (1)
[20:01:36] <mru> Kovensky: funny thing is, that sign is not used in japanese
[20:01:36] <BBB> Dark_Shikari, maybe... I don't have that much money though
[20:02:03] <BBB> Dark_Shikari, can't you move to alabama? I hear cost-of-living is much lower there
[20:02:07] <Dark_Shikari> lol
[20:02:22] <Dark_Shikari> I mean, if I'm not going to get paid at least $80-100 an hour, it's not worth working for money
[20:02:29] <BBB> true
[20:02:37] <BBB> what if we did it grant-wise?
[20:02:52] <BBB> if I paid $1000 in advance, how much more would you be motivated to work on it than now?
[20:03:00] <BBB> and what's the current level?
[20:03:14] <Dark_Shikari> advance or after doesn't matter
[20:03:15] <kierank> we should get EU grants
[20:03:21] <Dark_Shikari> and also, you overestimate how much I could go do :/
[20:03:40] <Kovensky> Dark_Shikari: you could what?
[20:03:51] <BBB> well, you complain most, I'm hoping that means that you have direct means of improving it
[20:04:03] <BBB> Dark_Shikari, also, you are into x264, so you have some pre-existing knowledge
[20:04:06] <av500> Dark_Shikari: couldnt u use like char for loop vars to speed it up :)
[20:04:29] <BBB> like, if you gave me 1 week, I would probbaly be able to figure out the main 10 functions in decoding of the first packet + extradata
[20:04:32] <Kovensky> wouldn't that make it slower, or at best not change speed at all?
[20:04:37] <BBB> that would be 1 week of work, with no change in code at all
[20:05:18] <j-b> it shouldn't be 1000$ per %, but 1000$ per 5%, then per 4%, then 3%, then 2% etc... since the more it goes, the more difficult it would be, no?
[20:05:57] <Honoome> j-b: depends if you change the reference point or not
[20:06:28] <j-b> indeed.
[20:06:31] <j-b> lu_zero: ping
[20:07:14] <Dark_Shikari> j-b: I was thinking compounded
[20:07:20] <Dark_Shikari> e.g. 1% faster means 100fps -> 101fps
[20:07:27] <Dark_Shikari> and 1% again means 101 -> 102.01fps
[20:07:32] <Dark_Shikari> etc
[20:07:32] <lu_zero> j-b: pong
[20:07:51] <Honoome> lu_zero: I'll send you the bill as personal assistant, don't worry :P
[20:08:53] <lu_zero> Honoome: I guess I'll bill the support calls as well
[20:08:57] <j-b> lu_zero: sorry, do you still have this pthread/mutex patch for mozilla??
[20:09:03] <lu_zero> somewhere sure
[20:09:05] <lu_zero> why?
[20:09:15] <lu_zero> it got missed?
[20:09:37] <j-b> I was in holiday :)
[20:09:47] <lu_zero> ah
[20:09:53] <Honoome> lu_zero: j-b was almost _here_ and didn't tell us >_<
[20:10:04] <lu_zero> Honoome: venice I guess
[20:10:12] <lu_zero> and you went there by train?
[20:10:16] <j-b> nope
[20:10:19] <j-b> Flew
[20:10:29] <lu_zero> ah, then just next to diego =P
[20:10:40] <lu_zero> I hope you had a good time
[20:11:20] <j-b> well, yeah. It was a bit hard to not have Internet for a week though...
[20:11:28] * j-b feels like a drug addict
[20:11:32] <lu_zero> thehehe
[20:11:51] * Honoome spent €50 to have internet in Bruxelles -- that's definitely a fix! :P
[20:12:14] <Honoome> [given I left the phone connected since Friday night till we left, with Google Maps, it's not even a bad price, I'd say]
[20:12:17] * Kovensky once had to spend 3 days without internet while at home
[20:12:30] <Kovensky> it's a good way to watch anime though
[20:12:32] <Kovensky> :D
[20:12:35] <lu_zero> pff
[20:12:41] <Honoome> Kovensky: yeah if you have them already downloaded ;_;
[20:12:47] <Kovensky> Honoome: indeed ;_;
[20:12:55] <lu_zero> j-b: I guess I'll have to redo it ^^'
[20:13:17] <twnqx> that's why i'm always havona at least a TB unwatched!
[20:13:17] <Honoome> lu_zero: which patch are we talking about?
[20:13:19] <Kovensky> <Chii[AR]> MYSTATS: kovensky (uid: 285185) (10h old) - 155 anime, 1924 eps (1410 / 73.28% watched) and 2070 files in mylist (545 GB, 1.8% of AniDB, 1.3% watched, taking 23 days, 6 hours and 4 minutes of your life).
[20:13:20] <j-b> Honoome: lu_zero: but it was for a good cause: Surprise holidays for Carneval in Venise for my g/f => 6 month of geeking without problem!
[20:13:30] <twnqx> cute, kov, cute
[20:13:42] <lu_zero> j-b: =)
[20:14:01] <lu_zero> I fully approve the choice ^^
[20:14:22] <Honoome> j-b: for when I'll find a gf, you'll have to organise something else in Paris then ;)
[20:14:23] * Kovensky has 36GB of unwatched anime on the craptop, 16GB elsewhere
[20:14:34] <Honoome> (okay s/when/if/ :P)
[20:14:52] <lu_zero> Honoome: work hard on it, she'll capitulate soon
[20:15:13] <Honoome> lu_zero: yeah sure, I'll keep on trying… =_=
[20:15:17] <j-b> Honoome: why do you think VDD were in Paris? :)
[20:15:46] <KotH> Kovensky: i can top that
[20:16:04] <Honoome> j-b: eh okay, but I couldn't use the excuse last december! :P make sure you do it again next december and that'd be fine ;)
[20:16:10] <Kovensky> KotH: I'm not comparing sizes, just stating my situation =p
[20:16:14] * lu_zero is pondering of having a FFconf in Torino sooner or later
[20:16:26] <Kovensky> I used to have 200GB of unwatched stuff when I had a big HDD, but that's still much less than twnqx
[20:16:31] <lu_zero> maybe september could be an useful time hmmm
[20:17:23] <KotH> lu_zero: it looks like it'll be paris this time :)
[20:18:57] <lu_zero> KotH: paris, when?
[20:19:06] <KotH> lu_zero: ask j-b
[20:19:07] <KotH> :)
[20:19:13] * lu_zero should catch up with emails =_-
[20:19:23] <KotH> this has not been announced yet
[20:19:27] <lu_zero> today I spent most of my time in a looong meeting
[20:19:33] <KotH> so, no way you can catch up on that :)
[20:20:20] <lu_zero> ahh
[20:20:22] <lu_zero> hmm
[20:20:35] <mru> w/o faster-than-light broadband at least
[20:20:53] <mru> fiber is so slow ;-)
[20:21:27] <Honoome> mru: does BT offer that now?
[20:21:44] <av500> FF fast fiber
[20:21:48] <KotH> Honoome: dont believe mru
[20:21:53] <Honoome> [note: offer, not provide]
[20:21:53] <KotH> Honoome: he uses out of band communication
[20:22:03] <kierank> [20:21] <Honoome> mru: does BT offer that now? --> not yet
[20:22:16] <j-b> lu_zero: I believe that we will try to have VDD at the same time of the FFconf
[20:22:43] <KotH> j-b: try? well, why not just do it? :)
[20:22:55] <lu_zero> ah
[20:23:10] <av500> ffconf in paris, i am all for it!
[20:23:25] <Honoome> lu_zero: next time I'll try to board the plane at the Marco Polo though :P
[20:23:29] <j-b> KotH: because I am not an omnipotent chief
[20:23:34] <KotH> av500: btw: would your employer be a sponsor? :)
[20:23:34] <mru> guys, it's FFcon
[20:23:37] <KotH> j-b: who is?
[20:23:41] <j-b> KotH: noone
[20:23:48] <KotH> ^^'
[20:23:52] <mru> ffconf is the replacement-to-be for autoconf
[20:23:55] <av500> j-b: what do we need to get lacantine?
[20:23:57] * Honoome hands j-b pointy hairs
[20:23:58] <j-b> decisions are collegial
[20:24:02] <Honoome> mru: oooh! I want that! :D
[20:24:15] <KotH> j-b: well..
[20:24:36] <j-b> KotH: anyway, more seriously, I don't see anyone opposing the coupling from our part
[20:24:40] <KotH> j-b: so you mean i have to bring chocolate for all vlc devels for them to do something in my favor? :)
[20:24:44] <j-b> and if it is not in Paris, we will follow
[20:24:53] <KotH> .o0(that'll be expensive)
[20:24:58] * kierank votes south of france
[20:25:06] <KotH> hmm..
[20:25:10] <lu_zero> KotH: I could help you a bit =P
[20:25:14] <j-b> kierank: doable, at my place, if we are less than 20
[20:25:31] <KotH> j-b: is your place at the sea?
[20:25:36] <j-b> no :'(
[20:25:40] <KotH> doh!
[20:25:47] <lu_zero> j-b: where are you?
[20:25:54] <KotH> if south france, we need a beach! with girls and everything! ;)
[20:26:11] <j-b> Paris, but my parents have a place near Orange.
[20:26:39] <av500> j-b: would LaCantine be an option? they love like open source, no?
[20:26:46] <lu_zero> ah
[20:26:59] <j-b> av500: well, of course, like for this year, we would use LaCantine for a night
[20:27:21] <av500> j-b: I alsmost feel like at home there now :)
[20:27:33] <j-b> av500: :)
[20:27:54] <j-b> KotH: if it is expensive, only the important devs will go, then.
[20:28:00] <av500> and there are nice pubs close by....
[20:28:08] <KotH> j-b: there is always sponsoring :)
[20:28:50] <j-b> KotH: well, this year, we didn't pay the hotel and LaCantine, this helped quite abit
[20:29:22] <KotH> j-b: i talked with linuxhotel and they gave quite nice numbers
[20:29:52] <KotH> j-b: 1keur for 20 people, all inclusive
[20:29:57] <j-b> nice.
[20:30:00] <Dark_Shikari> plane flights are more of an issue.
[20:30:02] <av500> KotH: nice
[20:30:13] <j-b> food is an issue too.
[20:30:13] <av500> Dark_Shikari: there is always bug bounties :)
[20:30:14] <Dark_Shikari> imagine the cost of a round trip flight from the US
[20:30:14] <KotH> but they have only 20 beds...
[20:30:27] <av500> j-b: Aldi ftw!
[20:30:28] <KotH> another 20 can be accomodated using cheaper means, but that depends on the people
[20:30:29] <j-b> KotH: how many ffmpeg folks are from the US ?
[20:30:34] <lu_zero> Dark_Shikari: around 200e from time to time
[20:30:45] <KotH> j-b: hmm... 5-6 i suppose
[20:30:49] <lu_zero> (well depends on the coast)
[20:30:52] <KotH> j-b: very few
[20:30:55] <kierank> 200 euro flight!
[20:31:01] <mru> lu_zero: east coast? west coast?
[20:31:02] <peloverde> Dark_Shikari: I booked my FOSDEM flight fairly last minute and it was pretty cheap
[20:31:06] <Dark_Shikari> 200 EUROS?!
[20:31:06] <KotH> j-b: so far, i got only word from peloverde that he'd come
[20:31:10] <lu_zero> east coast -> ita
[20:31:10] <Dark_Shikari> isn't that cheaper than the fuel?
[20:31:16] <j-b> KotH: ok
[20:31:26] <Dark_Shikari> where did you find a flight for 200 euros?
[20:31:28] <lu_zero> alitalia from time to time had such offers
[20:31:38] <kierank> no wonder they're bust
[20:31:39] <Dark_Shikari> of course we'd need very advance notice to get our passports in line
[20:31:50] <lu_zero> kierank: the other way round
[20:31:53] <peloverde> Dark_Shikari: get your passport now, it's worth having
[20:31:56] <Dark_Shikari> I have one
[20:31:59] <Dark_Shikari> I just don't have it _with_ me
[20:32:00] <mru> I can usually find flights anywhere in europe for less than £200
[20:32:13] <Dark_Shikari> mru: cross atlantic costs more
[20:32:16] <Dark_Shikari> also lol
[20:32:18] <Dark_Shikari> we started at $
[20:32:19] <Dark_Shikari> then we went to euros
[20:32:21] <Dark_Shikari> then pounds
[20:32:29] <j-b> KotH: anyway, December in Paris was not the best idea we have had.
[20:32:35] <lu_zero> eu is about 99e now
[20:32:51] <j-b> KotH: and I guess that if we can do a VDD one day after/before the FFconf, that would be cool.
[20:32:55] <lu_zero> right now
[20:33:00] <Dark_Shikari> well at the rate things are going the euro will be worthless in a few months
[20:33:01] <lu_zero> Milano - Chicagod€ 418e
[20:33:03] <Dark_Shikari> so we better buy tickets soon
[20:33:25] <lu_zero> Milan - Boston is 417e
[20:34:05] <KotH> j-b: ffcon is planned to be 2-3 days on a long weekend
[20:34:10] <KotH> j-b: so that we have time
[20:34:23] <KotH> j-b: what is the purpose of vdd?
[20:34:29] <mru> BEER
[20:34:37] <lu_zero> mru troll beer?
[20:34:40] <lu_zero> theheh
[20:34:40] <KotH> j-b: wouldnt it make sense to just merge both ffcon and vdd?
[20:34:50] <mru> I think it would
[20:35:01] <Dark_Shikari> I just did a bing travel search, the cheapest flight from LA area to Paris and back, June 20th-25th (random dates), is $1506
[20:35:12] <lu_zero> Dark_Shikari: ??
[20:35:16] <lu_zero> bing is broken
[20:35:28] <kierank> that price sounds about right
[20:35:30] <j-b> KotH: we just need to have a room for 1 day for VLC folks to take decisions about how we can improve porn searching in VLC
[20:35:40] <KotH> j-b: ah...
[20:36:12] <KotH> j-b: that'd be a 10min task if you'd had the right people from mplayer around to show you how it is done correctly ;->
[20:36:13] <j-b> and that can be at the same time as other discussions are around
[20:36:21] <Dark_Shikari> lu_zero: what would you recommend?
[20:36:31] <lu_zero> uhmmm
[20:36:43] <lu_zero> seems that is the offer now
[20:36:46] <mru> KotH: invite gabu?
[20:36:49] * lu_zero is a bit puzzled
[20:37:18] <KotH> mru: uh.. that'd be fun
[20:37:30] <lu_zero> which is the dollar/eu exchange rate now?
[20:37:37] <peloverde> CLE<->Paris 8/5 to 8/10 from $1315, my guess is that prices will drop then come back up. CLE to BRU cost me less than $800
[20:37:38] <KotH> mru: though, it might be a good idea to invite some people from the old team :)
[20:37:46] <Dark_Shikari> 1.36 dollars to the euro
[20:37:49] <KotH> mru: like arpi, .so, lgb...
[20:38:12] <Dark_Shikari> I tried over a weekend and it's "only" $1394
[20:38:13] <av500> peloverde: BRU to Paris is cheap too :)
[20:38:14] <Dark_Shikari> using british airways
[20:38:35] <KotH> Dark_Shikari: try swiss, they are cheaper :)
[20:38:47] <KotH> if not by price, then by service
[20:39:08] <lu_zero> west coast seems about a 2x the price
[20:39:25] <lu_zero> alitalia.it has an USA website I'm watching
[20:39:30] <Dark_Shikari> swiss was more
[20:39:34] * mru likes flying with virgin atlantic
[20:39:38] <Dark_Shikari> lu_zero: I can get very cheap cross country flights if necessary
[20:39:45] <Dark_Shikari> if they're far enough in advance
[20:39:47] <av500> Dark_Shikari: for when did you search?
[20:39:51] <KotH> j-b: btw: the alternative would be to make ffcon/vdd in .jp :)
[20:39:52] <av500> I find 737€
[20:39:56] <Dark_Shikari> av500: june 18-21
[20:40:00] <Dark_Shikari> as a random example of a cross-weekend set of dates
[20:40:05] <Dark_Shikari> KotH: lol
[20:40:11] <kierank> british airways will lose your bags
[20:40:13] <kierank> fact
[20:40:25] <mru> av500: 737 is a bit small for transatlantic flights
[20:40:26] <j-b> KotH: would be great.
[20:40:31] <mru> 747 or 777 is much better
[20:40:38] <av500> Dark_Shikari: june cost more, I get 1052
[20:40:42] <av500> with UA
[20:40:46] <Dark_Shikari> lol, to tokyo is even more
[20:40:48] <Dark_Shikari> or wait
[20:40:49] <Dark_Shikari> no it isn't
[20:40:52] <Dark_Shikari> $887 to tokyo, roundtrip
[20:41:01] <Dark_Shikari> Guess it makes sense, given I'm on the west coast.
[20:41:02] <lu_zero> $ 628 ny-rome
[20:41:04] <KotH> j-b: if we want that, i could organize places to stay
[20:41:07] * mru avoids flying with us-based airlines
[20:41:23] <Dark_Shikari> I like united because I have a free premiere membership
[20:41:27] <av500> mru: I guess it is a codeshare with 15 different airlines....
[20:41:27] <KotH> j-b: but then i'd need to start asking the right people soon
[20:41:29] <j-b> KotH: Dubai  has lots of hotels too.
[20:41:31] <Dark_Shikari> and my mother has so many free upgrades that she just gives them to me
[20:41:39] <Dark_Shikari> because she doesn't fly enough inside the country to use them
[20:41:40] <KotH> j-b: who cares about arabs?
[20:41:41] <KotH> ;)
[20:41:48] <Dark_Shikari> so I get first class for nothing
[20:41:53] <mru> JIHAD
[20:41:58] <av500> lol
[20:42:11] <KotH> mru: when did you join gods army? ;)
[20:42:24] <lu_zero> hmm
[20:42:25] <av500> KotH: I try to image how he looks like when he yells it....
[20:42:38] <lu_zero> vdd+ffcon in tokyo sounds interesting
[20:42:46] <mru> not realisting imo
[20:42:50] <KotH> lu_zero: nah.. tokyo is too expensive
[20:42:57] <lu_zero> better kyoto =P
[20:42:59] <av500> j-b will have an issue to get his posse over there, no?
[20:42:59] <Dark_Shikari> lol
[20:43:01] <KotH> lu_zero: ok, not as expenisve as zürich, but still
[20:43:05] <lu_zero> pff
[20:43:11] <Dark_Shikari> we could go into germany, then I might have a chance of understanding people
[20:43:14] <lu_zero> tokyo about lodging?
[20:43:19] <av500> Dark_Shikari: I am all for it
[20:43:30] <KotH> lu_zero: there are a few nice places in shikoku, that are quite cheap
[20:43:30] <lu_zero> we could takeover a internet/book cafe
[20:43:34] <mru> I think europe is the only realistic option
[20:43:36] <av500> thats KotHs linuxhotel
[20:43:49] <kierank> that linuxhotel looks quite nice
[20:43:54] <av500> old europe ftw!
[20:44:00] <lu_zero> thehe
[20:44:01] <kierank> we should have the conference at cern
[20:44:02] <Dark_Shikari> lol wait what
[20:44:06] <Dark_Shikari> a flight march 18-21
[20:44:09] <Dark_Shikari> is cheaper than june 18-21
[20:44:12] <av500> yes
[20:44:16] <Dark_Shikari> what the fuck is with that
[20:44:18] * lu_zero loves Geneve
[20:44:20] <av500> june is holidays, no?
[20:44:31] <KotH> kierank: difficult to arange
[20:44:35] <lu_zero> Dark_Shikari: depends on their statistics
[20:44:37] <av500> Dark_Shikari: I got 737€ for a random May weekend
[20:44:57] <twnqx> you know Dark_Shikari... a flight riyadh -> frankfurt -> riyadh with lufthansa is like half the cost of frankfurt -> riyadh ->  frankfurt...
[20:45:05] <twnqx> don't ask me for the reasons
[20:45:06] <Dark_Shikari> twnqx: lol
[20:45:16] <BBB> what is vdd?
[20:45:19] <BBB> videolan dev days?
[20:45:20] <kierank> welcome to the world of international airfare pricing
[20:45:28] <twnqx> it's like that for the past 5 years
[20:45:37] <mru> BBB: Vss and Vdd, didn't you take electronics in school?
[20:45:41] <Dark_Shikari> kierank: it's bad enough in the US
[20:45:46] <Dark_Shikari> so, here's my insane airline story
[20:46:07] <Dark_Shikari> 1) $1600+ for a direct flight across the country, economy class
[20:46:16] <kierank> at least in eu there are very strong consumer protection laws when the airlines bump you off
[20:46:17] <iive> Dark_Shikari: i've heard about some ticket system, where you can get extremely cheap price for air traver, something around 30-40euro, for flight in europe. the trick is that you don't have exact date for travel, and they stuff you in when there are unused seats.
[20:46:23] <Dark_Shikari> iive: ryanair?
[20:46:29] <Dark_Shikari> but there is one good story I had
[20:46:31] <kierank> he means standby
[20:46:34] <Dark_Shikari> due to the huge snow storm like 2 months ago
[20:46:41] <Dark_Shikari> my flight was cancelled and I had to stay overnight
[20:46:46] <Dark_Shikari> then, the next day, when I got on my replacement flight
[20:46:48] <Dark_Shikari> they had swapped the planes
[20:46:54] <Dark_Shikari> an international 777 was doing the domestic flights
[20:46:56] <BBB> mru: I sucked at most things non-math/bio
[20:46:59] <Dark_Shikari> and as it happened, I had a free upgrade for that flight
[20:47:07] <Dark_Shikari> but the original flight was only economy/first
[20:47:09] <BBB> bleh $1200 for a weekend munich
[20:47:10] <BBB> darn
[20:47:14] <BBB> too expensive
[20:47:14] <Dark_Shikari> and upgrades are supposed to be one +class
[20:47:19] <Dark_Shikari> But they made a mistake
[20:47:26] <Dark_Shikari> and I got a first-class seat on an international 777, for free
[20:47:44] <Dark_Shikari> I don't think I'll ever end up in one of those for the rest of my life
[20:48:12] <Dark_Shikari> You know, the kind with the chairs that recline all the way back
[20:48:26] <Dark_Shikari> apparently one of those on an international flight is $19000
[20:48:34] <BBB> even my sister had those
[20:48:37] <kierank> yes, something ludicrous like that
[20:48:44] <BBB> I flew for 6 years, 10 times/year, and never had an upgrade :(
[20:48:57] <mru> likewise
[20:49:00] <Dark_Shikari> BBB: my mother has 100K club (100,000 miles per year) with united
[20:49:02] * BBB decides he hates flying
[20:49:04] <j-b> KotH: sorry to be stupid, but why not during the OpenVideoConference?
[20:49:05] <Dark_Shikari> they give her like 8 free domestic upgrades per year
[20:49:07] <mru> you need to fly 10 times/month to get that
[20:49:07] <Dark_Shikari> plus various other things
[20:49:12] <Dark_Shikari> and a ton of systemwide upgrades
[20:49:12] <Compn> Dark_Shikari : oh  , thats a hell of a plane
[20:49:16] <BBB> mru: so I heard
[20:49:17] <Dark_Shikari> mru: 4 times a year is enough
[20:49:20] <BBB> mru: I'm not that rich :)
[20:49:26] <Dark_Shikari> as long as it's east coast of the US -> china and back each time ;)
[20:49:30] <peloverde> I got a free business upgrade on a 777 from Heathrow to BRU
[20:49:31] <Compn> Dark_Shikari : did it have tvs in every seat or is that still a damn myth they put on every air travel advertisement?
[20:49:45] <kierank> economy has that too
[20:49:47] <Dark_Shikari> Compn: they have that for the international planes
[20:49:50] <mru> 777 from LHR to BRU?
[20:49:53] <mru> wtf?
[20:49:56] <kierank> most of them run linux as well
[20:50:02] * Compn has flown internationally, still on 727/737 tho, no tv, no completely flat chairs etc
[20:50:14] * Kovensky gets mighty confused when people use three-letter code for airports
[20:50:17] * Kovensky is used to ICAO notation
[20:50:26] <Compn> well its kind of easy
[20:50:26] <mru> every intercontinental flight I've been on in recent years had seatback screens
[20:50:29] <Kovensky> <-- flight sim fag
[20:50:29] <Compn> bru = brussles ?
[20:50:31] <Compn> ;p
[20:50:50] <Compn> lhr = london heathrow ? ;p
[20:51:04] <Compn> mru : must be a european thing
[20:51:24] <kierank> not really
[20:51:26] <Compn> what am i jealous of anyhow, nothing on tv to be sure
[20:51:32] <mru> I fly intercontinental only with european or asian airlines
[20:51:56] <mru> any decent airline has ~50 films or so to choose from
[20:52:14] <Compn> in usa flights, they try to terrorize people with julia roberts movies
[20:52:23] <Compn> i was on one flight ... three julia roberts movies in a row.
[20:52:30] <mru> you don't get to choose?
[20:52:37] <Compn> one big screen at the front
[20:52:38] <KotH> j-b: the idea was to have something seperate from the usual conferences, so that the developers have time to talk with each other, instead of other people
[20:52:39] <Compn> projector
[20:52:47] <mru> that's so 1980s
[20:52:57] <Compn> KotH : i think people get confused because you call it a conference, call it a meetup
[20:53:19] <KotH> ffmeet then?
[20:53:22] <kierank> google could host us
[20:53:23] <KotH> or ffmeat? ;)
[20:54:17] <peloverde> conference sounds more formal, I like it better
[20:54:30] <Compn> but then you get into arguing about booths
[20:54:32] * kierank prefers pissup
[20:54:41] <Compn> there wont be any booths!
[20:54:41] <av500> Compn: like a booth per codec?
[20:54:43] <Compn> heh
[20:54:54] <av500> flamebooth
[20:55:01] <mru> what, no booths?  can we at least keep the booth babes?
[20:55:20] * Compn daydreams of showing up to ffmpeg booth and trolling about including ffmpeg in some proprietary software without giving credit or source
[20:55:23] <av500> yv: ?
[20:56:08] <mru> av500: did you learn about plural forms in school?
[20:56:22] <av500> mru: I can try to recruit the former Archos booth babe...
[20:56:27] <av500> prularl?
[21:03:28] <iive> Compn: julia roberts... I wonder if that have some connection with that man whom underwear ignited.
[21:15:36] <CIA-34> ffmpeg: cehoyos * r21840 /trunk/libavformat/asfdec.c:
[21:15:36] <CIA-34> ffmpeg: workaround for broken files created by previous versions of asfenc.
[21:15:36] <CIA-34> ffmpeg: Patch by Anton Khirnov, wyskas gmail
[21:20:32] <CIA-34> ffmpeg: cehoyos * r21841 /trunk/libavformat/asfenc.c:
[21:20:32] <CIA-34> ffmpeg: Strings in extended content header are UTF16,
[21:20:32] <CIA-34> ffmpeg: so terminating NULLs are 2 bytes long, not 1.
[21:20:32] <CIA-34> ffmpeg: Patch by Anton Khirnov, wyskas gmail
[21:42:58] <av500> j-b: #meego was so much fun today...
[21:43:14] <j-b> av500: trolling? :D
[21:43:19] <av500> just a little
[21:43:28] <j-b> av500: so, it is going to be Qt based? .deb or .rpms?
[21:43:35] <av500> but no need, they all run around like headless chickens by themselves....
[21:43:40] <av500> qt, rpm
[21:43:47] <av500> instread of gtk, debian and .deb
[21:43:59] <av500> so, world ends, epxect cheap N900s soon
[21:44:24] <j-b> based on what then? Fedora?
[21:44:41] <av500> moblin
[21:44:45] <av500> based on mobin
[21:44:48] <av500> based on moblin
[21:44:48] <j-b> ok
[21:44:56] <j-b> ok, I'll live with it
[21:52:07] <BBB> everybody that cares already has an android phone
[21:52:14] <BBB> everybody else already has an iphone
[21:52:19] <BBB> you are TOOO FUCKING LATE
[21:52:26] <BBB> ^d^d
[21:52:57] <av500> BBB: join in :)
[21:53:03] <BBB> no thanks
[21:53:08] <BBB> my time is more precious than that :)
[21:53:21] <BBB> let them waste money
[21:53:31] <av500> yes
[21:53:55] <BBB> also, check the new york subway around rush hours
[21:54:14] <BBB> the ads are crazy
[21:54:25] <BBB> for samsung SMARTphones, for LG SMARTphones
[21:54:30] <BBB> there's only two not being advertised
[21:54:37] <BBB> iphone and nexus
[21:54:46] <av500> no use to advert iphone in nyc I guess :)
[21:54:49] <BBB> and, conveniently, that's the only ones you see in people's hands
[21:54:58] <BBB> actually, nexus not that much
[21:55:01] * av500 should move to ny...
[21:55:01] <BBB> but I've seen 1 or 2
[21:55:17] <BBB> the kindle and barnes and nobles ebook reader are quite popular
[21:55:25] <BBB> but 95% of devices are iphones or blackberry
[21:55:51] <j-b> BBB: agreed.
[21:55:52] <BBB> but blackberry is losing quickly, it used to be just blackberry :)
[21:56:57] <j-b> av500: anyway, it will be fun to watch
[21:57:05] <av500> yes
[21:57:16] <av500> I watch it all day, cant take much more
[21:57:21] <j-b> av500: I just hope we don't loose the OpenMax IL layer
[21:57:32] <av500> j-b: i dont think so
[21:57:43] <av500> I guess the guts will mostly stay the same
[21:57:56] <j-b> I'll pray for it
[21:58:03] <j-b> but I won't cry over Moblin
[21:58:28] <av500> I stopped following moblin changes of ui, distro etc...
[21:58:53] <j-b> but, that is a bad day for GTK
[21:59:08] <av500> and is that bad?
[21:59:17] <j-b> no.
[21:59:23] <j-b> :)
[21:59:26] <av500> :)
[21:59:45] <j-b> but loosing Nokia, Google and Intel support in a few month is quite bad
[22:00:16] <av500> dragging your heels for years doing nothing is deserves that
[22:07:56] <CIA-34> ffmpeg: michael * r21842 /trunk/libavcodec/ (h264.h h264_cavlc.c h264_cabac.c):
[22:07:56] <CIA-34> ffmpeg: Split setting neighboring MBs from fill_decode_caches()
[22:07:56] <CIA-34> ffmpeg: no speed change.
[22:10:39] <J_Darnley> Um...
[22:10:54] <J_Darnley> I think there might me a missing dependency for wmavoice
[22:12:00] <J_Darnley> When building with lots of stuff disabled, this happens:
[22:12:03] <J_Darnley> libavcodec/wmavoice.c:971: undefined reference to `ff_acelp_lspd2lpc'
[22:17:09] <astrange> missing lsp.o dep
[22:17:22] <astrange> you could add it to the makefile and send a patch
[22:17:27] <TheFluff> 20:30:36 < Kovensky> meanwhile, every time I compile ffms2 I get warnings about a deprecated function :( <-- which one, the av_get_pixfmtwhatever one?
[22:17:27] <astrange> too bad those deps aren't transitive
[22:17:41] <TheFluff> what is it supposed to have been replaced with?
[22:18:28] <J_Darnley> So be it
[22:36:12] <J_Darnley> Is there also something in configure which controld dependencies?
[22:37:38] <J_Darnley> *controls
[22:40:37] <astrange> yes, for things that can be --disabled
[22:44:36] <mru> sounds like wmavoice_decoder_select="lsp" in configure is needed
[22:45:04] <mru> hmm no
[22:45:07] <mru> that's lpc
[22:45:15] <mru> damn, this stuff needs cleaning up
[22:46:35] <J_Darnley> Should I still send an email with what I did and the first change?
[22:49:39] <BBB> I can add stuff that's missing
[22:49:41] <BBB> give me a second
[22:49:52] <mru> andoma: http://www.youtube.com/watch?v=r9X3K2dRMdM
[22:50:07] <BBB> it depends on lzo and lsp
[22:50:13] <BBB> lzo is in libavutil
[22:50:18] <BBB> is that a dependency or is that always on?
[22:50:51] <mru> I think it's always on
[22:51:30] <BBB> ok
[22:51:35] <BBB> then we only need lsp.c in Makefile
[22:52:02] <BBB> anyone can commit that, so I don't have to change trees or quilt remove add all my patches
[22:52:09] * BBB still grumbles at his failed git experiment
[23:05:02] <CIA-34> ffmpeg: michael * r21843 /trunk/libavcodec/h264_cabac.c:
[23:05:02] <CIA-34> ffmpeg: Drop compute_mb_neighbors() and move fill_decode_neighbors() up to take its
[23:05:02] <CIA-34> ffmpeg: role.
[23:05:02] <CIA-34> ffmpeg: Should be faster as this is a strict code removial.
[23:10:37] <Dark_Shikari> Direct temporal skiped MBs dont need fill_decode_caches() at all so dont call it
[23:10:40] <Dark_Shikari> for them.
[23:10:42] <Dark_Shikari> wow, I didn't even realize that
[23:10:47] <Dark_Shikari> they really require _no_ neighbor information
[23:15:32] <mru> "skiped", you talk like michael now?
[23:15:43] <Dark_Shikari> that was a copy paste
[23:15:45] <Dark_Shikari> from the commit message
[23:16:13] <mru> he did say other's english had got worse...
[23:16:46] <Dark_Shikari> no, his is still just bad
[23:19:25] <Honoome> mru: you definitely are picking up the British way ;)
[23:28:29] <mt> Are atrac3 streams in wav always stereo, or could they be joint-stereo ?
[23:39:07] <Yuvi> astrange: ideally it should allocate a blank frame instead of requiring a keyframe to be seen, especially since it looks like they're hacking in PIR to theora
[23:39:54] <Yuvi> but I haven't really looked at copy buffers / the mpegvideo buffer model
[23:42:13] <Dark_Shikari> lol they're adding PIR to theora?
[23:42:23] <Dark_Shikari> and they don't even have proper VBV management yet?
[23:42:36] <Dark_Shikari> or threading
[23:43:07] <mru> pir?
[23:44:05] <Dark_Shikari> periodic intra refresh
[23:44:08] <Dark_Shikari> aka keyframeless
[23:44:12] <mru> ah
[23:44:47] <Yuvi> best part is that it breaks nearly seeking algorithms in ogg
[23:44:55] <Yuvi> *nearly all
[23:45:03] <Dark_Shikari> I wonder if they'll enable it by default
[23:45:07] <Dark_Shikari> since they insist on having no options whatsoever
[23:45:16] <Dark_Shikari> which makes it practically impossible to construct any real streaming system
[23:45:19] <mru> self-destruct by default, I like it
[23:49:55] <iive> no options?