IRC Logs

*******************************************************************
[03:00] <Company> do an if (delay < 0) delay = 0;
[03:00] <Company> that if ((guint64) delay > osssink->handled) part sounds seriously wrong, too
[03:01] <Company> though that should be ok if you check for < 0 before
[03:01] <BBB> mk
[03:01] Action: BBB tries
[03:02] <BBB> ah
[03:02] <BBB> yes
[03:02] <BBB> that part is screwing up
[03:02] Action: BBB tries again
[03:02] <BBB> the cast is screwed
[03:02] Action: BBB hugs company
[03:02] <BBB> finally
[03:03] <BBB> and whaddayathink?
[03:03] <BBB> it works! \o/
[03:03] <BBB> I don't actually need to check for < 0, I think... well, I can add it, but it's not needed to make it work
[03:03] <BBB> (here)
[03:05] <Company> it should check
[03:06] <BBB> ok
[03:06] <BBB> this is cute :)
[03:06] Action: BBB moves on to fixing MPEG
[03:06] smoke ([email protected]) got netsplit.
[03:09] <Company> hm, nmm is very similar to gstreamer
[03:09] <Company> 0.3
[03:09] <BBB> it is... however, they're trying to be network transparent
[03:09] <BBB> not sure what that means for a media framework
[03:10] <BBB> so, next stop = MPEG
[03:12] smoke ([email protected]) got lost in the net-split.
[03:13] Nick change: kh_zZz -> khalek
[03:13] <Company> http://graphics.cs.uni-sb.de/NMM/dist-0.4.1/Docs/Doxygen/html/classNMM_1_1PlaybackNode.html#b18
[03:13] <Company> ;)
[03:13] <BBB> hm
[03:13] <BBB> delay starts returning 0
[03:14] <BBB> is that what you meant with your 'deadlock' in the clocking?
[03:14] <Company> "delay starts returning 0" ?
[03:14] <BBB> get_delay() returns 0 (GET_ODELAY)
[03:14] <BBB> the ODELAY ioctl thing
[03:15] <BBB> I suppose that means there's no audio queued
[03:15] <Company> it's not very bad if it returns 0
[03:15] <Company> but yeah, it means there's no audio queued
[03:15] <Company> GST_ODELAY returns the number of bytes still in the soundcard buffer
[03:17] <BBB> and if ODELAY returns negative numbers?
[03:17] <Company> ok, so apart from the alsalib assertion all blocker like bugs are fixed for 0.8 it seems
[03:17] <BBB> not yet...
[03:17] <Company> that probably menas you have an underrun
[03:17] <BBB> MPEG doesn't work yet
[03:17] <Company> all the bugs people blame me for ;)
[03:18] <BBB> *g*
[03:18] <BBB> go fix alsa in gnome-volume-control :p
[03:18] jcsston` ([email protected]) joined #gstreamer.
[03:19] <BBB> hm
[03:19] <BBB> it still slows down after 1:30
[03:20] <BBB> so it's partially, but not fully fixed
[03:22] <BBB> basically it still dies mostly after ODELAY returns <0
[03:22] <BBB> there must be something like having to reset the device or so
[03:22] Action: BBB wonders
[03:25] Action: BBB continues debugging
[03:25] <Company> a/v out of sync and one queue full?
[03:26] <BBB> I suspect something else, actually
[03:26] <Company> lol, nmm and gstreamer are so similar in design, it's boring to even compare them ;)
[03:26] <BBB> heh, I see what's wrong, I think
[03:26] <BBB> if you drop a frame, you don't actually update osssink->handled
[03:26] <BBB> so the clock gets stuck
[03:28] <BBB> hm, if I do that, sounds goes complete away
[03:28] <BBB> but video keeps playing now
[03:28] <BBB> so I'm making progress
[03:29] <Company> if i drop sound i drop it because it's late
[03:29] <Company> s/sound/frames/
[03:30] <BBB> yes
[03:30] <Company> ->handled should only be updated if you send data to the card
[03:30] <BBB> ->handled is used by _get_time()
[03:30] <BBB> so I think it needs to know if I drop samples
[03:30] <BBB> else, the time will be invalid
[03:30] <BBB> (maybe -> handled needs to be renamed to something else?)
[03:32] <Company> yeah, handled is kinda wrong as name for that variable
[03:32] Action: BBB sort of agrees
[03:32] <BBB> yay
[03:32] <BBB> another unsigned variable overflow
[03:32] <BBB> Drop 4256470528 bytes
[03:32] <BBB> :p
[03:32] <BBB> no wonder sound stops
[03:38] jcsston ([email protected]) left irc: Read error: 110 (Connection timed out)
[03:42] NFusi0n ([email protected]) left irc: Read error: 110 (Connection timed out)
[03:50] Action: BBB has video continuously working now
[03:50] <BBB> now to find why audio 'stops'
[03:50] <BBB> probably still an underrun
[03:50] <BBB> but ohwell
[04:00] Nick change: jcsston` -> jcsston
[04:33] sub_pop ([email protected]) joined #gstreamer.
[05:05] <Misirlou> Company: ping
[05:09] <Company> Misirlou: pong
[05:09] <Misirlou> k
[05:21] <BBB> do you guys think it's a "good idea" if I work on further ffmpeg integration?
[05:21] <BBB> or are there more important blocker issues?
[05:22] <Company> i'm out of the loop
[05:22] <Company> so i have no idea
[05:24] jcsston ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[05:25] <BBB> why are you out of the loop?
[05:25] <BBB> I really don't get you sometimes
[05:34] jcsston ([email protected]) joined #gstreamer.
[05:39] <Company> because i only read the lists and don't hang out on irc because of work
[05:39] <Company> so i miss all the discussions about current issues
[05:40] <BBB> :(
[05:40] <BBB> same for me, actually
[05:40] <BBB> which is why I'm trying to use the list for "important" issues nowadays
[05:41] <Company> gnight
[05:41] Company ([email protected]) left irc: Remote closed the connection
[05:48] bitshifter ([email protected]) joined #gstreamer.
[06:30] joeforker (dholth@3ffe:bc0:8000:0:0:0:0:262b) joined #gstreamer.
[06:30] <joeforker> hi
[06:30] <BBB> hi
[06:31] <joeforker> I'm looking for a gstreamer ogg muxer.
[06:31] <BBB> we don't have one
[06:31] <BBB> what do you want to do?
[06:31] <joeforker> and I'm not scared of figuring out how to write one.
[06:31] <BBB> ah, that's good :)
[06:31] bforbes ([email protected]) joined #gstreamer.
[06:31] <BBB> avimux or matroskamux are good muxer examples
[06:32] <joeforker> I want to encode a live theora stream from an interlaced v4l source.
[06:32] <BBB> (if you need examples)
[06:32] <joeforker> gstreamer is all about that, right?  hook up a v4l to a deinterlacer to a theora encoder (whoops, more coding), to an ogg muxer, and put some vorbis in there while you're at it.
[06:32] <joeforker> how does gstreamer handle chained ogg streams?
[06:32] <BBB> what do you mean with 'chained'?
[06:33] <joeforker> I mean cat theora1.ogg theora2.ogg > chained.ogg
[06:33] <BBB> oh... don't know
[06:33] <joeforker> ok
[06:33] <BBB> is that part of the ogg specs?
[06:33] <BBB> then we handle it
[06:33] <BBB> else we don't
[06:33] <joeforker> yes it is part of the spec.
[06:33] <joeforker> it's how radio works.
[06:33] <BBB> we simply use libogg for demuxing ogg streams
[06:33] <joeforker> okay.
[06:34] <joeforker> all it means is that the media player has to be able to change sample rate, video dimensions, and so on without actually opening a new file.
[06:34] <BBB> we don't have a theora encoder or a ogg muxer
[06:34] <BBB> we have all other elements
[06:34] <joeforker> yep. I noticed ;-)
[06:34] <BBB> theora encoder should be easy
[06:34] <BBB> the ogg muxer is a bit more tricky
[06:34] <BBB> but if you take one of the examples, it should be straightforward :)
[06:34] <joeforker> super.
[06:34] <BBB> (avimux or matroskamux being examples)
[06:35] <joeforker> how about the python bindings?  are they healthy and awe-inspiringly useful?
[06:35] <BBB> they are very healthy and actively maintained
[06:35] <joeforker> neat.
[06:35] <BBB> though I don't use them myself
[06:35] <BBB> ;)
[06:35] <BBB> but I regularly see commits
[06:35] <joeforker> thanks.
[06:36] <BBB> np :)
[06:36] <joeforker> why does the irc subject say 0.7.5 when the web page says 0.6.5?
[06:38] <BBB> 0.6.5 is a maintainance of stable
[06:38] <BBB> 0.7.5 is unstable
[06:38] <BBB> :)
[06:39] <joeforker> ok
[06:45] mxpxpod ([email protected]) left irc: Read error: 113 (No route to host)
[06:54] <joeforker> does gstreamer still have a wiki?
[06:55] <BBB> yes, but unmaintained
[06:56] <joeforker> ok
[07:00] joeforker (dholth@3ffe:bc0:8000:0:0:0:0:262b) left irc: "Client exiting"
[07:04] camh ([email protected]) left irc: "Client exiting"
[07:05] pr1ze ([email protected]) left irc: Remote closed the connection
[07:21] camh ([email protected]) joined #gstreamer.
[07:28] BBB ([email protected]) left irc: "Client exiting"
[07:28] jcsston` ([email protected]) joined #gstreamer.
[07:32] jcsston ([email protected]) left irc: Read error: 60 (Operation timed out)
[07:37] ChrisHJW ([email protected]) joined #gstreamer.
[07:53] Nick change: jcsston` -> jcsston
[07:53] jcsston ([email protected]) left irc: Nick collision from services.
[07:54] jcsston ([email protected]) joined #gstreamer.
[07:57] escii ([email protected]) joined #gstreamer.
[07:57] <escii> hullo
[07:58] <escii> is pngenc used by nautalis a part of gstreamer ?
[08:04] ChriHJW_log ([email protected]) left irc: Read error: 110 (Connection timed out)
[08:04] ChrisHJW ([email protected]) left irc: Read error: 110 (Connection timed out)
[08:35] <escii> HARDASS
[08:58] markey ([email protected]) joined #gstreamer.
[09:03] walters ([email protected]) left irc: Remote closed the connection
[09:03] Nick change: mathrick|sleep -> mathrick
[09:03] <mathrick> g'morning
[09:11] <markey> moin
[09:30] mathrick ([email protected]) left irc: Read error: 110 (Connection timed out)
[09:43] ChrisHJW ([email protected]) joined #gstreamer.
[09:49] mathrick ([email protected]) joined #gstreamer.
[09:53] aeyakovenko ([email protected]) joined #gstreamer.
[09:53] <aeyakovenko> hello
[09:56] aeyakovenko ([email protected]) left #gstreamer.
[09:56] somex1 ([email protected]) left #gstreamer.
[09:58] Nick change: wingo_ -> wingo
[09:58] <wingo> mornin, folks
[09:59] Action: wingo lost his shoes last night and is irritated with himself
[10:01] <mathrick> yo wingo
[10:03] <wingo> howdy mathrick
[10:03] <mathrick> wingo: /me has evil idea, to make elisp bindings, but that's distant future yet :)
[10:03] khalek ([email protected]) left irc: Nick collision from services.
[10:04] <wingo> ak! it would be better to work on converting emacs to use guile ;)
[10:04] <mathrick> wingo: any guesstimates if existance of guile bindings will help?
[10:05] <wingo> dunno, they're highly based on GOOPS, an object system
[10:05] <mathrick> wingo: heh, you really wanna mess with RMS, don't ya ;P
[10:05] <wingo> although the g-wrap layer could help on a low level, for a less OO binding set
[10:05] <wingo> i don't think he's categorically opposed -- it would take a lot of work though
[10:06] <wingo> s/would/will/ ;)
[10:06] <mathrick> wingo: ever read about why GNU Emacs / XEmacs schism occured?
[10:07] <mathrick> wingo: that's what ones involved say - RMS will gladly accept your vision, provided it matches exactly his own
[10:07] <wingo> no, although i've been meaning to google about that for a while
[10:08] <mathrick> wingo: lemme help you :)
[10:09] <mathrick> http://www.emacswiki.org/cgi-bin/emacs.pl/EmacsSchism
[10:12] <wingo> thanks ;)
[10:13] <mathrick> wingo: speaking of which - do you think guile is better than elisp?
[10:13] <mathrick> wingo: np :)
[10:13] <mathrick> wingo: i've been reading that recently myself :)
[10:13] Action: mathrick wonders if he already pointed wingo at semantic.el and ECB
[10:14] <wingo> i'm ignorant, to be honest. scheme is simpler and guile is multi-purpose. that's the real difference, as it seems to me. bust i only started to hack in scheme last year.
[10:15] <wingo> *but
[10:15] sub_pop ([email protected]) left irc: "Client exiting"
[10:15] <mathrick> wingo: ah, may I ask how is scheme simpler? after reading bit of Succesful Lisp, I'm mostly convinced language cannot get any simpler than lisp :)
[10:16] <mathrick> wingo: at least wrt syntax
[10:17] <wingo> naw, here i will defer to my ignorance ;) i haven't really hacked in a real lisp -- scheme only.
[10:17] Action: mathrick remembers something about lexical vs dynamic scoping
[10:17] <mathrick> wingo: I think that's the main difference, not that I know what it actually means ;)
[10:20] <wingo> see http://mitpress.mit.edu/sicp/full-text/book/book-Z-H-35.html#%_sec_5.5.6
[10:20] kwm ([email protected]) joined #gstreamer.
[10:21] <mathrick> wingo: k
[10:21] <mathrick> hmm, I sense some food being ready, bbl :)
[10:21] Nick change: mathrick -> mathrick|food
[10:27] Uraeus ([email protected]) joined #gstreamer.
[10:27] <Uraeus> morning
[10:28] <wingo> morning, U
[10:29] Nick change: markey -> afk
[10:35] <Uraeus> lots of nice bugfixin this weekend :)
[10:47] <Uraeus> wingo: I am a bit worried about the viability of coming to visit you, turns out that the stop in Africa will cost just as much as my three other stops together :) but I will ask a real travel agent to try and make me an offer :)
[10:50] bitshifter ([email protected]) left irc: "bye"
[10:57] Nick change: mathrick|food -> mathrick
[10:58] <mathrick> wingo: humm, interesting. I didn't know that book was available online, thanks for enlightening me :)
[11:12] <wingo> Uraeus: yeah, that's what most people find ;) it's a nice idea until you look at plane ticket prices
[11:13] <mathrick> wingo: where do you live?
[11:13] <thomasvs-out> I'm getting really confused by the discussion we're having about our designs
[11:14] <thomasvs-out> Uraeus: why are you so silent about these issues anyway ? this isn't like you at all.
[11:14] Nick change: thomasvs-out -> thomasvs
[11:15] <wingo> mathrick: namibia
[11:15] <wingo> hi thomasvs
[11:17] <mathrick> yo thomasvs
[11:17] <mathrick> wingo: wow, that's probably hot place?
[11:19] <Uraeus> thomasvs: sorry, busy fixing up my appartment for the sale, so I have mostly just been compiling and testing stuff and filing bug reports during my 'breaks'
[11:20] <Uraeus> thomasvs: but basically I agree with you, a regression is a bug in the eye of the application developers
[11:21] <thomasvs> Uraeus: well, I get the feeling I'm pretty much alone with my POV, and people here think it's not that important to be a part of GNOME at all, so I think I'll start shutting up from now on.
[11:22] <thomasvs> it gets really old really quick when you seem to be always complaining about stuff and nobody agrees at all, so I guess I'll just go with the flow.
[11:24] bitshifter ([email protected]) joined #gstreamer.
[11:25] <Uraeus> thomasvs: I will join the fray today :)
[11:26] <thomasvs> I will try to fix some more issues, not really sure why though in the weekend :/
[11:29] Nick change: afk -> markey
[11:33] sublett ([email protected]) joined #gstreamer.
[11:35] Action: wingo works on getting his sched fixes to not crash rb
[11:49] alley_cat ([email protected]) joined #gstreamer.
[11:52] <Uraeus> hi alley_cat
[11:52] <Uraeus> alley_cat: do you have a updated NAS patch? I didn't apply the last one I got from you since you said it was just a intermediate thing, but if it is the latest then I apply that one
[11:53] jcsston ([email protected]) left irc: Read error: 110 (Connection timed out)
[11:53] <alley_cat> let me check what i have right now, haven't had time to work on it recently
[11:58] <alley_cat> got some fixes and the mixer interface, but couldn't test the mixer interface yet, have to add property probe first
[11:59] <alley_cat> or is there something besides the gst-volume thing that uses the mixer interface now?
[11:59] <Uraeus> alley_cat: acme and the gnome-panel volume applet uses it
[12:00] <alley_cat> all with property probe?
[12:01] <Uraeus> alley_cat: don't ask me how they do it :)
[12:03] <alley_cat> what does the latest patch i sent you contain? if that's all the fixes i have here then it can be applied, will see if i find some time to test the mixer interface
[12:04] <Uraeus> alley_cat: hmm, I apply that patch then and tell you when I have done so, then you can do a diff to see what might/might not be missing, ok?
[12:05] <alley_cat> ok
[12:14] <alley_cat> i still wonder why i have all sorts of problems with current gstreamer that no one else seems to have ...
[12:16] <Uraeus> alley_cat: like what?
[12:16] <alley_cat> i can't play any avi files
[12:16] <alley_cat> either segfaults (always in different places) or hangs with 100% cpu usage
[12:17] <alley_cat> i can play mpeg, and most matroska files
[12:17] <alley_cat> seeking doesn't work anywhere
[12:18] <Uraeus> alley_cat: this is with both gst-launch and with gst-player/totem?
[12:19] <alley_cat> yes
[12:19] <alley_cat> i'm updating and recompiling everything right now
[12:20] <alley_cat> funny thing with avis is that it works if the file doesn't have an index
[12:20] teuf ([email protected]) joined #gstreamer.
[12:20] <alley_cat> the function where it hangs is a g_malloc, which is really odd
[12:20] <thomasvs> alley_cat: valgrind it ?
[12:21] <teuf> anyone know what BBB meant with "alsa buggering with gnome-volume-control" ?
[12:21] <alley_cat> never tried valgrind, is it easy to use?
[12:22] <Uraeus> teuf: not sure, maybe he was refering to the issue of the volume control not setting volume on both master and PCM which cause no sound even if volume is high in some cases?
[12:23] <teuf> oh yeah, I'm getting this pcm issue at work
[12:24] <teuf> apart from that, alsa+gnome-volume-control isn't buggy, it's only that some stuff needs to be implemented in alsasink
[12:25] <Uraeus> teuf: agreed, I use ALSA now and it works like a charm, except the irritating 0 sound until I manually increase PCM each reboot
[12:25] <Uraeus> alley_cat: ok, your NAS patch is commited to CVS
[12:27] <thomasvs> alley_cat: yeah, it is
[12:30] pb_ (~pb@2002:5160:45ef:0:240:95ff:fe30:aa84) joined #gstreamer.
[12:30] <alley_cat> Uraeus: ok, that's still missing some stuff, i'll prepare a new patch
[12:41] <alley_cat> hmm, seems like valgrind doesn't like nptl glibc
[12:47] <thomasvs> alley_cat: you sure ? works fine here
[12:48] <alley_cat> doesn't compile here
[12:50] LeRoutier ([email protected]) joined #gstreamer.
[12:51] <LeRoutier> hello
[12:51] <LeRoutier> who did the media-info-priv changes lately ?
[12:52] <thaytan> thomas
[12:52] <LeRoutier> that's thomas according to the changelog
[12:52] <LeRoutier> media-
[12:52] <thaytan> evening, all
[12:52] <LeRoutier> hi thaytan
[12:52] <LeRoutier> media-info-priv.c: In function `tag_flag_score':
[12:52] <LeRoutier> media-info-priv.c:162: warning: implicit declaration of function `gst_tag_get_flag'
[12:54] <thaytan> it needs to include <gst/gsttag.h>
[12:55] <thomasvs> no, you need to update core
[12:56] <thaytan> oh, right :)
[12:56] <thomasvs> hm
[12:56] Action: thomasvs wonders about having all those media info function take a GError ** as last argument
[12:57] <wingo> can someone do me a favor? i'm trying to debug the opt scheduler
[12:57] <LeRoutier> thomasvs, you mean my gstreamer core is out-dated ?
[12:58] <teuf> wingo: what do you need ? someone to bring you beers to make you feel better while doing that ? ;)
[12:58] <wingo> i need someone to run rhythmbox, but with the opt scheduler from a few days ago (at http://freedesktop.org/cgi-bin/viewcvs.cgi/*checkout*/gstreamer/gstreamer/gst/schedulers/gstoptimalscheduler.c?rev=1.61&content-type=text/plain)
[12:59] <wingo> teuf: mm, beer -- I might just go to the store now ;)
[12:59] <alley_cat> bah, libtool borks, tries to use the wrong glib :/
[12:59] <LeRoutier> thomasvs, nevermind, you're right, i missed a checkout on gstreamer core, including one with gsttag.h
[12:59] <teuf> wingo: how do I run rb with an older opt ?
[12:59] <teuf> cvs up -r x.xx on the appropraite file ?
[12:59] jcs ([email protected]) joined #gstreamer.
[12:59] <wingo> yeah
[12:59] <wingo> it's 1.61 i think
[13:00] <wingo> i need a --gst-debug-level=5 output (bzipped)
[13:00] <teuf> ok, I grabbed 1.61
[13:00] <wingo> they say you can get the problem by playing a song, then pressing next
[13:01] <wingo> so you don't have to play for long ;)
[13:01] <teuf> yep; it used to crash on that here
[13:01] <Uraeus> actually you don't even have to do that, just remove your .rhythmbox directory and try importing a folder
[13:01] <Uraeus> it will crash rb with that version of opt
[13:03] <teuf> wingo: the trace is full of ESC[ is that ok ?
[13:04] <wingo> that's fine, you can read it with less -R
[13:04] <thaytan> you can do --gst-debug-no-color to avoid that too
[13:04] <teuf> http://cfergeau.free.fr/opt.log.bz2
[13:04] <teuf> I ran rhythmbox, started to play a song, and double clicked on a different one
[13:05] <wingo> thanks, yo.
[13:06] Nick change: markey -> markey|afk
[13:06] <teuf> wingo: if you need backtraces, or if you need to check some stuff in gdb while running rb, just ask
[13:07] <wingo> thanks
[13:15] <bitshifter> teuf: btw, do the latest fixes from Company and yourself fix all your memleak problems when reading id3v2 metadata?
[13:15] <alley_cat> ok, had to get cvs valgrind
[13:15] <teuf> bitshifter: nope, not all, but it leaks less
[13:16] <teuf> bitshifter: importing my library eats 150MB, as opposed to the 600MB before company's fix ;)
[13:16] ChrisHJW ([email protected]) left irc: 
[13:16] <bitshifter> teuf: yeah, here too; was just wondering :)
[13:19] <alley_cat> great, what the hell is wrong here? gstreamer always links against the wrong glib
[13:19] <Uraeus> alley_cat: glib1 ?
[13:19] <thaytan> alley_cat: check the .la files 
[13:19] <alley_cat> no, the glib2 that's installed in /usr, though it should use the new one in $HOME/gst/lib
[13:20] <thaytan> libtool likes to play silly buggers with those
[13:20] <alley_cat> PKG_CONFIG_PATH is set correctly, and the -L/home/arwed/gst/lib gets added, but libtool still uses the stuff in /usr
[13:20] <thaytan> it'll do that if it
[13:20] <thaytan> is using a .la from /usr/lib that specifies the /usr/lib/libglib first
[13:21] <alley_cat> it shouldn't use any .la file from there
[13:21] <thaytan> probably not, but that's libtool for you
[13:22] <alley_cat> interesting is that it worked a week ago, and i didn't change anything
[13:23] Action: alley_cat does a make distclean and tries again
[13:23] Action: mathrick believes there are only two kinds of issues - unsolvable and self-solvable ;)
[13:23] <mathrick> alley_cat: wait another week, it will go away probably :)
[13:39] foser ([email protected]) joined #gstreamer.
[13:46] <alley_cat> hmm, the libgstreamer.la gets created with wrong info
[13:50] <alley_cat> *sigh* guess i have to remove the normal glib2 while building gstreamer
[13:50] <teuf> alley_cat: removing the development headers should be enough
[13:51] <alley_cat> it isn't
[13:51] <alley_cat> it uses the headers of the glib installed in my home, but links to the ones in /usr/lib
[13:51] <thaytan> you actually need to remove whichever dev package is being used to collect /usr/lib/libglib 
[13:52] <thaytan> ie, one of the ones referencing it in their .la files
[13:52] <thaytan> and make sure that the gstreamer .la files get rebuilt finding the correct libglib
[13:52] <alley_cat> the only .la file that gets used that contains them is the gstreamer la file
[13:59] <LeRoutier> ++
[13:59] LeRoutier ([email protected]) left irc: "Leaving"
[14:00] <thaytan> alley_cat: hard to diagnose from here, but libtool definitely should not pick up /usr/lib/libglib if there's an earlier one in the LD_LIBRARY_PATH 
[14:01] <alley_cat> i know it shouldn't, but for some reason it does
[14:01] <alley_cat> removing the system glib for the build helped
[14:02] <thaytan> I know that I'm buiding it fine with a system glib, but no -dev packages 
[14:02] <thaytan> I avoid having ANY .la file in /usr/lib/ that references libglib and friends tho
[14:15] Nick change: markey|afk -> markey
[14:36] benno ([email protected]) joined #gstreamer.
[14:41] camh ([email protected]) left irc: Read error: 110 (Connection timed out)
[14:43] camh ([email protected]) joined #gstreamer.
[14:49] <alley_cat> i think i give up on building gstreamer for now
[14:49] <alley_cat> too much breaks with the new glib
[14:57] aviateur ([email protected]) joined #gstreamer.
[14:58] <alley_cat> can i put a new glib in system location without breaking everything else?
[14:59] apoc_ ([email protected]) joined #gstreamer.
[14:59] <aviateur> hey everybody, i'm having a rather irritating problem with gstreamer 0.6.4 on an intel i8x0 soundcard.  anytime i try gst-launch sinesrc ! osssink i get nothing but static
[15:00] <aviateur> in fact, anytime i launch anything with sound, i get static, skipped sound horribleness
[15:00] <aviateur> what's funny is that it used to work on slackware 9.0, but slack 9.1 seems to have done it
[15:03] <wingo> did the 9.1 change the version of GStreamer?
[15:03] <aviateur> no
[15:03] <rvalles> aviateur: does the sound work for non-gstreamer apps?
[15:04] <aviateur> rvalles, yeah, it works fine.
[15:04] <rvalles> great sinewave :)
[15:04] camh ([email protected]) left irc: Read error: 110 (Connection timed out)
[15:04] <aviateur> i'm certain its a gstreamer only symptom, another weird thing, if i launch from a cdrom src, i get an error about 'clock is waaayy behind'
[15:05] <aviateur> so i thought maybe it was the clock speed of my soundcard, and i modified that with alsa, but it didn't change anything
[15:05] <rvalles> "Let's aim for 0.8.0 on monday"
[15:05] <rvalles> that fast?!?!?
[15:06] <rvalles> is anything interactive working?
[15:06] <rvalles> (dvd menus? flash?)
[15:08] <thaytan> yeah, they work
[15:09] <rvalles> thaytan: "hey"?
[15:09] <rvalles> thaytan: "they"?
[15:09] <rvalles> thaytan: both?
[15:10] <thaytan> well, DVD menus work
[15:10] <thaytan> I haven't tried flash, but I believe ds was working on it
[15:10] <rvalles> and, about decss?
[15:11] <thaytan> DVD menus need the overlays/highlights working, but I'll get back to that 
[15:11] <thaytan> they don't require holding 0.8 off
[15:11] <thaytan> decss works, because libdvdnav takes care of it 
[15:11] <aviateur> do soundcards use dma?  do you think it might be a dma issue?
[15:13] <rvalles> nice
[15:15] apoc ([email protected]) left irc: Read error: 110 (Connection timed out)
[15:30] <aviateur> alright, so i did some looking around and installed some slack packages for 0.7.4, and now it doesn't seem to find the plugin registry upon launch and says "Spinning.  Please run 'gdb gst-launch 18762' to continue debugging, Ctrl-C to quit, or Ctrl-\ to dump core.
[15:30] <aviateur> Caught interrupt."
[15:30] <thaytan> have you run gst-register?
[15:31] <aviateur> yes.  but it seems now 0.6.4 can't find the db either :)
[15:32] <aviateur> thaytan, where should that normally be, anyways?
[15:32] <thaytan>  /var/lib/gstreamer
[15:33] <aviateur> yeah, i have nothing there.
[15:33] <aviateur> is it shared across versions?
[15:33] <thaytan> it puts a 0.6 and 0.7 dir in there 
[15:34] <aviateur> ah ah, found them
[15:34] <aviateur> they look alright... 
[15:35] <thaytan> I don't know why apps won't find the registry then
[15:35] <aviateur> on launch it says : Caught SIGSEGV accessing address (nil)
[15:35] <thaytan> what are you launching?
[15:35] <aviateur> where normally i would get:  info (322233)  or something like that while it loads the xml
[15:36] <thomasvs> no, /var/cache
[15:36] <aviateur> gst-launch-0.7 sinesrc ! osssink
[15:36] <aviateur> thomasvs, huh?
[15:37] <aviateur> thomasvs, mine are in /var/lib/cache
[15:37] <thomasvs> aviateur: the registry lives in /var/cache/gstreamer-$majorminor
[15:37] <thomasvs> aviateur: hm, that's wrong then.
[15:37] Action: thomasvs goes to check
[15:38] <thaytan> and mine is in /var/lib/gstreamer/0.6 from the debian package
[15:38] <thaytan> and $prefix/var/cache/gstreamer-0.7 from CVS
[15:38] <thomasvs> why did taaz put it there ?
[15:38] <thomasvs> that's weird
[15:39] <thomasvs> anyway, it should be localstatedir/cache
[15:39] <wingo> catch ya'll later.
[15:39] <Uraeus> see ya wingo
[15:39] wingo ([email protected]) left irc: "I like core dumps"
[15:39] <aviateur> thomasvs, but i don't think that's necessarily the problem  here.  i mean, here i'll rephrase
[15:40] <thomasvs> thaytan: uh, gstreamer-properties doesn't use our gconf lib ?
[15:40] <aviateur> at first, with just gstreamer0.6.4, the registry worked fine, but the sound output was crrrap and crrrapy only for gst apps.  then i installed 0.7.4 and now neither the registry for 0.64 or 0.7 is working.
[15:41] <aviateur> so as far as the registry goes, its probably just a shitty slackware package
[15:41] <aviateur> but the real thing that bugs me, is that i know gst was working fine on this machine until a few months ago when i upgraded to slack 9.1 or maybe it was just a dropline upgrade, i can't really remember anymore.
[15:42] <thomasvs> hm
[15:42] <thomasvs> well, sounds hard to debug without knowing how slackware packages work or which one fucked it up
[15:42] <aviateur> thomasvs, yeah.  crap.
[15:43] <aviateur> anybody ever heard of any output problems with i8x0 soundcards?
[15:43] <aviateur> or specific output quality problems?
[15:43] <aviateur> google isn't really cooperating with me.
[15:44] <thaytan> thomasvs: no, but I don't remember why
[15:44] <thomasvs> thaytan: care to fix it ? :)
[15:44] <Uraeus> thomasvs: any idea on that missing -lgstinterfaces-0.7 I am suffering? I got it with my own RPMS and now with yours, and it hinders me from building most things
[15:44] <thomasvs> Uraeus: check where the lib is
[15:44] <Uraeus> thomasvs: whats the filename I should be looking for?
[15:45] <thomasvs> libgstinterfaces-0.7.so*
[15:45] <thomasvs> hm
[15:45] <Uraeus> ok, I investigate
[15:45] <thomasvs> I don't have the .so file
[15:45] <thomasvs> yep, it's missing from the -devel package for some reason
[15:45] <thomasvs> that's weird.
[15:46] <thaytan> thomasvs: I think it was because I wanted to be able to monitor the key for changes to update the UI
[15:46] <thomasvs> thaytan: hm- makes sense.  well, then this should be added in the gconf lib
[15:46] <thaytan> I should add an interface for that the gstgconf
[15:46] <thomasvs> thaytan: the gconf lib contains the right path, and it gets changed sometimes
[15:46] <thomasvs> thaytan: yeah, exactly
[15:46] <thomasvs> thaytan: well, there's still time, but it needs to be done today :)
[15:48] <thomasvs> Uraeus: rebuilding packages atm
[15:48] Action: thomasvs feels it is nice when you can use one of your own tools to make your life easier
[15:48] <Uraeus> thomasvs: I have the file actually
[15:48] <Uraeus> thomasvs: but I it seems it is called libgstinterface, while our apps look for gstinterface. Could the missing 'lib' be the problem?
[15:51] Action: aviateur 's gonna go smoke
[15:56] <thomasvs> nope, that's not it
[15:56] <thomasvs> Uraeus: which file do you have ?
[15:56] <thaytan> thomasvs: actually, it doesn't look like it monitors the key at all
[15:56] <Uraeus> thomasvs: libgstinterfaces-0.7.so.2 and libgstinterfaces-0.7.so.2.0.0
[15:56] <thaytan> so I don't know why I didn't use gst_gconf, especially since the pipeline tests _do_ use it
[15:57] <Uraeus> thomasvs: wit so.2 being a symlink to so.2.0.0
[15:57] <thomasvs> Uraeus: right, that's what I'm saying - it doesn't have the .so file (without numbers)
[15:57] <thomasvs> Uraeus: which is what it needs to link
[15:57] <thomasvs> Uraeus: try manually symlinking so.2 to so
[15:57] <Uraeus> thomasvs:ok thanks
[15:57] <thomasvs> thaytan: hm - well, in that case, can you fix it ? and maybe add a way to watch to gstgconf ? :)
[15:58] <Uraeus> thomasvs: should I update the bug or are you looking into it?
[15:59] <thomasvs> Uraeus: I'm rebuilding right now to check
[15:59] aviateur ([email protected]) left irc: "Leaving"
[16:00] <Uraeus> thomasvs: ok
[16:00] <thomasvs> Uraeus: feel free to check as well in your packages
[16:01] <thaytan> thomasvs: fixed it, I'll take a look at adding monitoring
[16:01] <thaytan> what are the rules about committing to gnome-media?
[16:01] <thomasvs> thaytan: it's your properties capplet, so you can if you want to
[16:01] <thomasvs> afaik at least
[16:02] <thaytan> great, I'll use the 'thomas said I could' defense if I get in trouble ;)
[16:03] <thomasvs> sigh, ok :)
[16:05] rvalles ([email protected]) left irc: "Changing server"
[16:12] <alley_cat> anyone able to compile gst-ffmpeg?
[16:13] <alley_cat> make[8]: *** No rule to make target `8bps.c', needed by `8bps.lo'.  Stop.
[16:14] <Uraeus> alley_cat: I try to compile it now
[16:20] camh ([email protected]) joined #gstreamer.
[16:25] <Uraeus> alley_cat: seems gst-ffmpeg need some disting love so I well need a bit more time testing on it :)
[16:26] <thaytan> thomasvs: Adding key monitoring would add more static vars to gstgconf, does it need to be threadsafe, do you think?
[16:29] <thaytan> at the moment, it's as threadsafe as gconf_client is. notification means adding some signal handling
[16:31] <alley_cat> hmm, deleting the ffmpeg dir and getting a new checkout of ffmpeg seems to have fixed it
[16:34] <alley_cat> getting a different error now :P
[16:39] <Uraeus> alley_cat: I am working on the gst-ffmpeg build now, many small issues it seems :)
[16:39] <Uraeus> alley_cat: I tell you when I have commited it
[16:42] <Uraeus> thomasvs: media-info-priv.c: In function `tag_flag_score':
[16:42] <Uraeus> media-info-priv.c:162: warning: implicit declaration of function `gst_tag_get_flag'
[16:42] <Uraeus> make[4]: *** [libgstmedia_info_0.7_la-media-info-priv.lo] Error 1
[16:45] <thomasvs> Uraeus: update core
[16:45] <Uraeus> ok
[16:45] <thaytan> auto-apt is a nice way to build gst-plugins on a fresh system
[16:46] escii ([email protected]) left irc: "changing servers"
[16:47] jdahlin ([email protected]) joined #gstreamer.
[16:47] <Uraeus> hi jdahlin
[16:48] <jdahlin> hello Uraeus
[16:48] <Uraeus> jdahlin: gsr does not work for me :(
[16:51] <jdahlin> Uraeus: hmm, bad
[16:52] <jdahlin> Uraeus: what's wrong?
[16:52] <Uraeus> jdahlin: well the first time i tried to run it just crashed when I pressed 'record' (trace in bugzilla)
[16:53] <Uraeus> jdahlin: the second times it seems to record but when trying to play back what I recorded I get:
[16:53] <jdahlin> Uraeus: bug #?
[16:53] <jdahlin> I assume you're using CVS HEAD
[16:53] <Uraeus> jdahlin: http://bugzilla.gnome.org/show_bug.cgi?id=136456
[16:53] <Uraeus> jdahlin: yes
[16:54] <Uraeus> jdahlin: anyway now it seems to record, but on playback I get:
[16:54] <Uraeus> record_state_changed_cb: entered: PAUSED->READY
[16:54] <Uraeus>  
[16:54] <Uraeus> ** (gnome-sound-recorder:26189): CRITICAL **: how to remove plugins?
[16:54] <Uraeus>  
[16:54] <Uraeus> (gnome-sound-recorder:26189): GStreamer-CRITICAL **: Factory for `wavparse' has no type
[16:54] <Uraeus> Segmentation fault
[16:55] benno ([email protected]) left #gstreamer ("Client Exiting").
[17:01] <thaytan> thomasvs: gstgconf changes aren't going to happen tonight, it's already 3am and well past my bedtime
[17:01] <thaytan> I'll commit gstreamer-propeties and sleep
[17:01] <Uraeus> thaytan: sleep well Jan
[17:02] <Uraeus> hmm, think we need to get Martin CVS access :)
[17:04] wheels ([email protected]) left #gstreamer.
[17:04] wheels ([email protected]) joined #gstreamer.
[17:07] <thomasvs> Uraeus: ok, made new packages, contain the file now
[17:08] <thomasvs> Uraeus: strange that it didn't give errors somehwere though
[17:08] <thomasvs> will upload packages now
[17:08] <jdahlin> Uraeus: if possible, can you rebuild gsr with debugging symbols?
[17:10] smoke ([email protected]) joined #gstreamer.
[17:14] <Uraeus> jdahlin: how do I do that again? pass something to configure?
[17:16] <jdahlin> Uraeus: or rebuild with make CFLAGS="-g"
[17:16] <jdahlin> (make clean first)
[17:22] <Uraeus> jdahlin: btw, I can only reproduce the second error, not the one in the bug
[17:24] <Uraeus> thomasvs: let me know when I can apt-get 
[17:28] Nick change: lilo -> TomSawyer
[17:29] Nick change: TomSawyer -> lilo
[17:42] <thomasvs> Uraeus: uploading
[17:47] Nick change: Uraeus -> Ura_dinner
[18:05] jcsston ([email protected]) joined #gstreamer.
[18:37] Nick change: markey -> niels_
[18:37] Nick change: niels_ -> markey
[19:05] jcsston ([email protected]) left irc: Read error: 60 (Operation timed out)
[19:05] somex1 ([email protected]) joined #gstreamer.
[19:09] jdahlin ([email protected]) left irc: "Leaving"
[19:15] <thomasvs> Ura_dinner: can you try updating ?
[19:29] <thomasvs> anyone knows what exactly to use now to get a running time value ?
[19:55] thomasvs ([email protected]) left irc: Remote closed the connection
[20:00] thomasvs ([email protected]) joined #gstreamer.
[20:02] benow ([email protected]) joined #gstreamer.
[20:02] <benow> heyas all
[20:03] <benow> ** (process:11849): CRITICAL **: how to remove plugins?
[20:03] <benow>  
[20:03] <benow> (process:11849): GStreamer-CRITICAL **: file gstscheduler.c: line 910 (gst_scheduler_factory_create): assertion `factory->type != 0' failed
[20:03] <benow>  
[20:03] <benow> GStreamer-ERROR **: Critical error: could not get scheduler "opt"
[20:03] <benow> any ideas?
[20:04] <teuf> what are you doing with what ersion of gstreamer ?
[20:05] <teuf> thomasvs: I talked a bit with the gst-ruby maintainer today, and he said he felt like you and the other gstreamer developers didn't like ruby much ;)
[20:05] <thomasvs> teuf: huh ? why does the think that ?
[20:05] <thomasvs> benow: you're registering plugins twice, it looks like
[20:05] <thomasvs> benow: long time no see btw
[20:05] <teuf> he seems to have mailed you several times without getting answers
[20:05] <thomasvs> teuf: to the mailing list ?
[20:06] <teuf> and it seems he sent release announcements fro gst-ruby which weren't approved
[20:06] <teuf> thomasvs: private mail for you I think
[20:06] <thomasvs> what's his name ?
[20:06] <thomasvs> approved to the list he means ? He's not on the list himself ?
[20:06] <thomasvs> I think the people adminning it are uraeus, wtay and omega :)
[20:06] <thomasvs> maybe I should get myself on there too
[20:06] <teuf> laurent sansonetti
[20:07] <teuf> and yeah, I meant approval to the list
[20:07] <teuf> I don't know whether he is on it or not
[20:07] <benow> thomasvs: hmmm, so nuke registry.xml and re-register?  this comes when spooling up a small test app (java-jni gst)
[20:09] <thomasvs> teuf: I guess he's not, otherwise it wouldn't need approval
[20:09] <thomasvs> there's no laurent sansonetti in my (quite full) INBOX :)
[20:10] <thomasvs> can you get him to mail me, but with my correct address ?
[20:10] md` ([email protected]) left irc: 
[20:13] <benow> meh, same thing when nuking registries (~/.gstreamer*,/usr/var/cache/gstreamer-0.7/*,~root/.gstreamer*)... pipelines from gst-launch work fine.
[20:13] <teuf> thomasvs: he tried to mail you 1 year ago because he wanted to use gstreamer cvs for his bindings
[20:13] <teuf> thomasvs: it's a bit late now :)
[20:13] <teuf> thomasvs: who is omegahacker ?
[20:13] <thomasvs> teuf: erik walthinsen, the original author
[20:13] <teuf> thomasvs: some of your mailing lists seem not to have an active moderator
[20:14] <thomasvs> benow: what I'm saying is, I think you have an element twice.  ie, two times the same .so flie
[20:14] <teuf> gstreamer-announce, gstreamer-bugs
[20:14] <thomasvs> teuf: hm, well, one year ago is a long time, but if he mailed more than once I would have known I think.
[20:14] <teuf> yeah, I dunno what happened
[20:14] <thomasvs> teuf: in any case, it was not intended, so if he mails me we'll take it up again
[20:15] <teuf> yeah, I told him that I'd be really surprised if you ignored his mails
[20:15] <benow> hmmm, would it be looking in both /usr/lib and /usr/local/lib... I'll go get a bit dirty
[20:15] md` ([email protected]) joined #gstreamer.
[20:17] jdahlin ([email protected]) joined #gstreamer.
[20:18] <thomasvs> teuf: well, keep me posted - let me know if he'll mail me, if not, get me his address and I'll mail him
[20:18] <benow> hmm, same thing.  gst is 0.7 new from cvs... remenance from early 0.6 install seem to be removed.  gst-launch works fine, so I'm geussing its something with the jni code... could these symptoms occur without proper initialization?
[20:19] <thomasvs> benow: it won't catch up 0.6 plugins.  what does gst-register say ?
[20:19] <thomasvs> does it complete without problems ?
[20:19] <benow> yep, loads alot
[20:19] <thomasvs> but no errors ?
[20:20] <benow> nope.... :
[20:20] <benow> rebuilding user_registry (/home/andy/.gstreamer-0.7/registry.xml)
[20:20] <benow> loaded 122 plugins with 327 features
[20:22] <thomasvs> hm, I don't know how the jni code works
[20:22] jimmy_dean ([email protected]) left irc: Read error: 110 (Connection timed out)
[20:22] <thomasvs> but if gst-launch for example is able to run fine, I can only assume the java code is doing something wrong
[20:23] <benow> the jni code is pretty straight fwd afa normal chain building goes, but does a wacky String[]->argc,argv conversion for gst_init(int,char*)
[20:24] <thomasvs> is this code that used to work fine before ?
[20:25] <benow> I don't think I've ever managed to get this codebase going... got it close, but not going.  it's an ant (build.xml) enabled jni tree I started a while ago.
[20:26] <benow> I've been doing jni wrappers for the required libs (libshout2, etc), but have decided to take another look at wrapped gst, as it would do all that I require (and more) after initial wrapping pain.
[20:28] <benow> self.getHand(Hand.RIGHT).equip(hammer).smack(Hand.LEFT);
[20:32] <thomasvs> well, I'd help you if you knew what you were talking about :)
[20:37] Nick change: Ura_dinner -> Uraeus
[20:39] <thomasvs> Uraeus: test the rpm update please :)
[20:39] <Uraeus> jdahlin: I discovered how to reproduce my first crash, the one with the trace in the bug, just set the recording quality to cd-lossy
[20:39] <Uraeus> thomasvs: its underway :)
[20:40] <Uraeus> jdahlin: setting it to voice makes it record as I told you, but it then crash when trying to play it back
[20:42] <jdahlin> Uraeus: okay
[20:42] <jdahlin> I will try to reproduce
[20:43] <teuf> thomasvs: ok, he sent you a mail, his address is [email protected]
[20:43] chris` ([email protected]) joined #gstreamer.
[20:44] <thomasvs> metacity
[20:45] <chris`> If I have the gstreamer alsa plugins installed and have /system/gstreamer/default/audiosink set to "alsasink sync=false", why would gstreamer apps say they can't find the audio sink?
[20:45] <jdahlin> thanks
[20:45] <Uraeus> thomasvs: ok, update worked, but using gst-launch-ext I am just getting static playing mp3 or ogg
[20:46] <Uraeus> hmm, able to play a wav with gst-launch at least :)
[20:47] <thomasvs> Uraeus: I don't think gst-launch-ext has been updated lately
[20:47] <thomasvs> chris`: how new is your gstreamer ?
[20:48] <Uraeus> thomasvs: I am getting lot of messages with mad and vorbisdec like this one: (process:21063): GStreamer-CRITICAL **: Factory for `vorbisdec' has no type
[20:48] <chris`> 0.6.4, latest in debian sid
[20:48] <Uraeus> thomasvs: that is the samme kinda message that makes gsr crash on playback also
[20:48] <Uraeus> thomasvs: any idea what it means?
[20:48] <thomasvs> Uraeus: can you run gst-register ?
[20:49] <Uraeus> thomasvs: yes, did so twice to try and get that message away
[20:49] <thomasvs> heh.
[20:49] <thomasvs> maybe it works if you do it four times ?
[20:49] <thomasvs> it works better if you're naked doing it :)
[20:49] <Uraeus> thomasvs: how do you know I am not naked?
[20:50] <thomasvs> I don't, but I know you didn't do it four times :)
[20:50] <Uraeus> thomasvs: gst-launch filesrc location=police.ogg ! oggdemux ! vorbisdec ! osssink <- does this work for you?
[20:50] <thomasvs> anyways, you sure you don't also have another install or something ?
[20:50] <Uraeus> thomasvs: I am sure
[20:50] <thomasvs> Uraeus: you forgot audioconvert
[20:50] <thomasvs> vorbisdec does float audio
[20:50] <jdahlin> if he's naked he wouldn't be able to type so fast, he would be busy with at least one hand
[20:53] <thomasvs> you can be naked without touching yourself
[20:53] Action: Uraeus doesn't understand anything atm
[20:53] <Misirlou> yeah, jeez
[20:53] Action: thomasvs looks across the table to make sure this is what's going on
[20:53] <Misirlou> jdahlin is such a perv!
[20:53] <Uraeus> [cschalle@3jq3n91 test]$ gst-launch filesrc location=test.mp3 ! mad ! audioconvert ! alsasink
[20:53] <Uraeus>  
[20:53] <Uraeus> ** (process:21077): CRITICAL **: how to remove plugins?
[20:53] <Uraeus>  
[20:53] <Uraeus> (process:21077): GStreamer-CRITICAL **: Factory for `mad' has no type
[20:53] <Uraeus> WARNING: erroneous pipeline: no element "mad"
[20:53] <Uraeus>          Trying to run anyway.
[20:53] <Uraeus> RUNNING pipeline ...
[20:53] <Misirlou> No respect for the millions of nudists. :(
[20:53] <Uraeus> Caught interrupt.
[20:53] <Uraeus> Execution ended after 10561 iterations (sum 383957000 ns, average 36356 ns, min 33000 ns, max 8202000 ns).
[20:53] <Uraeus> [cschalle@3jq3n91 test]$ gst-inspects|grep mad
[20:53] <Uraeus> bash: gst-inspects: command not found
[20:53] <Uraeus> [cschalle@3jq3n91 test]$ gst-inspect|grep mad
[20:53] <Uraeus> mad:  id3tag: id3 tag extractor
[20:53] <Uraeus> mad:  mad: mad mp3 decoder
[20:54] Last message repeated 1 time(s).
[20:54] <Uraeus> mad:  id3tag: id3 tag extractor
[20:54] <Uraeus> can you guys explan this?
[20:54] Action: chris` pokes his alsasink
[20:54] <Uraeus> I mean it reports no 'mad' element, but gst-inspect clear says it is there
[20:57] <Uraeus> argh ! I do have some old extra stuff installed
[20:57] <Uraeus> thomasvs: what do you call your packages? .fdr. or .lvn. ?
[21:09] <Uraeus> thomasvs: now I am 100% sure I am just using your packages, but mp3 and ogg playback is still borked
[21:09] <Uraeus> apoc_: both modplug and mikmod playback has issues
[21:15] <thomasvs> Uraeus: you have it more than once, so you borked something
[21:16] <thomasvs> Uraeus: can you remove your user registry ?
[21:17] sub_pop ([email protected]) joined #gstreamer.
[21:19] BBB ([email protected]) joined #gstreamer.
[21:19] <BBB> thomasvs: you're truely annoying. you propose a fix, we accept it, we all agree it's icky and I propose that during 0.9.x I'll work on a proper solution. what is your problem?
[21:20] <BBB> I truely don't see how anyone can disagree with the above solution
[21:25] wheels ([email protected]) left irc: Read error: 60 (Operation timed out)
[21:25] aeyakovenko ([email protected]) joined #gstreamer.
[21:25] <aeyakovenko> hello
[21:26] <BBB> (and for the .list - I think we separated that in order to distinct between gst-plugins and gstreamer. feel free to merge)
[21:26] <aeyakovenko> can a single process create more then one pipelien?
[21:27] <BBB> sure
[21:28] <aeyakovenko> well for some reason i cant get this pipleline to work, (fdsrc, ffdec_h263, xvideosink)
[21:29] <aeyakovenko> i know there is data being put into the fd, but its never pulled, so it just fills up to the system limit, then all my writes block
[21:30] <BBB> do you actually iterate the pipeline?
[21:30] <aeyakovenko> its on its own thread
[21:30] <BBB> is it in playing?
[21:30] <aeyakovenko> yep
[21:31] <BBB> you could try adding some debug code into fdsrc and ffdec and see whether each of these receives data, whether it processes data and whether it outputs anything
[21:31] <aeyakovenko> what is also strange, when i try to destroy that pipeline, it blocks when i set it to NULL
[21:32] <aeyakovenko> ok, ill try that
[21:34] <aeyakovenko> any idea why it would block on setting that pipeline to NULL?
[21:37] <thomasvs> BBB: my problem is that I don't understand why you guys claim the design is correct.
[21:37] <thomasvs> BBB: you seem to think that this issue is a new one, while I say that this is something we new from the start wouldn't work with our state redesign
[21:37] <mathrick> chris`: alsasink is to be considered broken in 0.6, you can spare some time by not even trying ;)
[21:37] <thomasvs> BBB: other than that, no problem.
[21:38] <chris`> mathrick: heh, ok then
[21:38] <thomasvs> BBB: it's just annoying to see us swing from left to right and back again, and still we have no design for how it really ought to work
[21:38] <BBB> aeyakovenko: not really
[21:38] <BBB> thomasvs: now, I agree with *that* part
[21:38] <thomasvs> BBB: it just happens to work the way some guys at one time decided to do it because it worked out best for the kind of stuff THEY were doing
[21:38] <BBB> thomasvs: but as for this one, just accept that it's a bug
[21:38] <BBB> and just accept that we'll work on it during 0.9.x
[21:39] <BBB> that sounds perfect to me, no?
[21:39] Action: BBB wants to add some sort of "unknown codec fallback" to gst-ffmpeg
[21:40] Action: mathrick has impression thomasvs agreed with BBB's solution, just not exactly with wording BBB used ;)
[21:40] <thomasvs> BBB: I'm not sure.  I'd rather hear us say "yes we care about gnome, no we don't want to be replaced by helix or xine, and yes we are going to work on some designs this time around"
[21:42] Nick change: mathrick -> mathrick|watchin
[21:43] wheels ([email protected]) joined #gstreamer.
[21:50] <BBB> thomasvs: ... that's just not my style
[21:51] <thomasvs> BBB: what isn't ?
[21:51] <BBB> crawling
[21:51] <BBB> I'd rather just fix stuff ;)
[21:51] <thomasvs> crawling ?
[21:51] <thomasvs> am I missing something ? :)
[21:51] <BBB> (15:40:33) thomasvs: BBB: I'm not sure.  I'd rather hear us say "yes we care about gnome, no we don't want to be replaced by helix or xine, and yes we are going to work on some designs this time around"
[21:51] <BBB> that
[21:52] <thomasvs> what do you mean by crawling ?
[21:52] <BBB> "door het stof gaan"
[21:52] <thomasvs> yeah, but how are the two related ?
[21:52] <thomasvs> I don't get it :)
[21:53] <BBB> n/m, I'll just fix stuff and you tell the gnome people how much both of us love them, ok? ;)
[21:53] <thomasvs> I think that's getting old to them :)
[21:53] Nick change: harshyOut -> harshy
[21:53] <BBB> and while you're at it, there's a cinepak patch in bugzilla that needs love (patch -p1 < .. && cvs commit)
[21:54] <BBB> :p
[21:54] <BBB> btw, thomasvs, where is dolphy?
[21:54] <BBB> I haven't seen him, and I've got a blocker bug for him
[21:55] <BBB> the switch one in libgstplay
[22:01] <BBB> and it'd rock if someone could tell me that totem works perfectly fine for him
[22:01] <BBB> because it doesn't here
[22:01] <BBB> and before yesterday, I haven't ever been able to play any movie
[22:01] <BBB> which is extremely weird
[22:01] <teuf> BBB: he left for 1 week for holidays iirc
[22:01] <BBB> hm
[22:01] <BBB> k
[22:02] <alley_cat> i haven't been able to play avi files for weeks
[22:04] <BBB> thank you for mentioning that!
[22:05] <BBB> so, what exactly doesn't work
[22:05] <alley_cat> Uraeus: how far have you gotten with fixing gst-ffmpeg?
[22:05] <alley_cat> BBB: it's quite random behaviour, so i want to valgrind it to find some clues
[22:05] <BBB> I fixed most of my AVI issues yesterday night
[22:05] <alley_cat> but without ffmpeg it's kinda hard to play anything
[22:05] <BBB> oh, yes
[22:06] jdahlin ([email protected]) left irc: "Leaving"
[22:06] Zeenix ([email protected]) joined #gstreamer.
[22:07] <Zeenix> hello
[22:08] <alley_cat> BBB: behaviour seems to be dependent on system load, sometimes it segfaults, sometimes it plays (especially with debugging output on), sometimes it hangs eating 100% cpu
[22:08] <alley_cat> the hangs happening in a g_malloc as far as i could tell from my gdb runs
[22:09] <Zeenix> does the kernel api provides a hashtable implementation?
[22:10] <BBB> no
[22:10] <BBB> alley_cat: with avi? I had that with quicktime, avi works fine here
[22:10] <BBB> I fixed a memory error in quicktime yesterday
[22:10] <BBB> avidemux plays fine here
[22:10] <BBB> afaik
[22:10] <alley_cat> BBB: with avi, odd thing is that it worked for files without an index
[22:10] <BBB> but I'm very interested in valgrind runs from that
[22:10] <BBB> :)
[22:10] <BBB> ah
[22:10] <BBB> interesting
[22:11] <BBB> as I said, please run it through valgrind
[22:11] <BBB> all my files have an index
[22:11] <Zeenix> oh sorry
[22:11] <BBB> but they all work
[22:11] <alley_cat> as soon as gst-ffmpeg compiles again i'll do it
[22:11] <BBB> zeenix: the 'no' was for you :)
[22:11] <BBB> alley_cat: what fails?
[22:11] <BBB> dsputil.h missing?
[22:11] <alley_cat> undefined references when linking libavformat
[22:12] <BBB> hm
[22:12] <BBB> libtool is rather tricky
[22:12] <BBB> I'm bad at that
[22:12] <BBB> I'll try to have a good look tonight, ok?
[22:12] <alley_cat> no, when linking output-example actually
[22:12] <Zeenix> BBB: no, i meant to throw this in #linux-kernel, anyone by 'no' you meant that it does'nt offer this implementation?
[22:12] <BBB> yes
[22:12] <BBB> alley_cat: er... ok, I probably want to disable compilation of output_example then
[22:12] <BBB> I'll see how to do that
[22:13] <BBB> if you disable that manually, does the rest compile fine?
[22:14] <Zeenix> s/anyone/anyway
[22:14] <BBB> and - to ask more - does anyone know how we playback 5:1 audio?
[22:14] <BBB> osssink only wants 2 channels
[22:14] <BBB> audioconvert only does mono-stereo or reverse
[22:14] <BBB> how to play 5:1?
[22:15] <alley_cat> disabling that worked, yay
[22:15] <BBB> file a bug and assign to me, I'll kick myself into fixing that
[22:15] <BBB> :)
[22:15] <alley_cat> wonder why i didn't even try that before :)
[22:16] <alley_cat> what's your bugzilla account?
[22:16] <BBB> [email protected]
[22:17] md` ([email protected]) left irc: 
[22:17] <Uraeus> BBB: is your auto* patch for ffmpeg not commited? current ffmpeg cvs is not even close to building, or do I need to update it myself?
[22:17] <BBB> Uraeus: "not even close to building" = "what errors do you get"?
[22:18] <Uraeus> BBB: the makefiles tries to dist files that doesn't exist etc.
[22:18] <aeyakovenko> is fdsrc part of core or gst-plugins?
[22:18] <BBB> it's all committed. the above from alley_cat seems like libtool funkiness, which I'm bad at
[22:18] <BBB> dist files that don't exist = incomplete ffmpeg checkout
[22:18] <BBB> aeyakovenko: core
[22:18] <BBB> gstreamer/gst/elements/gstfdsrc.c
[22:18] <aeyakovenko> cool thanks
[22:20] <Uraeus> alley_cat: rm -rf gst-libs in gst-ffmpeg and check out from cvs and try again, I think that is the problem for both of us based on what BBB just said
[22:20] <alley_cat> i removed the ffmpeg dir and let it checkout a fresh one, that fixed the other errors i was getting
[22:20] <Uraeus> BBB: trying to play mp3 and ogg I get errors about no factory information, any idea?
[22:21] <alley_cat> Uraeus: got another nassink patch for you, still some problems though
[22:21] <Uraeus> alley_cat: [email protected]
[22:21] <alley_cat> filesrc location=test.ogg ! spider ! nassink outputs garbled stuff
[22:21] <BBB> Uraeus: hm, I've found that once before and that appeared to be some plugin using a non-static plugin init function
[22:21] <BBB> apart from that, I'm not very sure
[22:21] <BBB> could you post it on bugzilla?
[22:22] <BBB> also, it'd be useful if you'd run with --gst-fatal-warnings and provide a backtrace on the warning where it cannot find the type
[22:22] <BBB> that's sometimes useful
[22:22] <BBB> I don't see such warnings, though...
[22:22] <Uraeus> BBB: sure, but could you just try this pipeline to see if you can reproduce: gst-launch filesrc location=test.mp3 ! mad ! osssink
[22:22] <BBB> I'm not at home ;)
[22:22] <BBB> I'll try tonight
[22:22] <Uraeus> BBB: ah ok :)
[22:23] <BBB> as I said, post on bugzilla, I'll post my findings tonight and you'll find it tomorrow ;)
[22:23] <Uraeus> cause that pipeline gives me the error and static in the speaker
[22:23] Action: BBB @ university
[22:24] <BBB> I'll be back tonight
[22:24] <BBB> bbl!
[22:24] BBB ([email protected]) left #gstreamer.
[22:27] <alley_cat> hmm, gst-player doesn't use either gstreamer/* gconf keys nor gstreamer-0.7/*
[22:28] <Uraeus> alley_cat: no, dolphy has been focusing on Totem lately, so gst-player doesn't work that well atm cause it isn't updated to latest changes
[22:29] Action: Uraeus thinks some commit in the last 12 hours have borken things up bad
[22:33] <ds> ?
[22:34] <Uraeus> ds: http://bugzilla.gnome.org/show_bug.cgi?id=136494
[22:37] <alley_cat> doh, totem 0.99.9 doesn't find any gconf keys either
[22:40] Action: alley_cat tries cvs totem again
[22:41] <alley_cat> hmm, guess the nas patch willl have to wait
[22:41] <alley_cat> for some reason it breaks nassink for ogg files
[22:42] <Uraeus> alley_cat: prolly cause the ogg plugins produce float audio?
[22:45] <alley_cat> bah, cvs totem need gtk 2.3
[22:46] <Uraeus> ok, time for me to head to bed, night
[22:46] Uraeus ([email protected]) left irc: "Client exiting"
[22:46] jcsston ([email protected]) joined #gstreamer.
[22:47] Shoragan ([email protected]) joined #gstreamer.
[22:48] <alley_cat> cool, avidemux also hangs with valgrind, without getting any output  from valgrind :/
[22:49] md` ([email protected]) joined #gstreamer.
[22:52] <aeyakovenko> how do i set the debug levels?  --gst-debug-level=THREADS doesnt do anything
[22:54] jimmy_dean ([email protected]) joined #gstreamer.
[22:55] Action: Zeenix tries porting ghashtable to the kernel...
[23:03] kmaraas ([email protected]) left irc: Read error: 110 (Connection timed out)
[23:07] lupusBE ([email protected]) joined #gstreamer.
[23:18] Action: sxpert disables dbus in carpc sound server.. lets see where that monster leak comes from
[23:23] <teuf> ah BBB left, I wanted to tell him that totem worked with gstreamer here ;)
[23:23] <sxpert> btw, I still get "** (process:28808): WARNING **: opt: unlink elements in same group: implement me"
[23:24] <sxpert> that's gstreamer from a few days ago...
[23:26] kwm ([email protected]) left irc: "Leaving"
[23:27] Zeenix ([email protected]) left irc: Read error: 60 (Operation timed out)
[23:27] Zeenix ([email protected]) joined #gstreamer.
[23:29] mathrick|watchin ([email protected]) left irc: Read error: 110 (Connection timed out)
[23:39] harshy ([email protected]) left irc: "I quit for now"
[23:41] LeRoutier ([email protected]) joined #gstreamer.
[23:41] <LeRoutier> hello
[23:46] mathrick ([email protected]) joined #gstreamer.
[23:46] <alley_cat> totem with gstreamer doesn't work at all here
[23:48] sub_pop ([email protected]) left irc: "Client exiting"
[23:51] lupusBE ([email protected]) left irc: Remote closed the connection
[23:53] <LeRoutier> what a beautifull summary for a bug : "head -number" is evil (and deprecated/unportable) (http://bugzilla.gnome.org/show_bug.cgi?id=136500)
[23:55] markey ([email protected]) left irc: "leaving"
[00:00] --- Mon Mar  8 2004
[00:06] jcs ([email protected]) left irc: Remote closed the connection
[00:06] harshy ([email protected]) joined #gstreamer.
[00:07] <alley_cat> well, it's correct, head -n <number> is correct posix
[00:10] <LeRoutier> yes, i know, that's why i filled this bug report
[00:12] <LeRoutier> just attached the 2 trivial patches that make it use the posix/portable syntax
[00:12] <alley_cat> the summary is really nice
[00:13] <alley_cat> gstreamer devels might even listen to that complaint (not like mplayer devs)
[00:13] <alley_cat> what i gathered from the discussion about this issue on mplayer-dev-eng some time ago several unices don't support head -n <number>
[00:13] chris` ([email protected]) left #gstreamer ("Client exiting").
[00:14] <LeRoutier> well, it's their problem (those unixes). i suppose the -number is a GNU extension. all unixes have to support POSIX/SUS or die
[00:15] <alley_cat> after my experiences with several solaris boxes i can only agree that most unices should probably die ;)
[00:16] <LeRoutier> i hated solaris 2.6 so much. and had so much problems with signals using perl on that
[00:20] <benow> jni 1 : 0 benow
[00:20] benow ([email protected]) left irc: "Guiding the meefalo to glorious meef free living"
[00:21] <sxpert> alley_cat: then those unices head(1) command is brokeb
[00:21] <sxpert> b0rken even
[00:24] <LeRoutier> anyone know when the newer bugzilla would be used on gnome site ? actual one is a pity
[00:26] <Misirlou> LeRoutier: It should be Coming Soon.
[00:26] <Misirlou> LeRoutier: I believe that Andrew Sobala is "in charge" of it.
[00:27] <LeRoutier> k, i suppose it would come close to gnome 2.6 release
[00:27] <LeRoutier> thx
[00:30] teuf ([email protected]) left irc: Remote closed the connection
[00:45] Zeenix ([email protected]) left irc: "The whole problem with the world is that fools and fanatics are always so certain of themselves, but wiser people so full of 
[00:54] thaytoo ([email protected]) joined #gstreamer.
[00:55] <LeRoutier> good morning australia
[00:58] <thaytoo> please tell me you've never seen that show
[01:00] <LeRoutier> well, if it is an australia TV show, there is little (mal)chance i saw it one day
[01:02] <thaytoo> it is, and a terrible example of our culture :)
[01:04] <LeRoutier> well, every 'civilized' country has its own TV garbage at any hour. more channels meaning more chances to see so different garbage
[01:15] <wheels> thaytoo: Ah, as opposed to say the crocodile wrangler or something?  :-)
[01:16] <sxpert> thaytoo: "culture" ?
[01:16] <sxpert> wheels: it's the "crocodile hunter" :D
[01:17] Action: wheels doesn't have a tv, so I'm going on third party information.  :-)
[01:17] sublett ([email protected]) left irc: "Nice of you to put me on the money"
[01:17] <LeRoutier> i prefer the 'crocodile dundee' personnaly. and even better, Amstrad's crocodile
[01:17] smoke ([email protected]) got netsplit.
[01:18] <sxpert> LeRoutier: we all know amstrad manufactured fake lacoste computers :D
[01:19] <LeRoutier> i miss my Amstrad CPC, sniff
[01:19] <sxpert> hehehe
[01:20] Action: sxpert still has a working SX64 (except the floppy dive that had a mecanical failure :( )
[01:20] <LeRoutier> sxpert: usually, those kind of failure could be fixed with a little elastic
[01:21] <sxpert> LeRoutier: not this time, broken foundry piece...
[01:21] <LeRoutier> oups dommage, me rappelle le nombre de fois où j'ai remplacé la couroie du lecteur 3" du CPC par un élastique. et hop, ça remarchait
[01:22] <sxpert> hehehe
[01:22] <sxpert> LeRoutier: en fait, on voit ou c'est casse, mais aucune idee de la forme que ca avait, donc, c'est mort...
[01:22] <sxpert> sinon, e-machine-shop.com c'est bien :D
[01:23] smoke ([email protected]) got lost in the net-split.
[01:23] <thaytoo> wheels: on a par with Steve Irwin, yah
[01:23] <LeRoutier> si c'est le même type de lecteur que le CPC, je dois avoir un schema pour pouvoir brancher un lecteur 3"1/2 standard dessus. dommage pour le contenu des vieilles disquettes bien sûr
[01:24] <sxpert> LeRoutier: non, c'est juste une mecanique de 5 1/4, le controlleur etait entierement programmable (les protections de jeux etaient redoutables, avec des pistes en diagonales...
[01:25] <LeRoutier> lets stop this nostalgia trip, it hurts too much :)
[01:25] <LeRoutier> (and it polutes the channel)
[01:25] spyder482 ([email protected]) joined #gstreamer.
[01:26] harshy ([email protected]) left irc: Read error: 113 (No route to host)
[01:26] <sxpert> LeRoutier: heh
[01:27] <sxpert> LeRoutier: http://www.zock.com/8-Bit/D_1541.HTML
[01:29] <thaytoo> ahr, sweet 1541
[01:29] <thaytoo> vice makes me happy
[01:36] <sxpert> hehehe
[01:41] thaytoo ([email protected]) left irc: Read error: 113 (No route to host)
[01:46] thaytoo ([email protected]) joined #gstreamer.
[01:49] foser ([email protected]) left irc: "[ I want to believe ]"
[02:02] harshy ([email protected]) joined #gstreamer.
[02:04] <LeRoutier> anyone with CVS access to apply my 2 trivial patches ? (http://bugzilla.gnome.org/show_bug.cgi?id=136500). or is it for thomasvs ?
[02:06] ramon ([email protected]) joined #gstreamer.
[02:07] ramon ([email protected]) left irc: Client Quit
[02:12] sub_pop ([email protected]) joined #gstreamer.
[02:29] pb_ (~pb@2002:5160:45ef:0:240:95ff:fe30:aa84) left irc: "Client exiting"
[02:45] <LeRoutier> good night everyone
[02:45] LeRoutier ([email protected]) left irc: "Leaving"
[02:50] jcsston ([email protected]) left irc: Read error: 60 (Operation timed out)
[02:55] sethn ([email protected]) joined #gstreamer.
[02:56] <sethn> Is there some sort of FFT plugin available for gstreamer?


-------------------------------------------------------
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