IRC Logs

*******************************************************************
[03:01] <BBB> http://freedesktop.org/cgi-bin/viewcvs.cgi/gstreamer/gst-plugins/gst/qtdemux/qtdemux.c.diff?r1=1.40&r2=1.41
[03:01] <BBB> that commit broke parsing
[03:01] <BBB> of MACE
[03:02] <BBB> still doesn't work if I move offset = 52; to inside the braces, though
[03:02] Action: BBB wonders why
[03:03] arana ([email protected]) joined #gstreamer.
[03:03] <ds-work> did I write that code?
[03:03] <ds-work> no
[03:03] benwei ([email protected]) joined #gstreamer.
[03:04] <benwei> is there a gstreamer plugin to allow mp3 encoding?  there is an option for it in sound-juicer, but it is greyed out for me
[03:04] <BBB> http://freedesktop.org/cgi-bin/viewcvs.cgi/gstreamer/gst-plugins/gst/qtdemux/qtdemux.c.diff?r1=1.36&r2=1.37
[03:04] <BBB> that one breaks stuff too
[03:04] <BBB> both are benjamin's
[03:04] <BBB> benwei: lame
[03:04] <BBB> you need liblame for that
[03:05] <ds-work> BBB: some of it seems wrong
[03:05] <benwei> BBB: i have lame and liblame installed
[03:05] <benwei> is there some kind of linking thing that ties it into gstreamer?
[03:06] <BBB> ds-work: exactly
[03:06] <BBB> I think that second commit breaks stuff
[03:06] <BBB> the first is just breaking it more
[03:06] <BBB> so... can you commit a fix? I'll test :p
[03:07] <ds-work> heh
[03:07] <ds-work> I've worked on gstreamer about 10 minutes in the last week
[03:07] Action: ds-work sucks
[03:09] <BBB> benwei: yes
[03:09] <BBB> benwei: read config.log on why it couldn't link to lame
[03:09] <BBB> benwei: search for the string lame in it
[03:09] <BBB> saves some time ;)
[03:10] <benwei> BBB: alright, thanks
[03:11] <BBB> ds-work: please just try a fix and send me a patch or so
[03:11] <BBB> I don't know the quicktime structure well enough to fix this
[03:11] spyder482 ([email protected]) left irc: Read error: 110 (Connection timed out)
[03:15] <thaytoo> BBB: I heard a rumour that Dave Singer singer at apple.com might be a good place to contact for mpeg4/quicktime stuff
[03:15] <Misirlou> Dave Singer Singer SING SING SINGER
[03:15] <BBB> thaytoo: ?
[03:16] <BBB> thaytoo: but we have no issues :) we just need some extensions
[03:16] <BBB> unless he can provide me with encoder sources
[03:16] <thaytoo> one of the chief engineers and involved on the mpeg standards committee...
[03:16] <Misirlou> BBB: Don't taint yourself!
[03:16] <BBB> ?
[03:16] <BBB> dude
[03:17] <BBB> I'm burried under tained code here
[03:17] <BBB> don't worry about me
[03:17] <BBB> I sacrifice myself for the better sake of other people
[03:17] Action: BBB throws some random_device win32 source code in the air
[03:17] Action: BBB also has halflife2 source code somewhere on his HD
[03:17] <BBB> anyone have win2k source code for me?
[03:18] <BBB> thaytoo: interesting... what could he do for us?
[03:18] <BBB> MPEG-4 specs are available via ISO
[03:18] <BBB> I can probably be nice at work and ask them to buy it for me
[03:19] <thaytoo> he may be able to get us into touch apple people to bless an open source mpeg4 encoder
[03:20] <thaytoo> "with" 
[03:20] <BBB> we already have that
[03:20] <BBB> ffmpeg ;)
[03:20] <BBB> that doesn't solve the patent issue
[03:20] <BBB> I'd love some better documentation on sorensen, though, but he can't give that (he's under an NDA about that himself)
[03:23] <thaytoo> no, he won't be able to give that
[03:23] <BBB> ...
[03:23] <thaytoo> so, I'm not sure that he can do anything here
[03:23] <thaytoo> it's just a name I heard that might be able to answer questions :)
[03:23] <BBB> heh :)
[03:23] <BBB> we have the quicktime specs
[03:24] <BBB> we can get the MPEG-4 specs if we want them
[03:24] <BBB> I'll keep it in my mind ;)
[03:32] benwei ([email protected]) left irc: "Leaving"
[03:49] Action: BBB awaits ds-work's patches
[03:49] <ds-work> well, that will be a number of hours
[03:49] <ds-work> if even tonight
[03:49] <BBB> I have two more hours before I go to bed
[03:49] <BBB> maybe 3
[03:50] <ds-work> as I have dinner plans in about an hour
[03:50] <ds-work> I need to get a gf, so I stop socializing uncontrollably
[03:50] <BBB> ?
[03:50] <ds-work> one that keeps me at home, hacking
[03:50] <BBB> I socialize even more with a gf, normally
[03:51] Action: BBB must be weird
[03:51] <BBB> especially if she has cute good-looking friends, too
[03:51] <BBB> erm
[03:51] Action: BBB hacking again
[03:51] <BBB> ;)
[03:58] Action: bitshifter stares at top while gstreamer is eating 827 MB and rising while importing metadata
[03:59] <ds-work> leakage
[03:59] <BBB> ugh
[03:59] <BBB> which element was unfixed?
[03:59] <ds-work> ogg seems to be ok
[03:59] <bitshifter> I think it's spider not unreffing elements it puts into the pipeline, but that's only a layman's guess
[04:00] <BBB> that doesn't take so much memory
[04:00] <bitshifter> if you read in a whole library of mp3 files with one pipeline for each file, then yes
[04:01] <BBB> hm...
[04:01] <BBB> bleh
[04:01] <BBB> unref spider :)
[04:01] <BBB> it's a bin
[04:01] <BBB> so it'll unref is contents
[04:01] <bitshifter> shouldn't it be enough to unref the pipeline?
[04:02] <ds-work> how do I change the allocator of a GType?
[04:02] <BBB> bitshifter:  yes
[04:09] <ds-work> ds-work: you can't
[04:09] <BBB> I gues snot
[04:10] <BBB> I guess you'd use boxed types for that
[04:11] <ds-work> it will use a GMemChunk if you specify n_preallocs, though
[04:21] ProtectR ([email protected]) joined #gstreamer.
[04:27] sgf ([email protected]) joined #gstreamer.
[04:45] ishamael ([email protected]) joined #gstreamer.
[04:45] Marsupilami23 ([email protected]) left irc: "I don't wanna grow up, I'm a Toys R Us Kid. There's a million toys at Toys R Us that I can play with!"
[04:46] ishamael ([email protected]) left irc: Remote closed the connection
[04:46] ishamael ([email protected]) joined #gstreamer.
[04:48] <ishamael> hi.  i'm having a problem with gstreamer 0.8.  i compiled from source, and everything appears to install OK, but when i run gst-register-0.8 none of the plugins from gst-plugins gets registered.  i see the shared objects in /usr/lib/gstreamer-0.8....  so why doesnt gst-register pick them up? is there a new installation step i missed or...?
[05:02] <bitshifter> ishamael: tried running 'ldconfig'?
[05:02] <ishamael> yeah
[05:22] arana ([email protected]) left irc: "Leaving"
[05:40] jimmy_dean ([email protected]) left irc: Remote closed the connection
[06:00] smoke ([email protected]) got netsplit.
[06:06] smoke ([email protected]) got lost in the net-split.
[06:30] BBB ([email protected]) left irc: "Client exiting"
[06:47] jimmy_dean ([email protected]) joined #gstreamer.
[07:10] aeyakovenk1 ([email protected]) joined #gstreamer.
[07:10] <aeyakovenk1> nothing works 
[07:23] mx|ibook ([email protected]) joined #gstreamer.
[07:23] thaytoo ([email protected]) left irc: "leaving"
[07:45] <aeyakovenk1> is there something funky with tee's in the new release?
[07:47] ChrisHJW ([email protected]) joined #gstreamer.
[07:48] <ProtectR> anyway experienced with OpenAL?
[07:49] <ProtectR> s/anyway/anybody
[07:54] ProtectR ([email protected]) left irc: "Leaving"
[07:55] <aeyakovenk1> whats OpenAL?
[07:56] <aeyakovenk1> blah, i left myself loged in at werk
[08:07] <ds> moo
[08:11] Nick change: kh_out -> khalek
[08:19] sjoerd ([email protected]) left irc: Read error: 110 (Connection timed out)
[08:22] ChrisHJW ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[08:23] <aeyakovenk1> is there something new about tee's in the new release?
[08:25] mx|ibook ([email protected]) left irc: "Leaving"
[08:25] Rotty ([email protected]) joined #gstreamer.
[08:30] ishamael ([email protected]) left irc: Read error: 110 (Connection timed out)
[08:34] sri_ ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[08:43] cebit2004 ([email protected]) joined #gstreamer.
[08:44] Nick change: cebit2004 -> ensonic
[08:44] <ensonic> hi hi
[08:46] <aeyakovenk1> hello
[08:48] Action: ensonic will be at the cebit/hannover/germany from 18/03 until 21/03 (Halle 11/D31) - so if you are there come around
[08:50] elephantum ([email protected]) joined #gstreamer.
[08:52] <ensonic> okay, I have to shutdown and pack the computer
[08:52] ensonic ([email protected]) left #gstreamer ("Leaving").
[08:52] <elephantum> hi all
[08:53] <elephantum> what are the *sink and *src modules and what is the difference?
[08:54] <ds> source elements are sources of data.  sink elements are sinks of data
[08:55] <elephantum> ds: ok, what is osssink function?
[08:58] <ds> to output sound data to an OSS device
[08:58] <elephantum> oh... I wanted to write osssrc
[08:58] <elephantum> I cant understand it function
[08:59] <ds> osssrc records sound from an OSS device
[09:00] markey ([email protected]) joined #gstreamer.
[09:00] <elephantum> ds: ah, thanks... and why can`t I pipe queue with osssink ?
[09:00] mathrick|sleep ([email protected]) left irc: Remote closed the connection
[09:00] <ds> you can
[09:02] <elephantum> I write "queue ! osssink" in /system/gstreamer/default/audiosink
[09:02] <ds> why?
[09:03] <elephantum> I want to get rid of lags during play in rhythmbox
[09:03] <ds> queue won't help with that
[09:04] <elephantum> oh, then someone disinformed me.
[09:04] <elephantum> what can help me?
[09:08] kmaraas ([email protected]) left irc: Client Quit
[09:13] somex1 ([email protected]) left #gstreamer.
[09:17] <elephantum> then what is the queue function?
[09:19] <ds> to decouple threads
[09:20] <elephantum> decouple?
[09:21] <ds> elements in different threads need to synchronize somehow
[09:22] <elephantum> I get it, thanks
[09:31] <elephantum> ds: hm... seems enlarging buffer-size in osssink helped
[09:33] <aeyakovenk1> is there an easy way to pull and push raw data into and out of a pipeline, without using a pipe?
[09:34] <ds> fakesrc and fakesink
[09:35] <ds> but it's probably smarter to use a full gstreamer pipeline and create custom elements
[09:35] <aeyakovenk1> how do i pull those buffers out of fake src and fake sink?
[09:36] <ds> connect to the handoff signal
[09:41] <aeyakovenk1> ds: is that in some docs i can read?
[09:41] <ds> don't know
[09:46] sri ([[email protected]) left irc: Read error: 104 (Connection reset by peer)
[09:50] ChriHJW_log ([email protected]) joined #gstreamer.
[09:50] elephantum ([email protected]) left #gstreamer.
[09:55] ChrisHJW_log ([email protected]) left irc: Read error: 110 (Connection timed out)
[10:01] jcsston ([email protected]) left irc: Read error: 60 (Operation timed out)
[10:04] thomasvs ([email protected]) left irc: Read error: 110 (Connection timed out)
[10:22] jdahlin ([email protected]) joined #gstreamer.
[10:30] thomasvs ([email protected]) joined #gstreamer.
[10:36] Shoragan ([email protected]) joined #gstreamer.
[10:37] <taaz> jdahlin: um. i didn't respond to your email yet. mind if i just forward repsonse to devel list?
[10:40] <jdahlin> taaz: not at all
[10:40] dolphy ([email protected]) joined #gstreamer.
[10:41] mathrick ([email protected]) joined #gstreamer.
[10:41] <mathrick> ho
[10:42] <thomasvs> ds-work: still there ?
[10:45] aeyakovenk1 ([email protected]) left irc: Remote closed the connection
[10:45] aeyakovenko ([email protected]) left irc: Remote closed the connection
[10:53] <thomasvs> apoc__: ping ?
[10:53] thaytan ([email protected]) joined #gstreamer.
[10:54] sublett ([email protected]) joined #gstreamer.
[10:59] Action: taaz tries first pass of core debs
[11:05] <taaz> hmm.  they mostly worked
[11:20] <taaz> not registering elements on install... blah.  time for bed
[11:43] LeRoutier ([email protected]) joined #gstreamer.
[11:43] <LeRoutier> hello
[11:44] <LeRoutier> anyone has problems accessing anon CVS ?
[11:44] <jdahlin> yes
[11:44] <jdahlin> we're working on it
[11:44] <LeRoutier> ok, i'll wait so
[11:44] <LeRoutier> thx
[11:46] pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) joined #gstreamer.
[11:57] Nick change: markey -> afk
[12:10] jonathang ([email protected]) joined #gstreamer.
[12:13] sxpert_work ([email protected]) left irc: "Leaving"
[12:16] sxpert_work ([email protected]) joined #gstreamer.
[12:18] foser ([email protected]) joined #gstreamer.
[12:30] KoRnouille ([email protected]) left irc: No route to host
[12:38] Mithrandir ([email protected]) joined #gstreamer.
[12:43] jonathang_ ([email protected]) joined #gstreamer.
[12:44] jonathang ([email protected]) left irc: Nick collision from services.
[12:45] Nick change: jonathang_ -> jonathang
[12:50] iain ([email protected]) joined #gstreamer.
[12:59] smoke ([email protected]) joined #gstreamer.
[13:01] <LeRoutier> bbl
[13:01] LeRoutier ([email protected]) left irc: "Leaving"
[13:20] <thomasvs> jdahlin: anoncvs is fixed
[13:20] wheels ([email protected]) left irc: "using sirc version 2.211+KSIRC/1.3.9"
[13:20] <thomasvs> jdahlin: can you add gst-player and gst-recorder while you're at it ?
[13:22] <jdahlin> thomasvs: let me fix gst-python first
[13:22] Nick change: iain -> {
[13:22] <{> The University is closed for Saint Patrick's Day 17 March 2004 and will
[13:22] <{> reopen on Thursday 18 March 2004.
[13:22] <{> Have a Happy Christmas and and even better New Year.
[13:23] Nick change: { -> iain
[13:23] <iain> (an auto-reply)
[13:32] <iain> I think I need ds-work to look at this caps problem
[13:36] <thomasvs> I want to add back pages that say how to get gstreamer for various distros
[13:37] <thomasvs> an overview, and one page per distro
[13:37] <thomasvs> any suggestions on how to nicely do this in the website ?
[13:37] <thomasvs> ie, how to name the button on the left, and organise files ?
[13:37] <thaytan> "download" is a popular one
[13:37] <thomasvs> yeah, I know
[13:37] wingo ([email protected]) joined #gstreamer.
[13:38] <thomasvs> almost was going with that one
[13:38] <thomasvs> not sure yet though
[13:38] <thomasvs> hey africa
[13:38] <thaytan> yo wingo
[13:38] Mithrandir ([email protected]) left #gstreamer.
[13:39] <wingo> hey folks
[13:40] Action: wingo is at a free software conference -- in africa! who knew?
[13:40] <wingo> http://tacticaltech.org/africasource
[13:40] <thomasvs> wingo: rock !
[13:40] <thomasvs> promote gst a little
[13:41] LeRoutier ([email protected]) joined #gstreamer.
[13:41] <jdahlin> "a little" :-)
[13:41] <LeRoutier> re
[13:42] <jdahlin> thomasvs: when you have time, can you look why make check in gst-python fails in the documentation?
[13:42] Action: jdahlin is scared of the documentation
[13:42] <thomasvs> jdahlin: sure
[13:42] <LeRoutier> rebuiling complete gstreamer CVS. i'll try to compile gst-ffmpeg with -O0 -g instead of -ggdb to see if it still fails
[13:43] <LeRoutier> (ronald wrote that -ggdb is a registry eater)
[13:43] <thaytan> wingo: that looks like great fun
[13:43] <wingo> it's pretty neat, although it ends up you get more ngo / organizer people than real hackers
[13:43] <wingo> but fun nonetheless.
[13:44] <thaytan> linux.conf.au in australia is good for solid hacker attendance
[13:45] <thaytan> roll up, roll up! ;)
[13:48] <jdahlin> thomasvs: thanks
[13:50] kmaraas ([email protected]) joined #gstreamer.
[13:55] <iain> hey wingo
[13:56] <wingo> howdy iain
[13:58] <iain> dolphy: do you want a patch that makes gst-player use the new file chooser?
[13:58] <thomasvs> iain: not yet, busy releasing a version first
[13:58] <thomasvs> iain: which can still work with gtk 2.4
[13:58] <thomasvs> iain: unless you can make it a configure detection
[13:58] <bilboed> Is there a picture/graph of gstreamer (& plugins) class dependencies ?
[13:58] <iain> thomasvs: you mean 2.2?
[13:58] <thomasvs> iain: uh, yeah, sorry
[13:59] <thomasvs> iain: I want to release something people can test 0.8 with
[13:59] <thomasvs> iain: after that, it's probably fine to use it, preferably with a compiletime check
[13:59] <iain> umm, I don't do compile time checks :)
[13:59] Action: iain shrugs
[13:59] <iain> its like 5 lines changed :)
[14:00] <thomasvs> iain: ok, file the bug, with patch, I'll do the compile time check
[14:01] <iain> ok
[14:01] <dolphy> iain: you have one ?
[14:01] <iain> yes
[14:02] <dolphy> iain: well just after thomasvs release of the player why not
[14:02] <dolphy> for the moment it's interesting that gst-player is still working with gnome2.4
[14:06] <iain> http://bugzilla.gnome.org/show_bug.cgi?id=137488
[14:06] <thomasvs> iain: assign it to me
[14:07] <iain> what email address?
[14:07] <thomasvs> thomas at apestaart dot org
[14:08] <iain> done
[14:09] <dolphy> iain: thx
[14:09] <dolphy> iain: nice patch..
[14:13] <thomasvs> Subject: 	New Afrikaans PO file for `gstreamer'
[14:13] <thomasvs> rock :)
[14:19] <wingo> goed :)
[14:20] <wingo> das is lekker :)
[14:21] <iain> caaaanat plaaaaay file
[14:25] ProtectR ([email protected]) joined #gstreamer.
[14:28] ProtectR ([email protected]) left irc: Remote closed the connection
[14:29] kmaraas ([email protected]) left irc: Client Quit
[14:36] Nick change: afk -> markey
[14:45] Nick change: khalek -> kh_zZz
[14:53] <LeRoutier> bbl
[14:53] LeRoutier ([email protected]) left irc: "Leaving"
[14:54] jimmy_dean ([email protected]) left irc: "Carrier Signal Lost"
[15:08] wingo ([email protected]) left irc: Read error: 110 (Connection timed out)
[15:14] markey ([email protected]) left irc: "leaving"
[15:21] pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) left irc: "later"
[15:47] <sgf> Hi. Is there anyone here who does the gstreamer architecture stuff?
[15:48] <thomasvs> sgf: heh, that's an open question :)
[15:49] <sgf> I've got a few ideas, but I'm not sure whether they'd add flexibility, or just bloat. :)
[15:50] <thomasvs> sgf: well, the best idea is to mail the list
[15:50] <thomasvs> people like discussing ideas a lot there, so
[15:51] <sgf> Ah. Ok. The website said something like 'Noone uses the list, it's all irc' (almost). :)
[15:51] <sgf> I'll go look at the archives and have a think, then.
[15:51] <thomasvs> where did it say that ?
[15:55] <sgf> http://www.freedesktop.org/~gstreamer/data/doc/gstreamer/head/faq/html/chapter-developing.html#developing-mail-gstreamer-devel <- Admittedly, it does say that mail is probably a good idea for general consensus. :)
[15:55] <sgf> I think I just misread it a bit...
[15:59] alley_cat ([email protected]) left irc: Read error: 60 (Operation timed out)
[16:03] paguib ([email protected]) joined #gstreamer.
[16:07] markey ([email protected]) joined #gstreamer.
[16:09] jimmy_dean ([email protected]) joined #gstreamer.
[16:11] alley_cat ([email protected]) joined #gstreamer.
[16:11] markey ([email protected]) left irc: Client Quit
[16:17] teuf ([email protected]) joined #gstreamer.
[16:21] Marsupilami23 ([email protected]) joined #gstreamer.
[16:26] wingo ([email protected]) joined #gstreamer.
[16:32] jimmy_dean ([email protected]) left irc: "Carrier Signal Lost"
[16:38] yippi ([email protected]) joined #gstreamer.
[16:39] teuf ([email protected]) left irc: Read error: 60 (Operation timed out)
[16:41] <bitshifter> gst-inspect says the lame plugin should encode to VBR by default, yet it doesn't, even if I set the properties explicitely. Any hints?
[16:42] <thomasvs> bitshifter: hm, good question.  check the source, I'd say.
[16:42] <thomasvs> bitshifter: could be that someone miseed implementing that
[16:43] <bitshifter> k, thanks.  There isn't an ogg muxer yet, is there?
[16:44] <bitshifter> thomasvs: btw, APE decoding works fine for me with current CVS gst-monkeysaudio 
[16:47] <thomasvs> bitshifter: ah, nice.  have an ape file somewhere ?
[16:47] paguib ([email protected]) left irc: Remote closed the connection
[16:47] <bitshifter> yep, but it's 279 MB :)
[16:48] <thomasvs> yikes.
[16:48] <thomasvs> n/m then _:)
[16:51] jcsston ([email protected]) joined #gstreamer.
[16:58] wingo_ ([email protected]) joined #gstreamer.
[16:59] wingo ([email protected]) left irc: Read error: 110 (Connection timed out)
[17:00] markey ([email protected]) joined #gstreamer.
[17:03] jimmy_dean ([email protected]) joined #gstreamer.
[17:08] ProtectR ([email protected]) joined #gstreamer.
[17:08] ProtectR ([email protected]) left irc: "Client exiting"
[17:10] jcsston ([email protected]) left irc: Read error: 60 (Operation timed out)
[17:16] mx|ibook ([email protected]) joined #gstreamer.
[17:18] wingo_ ([email protected]) got netsplit.
[17:18] sxpert_work ([email protected]) got netsplit.
[17:18] bitshifter ([email protected]) got netsplit.
[17:18] joem ([email protected]) got netsplit.
[17:20] wingo_ ([email protected]) returned to #gstreamer.
[17:20] sxpert_work ([email protected]) returned to #gstreamer.
[17:20] bitshifter ([email protected]) returned to #gstreamer.
[17:20] joem ([email protected]) returned to #gstreamer.
[17:21] sack ([email protected]) joined #gstreamer.
[17:30] jcsston ([email protected]) joined #gstreamer.
[17:31] wingo_ ([email protected]) left irc: Read error: 110 (Connection timed out)
[17:34] sjoerd ([email protected]) joined #gstreamer.
[17:35] KoRnouille ([email protected]) joined #gstreamer.
[17:42] ProtectR ([email protected]) joined #gstreamer.
[17:43] markey ([email protected]) left irc: "leaving"
[17:50] jcsston ([email protected]) left irc: Read error: 60 (Operation timed out)
[17:50] ProtectR ([email protected]) left irc: "Client exiting"
[17:54] ishamael ([email protected]) joined #gstreamer.
[17:54] <ishamael> hi.  i'm having a problem with gstreamer 0.8.  i compiled from source, and everything appears to install OK, but when i run gst-register-0.8 none of the plugins from gst-plugins gets registered.  i see the shared objects in /usr/lib/gstreamer-0.8....  so why doesnt gst-register pick them up? is there a new installation step i missed or...?
[17:56] <mathrick> ishamael: perhaps run ldconfig?
[17:58] <ishamael> tried that
[17:59] BBB ([email protected]) joined #gstreamer.
[17:59] <BBB> hello
[17:59] <mathrick> yo BBB
[17:59] <BBB> anyone wanna debug libmpeg2 for me?
[17:59] <BBB> so I can focus on gst-rec again
[17:59] <mathrick> ishamael: any errors it gave?
[17:59] <ishamael> ldconfig? no
[18:00] <mathrick> ishamael: no, gst-register
[18:00] <ishamael> gst-register gives none either
[18:01] <sack> ishamael, did add the path parameter , i mean for instance ./gst-register-0.8 --gst-plugin-path=/usr/local/gstreamer/plungins/
[18:01] jcsston ([email protected]) joined #gstreamer.
[18:02] <sack> ishamael, i just compile both ( core , plugins ) and it works fine
[18:02] <ishamael> still nothing
[18:02] <sack> well some problems with po files :-/
[18:03] <mathrick> ishamael: run it with --gst-plugin-spew, I think that's the option name
[18:03] <ishamael> hrmph
[18:03] <ishamael> with --gst-plugin-spew it doesnt give any more information that normal?
[18:04] <ishamael> er, than normal.
[18:05] <BBB> brb
[18:05] BBB ([email protected]) left irc: "Download Gaim: http://gaim.sourceforge.net/"
[18:05] hyriand ([email protected]) left irc: Read error: 60 (Operation timed out)
[18:06] <mathrick> ishamael: well, it should
[18:07] mx|ibook ([email protected]) left irc: "Leaving"
[18:08] <mathrick> ishamael: does it print all these "adding plugin foo... " lines?
[18:10] <sack> thomasvs, gstreamer-0.8 compile ( core , plugins ) without problems ( po files problem was due a mess in cvs update ). :-)
[18:16] <dolphy> sack: hey :)
[18:17] <sack> dolphy, :-) How was your vacations ?
[18:17] <dolphy> sack: cloudy
[18:17] <dolphy> sack: good snow though
[18:17] <dolphy> sack: and good cheesy diners :)
[18:17] <sack> dolphy, great !
[18:18] <sack> dolphy, great work with gstreamer ... not is growing really fast :-)
[18:18] <dolphy> sack: well you better congratulate thomasvs and johan :)
[18:19] <sack> dolphy, heh i know :-) 
[18:21] markey ([email protected]) joined #gstreamer.
[18:23] ishamael ([email protected]) left irc: Read error: 110 (Connection timed out)
[18:27] sri ([[email protected]) joined #gstreamer.
[18:30] <thomasvs> sack: hey :) for the .po file, I managed to get the GNU translation project to translate us
[18:30] <thomasvs> sack: which is good, but it means we shouldn't touch them anymore.
[18:30] <thomasvs> sack: if you want to still translate, I think you should join them.
[18:31] Action: thomasvs should send out a mail about that.
[18:31] <sack> thomasvs, wohhhhh ! cool :-)
[18:31] <thomasvs> yeah, I already got serbian and afrikaans :)
[18:31] <sack> thomasvs, i like it ... no more divide and conquer ... now is time to -> " Join and become Stronger "
[18:32] <sack> thomasvs, well i guess i'll bother a bit more to Ronald about v4l/v4l2 :-D
[18:34] <thomasvs> heh, yeah, do :)
[18:36] jcsston ([email protected]) left irc: "Pong Timeout"
[18:37] somex1 ([email protected]) joined #gstreamer.
[18:43] LeRoutier ([email protected]) joined #gstreamer.
[18:43] <LeRoutier> hello
[18:45] LeRoutier ([email protected]) left irc: Client Quit
[18:49] <thomasvs> [gst-head] [thomas@otto po]$ gst-launch
[18:49] <thomasvs> FEL: rörledningen kunde inte konstrueras: tom rörledning är inte tillåtet.
[18:50] <jdahlin> rörledningen, lol
[18:50] <jdahlin> correct translation, but it sounds very silly
[19:01] <thomasvs> yeah :)
[19:01] <thomasvs> hard to avoid that in translations
[19:01] <thomasvs> what is it, like plumbing ?
[19:02] <jdahlin> no, just the translation for "the pipeline"
[19:05] dolphy ([email protected]) left irc: "Network down, IP Packets delivered via UPS"
[19:14] pb_ (~pb@2002:5169:e72f:1:240:95ff:fe30:aa84) joined #gstreamer.
[19:30] Nick change: apoc__ -> apoc
[19:31] <apoc> hi
[19:31] Action: iain suspects that audioconvert is the source of his problem
[19:32] <iain> man, the new indent rules make the code look like ass
[19:37] kwm ([email protected]) joined #gstreamer.
[19:41] ChrisHJW ([email protected]) joined #gstreamer.
[19:48] <ds-work> bouncy bouncy
[19:49] steve_b ([email protected]) left irc: "I don't think I'm going to use hamster style any more."
[19:49] <iain> yey, ds-work is here to brighten my day
[19:49] sack ([email protected]) left irc: Read error: 60 (Operation timed out)
[19:50] jonathang ([email protected]) left irc: "Leaving"
[19:53] kmaraas ([email protected]) joined #gstreamer.
[19:53] <thomasvs> ds-work: so, are you going to reply to jeff's mail ?
[19:54] <ds-work> jeff wrote email?
[19:55] sublett ([email protected]) left irc: "Nice of you to put me on the money"
[19:55] <thomasvs> ds-work: did you drop off the planet or something ?
[20:01] dolphy ([email protected]) joined #gstreamer.
[20:03] lupusBE ([email protected]) joined #gstreamer.
[20:04] <ds-work> should we avoid making string changes?
[20:05] ishamael ([email protected]) joined #gstreamer.
[20:09] Shoragan ([email protected]) left irc: "Leaving"
[20:13] yippi ([email protected]) left irc: "Client exiting"
[20:14] <bitshifter> thomasvs: when you get that ape sample from Jeremy, could you put it into the media folder?
[20:15] <iain> ds-work: could you look at a gst-debug log for me and see if you can see why mad won't connect to audioconvert?
[20:16] yippi ([email protected]) joined #gstreamer.
[20:16] thomasvs ([email protected]) left irc: Read error: 113 (No route to host)
[20:16] kwm ([email protected]) left irc: Remote closed the connection
[20:16] <ds-work> iain: sure
[20:18] sublett ([email protected]) joined #gstreamer.
[20:19] <iain> ds-work: cool, thanks. http://62.188.60.41/~iain/log.gz
[20:21] sgf ([email protected]) got netsplit.
[20:22] <jdahlin> ds-work: definitely no string changes
[20:24] <ds-work> jdahlin: what about bugs from translators saying "fix your strings"?
[20:24] teuf ([email protected]) joined #gstreamer.
[20:25] <jdahlin> ds-work: I'm not sure, hmm.
[20:25] <jdahlin> ds-work: perhaps we should ask them
[20:27] sgf ([email protected]) got lost in the net-split.
[20:32] steve_b ([email protected]) joined #gstreamer.
[20:42] <ds-work> iain: you're using filtercaps audio/x-raw-float, buffer-frames=(int)256
[20:43] <ds-work> audioconvert only supports buffer-frames=0
[20:43] <ds-work> don't know why
[20:44] <iain> oh, ok
[20:44] <iain> it used to work
[20:45] <ds-work> ask Company
[20:45] <ds-work> I would imagine it's because audioconvert can't guarantee properly framed data
[20:45] <iain> ok
[20:47] thomasvs ([email protected]) joined #gstreamer.
[20:51] jdahlin ([email protected]) left irc: "Leaving"
[20:56] kwm ([email protected]) joined #gstreamer.
[20:57] sublett ([email protected]) left irc: "Nice of you to put me on the money"
[21:00] yippi ([email protected]) left irc: Remote closed the connection
[21:01] aflinta ([email protected]) joined #gstreamer.
[21:01] yippi ([email protected]) joined #gstreamer.
[21:11] thomasvs ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[21:30] Action: ds-work considers the possibility of taking 3 months off and completely rewriting GStreamer core
[21:31] sublett ([email protected]) joined #gstreamer.
[21:33] thomasvs ([email protected]) joined #gstreamer.
[21:34] <ds-work> thomasvs: which Jeff email are you talking about?
[21:40] sgf ([email protected]) joined #gstreamer.
[21:44] iain ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[21:50] <thomasvs> ds-work: Jeff waugh ?
[21:50] <thomasvs> gnome release manager ?
[21:58] alexthelion ([email protected]) joined #gstreamer.
[22:02] <alexthelion> Hi. I have an alsasink here which is refusing to collaborate. Has anybody a minute to help me?
[22:11] kmaraas_ ([email protected]) joined #gstreamer.
[22:13] kmaraas ([email protected]) left irc: Read error: 110 (Connection timed out)
[22:15] Nick change: kmaraas_ -> kmaraas
[22:21] <thomasvs> alexthelion: shoot
[22:21] kmaraas ([email protected]) left irc: Connection reset by peer
[22:23] <alexthelion> ok, it's a 0.7.6. if I type gst-launch sinesrc ! alsasink, it just says ERROR: pipeline doesn't want to play.
[22:23] <alexthelion> osssink works. alsa with other players works, too.
[22:24] <taaz> ds-work: what would you rewrite?
[22:24] <thomasvs> alexthelion: what more does it say beside that ?
[22:25] <alexthelion> unfortunately nothing more. except "running pipeline" just before that.
[22:28] Nick change: kh_zZz -> kh_out
[22:29] <ds-work> taaz: whatever is necessary.  Everything would be reviewed
[22:29] <alley_cat> alexthelion: use sinesrc ! audioscale ! audio/x-raw-int,rate=48000 ! alsasink
[22:30] <alley_cat> or rate=44100, whatever your card works best with
[22:30] <alexthelion> alley_cat, indeed...
[22:33] <alexthelion> I set the default audiosink to alsasink in gconf-editor. is there a way to verify that e.g. rhythmbox now indeed uses alsasink?
[22:34] <alexthelion> or do client programs not even need to read that value?
[22:36] yippi ([email protected]) left irc: Remote closed the connection
[22:37] yippi ([email protected]) joined #gstreamer.
[22:38] markey ([email protected]) left irc: "leaving"
[22:39] <thomasvs> alexthelion: no, they read it
[22:41] yippi ([email protected]) left irc: Client Quit
[22:41] yippi ([email protected]) joined #gstreamer.
[22:42] m_wheels ([email protected]) joined #gstreamer.
[22:52] <alexthelion> thanks, everybody, and bye.
[22:52] alexthelion ([email protected]) left irc: "Leaving"
[23:00] kmaraas ([email protected]) joined #gstreamer.
[23:00] <thomasvs> do avi files not have the concept of "play me back with this aspect ratio" ?
[23:02] <thomasvs> hm, it seems they don't
[23:02] <thomasvs> bummer
[23:02] <thomasvs> ChrisHJW: why is #matroska inviteonly ?
[23:02] <Misirlou> thomasvs: ha ha ha ha ha
[23:03] <ds-work> keep out the trolls^Wlusers, perhaps
[23:04] dolphy ([email protected]) left irc: "Network down, IP Packets delivered via UPS"
[23:04] <Misirlou> Gotta love that free software camaraderie.
[23:04] Nick change: m_wheels -> wheels
[23:04] <thomasvs> well, I'm bummed because I'm about to convert some dvd's, and I would like to convert them to some format that can keep the non-square pixel stuff
[23:04] <thomasvs> because I intend to playback on my home system that is displaying in non-square pixels soon
[23:05] <thomasvs> on the side, I'm also wondering how we will want to do this in gst - use a caps field ?
[23:07] <ds-work> thomasvs: pixel_width and pixel_height
[23:07] <ds-work> thomasvs: it's not supported very well right now
[23:07] <thomasvs> I assume not, but I hope to focus a little on what gst needs to change to be used in something like dave/dina
[23:07] <ds-work> thomasvs: MPEG video streams should be able to handle it fine
[23:07] <thomasvs> so I guess that's my spare time project for the next dev cycle :)
[23:07] <thomasvs> yeah, MPEG can do it.
[23:08] <thomasvs> but then I'm stuck with big files
[23:08] <ds-work> you can put MPEG video into AVIs
[23:08] <thomasvs> yeah, I'm saying MPEG is not that useful as a compression codec.
[23:08] <ds-work> ah, ok
[23:08] <thomasvs> unless mpeg-4 has this too ?
[23:08] <ds-work> probably
[23:08] <thomasvs> hm
[23:08] <thomasvs> although I think the pixel aspect is stored in the container format, even for mpeg
[23:09] <thomasvs> ie, mpeg in avi just has the video stream in the avi, without the info of the pixel aspect
[23:09] <thomasvs> maybe I'm better off going to square pixels for now.
[23:09] <thomasvs> seems like a waste though
[23:09] <ds-work> no, pixel aspect is part of the video stream
[23:10] ishamael ([email protected]) left irc: Read error: 60 (Operation timed out)
[23:11] thaytan ([email protected]) left irc: Read error: 113 (No route to host)
[23:12] Action: ds-work thinks about splitting caps into "negotiation", "preferences", "informative", and "codec metadata"
[23:15] BBB ([email protected]) joined #gstreamer.
[23:16] _axel ([email protected]) joined #gstreamer.
[23:17] kmaraas ([email protected]) left irc: "Leaving"
[23:18] <_axel> hi, i built gst from cvs and the command 'gst-launch-0.8 videotestsrc ! ximagesink' works perfectly, but 'gst-launch-0.8  videotestsrc ! vertigoTV ! ximagesink' doesnt work, i get 'ERROR: pipeline doesnt want to play'
[23:18] <_axel> gst-inspect-0.8 shows vertigoTV is installed
[23:18] <_axel> any help?
[23:19] <BBB> gst-inspect vertigoTV
[23:19] <BBB> see the pads
[23:19] <BBB> it might only work on YUV video
[23:19] <ds-work> works here
[23:19] <BBB> in which case you need a ffcolorspace element between ximagesink and vertigoTV
[23:19] <BBB> ds-work: it probably works only on 32bbp RGB or so then
[23:19] <thomasvs> _axel: add ffcolorspace
[23:19] <BBB> he might have a 16bpp display
[23:19] <thomasvs> BBB: can you help me with my square pixel problem ?
[23:19] <_axel> i do have 16bpp display
[23:19] <_axel> the rest of what you said sounds like Russian to me
[23:19] <BBB> then add a colorspace :)
[23:20] <_axel> BBB: how?
[23:20] <BBB> videotestsrc ! vertigoTV ! ffcolorspace ! ximagesink
[23:20] Action: _axel tries
[23:20] <_axel> WARNING: erroneous pipeline: no element "ffcolorspace"
[23:20] Action: thomasvs tried fixing the bonobo control today so it worked in evolution
[23:20] <BBB> you need gst-ffmpeg
[23:20] <thomasvs> _axel: you don't have 0.8 do you ?
[23:20] <thomasvs> oh
[23:21] <_axel> thomasvs: i think so
[23:21] <BBB> what's the gst-plugin one called?
[23:21] <thomasvs> _axel: use ffmpegcolorspace
[23:21] <thomasvs> _axel: that'll work
[23:21] <thomasvs> BBB: so, no love from you ?
[23:21] Action: thomasvs sulks
[23:21] <BBB> thomasvs: what problem?
[23:21] <_axel> ok i need to rebuild plugins cause i didnt have ffmpeg-devel at build time
[23:21] <thomasvs> BBB: it's simple really
[23:21] <BBB> _axel: you don't need it
[23:21] <thomasvs> _axel: you don't.  you have ffmpegcolorspace.  run gst-inspect ffmpegcolorspace
[23:21] <BBB> _axel: it's always there
[23:21] <thomasvs> BBB: so I have a bunch of friends dvds, at 720x576, with nonsquare pixels, right ?
[23:22] <_axel> uhm
[23:22] <_axel> gst-inspect-0.8 ffmpegcolorspace
[23:22] <_axel> no such element or plugin 'ffmpegcolorspace'
[23:22] <BBB> thomasvs: yes
[23:22] <thomasvs> BBB: directfb can display to pal tv natively, at this resolution
[23:22] <thomasvs> _axel: the nyou don't have 0.8.0, but 0.7.6
[23:22] <thomasvs> BBB: so ...
[23:22] <_axel> ok, /me updates
[23:22] <thomasvs> BBB: apparently I cannot transcode to avi and keep the information that the pixels are nonsquare
[23:22] <_axel> should i rebuild gst or gst-plugins?
[23:22] <BBB> indeed, you can't
[23:22] <thomasvs> (unless, says ds, if I use MPEG as the video codec inside the avi)
[23:23] <thomasvs> BBB: it looks to me like matroska can actually do this
[23:23] <thomasvs> BBB: and I'm wondering if ogm can.
[23:23] <BBB> yes, but matroskamux doesn't implement it
[23:23] <BBB> dunno about ogm
[23:23] <thomasvs> BBB: but anyway, what would you do in my situation ?
[23:23] <BBB> not care and just rip to AVI
[23:23] <BBB> and resize during playback
[23:23] <ds-work> could we replace 'colorspace' with 'ffmpegcolorspace'?
[23:23] <thomasvs> ie, I want to be able to make a file with the nonsquare pixels, but so that a player can figure it out automatically
[23:23] <thomasvs> ds-work: we conceivably could.
[23:23] <thomasvs> ds-work: if you implement plugin aliasing somehow :)
[23:24] <ds-work> say, by registering an element twice? :)
[23:24] <thomasvs> ds-work: doesn't work :)
[23:24] <thomasvs> BBB: so, you'd keep it nonsquare too then ?
[23:24] <thomasvs> BBB: why does everyone transcode to 4:3 btw ?
[23:25] <ds-work> thomasvs: why not?
[23:25] <thomasvs> ds-work: why it doesn't work ?
[23:25] Action: BBB wants a UI to do all this :(
[23:25] <thomasvs> ds-work: oh, you mean use the exact same code twice ?
[23:25] <thomasvs> ds-work: or is there a way to register the same plugin code under two diff names ?
[23:25] <thomasvs> hm, yeah
[23:25] <thomasvs> that probably works :)
[23:26] <ds-work>   gst_element_register (plugin, "ffmpegcolorspace",
[23:26] <ds-work>       GST_RANK_NONE, GST_TYPE_FFMPEGCOLORSPACE);
[23:26] <ds-work>   gst_element_register (plugin, "colorspace",
[23:26] <ds-work>       GST_RANK_NONE, GST_TYPE_FFMPEGCOLORSPACE);
[23:26] <thomasvs> yeah
[23:26] <thomasvs> stupid me
[23:27] <thomasvs> BBB: also, suppose I'm ripping to xvid, 350 mb per file.  Should I keep the same pixel size, or should I cut the size in half or four ?
[23:27] Action: BBB will work on gst-rec for the next few days
[23:27] <BBB> ds-work: did you fix qtdemux yet?
[23:27] <thomasvs> BBB: I'm assuming making it smaller will make the result better due to the codec having more room to play with ?
[23:27] <BBB> how long per movie?
[23:27] <ds-work> BBB: no, but I reproduced a bunch of qtdemux problems last night
[23:27] <thomasvs> BBB: 22 minutes for 320 mb of video
[23:27] <BBB> ds-work: d'oh :D
[23:27] yippi ([email protected]) left irc: Remote closed the connection
[23:27] <BBB> I'd use half-PAL then
[23:28] Action: ds-work should work on gst-mozilla-plugin
[23:28] <BBB> full-PAL would look awful
[23:28] <thomasvs> BBB: which is ? that's not 360x288 is it ?
[23:28] <BBB> 384x288 (squarepix) or 352x288 (you)
[23:28] <BBB> you=CCIR601
[23:29] <BBB> or so
[23:29] Action: BBB is bad at names
[23:29] <thomasvs> huh ? full pal is 720x576 no ?
[23:29] <BBB> yes\
[23:29] <BBB> but it's always a multiple of 16
[23:29] <thomasvs> eek
[23:29] lupusBE ([email protected]) left irc: "Bezig met verlaten"
[23:29] <BBB> so (720/2)~16
[23:29] <_axel> gst autogen told me i suck because my 'Atomic reference counting is out of date'... do i really suck?
[23:29] <thomasvs> well, then I have to crop ?
[23:29] <BBB> use videocrop to  cut off the left/right 8 pixels
[23:29] <thomasvs> ugh
[23:30] <thomasvs> I don't want to do that :)
[23:30] <thomasvs> what's this 16 crap ?
[23:30] <BBB> it's video
[23:30] yippi ([email protected]) joined #gstreamer.
[23:30] <BBB> DCT blocks (I think) are always 16x16
[23:30] <thomasvs> hm, yeah
[23:30] <thomasvs> argh, this sucks
[23:30] <ds-work> 8x8, but Cb and Cr are subsampled
[23:30] <BBB> hi yippi
[23:30] <BBB> ds-work: eh, yes, that
[23:31] <_axel> what is this 'Atomic reference counting' thing? a subtle way of telling me to update to kernel 2.6?
[23:31] <yippi> heya
[23:31] <BBB> _axel: no
[23:31] <_axel> ah
[23:31] <BBB> _axel: don't mind it, just continue
[23:31] <BBB> it didn't error out, did it?
[23:31] <_axel> no, i just continued anyway
[23:31] <_axel> just wanted to make sure
[23:31] <thomasvs> BBB: how about 320x256 ?
[23:32] <BBB> ugh
[23:32] <BBB> no
[23:32] <BBB> 352x288 is really good
[23:32] <thomasvs> BBB: if I have to crop, it means I lose stuff.  I want to not lose stuff.  otherwise I can just as well rip to square pixels.
[23:33] <BBB> then rescale it
[23:33] <thomasvs> man
[23:33] <thomasvs> digital video was designed badly from the start it seems :)
[23:33] <BBB> 352x288 is the official VCD resolution, you really don't lose that much
[23:33] Action: ds-work notes that we don't use asm/atomic.h at all
[23:33] <BBB> in case you didn't know, the border pixels are always trash anywya
[23:33] <thomasvs> what's wrong with 320x256 then ?
[23:33] <BBB> too small, and your DVD player won't like it
[23:33] <thomasvs> what DVD player ?
[23:34] <thomasvs> ok, so how about 400x320 ?
[23:34] <BBB> why?!?
[23:34] <thomasvs> to keep 5:4
[23:34] <BBB> use 384x288 or 352x288
[23:34] <BBB> that's official resolutions
[23:34] <BBB> it's not 5:4
[23:34] <BBB> just rip to sqpx then :)
[23:35] <thomasvs> gah
[23:35] <BBB> 5:4 is the data aspect ratio
[23:35] <thomasvs> I hate sqpx :)
[23:35] <BBB> not the video aspect ratio
[23:35] <BBB> because/
[23:35] <thomasvs> BBB: I know that 5:4 is the data aspect ratio
[23:35] aeyakovenko2 ([email protected]) joined #gstreamer.
[23:35] <aeyakovenko2> hello
[23:35] <thomasvs> BBB: I'm just trying to figure out if there is no decent way of not having to do a scaling to square then back to nonsquare
[23:37] <ds-work> who uses --enable-plugin-builddir?
[23:37] <BBB> thomasvs: dunno
[23:37] <BBB> ds-work: autogen.sh
[23:37] <thomasvs> ds-work: I do
[23:37] <thomasvs> wait
[23:37] <thomasvs> I don't
[23:37] <thomasvs> I stopped using it at some point
[23:38] <ds-work> thomasvs: I think you were the only one.  I'm gonna rip it out unless someone complains.  (after we branch for 0.9)
[23:39] <thomasvs> ds-work: you "think" ? are you clairvoyant or smth ? :)
[23:39] <ds-work> btw, who wants to go through the bug list and ping all the 0.6 bugs?
[23:39] <ds-work> thomasvs: I've asked before
[23:39] <ds-work> and since you asked, yes, I am clairvoyant
[23:39] <BBB> ds-work: close all that don't appl
[23:39] <BBB> y0.8is out
[23:39] <BBB> 0.6is now deprecated and unsupported
[23:40] <BBB> imo
[23:40] <BBB> oh, btw
[23:40] <BBB> even if we branch
[23:40] <ds-work> BBB: I was thinking "please reproduce with 0.8, NEEDINFO"
[23:40] <BBB> please please please keep supporting it
[23:40] <BBB> I'm intending to long-time work on 0.8.x
[23:40] <BBB> ds-work: ok
[23:41] <ds-work> *I'm* not the one who dropped 0.6 like a lead brick
[23:41] <BBB> benjamin is ;)
[23:41] jcsston ([email protected]) joined #gstreamer.
[23:41] <BBB> and I don't want that again
[23:41] <BBB> we need 0.8.x for the next few months
[23:41] <BBB> maybe even a full year or more
[23:41] <ds-work> I developed against 0.6 until a good deal after the debug changes
[23:41] Action: BBB whistles qtdemux
[23:42] <ds-work> that's when it became too much of a pain to deal with the differences
[23:42] <BBB> did you read my email?
[23:42] <BBB> stsd_data and esds_data fine with you?
[23:42] <ds-work> re?
[23:42] <thomasvs> BBB: what about svcd anamorph, 480x576 ?
[23:42] <BBB> for respectively svq3 and mpeg4
[23:42] <thomasvs> BBB: does that get used ?
[23:42] <ds-work> I'd prefer something generic, like "codec_data"
[23:43] <BBB> thomasvs: yes
[23:43] <thomasvs> BBB: does it make sense to use it in my case ?
[23:43] <BBB> ds-work: some codecs might have two, and I want to separate per atom if possible
[23:43] <ds-work> BBB: what do the two mean?
[23:43] <BBB> thomasvs: don't think so
[23:43] <BBB> divx is better
[23:43] <BBB> ds-work: dunno
[23:43] <ds-work> BBB: btw, I already have ideas on how to do this correctly
[23:44] <BBB> explain
[23:44] <ds-work> BBB: well, correctly in 0.9
[23:44] <BBB> yes, yes
[23:44] <BBB> we can always backport
[23:44] <ds-work> what you are doing is correct in 0.8
[23:44] <BBB> how?
[23:45] <ds-work> basically, I want to split caps into negotiation and informative parts
[23:45] <ds-work> perhaps more
[23:45] <BBB> ?
[23:45] <ds-work> pixel aspect is "informative"
[23:45] <BBB> hm
[23:45] <ds-work> codec data is informative
[23:45] <ds-work> preferred caps is informative
[23:45] <BBB> width is informative
[23:45] <BBB> for mpegdemux
[23:46] <BBB> but not for videotestsrc
[23:46] <BBB> ...
[23:46] <BBB> ?
[23:46] <ds-work> yeah, there are some things that are ambiguous
[23:46] <BBB> that's variable and fixed values in getcaps reply
[23:46] <BBB> don't bother
[23:46] <BBB> let's fix real bugs first
[23:46] <ds-work> for raw video, width is necessary
[23:46] <BBB> autoplugger
[23:46] <BBB> dvd
[23:46] <BBB> scheduler
[23:47] <ds-work> well, if I start doing core work, I'm going to do a *lot* of core work
[23:47] <BBB> I know... start with the important things :)
[23:48] Action: BBB should start profiles some day
[23:48] <BBB> s/start/finish
[23:48] <BBB> as soon as we branch off 0.9.x, basically
[23:49] <ds-work> on the way to work today, I was thinking of really basic scheduling changes
[23:50] <ds-work> like potentially concurrently running videotestsrc ! xvimagesink
[23:50] <ds-work> with no "put this in a separate thread" hints
[23:50] <ds-work> and getting rid of queue
[23:51] kwm ([email protected]) left irc: "Leaving"
[23:51] <BBB> queue is still useful if apps explicitely want something
[23:51] <BBB> maybe
[23:51] <ds-work> yes, but it would no longer be a special element
[23:52] <BBB> ok
[00:00] --- Thu Mar 18 2004
[00:02] yippi ([email protected]) left irc: Remote closed the connection
[00:06] kmaraas ([email protected]) joined #gstreamer.
[00:09] <ChrisHJW> <thomasvs> BBB: apparently I cannot transcode to avi and keep the information that the pixels are nonsquare   <----- you can, but only with MPEG4 video
[00:09] <ChrisHJW> or RV9
[00:09] <ChrisHJW> both codecs have AR info in the elementary video stream
[00:09] <ChrisHJW> if you go for a different encoding, MPEG, RM or MKV container are your choice
[00:09] teuf ([email protected]) left irc: "Client exiting"
[00:10] <ChrisHJW> but MPEG, as we know, can normally only contain MPEG1/2 video
[00:10] <ChrisHJW> and RM can only work with RV stuff
[00:10] <ChrisHJW> so, for a general compression format, MKV is indeed your choice
[00:10] <ChrisHJW> not sure about MOV :/
[00:11] <BBB> MPEG-1 contains it inside the ES
[00:11] <BBB> so MPEG-1 inside AVI would work, too
[00:11] <BBB> and yes, that's valid
[00:11] <BBB> fourcc = MPGI
[00:11] <BBB> (18:09:55) ChrisHJW: so, for a general compression format, MKV is indeed your choice <- *sigh* we know :P
[00:12] <ChrisHJW> :)
[00:12] Action: BBB still has chris' post queued in his inbox
[00:12] <ChrisHJW> OGM doesnt have AR flag in the container
[00:12] Action: BBB will respond tonight
[00:12] wheels ([email protected]) left irc: "2.6"
[00:12] <BBB> brb
[00:12] <BBB> goign home
[00:12] BBB ([email protected]) left #gstreamer.
[00:16] Action: _axel still compiling 18 hours later
[00:16] <_axel> i can feel my beard growing
[00:26] Rotty ([email protected]) left irc: Remote closed the connection
[00:26] Nick change: kh_out -> khalek
[00:27] <_axel> oh, finally
[00:34] _axel ([email protected]) left irc: "Leaving"
[00:49] thaytoo ([email protected]) joined #gstreamer.
[00:51] wheels ([email protected]) joined #gstreamer.
[00:55] asdlfjkklsj ([email protected]) joined #gstreamer.
[00:55] <asdlfjkklsj> what's a video player that uses gstreamer?
[00:55] <ds-work> totem
[00:55] <ds-work> gst-player
[01:00] <thaytoo> morning all
[01:01] ChrisHJW ([email protected]) left irc: Read error: 110 (Connection timed out)
[01:07] Shoragan ([email protected]) joined #gstreamer.
[01:17] Nick change: mathrick -> mathrick|sleep
[01:17] <mathrick|sleep> g'nitght
[01:17] _axel ([email protected]) joined #gstreamer.
[01:20] iain ([email protected]) joined #gstreamer.
[01:20] <iain> ds-work: thanks, setting buffer frames to 0 worked
[01:20] bilboed_ ([email protected]) joined #gstreamer.
[01:21] pb_ (~pb@2002:5169:e72f:1:240:95ff:fe30:aa84) left irc: "Client exiting"
[01:21] <_axel> what would be the right pipeline for applying a vertigo filter to an avi file? im trying filesrc location="bla" ! typefind ! spider ! vertigoTV ! ffmpegcolorspace ! xvvideosink but it doesnt seem to work
[01:21] <_axel> it works without the vertigo filter though
[01:23] <ds-work> you probably want ximagesink or xvimagesink
[01:23] <_axel> uh yes sorry
[01:23] <_axel> im using imagesink
[01:23] <ds-work> and put a ffmpegcolorspace before and after the vertigoTV
[01:23] <_axel> k
[01:24] <iain> *sigh* deinterleave still doesn't deinterleave correctly
[01:24] bilboed ([email protected]) left irc: Read error: 110 (Connection timed out)
[01:24] <_axel> NEAT :D
[01:24] <iain> and audioconvert still converts everything to mono unless told otherwise :/
[01:25] <ds-work> iain: then tell it otherwise
[01:25] <iain> ds-work: I can't, because I don't know when "otherwise" is
[01:25] <ds-work> iain: add a fixate signal callback
[01:26] <iain> ds-work: surely it should be doing it itself
[01:26] <iain> ds-work: if stereo data is coming in, stereo data should be going out
[01:26] <ds-work> iain: no
[01:26] <iain> (unless it is told otherwise)
[01:26] <ds-work> only if the formats are exactly the same
[01:27] <iain> why?
[01:27] <ds-work> because a converter prefers the caps of an existing pad
[01:28] <ds-work> which may or may not be useful
[01:28] <_axel> ds-work: and once i get the warped video output working, how would i save it to an (encoded, not raw frames) file?
[01:28] <iain> that essentially makes audioconvert useless
[01:29] <ds-work> iain: no, because it's _always_ up to the application to encourage a pipeline to pick a good format
[01:29] <iain> there are cases when the application cannot know
[01:29] <iain> spider ! audioconvert ! deinterleave
[01:30] <iain> spider can output either stereo or mono, I can filer the audioconvert ! deinterleave to say "give me stereo", but I don't know what spider is going to output
[01:30] <iain> the default should be that what is expected
[01:31] <ds-work> what is expected?
[01:31] <iain> that stereo data in gives stereo data out
[01:31] <ds-work> not if the output has already specified that it prefers mono
[01:32] <iain> indeed, but the output has not specified that
[01:32] <ds-work> yes, it did
[01:32] <iain> the output in this case has said "I don't give a shit"
[01:32] <ds-work> right, so the core chose
[01:32] <ds-work> you lose
[01:33] <iain> the core should choose a sensible "Oh, its stereo in, it probably wants stereo out"
[01:33] <iain> otherwise its fucking useless
[01:33] <ds-work> there is ABSOLUTELY NO WAY TO KNOW THAT when the question is first asked
[01:33] Shoragan ([email protected]) left irc: "Leaving"
[01:34] <iain> sink caps has said "I'm stereo", src caps has said "I don't care"
[01:35] <ds-work> no
[01:35] <ds-work> sink caps said "i don't care", src caps said "i don't care"
[01:36] <iain> surely it renegotiates once sink caps has worked out what it wants?
[01:36] <ds-work> yes, but it's obviously still unconstrained if it's choosing mono
[01:37] <iain> if its unconstrained once it works out that its a stereo stream surely there's a bug somewhere
[01:38] <ds-work> no
[01:38] <iain> why?
[01:39] <ds-work> if you have sinesrc ! audioconvert ! audioconvert ! audioconvert ! audioconvert ! osssink, you're application better be prepared to tell all those audioconverts what to do
[01:39] sublett ([email protected]) left irc: "my friend the end"
[01:39] <iain> I would expect them to do nothing
[01:39] <iain> act like identity elements by default
[01:39] <iain> unless I tell them to do something
[01:40] <ds-work> sorry, that's not how it works
[01:41] <iain> so what am I actually supposed to do then?
[01:41] <ds-work> use a fixate signal callback to choose the format you want
[01:41] <iain> fixate signal callback on what
[01:41] <iain> when
[01:41] <iain> how?
[01:41] <ds-work> each pad
[01:41] <iain> how do I know the format I want, its locked tightly inside spider
[01:42] <ds-work> you'll get a signal with an unfixed caps, and you return the caps you want
[01:42] <iain> but I don't know the caps I want
[01:43] <iain> sorry to be a dick about this, but it seems like something quite fundamentally obvious and that I have to jump though a lot of hoops to get it to work
[01:44] <ds-work> ok
[01:44] <ds-work> you have an uncontrained pipeline
[01:44] <ds-work> it is _your design_ that you have an unconstrained pipeline
[01:44] apoc_ ([email protected]) joined #gstreamer.
[01:45] <iain> I think its a design forced on me...
[01:45] <ds-work> if you don't decide, in some way, to constrain it, the core will do it for you
[01:45] <iain> I'm asking HOW to do this
[01:45] <iain> I don't know how many channels I have
[01:45] <iain> I might have 1, I might have 2, I might have 46
[01:45] <ds-work> why don't you know?
[01:46] <iain> because te audio could be coming from anything
[01:46] <thaytoo> iain:  you will know in the fixate function
[01:46] <thaytoo> it will give caps that have 'channels' min:1, max:2
[01:46] <thaytoo> or somesuch
[01:46] <iain> that doesn't tell me how many I have
[01:46] <iain> that tells me I have either 1 or 2
[01:46] <ds-work> iain: spider doesn't give you that information, if you're using spider
[01:46] <thaytoo> it tells you that it can convert it to 1 or 2
[01:47] <iain> I know
[01:47] <iain> thats what I'm trying to say
[01:47] <iain> so I can't know how many channels I have, then I can't constrain it
[01:49] <iain> fuck it, I'll just include int2float and float2int in my code, they worked
[01:49] <iain> sort of
[01:50] <ds-work> what is your pipiline?
[01:50] <iain> filesrc ! spider ! audioconvert ! audio/x-float-raw ! deinterleave ! sink
[01:51] <ds-work> you know that spider adds an audioconvert automatically, right?
[01:51] <iain> are you sure?
[01:51] <iain> because company said it didn't, and my poking around with it hasn't had one added
[01:52] <ds-work> ok, it did for a while
[01:52] <iain> all I need to know, is how many channels are going into the audioconvert, then I can add channels=x to the filtercaps, and we'll be happy
[01:54] <ds-work> (/me notes that this is one of the reasons I wanted audioconvert to be simpler)
[01:55] <aeyakovenko2> i cant seem to get my pipeline to work, (v4lsrc -> tee (teepad1 -> ffenc_h263p -> fdsink) (teepad2 -> xvimagesink))
[01:55] <aeyakovenko2> without the tees all those things work on their own
[02:00] <iain> so is there anything I can do?
[02:01] apoc ([email protected]) left irc: Read error: 110 (Connection timed out)
[02:02] <ds-work> iain: you could complain about audioconvert doing too much, I suppose
[02:03] <iain> is that likely to do anything other than make me annoyed?
[02:05] <ds-work> probably
[02:05] <ds-work> as I warned about this problem as people were adding features to audioconvert
[02:06] <iain> so what features should be in audioconvert and what ones shouldn't?
[02:06] <iain> I'm using it as the replacement for int2float
[02:06] <iain> and float2int
[02:07] <ds-work> well, the obvious thing is that number of channels is a fundamentally different type of stream property than width, depth, or float/int
[02:09] <iain> so, keep width/depth/type and lose channel changing
[02:09] <iain> okay, I'll write a mail tomorrow complaining about it, and maybe even fix it
[02:09] smoke ([email protected]) got netsplit.
[02:09] <ds-work> iain: arguably, it should be split into several elements, with "audioconvert" being the combination of them
[02:10] <ds-work> iain: it could probably be done with one source file
[02:13] smoke ([email protected]) joined #gstreamer.
[02:13] <ds-work> or even with properties in audioconvert
[02:16] smoke ([email protected]) left irc: Client Quit
[02:19] <iain> oh well, I'll go and sleep on it
[02:19] <iain> sorry for being a dick
[02:19] iain ([email protected]) left irc: "goodnight"
[02:21] <_axel> what kind of foo do i need to add to filesrc bla ! typefind ! spider ! ffmpegcolorspace ! divxenc ! filesink bla so it doesnt segfault :/ 
[02:21] <_axel> or any other encoder i have available for that matter
[02:25] <ds-work> _axel: you're probably looking for a non-existant bug fix
[02:25] <ds-work> _axel: filing a bug report would help
[02:35] <_axel> but im really clueless about what happens
[02:35] <_axel> i just run the pipe and boom
[02:35] <_axel> i think that'd be a crap bug report
[02:35] <_axel> and i cant attach the file cause its 400mb ;)
[02:41] <ds-work> does it fail in gst-player?
[02:42] <ds-work> if it fails right at the beginning, you can probably just attach the first N MB or so
[02:43] <_axel> i just backtraced, it crashes in av_free ()
[02:43] <_axel> av_freep
[02:43] <ds-work> that's not useful
[02:43] <ds-work> need entire backtrace, plus the media file
[02:44] <_axel> i suspect my pipeline is not correct anyway


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click