IRC Logs

*******************************************************************
[03:00] <BBB> it's a bug in gstreamer
[03:00] <BBB> by default, it should passthrough unless otherwise requested
[03:00] <ds-work> it has nothing to do with fixating
[03:00] <ds-work> actually
[03:00] <BBB> so why do all my movies play at 16x16 then?
[03:00] <ds-work> it _does_ do passthrough unless requested
[03:00] <BBB> let me say it differently: my movies play at 16x16: is that a bug?
[03:00] <ds-work> how many times do I have to tell you?
[03:00] <BBB> it doesn't. my movies play at 16x16
[03:00] <BBB> it might passthrough in theory
[03:01] <BBB> but it doesn't in practiec
[03:01] <BBB> *my movies play at 16x16*
[03:01] <ds-work> something is forcing it to 16x16
[03:01] <ds-work> and it's not the plugins
[03:01] <BBB> it's not the totem window... totem's window is nicely at the default size (~400x300)
[03:01] <ds-work> and it's not the core
[03:01] <BBB> ok, but could you please help me debugging this? this is fairly annoying for end users
[03:02] <ds-work> it's likely due to the X window getting a configure event for that size
[03:02] <BBB> I don't care much where the bug is... however, in the end, there's a bug somewhere
[03:02] smoke ([email protected]) got netsplit.
[03:02] <ds-work> in the case of 16x16, the pipeline originally fixates to 16x16 as a guess, but then gets corrected before data flows
[03:03] <BBB> and could you explain me why ../gstreamer/tools/gst-launch videotestsrc ! ffcolorspace ! videoscale ! ximagesink plays at 16x16, even though videotestsrc:src fixates to 384x288 and ximagesink:sink fixates to 320x240?
[03:03] sub_pop ([email protected]) left irc: "Client exiting"
[03:04] <ds-work> I'd guess that for some reason, ximagesink decides to re-resize to 16x16 after the format is forced to XxY
[03:04] <BBB> so it *is* an element bug
[03:04] <ds-work> or an interaction with how it is used in gst-player
[03:05] <BBB> no, this is a gst-launch pipeline
[03:05] <BBB> so this is not in player
[03:05] <BBB> the gst-launch pipeline plays at 16x16
[03:06] <ds-work> your pipeline fixates to 16x16 because fixate functions do not travel down the pipeline.  What happens is that ffcolorspace ! videoscale is the last to fixate, which gives 16x16
[03:06] <ds-work> we need an extra layer of complexity in the negotiation code to handle preferred formats
[03:06] <ds-work> I'll probably do that in 0.9
[03:06] <BBB> but shouldn't ffcolorspace and videoscale 'forward' the getcaps() function to videotestsrc/ximagesink?
[03:06] <BBB> and 'get' the preferred size?
[03:06] <ds-work> fixate has nothing to do with getcaps
[03:07] <ds-work> getcaps is _allowed_ caps
[03:07] <BBB> I know
[03:07] <BBB> but those elements don't have a fixate function, right?
[03:07] <ds-work> not preferred or set caps
[03:07] <BBB> so it uses the default one
[03:07] <ds-work> yes
[03:07] <BBB> right?
[03:08] smoke ([email protected]) got lost in the net-split.
[03:08] <ds-work> yes, the ffcolorspace ! videoscale uses the default fixate functions
[03:08] <BBB> ok
[03:08] <BBB> so the formats going "in" this fixate functions are the return values of the getcaps functions on ffcolorspace:src and videoscale:sink, right?
[03:09] <ds-work> yes
[03:09] <BBB> ffcolorspace:src "forwards" the getcaps function to its other pad's peer (videotestsrc:src), and then adds 'any' format because that's what it does, right?
[03:09] <ds-work> yes
[03:10] <BBB> and the formats that videotestsrc supports are prepended to the list of formats that ffcolorspace can convert to, so that it "prefers" to be a passthrough, right?
[03:10] <ds-work> no
[03:11] <ds-work> passthrough is determined by link functions
[03:11] <ds-work> in a pad link function, you attempt to set the same caps on the otherpad.  if it works, you use paddthrough
[03:11] <BBB> yes, but the formats as given by getcaps determine what will "likely" or "unlikely" go into the link function
[03:12] <ds-work> well, it determines what can or cannot go to the link function
[03:12] <BBB> the core doesn't know when elements use passthrough, so elements need to say "I prefer this format because it'll give me passthrough" as part of their getcaps return value
[03:12] <BBB> and they do this by placing a gststructure earlier or later in the list of structs in the GstCaps
[03:13] <ds-work> no
[03:13] <ds-work> GstCaps has no preferences
[03:13] <BBB> why not? it used to work this way
[03:13] <ds-work> because it doesn't actually work
[03:13] <BBB> because...?
[03:13] <ds-work> at most, it only works with rough format
[03:14] <ds-work> and to keep ordering in gstcaps would have made the code a lot more complicated
[03:14] <BBB> yes, but it's required
[03:14] <ds-work> and literally impossible for intersection
[03:14] <ds-work> explain how it's required
[03:14] <ds-work> we've never had it
[03:15] <BBB> right now, it's mere luck if you end up doing passthrough... we need to actively prefer passthroughing
[03:15] <ds-work> we do actively prefer passthrough
[03:15] <BBB> and if you don't believe me: videotestsrc ! ... plays at 16x16 even though we fixated to other sizes... my file ! ... plays at 320x240 even though the file is 768x576... those are *bugs*
[03:16] <ds-work> at most, it's a bug in ximagesink
[03:16] <BBB> ok, so it's still a bug
[03:17] <BBB> and ximagesink causes the second... not the first
[03:17] <ds-work> the first is because we don't have a good mechanism for preferred caps
[03:17] <BBB> ok, so it's still a bug
[03:17] <ds-work> we have a method that doesn't suck too much
[03:17] <ds-work> a wishlist bug
[03:18] <ds-work> 16x16 is not _wrong_
[03:18] <BBB> I'm not saying this to annoy you. I'm saying this because after we've released 0.8.0, people will come in here and blame us that their video plays at 16x16 in totem
[03:18] <ds-work> as we have just discussed, that's a different bug
[03:18] <BBB> yes... and I'd still like to know which bug ;)
[03:19] <BBB> maybe it is a bug in ximagesink
[03:19] <BBB> that's very well possible
[03:19] <ds-work> I have never seen this bug
[03:19] <BBB> probably because you run with xvimagesink as output
[03:19] <BBB> (that's the default, IIRC)
[03:19] <ds-work> gst-player and gst-lauch always comes out with the right size
[03:19] <BBB> what bpp is your display running?
[03:19] <ds-work> no, i tend to use ximagesink, because swfdec looks better
[03:19] <ds-work> 32
[03:20] <BBB> I run at 16... maybe that matters? (I don't see how it would, but it might)
[03:20] <ds-work> it shouldn't
[03:21] <BBB> I know
[03:21] <BBB> I'll switch to 24 and try if that fixes something
[03:21] <ds-work> one of my laptops runs at 16 bit, and I never saw the bug there
[03:21] <BBB> hm...
[03:22] <ds-work> what do you get in the gst-play demo?
[03:23] <BBB> where is that?
[03:23] <BBB> you mean the player example?
[03:23] <ds-work> examples/gstplay/
[03:23] <BBB> [rbultje@shrek gst-plugins]$ examples/gstplay/player /media/clips/Trance\ Energy\ 2002\ DvDSebRip.avi
[03:23] <BBB> Segmentation fault
[03:23] <BBB> [rbultje@shrek gst-plugins]$
[03:24] <ds-work> cute
[03:24] <ds-work> how hard would that be to transfer to me?
[03:24] Action: ds-work needs to collect clips
[03:24] <BBB> I'll fix it, hold on
[03:25] <BBB> it's a few 100 MB, so we'll do that at the next gettogether, ok?
[03:26] <ds-work> ok, nm
[03:26] <taaz> http://www.insecure.org/nmap/images/haxxxor/video/HaXXXor-vol1-Nmap-high.mov
[03:26] <taaz> there's a clip for you ;)
[03:27] <ds-work> do you have fullpal.avi and halfpal.avi?
[03:27] Action: ds-work f33rs
[03:27] <BBB> [rbultje@shrek gstplay]$ ./player /media/clips/Trance\ Energy\ 2002\ DvDSebRip.avi
[03:27] <BBB> got video size 16, 16
[03:27] <BBB> :)
[03:27] <BBB> it runs though
[03:30] <BBB> ds-work: me?
[03:30] <BBB> er, dunno
[03:30] <BBB> probably not
[03:30] alley_cat ([email protected]) left irc: "May the Source be with you!"
[03:31] jcsston ([email protected]) joined #gstreamer.
[03:31] <BBB> Do you have any idea why quicktime movies are 'preferably' played through ffmpeg's quicktime demuxer? Shouldn't it prefer qtdemux?
[03:32] <BBB> Maybe a local error, but still
[03:32] <ds-work> rank problem
[03:32] sub_pop ([email protected]) joined #gstreamer.
[03:32] <BBB> Hm
[03:33] Action: BBB pulls down default rank reply from ffmpeg to 1
[03:33] <ds-work> qtdemux is RANK_PRIMARY
[03:38] Action: BBB f1x0rs
[03:38] <BBB> (gst-ffmpeg bug from me)
[03:39] Action: BBB found bug
[03:50] <ds-work> don't you love it when gdb tells you your plugin crashed in the middle of a big block of comments?
[03:54] <thaytoo> that means that your comments have a bug, yup
[03:54] <thaytoo> the easiest way to avoid comment bugs is never to write any comments
[03:54] <thaytoo> yup
[03:55] <ds-work> thaytoo: yer ma
[03:57] <thaytoo> heh :)
[03:58] <thaytoo> oh fine, it's either dirty or compiled with optimisations then
[03:58] <thaytoo> is there a third option?
[03:58] <BBB> you forgot to recompile after checking out?
[03:58] <thaytoo> that's 'dirty' 
[03:59] <ds-work> gdb is buggered?
[03:59] <BBB> oh
[03:59] Action: BBB doesn't know
[03:59] <BBB> shall I send our patches to the ffmpeg devels?
[03:59] Action: BBB feels like doing a flamewar
[03:59] <thaytoo> ds-work: hrmn, yeah.. could be gdb too
[03:59] <thaytoo> although it is usually pretty good about telling you when it has inconsistant internal state.
[04:05] <ds-work> oops
[04:05] Action: ds-work just rm'd a Makefile
[04:05] <ds-work> time to autogen again
[04:15] <Misirlou> ds-work: way 2 go newb
[04:17] Action: ds-work wonders if the many unchecked-in changes include any bugfixes
[04:22] <BBB> how do I get a backtrace of X errors?
[04:22] <BBB> X Error of failed request:  BadAccess (attempt to access private resource denied)
[04:22] <BBB> [..]
[04:23] <BBB> (gdb) bt
[04:23] <BBB> No stack.
[04:23] <BBB> ?
[04:23] <ds-work> use a powerpc
[04:23] Action: ds-work ducks
[04:24] Action: BBB doesn't kick anyone
[04:24] Action: BBB had good thoughts for the new month
[04:24] <ds-work> for that matter, anything !i386 will work
[04:24] <BBB> shut up
[04:24] <BBB> how do I get a backtrace? :p
[04:24] <ds-work> dunno
[04:25] <ds-work> I was just commenting about the fact that i386 is most likely to lose it's stack pointer
[04:26] <ds-work> anyone have an AAC file?
[04:26] <BBB> yes
[04:26] <BBB> one in matroska
[04:26] <ds-work> preferably one that we can put on fd.o and not get sued?
[04:26] <BBB> so you need to play it back using matroskademux .audio_00 ! faad ! osssink :)
[04:26] <BBB> it's Starwars_5_1_AAC thing
[04:26] <BBB> you have that, don't you?
[04:26] <ds-work> no
[04:27] <BBB> ehm
[04:27] <BBB> DCC?
[04:28] <BBB> wait
[04:28] <BBB> samples.matroska.org
[04:28] <BBB> :)
[04:28] <BBB> download it there
[04:29] <bluejay> playing oggs with spider is playing static, but using `oggdemux ! vorbisdec ! audioconvert` seems to work. Running with -v, shows that with spider is using half the channels, width, and depth after audioconvert. Any ideas?
[04:30] <BBB> not unless you can provide a list of the elements that it is using
[04:30] <BBB> to do that, use -o, output to an XML file and see what elements are put inside spider
[04:30] <BBB> :)
[04:31] <bluejay> with -v, it looks like spider's using the same elements.
[04:31] <ds-work> does that work?  or does it dump before spider plugs anything?
[04:32] <BBB> I think it dumps right after going into playing
[04:37] <bluejay> It looks like it didn't expand the spider in the xml dump.
[04:37] <ds-work> does it contain spideridentities?
[04:38] Action: ds-work fixes a number of bugs in faad
[04:39] <bluejay> ds-work: yes, two of them
[04:39] <BBB> that's the start state
[04:39] <BBB> so it's not far enough yet...
[04:39] <BBB> can you run it in gst-editor and see what spider does there?
[04:40] Action: BBB sees interesting crashes in ximagesink
[04:40] <bluejay> -v shows stuff inside spider
[04:40] <bluejay> I don't have gst-editor. It's another CVS module?
[04:40] <ds-work> yes
[04:42] <ds-work> how cute
[04:42] <ds-work> er, nm
[04:42] Action: ds-work is crazy
[04:42] <ds-work> nothing to see here
[04:50] Nick change: kh_out -> khalek
[04:53] <ds-work> could someone play a wav file and tell me if it's broken?
[04:55] <BBB> works fine using gst-launch
[04:55] <BBB> and using spider
[04:55] <ds-work> ok
[04:55] <BBB> interestingly, libgstplay fails to play it properly
[04:56] <BBB> I hear the sound, but there's loads of "ticks" in it
[04:56] <BBB> like someone is riding a motorbike
[04:57] <ds-work> ok, that's what I hear
[04:58] Action: ds-work wishes mpeg would work
[04:58] <ds-work> undefined symbol: mpeg2_pts
[04:58] <BBB> MPEG plays at one FPS and 16x16?
[04:58] <bluejay> ... the gst-editor in CVS seems to be for gst-0.6 :(
[05:01] <ds-work> it;s for 0.7
[05:02] Action: ds-work =~ s/-work//;
[05:02] <bluejay> oops, forgot to gst-register
[05:02] <ds-work> er
[05:02] Action: ds-work ~= s/-work//;
[05:06] <bluejay> in gst-editor, when I play the ogg, it makes the right elements, but doesn't connect them to anything. (It doesn't connect mad either) But it still plays static
[05:13] jcsston ([email protected]) left irc: Read error: 110 (Connection timed out)
[05:18] <BBB> ds-work: update to newer libmpeg2?
[05:45] <ds> BBB: yeah.  taaz?
[05:48] jcsston ([email protected]) joined #gstreamer.
[05:53] <BBB> *g*
[06:16] bitshifter ([email protected]) left irc: Remote closed the connection
[06:16] bitshifter ([email protected]) joined #gstreamer.
[06:30] harshy ([email protected]) joined #gstreamer.
[06:57] thaytoo ([email protected]) left irc: "leaving"
[07:06] BBB ([email protected]) left irc: "Client exiting"
[07:11] aeyakovenko ([email protected]) joined #gstreamer.
[07:11] <aeyakovenko> hello
[07:13] <aeyakovenko> can i only put elements into a thread or can i put bins into a thread
[07:37] <aeyakovenko> oh, ic, i need to make ghost pads
[07:42] ChrisHJW ([email protected]) joined #gstreamer.
[07:58] Action: bitshifter wonders why gst-plugins/debian/*  hasn't been updated since 0.4.1
[08:02] <ds> because it's never been maintained there
[08:16] harshy ([email protected]) left irc: "I quit for now"
[08:17] mathrick|sleep ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[08:22] ChrisHJW ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[08:29] mathrick ([email protected]) joined #gstreamer.
[08:30] markey ([email protected]) joined #gstreamer.
[08:47] ds ([email protected]) left irc: Read error: 60 (Operation timed out)
[08:57] sjoerd_ ([email protected]) left irc: Remote closed the connection
[09:05] kmaraas ([email protected]) left irc: Client Quit
[09:17] jimmy_dean ([email protected]) got netsplit.
[09:17] sxpert ([email protected]) got netsplit.
[09:18] sjoerd ([email protected]) joined #gstreamer.
[09:18] ds ([email protected]) joined #gstreamer.
[09:18] somex1 ([email protected]) left irc: "Download Gaim: http://gaim.sourceforge.net/"
[09:20] jimmy_dean ([email protected]) returned to #gstreamer.
[09:20] sxpert ([email protected]) returned to #gstreamer.
[09:56] thomasvs ([email protected]) joined #gstreamer.
[09:57] Nick change: markey -> markey|afk
[10:05] aeyakovenko ([email protected]) left irc: "Download Gaim: http://gaim.sourceforge.net/"
[10:05] irc ([email protected]) left irc: Remote closed the connection
[10:18] sublett ([email protected]) joined #gstreamer.
[10:18] harshy ([email protected]) joined #gstreamer.
[10:20] jcsston ([email protected]) left irc: Read error: 110 (Connection timed out)
[10:21] sxpert_work ([email protected]) joined #gstreamer.
[10:21] sjoerd ([email protected]) left irc: "Lost terminal"
[10:23] dolphy ([email protected]) joined #gstreamer.
[10:35] jdahlin ([email protected]) joined #gstreamer.
[10:40] wheels ([email protected]) left irc: "using sirc version 2.211+KSIRC/1.3.9"
[10:46] casaxno ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[10:47] casaxno_ ([email protected]) joined #gstreamer.
[10:51] Nick change: mathrick -> mathrick|Uni
[10:56] <KoRnouille> hey all, how can I figure out if I have an oss or esd sound card ?
[10:57] <thomasvs> eh ? :)
[10:57] <thomasvs> your soundcard is unrelated to the access method
[10:57] <thomasvs> ie, esd is a sound server
[10:58] <thomasvs> oss an api/module to directly access your card
[10:58] <sublett> you mean how to figure out the default sink? (gconftool-2 -g /system/gstreamer/default/audiosink)
[10:59] <thomasvs> ah, if that's what you meant, then yes, sublett is right
[11:00] <KoRnouille> ok
[11:00] <KoRnouille> osssink
[11:00] <KoRnouille> okay, thanks
[11:01] <KoRnouille> so I do have a problem... when I decode my video with queue / mad / osssink, the video playback gets everything slow
[11:01] <KoRnouille> and I can't hear anything from the sound card
[11:01] <KoRnouille> unfortunatly, I'm unable to know if my sound card work
[11:02] <KoRnouille> it's an onboard SoundMax
[11:02] <KoRnouille> on an intel board
[11:04] <thomasvs> why are you unable to know if it works ?
[11:04] <KoRnouille> I tried to use a XMMS player to read an mp3. But it reads my 5 minutes mp3 in a few seconds (no sound of course) (like if it was fastforward)
[11:04] <thomasvs> just play an audio file only
[11:04] <thomasvs> KoRnouille: you probably odn't have an mp3 plugin for xmms
[11:05] <KoRnouille> what can xmms read nativly ?
[11:05] <KoRnouille> .wav ?
[11:07] <KoRnouille> gst-launch-0.7 filesrc location="../Louie Vega ft Anane - Nos Vida.mp3" ! mad ! osssink
[11:07] <KoRnouille> but nothing comes out
[11:09] <thomasvs> KoRnouille: try an ogg in xmms
[11:09] <thomasvs> KoRnouille: are you sure your volume is ok ?
[11:09] <KoRnouille> hummm
[11:09] <thomasvs> try gst-launch sinesrc ! osssink
[11:09] <KoRnouille> software volume control
[11:10] <KoRnouille> gst-launch sinesrc ! osssink, RUNNING pipeline ...
[11:10] <thomasvs> does it make sound ?
[11:10] <KoRnouille> thomasvs: am I supposed to get sound ?
[11:10] <KoRnouille> aah non
[11:10] <KoRnouille> I don't
[11:10] <thomasvs> heh
[11:10] <thomasvs> check your volume settings on the card
[11:11] <KoRnouille> thomasvs: where can I see that on RH9 ?
[11:11] <KoRnouille> Im not used to X stuff
[11:11] <sublett> aumix
[11:14] <KoRnouille> I have a slider there... seems to be at around 70%
[11:15] alley_cat ([email protected]) joined #gstreamer.
[11:16] <bitshifter> KoRnouille: make sure the 'PCM' slider is also > 0
[11:24] kwm ([email protected]) joined #gstreamer.
[11:30] markey|afk ([email protected]) left irc: "bbl"
[11:34] fcrozat ([email protected]) joined #gstreamer.
[11:38] <KoRnouille> bitshifter: the PCM slider ? I don't see it
[11:39] <KoRnouille> and 1st of all, the slider I have looks like this : aumix +++++++++++++++++++++o++++++<Mic
[11:39] <KoRnouille> is this slider for the Mic ?
[11:39] <KoRnouille> that's the only slider I have
[11:44] sublett ([email protected]) left irc: "Nice of you to put me on the money"
[11:44] <thomasvs> KoRnouille: make a screenshot, you're confusing us :)
[11:44] <KoRnouille> alright
[11:44] <KoRnouille> :)
[11:45] sublett ([email protected]) joined #gstreamer.
[11:46] <KoRnouille> http://aconrad.nerim.net/debug/Sans%20titre-1%20copier.gif
[11:46] <fcrozat> ok, so, it seems v4l2 build is broken..
[11:46] <thomasvs> KoRnouille: hm, is that all you have ?
[11:46] <fcrozat> gstv4l2src.h doesn't include <linux/videodev2.h>
[11:46] <KoRnouille> yes
[11:46] <thomasvs> KoRnouille: are you sure you configured sound on your machine ?
[11:47] <thomasvs> KoRnouille: have you tried redhat-config-sound ?
[11:47] <thomasvs> fcrozat: highly possible, there's only one person here who works with the v4l2 stuff
[11:47] <KoRnouille> thomasvs: I installed RH9 and never touched my sound card before
[11:47] <thomasvs> KoRnouille: ok, first get it upaand running then.
[11:47] <fcrozat> hmm, strange, it is included in gstv4l2element.h..
[11:47] <fcrozat> 2s
[11:47] <thomasvs> it looks to me like you never configured sound at all
[11:47] <thomasvs> fcrozat: is this for 0.6 or 0.7 ?
[11:48] <fcrozat> CVS HEAD
[11:48] <fcrozat> jhbuild
[11:48] <KoRnouille> well, I don't really know how to do that. I though it was kinda plug and play ?
[11:49] <KoRnouille> just install RH9, and it will do the setup job for you
[11:49] <fcrozat> oh...
[11:49] <KoRnouille> my sound card it pretty standart I belive
[11:50] <KoRnouille> ill make sure everything it setup correctly in my bios
[11:50] <thomasvs> KoRnouille: redhat-config-sound, as I said
[11:53] <thomasvs> fcrozat: and ?
[11:53] <KoRnouille> okay thomasvs, I missed that line about redhat-config-sound. I'll try right away
[11:55] <KoRnouille> hummf... that's all I have : redhat-config-proc                 redhat-update-gnome-font-install2
[11:55] <fcrozat> thomasvs: nothing atm :((
[11:55] <fcrozat> still searching..
[11:55] <fcrozat> gstv4l2src.h:55: error: field `buffer' has incomplete type
[11:56] <thomasvs> KoRnouille: ok, you need to find the package that configures your sound card first
[11:59] Shoragan ([email protected]) joined #gstreamer.
[12:00] <fcrozat> thomasvs: ok, found the problem : this struct needs __KERNEL__ define
[12:00] <fcrozat> I'm not sure this is the right structure to use..
[12:01] Action: KoRnouille is installing redhat-config-soundcard
[12:01] <thomasvs> fcrozat: best is to file a bug and mail ronald about it
[12:01] bilboed_ ([email protected]) joined #gstreamer.
[12:02] <sublett> if it needs __KERNEL__ then it's probably not what you want
[12:02] <KoRnouille> hummmf... "no sound card detected"
[12:02] <fcrozat> sublett: indeed :)
[12:03] <fcrozat> thomasvs: I'll fill a bug and add him in cc
[12:03] <jdahlin> KoRnouille: I had a look at your example yesterday
[12:03] <KoRnouille> jdahlin: ah ? :)
[12:04] <jdahlin> KoRnouille: It's not possible to write something like that at the moment with the python bindings
[12:04] <KoRnouille> uurgh.
[12:04] <jdahlin> KoRnouille: We need to support the intefaces first (specifically the XOverlay) interface
[12:04] <KoRnouille> all this work for nothing
[12:04] <KoRnouille> how can I help ?
[12:05] <KoRnouille> do you have an idea when this will be available ?
[12:05] <KoRnouille> a few weeks ? a few months ? a few years ?
[12:05] <jdahlin> hopefully within next week
[12:05] <KoRnouille> whoaaah ! cool !
[12:05] <KoRnouille> will there be more docs ?
[12:05] <KoRnouille> well, I don't wanna ask for the moon neither
[12:07] <KoRnouille> jdahlin: is there a list somewhere where I could see what available / NA with python bindings ?
[12:07] <taaz> jdahlin: copying codegen code to gst-python?  why?
[12:08] <KoRnouille> jdahlin: how does my application looks like ? is it correctly written ? (as Im a newby in programming)
[12:08] <jdahlin> taaz: I don't want to depend on pygtk cvs head
[12:08] <jdahlin> :-)
[12:08] <jdahlin> KoRnouille: sorry, not at the moment
[12:08] <taaz> well shit.  copy in gst head too
[12:08] teuf ([email protected]) joined #gstreamer.
[12:08] <jdahlin> KoRnouille: Yeah, we'll probably write some at the moment
[12:09] <jdahlin> taaz: well, thomas wants to release as soon as possible so
[12:12] <KoRnouille> jdahlin: FMI, could you explain how binding works ? Maybe you could send me a src code to see how it goes all together ? Trying to understand how to write bindings... because I'm ready for getting things moving with python...
[12:12] ensonic ([email protected]) joined #gstreamer.
[12:12] bilboed ([email protected]) left irc: Read error: 110 (Connection timed out)
[12:12] <thomasvs> KoRnouille: if you don't know C, writing bindings is a bad idea :)
[12:12] <taaz> KoRnouille: just check out the code from cvs.  that's all there is.
[12:13] <KoRnouille> thomasvs: I know how to read C now. ;)
[12:13] <KoRnouille> in which directory would this be ?
[12:14] <thomasvs> KoRnouille: trust me - you cannot know enough C in two days to be able to write bindings
[12:14] <KoRnouille> thomasvs: I know. I just wanted to see how this works.
[12:14] <KoRnouille> thomasvs: but I'm looking forward to develop a DVB pluging for gstreamer.
[12:14] <taaz> jdahlin: do you really have a plan for supporting interfaces properly in the next week?  i can only assume this will be a hack.  i think proper support depends on pygtk code doing the proper thing such as adding wrapped interfaces to __bases__ on object creation.  which seems non-trivial.
[12:15] <KoRnouille> wouldn't that be cool ? decode mpeg2 and mpeg1 in hardware
[12:15] <jdahlin> taaz: there are patches for that bugzilla...
[12:15] <KoRnouille> + read all other video on the TV out of the DVB card (full featured card of course)
[12:16] <taaz> jdahlin: oh? do they work? ;)  i'm sick of wingo telling me things like that work in guile already ;)
[12:16] <KoRnouille> would it be in gstinterfaces ?
[12:16] <jdahlin> taaz: not really
[12:19] <fcrozat> thomasvs: another build problem : speex doesn't build :(
[12:23] sjoerd ([email protected]) joined #gstreamer.
[12:33] <thomasvs> fcrozat: here it builds fine
[12:33] <thomasvs> fcrozat: it doesn't work though
[12:33] kmaraas ([email protected]) joined #gstreamer.
[12:34] <fcrozat> thomasvs: hmm, with 1.1.4 ? :)
[12:34] <thaytan> I have some speex stuff I should commit
[12:34] <thaytan> it doesn't handle the content, but it adds typefinding
[12:34] <thaytan> a work-in-progress
[12:36] <thomasvs> I wanted to hear jdub's comment about ogo yesterday, so I tired :)
[12:36] <thomasvs> fcrozat: no, 1.0.3
[12:36] <thaytan> ogo?
[12:37] <thomasvs> oGALAXYo
[12:37] <thaytan> oh
[12:37] <fcrozat> thomasvs: you should try with 1.1.4 then :)
[12:37] <thaytan> fcrozat: which distro is that?
[12:37] <fcrozat> thaytan: Mdk 10.0 :)
[12:38] <KoRnouille> jdahlin: would you please notify me when ever the bindings I want to use are available please ?
[12:39] <jdahlin> KoRnouille: don't worry, you'll see the annonucements
[12:39] <jdahlin> KoRnouille: radio, tv, newspapers etc
[12:39] <KoRnouille> SMS on my phone, okay. :)
[12:41] <KoRnouille> also, I just wanna say you guys a doing a great job. I feel that gstreamer is moving on. A new release every few weeks. That's really nice.
[12:44] markey ([email protected]) joined #gstreamer.
[12:44] <jdahlin> taaz: do you have a link to wingos guile bindings
[12:48] <thomasvs> KoRnouille: thanks, we aim to please
[12:50] <KoRnouille> thomasvs: specially you, I think your kinda motivating. Having release deadlines, etc... this gives a good image to gstreamer
[12:50] iain ([email protected]) joined #gstreamer.
[12:50] <KoRnouille> it's not just "sometime in the next week", or "soon"
[12:50] <KoRnouille> it's on next monday. :)
[12:50] <thomasvs> KoRnouille: well, it can always slip :)
[12:51] <KoRnouille> thomasvs: I know, but at least, your giving deadlines, and people know what to hang on
[12:51] <thaytan> damnit
[12:51] <KoRnouille> I can see you guys are working great
[12:51] <KoRnouille> anyway, enough
[12:52] <thaytan> and if this stupid machine would DO SOMETHING I'd help hit the deadline
[12:52] <thaytan> gah gah gah
[12:52] <thomasvs> thaytan: what's the problem ?
[12:52] <thaytan> tomorrow we're off to replace my laptop, which will help
[12:52] <thaytan> thomasvs: the problem is I need to clear space on the hard disk by burning some of the 100G of DV footage I have on here
[12:52] <thomasvs> thaytan: well, I'm planning on doing an allweekend hacking marathon
[12:52] <thaytan> but 100Gig doesn't move too well
[12:52] <thomasvs> thaytan: aw, just remove it :)
[12:52] <thaytan> s/well/quickly/
[12:53] <thaytan> thomasvs: there is a local computer charity that refurbishes donated computers, installs linux and provides them to local charities etc
[12:53] <thaytan> they just received 10 palette loads of computers, so I'm helping them sort parts most of tomorrow
[12:54] <thaytan> and buying my laptop, but Sunday has Gstreamer all over it
[12:58] <thomasvs> anyone know why apparently it's ok to link vorbisdec and osssink even though their mimetypes differ ?
[13:03] <dolphy> what's the interset debug telling ?
[13:03] mathrick|Uni ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[13:03] <thomasvs> dolphy: it seems to have a state error during nego that it ignores, strange
[13:04] Action: dolphy rebuils gst's HEAD
[13:04] <dolphy> it's been a long time :)
[13:04] <bitshifter> speaking of vorbisdec: vorbisdec.c:55-59 looks like it might need fixing rather than uncommenting
[13:27] Nick change: fcrozat -> fcrozat|lunch
[13:37] <iain> so what stuff do I need to check out of cvs to get all the useful plugins?
[13:37] <iain> gst-plugins and gst-ffmpeg?
[13:42] <thomasvs> iain: yep
[13:42] <thomasvs> jdahlin: WARNING: No declaration for: gst_data_free
[13:42] <thomasvs> WARNING: No declaration for: gst_buffer_free
[13:43] <thomasvs> jdahlin: if these are the symbols you added, could you document them please ?
[13:44] <thaytan> gst-ffmpeg is still going to need massaging to build, right?
[13:44] <thaytan> or did I miss some fixes commits?
[13:44] <iain> thomasvs: and will gst-ffmpeg work from cvs now or do I need to do stuff by hand still?
[13:58] <jdahlin> thomasvs: I think they were removed
[13:59] <thomasvs> jdahlin: you sure ?
[14:00] <thomasvs> iain: I haven't checked it yet, bbb claims it should work
[14:00] <jdahlin> thomasvs: no
[14:00] <thomasvs> I can work on it after 0.8
[14:04] <bitshifter> thomasvs, jdahlin: http://sourceforge.net/mailarchive/message.php?msg_id=7251508
[14:05] <thomasvs> bitshifter: ah, thanks
[14:05] Action: thomasvs fixes doc
[14:07] <thaytan> are we going to kill xvideosink before 0.8?
[14:08] <thaytan> or maybe make it a bin autoplugger thing that tries xvimagesink then ximagesink?
[14:15] thaytan ([email protected]) left irc: Remote closed the connection
[14:19] <KoRnouille> should I install the alsa stuff for my sound card ?
[14:20] <KoRnouille> grep sound /var/log/messages
[14:20] <KoRnouille> Mar  5 10:27:34 STARS kudzu: aliased sound-slot-1 as audio
[14:20] psyduck ([email protected]) joined #gstreamer.
[14:29] markey ([email protected]) left irc: "bbl"
[14:53] thaytan ([email protected]) joined #gstreamer.
[14:59] <thomasvs> thaytan: I'm killing it off if nobody does the autoplugger, yeah
[14:59] bforbes ([email protected]) joined #gstreamer.
[15:03] mxpxpod ([email protected]) left irc: Read error: 110 (Connection timed out)
[15:13] bforbes ([email protected]) left irc: "*sigh* I don't want to come up with a quit message..."
[15:26] <ensonic> thomasvs: why have you marked bug 134863 as an enhancement?, the code has changed and therefore the docs should reflect that
[15:28] thaytan ([email protected]) left irc: Remote closed the connection
[15:31] <ensonic> thomasvs: btw. compiling gst has become much better as compared to one month ago! (about two weeks ago we talk about the makefile issue regarding to gdkpixbuf - have you had time looking into it?)
[15:31] psyduck ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[15:32] psyduck ([email protected]) joined #gstreamer.
[15:37] Action: dolphy will debug ac3
[15:46] wingo ([email protected]) joined #gstreamer.
[15:46] <wingo> hey folkses
[15:50] kwm ([email protected]) left irc: Remote closed the connection
[15:53] markey ([email protected]) joined #gstreamer.
[15:57] <thomasvs> ensonic: that ought to have been fixed, I know I fixed it correctly
[15:58] <thomasvs> ensonic: are you sure I changed it (the bug) ?
[16:16] <wingo> check ya'll later this weekend.
[16:17] <thomasvs> wingo: some of your patches got reverted, FYI
[16:17] <iain> wingo: is there a reason why audioconvert might always output 1 channel audio unless explicitly told to out 2?
[16:18] <iain> (when doing int to float conversion)
[16:23] fcrozat|lunch ([email protected]) left irc: "Client exiting"
[16:26] <dolphy> thomasvs: check your icq
[16:35] <ensonic> thomasvs: about 134863 the bughistory looks strange, [email protected] has changed it, so thats not you ;-) maybe you can just commit and close it
[16:37] <thomasvs> dolphy: use freedesktop.org/~gstreamer/ instead
[16:37] <dolphy> thomasvs: sam
[16:37] <dolphy> same
[16:38] <thomasvs> ugh, I get internal server errors now ?
[16:38] Action: thomasvs starts hating fdo
[16:38] wingo ([email protected]) left irc: Read error: 110 (Connection timed out)
[16:39] <ensonic> thomasvs: about the gst-plugins/gdkpixbuf stuff, I have updated cvs yesterday and the issue is still there
[16:40] <dolphy> thomasvs: don't you have a direct download link ?
[16:40] <thomasvs> http://freedesktop.org/~gstreamer/src/gst-ffmpeg/
[16:41] foser ([email protected]) joined #gstreamer.
[16:41] <dolphy> thomasvs: thx
[16:41] <thomasvs> ensonic: ok, can you make a full bug report detailing the problem, in bugzilla, so I can check again ?
[16:42] <ensonic> thomasvs: I do
[16:44] mathrick ([email protected]) joined #gstreamer.
[17:03] nturner|work ([email protected]) joined #gstreamer.
[17:10] kwm ([email protected]) joined #gstreamer.
[17:10] <ensonic> thomasvs: uhm, it will take a while - I need to upgrade libtool :-(
[17:21] markey ([email protected]) left irc: "bbl"
[17:30] kmaraas ([email protected]) left irc: "Leaving"
[17:33] <iain> why does audioconvert decide it wants to use 1 channel...why????
[17:34] Action: iain uses multiple ? to show the importance of this question like leet AOL USERS!!!!
[17:35] <teuf> iain: you forgot to insult the developers and say gstreamer is crap, you get answers much faster when you don't forget to do that 
[17:36] <dolphy> ac3 works 
[17:36] Action: dolphy kicks matthias
[17:36] <sxpert_work> lol
[17:36] <jdahlin> dolphy: really?
[17:36] <iain> teuf: oh yeah...GSTREAMER SUX! UR ALL DICKS!
[17:37] <KoRnouille> hummmf... The Intel Corp.|82801EB AC'97 Audio is not supported.
[17:39] <KoRnouille> <teuf> iain: you forgot to insult the developers.... // teuf always has the right solutions
[17:40] <dolphy> iain: here is an AOL user answer : because...
[17:42] <iain> dolphy: damn, them AOL users know everything
[17:43] <KoRnouille> jdahlin: In videowall setup, I will be reading multiple files from the server. Is there a way I can reset the client's pipeline (PLAYING / NULL / PLAYING) when a new file from the server come's in to play ?
[17:43] <KoRnouille> how can I tell the client to reset it's pipeline ?
[17:43] <jdahlin> KoRnouille: don't ask me, ask someone with clues :)
[17:44] <KoRnouille> okay : someone with clues, Im asking you the same question as jdahlin
[17:44] Nick change: khalek -> kh_zZz
[17:45] <KoRnouille> thomasvs: any idea ?
[17:54] casaxno_ ([email protected]) left irc: Remote closed the connection
[18:10] markey ([email protected]) joined #gstreamer.
[18:18] <dolphy> weeeeee
[18:20] sxpert_work ([email protected]) left irc: Remote closed the connection
[18:23] Action: dolphy loves sockets now :)
[18:24] psyduck ([email protected]) left irc: "Leaving"
[18:28] sub_pop ([email protected]) left irc: "Client exiting"
[18:29] sxpert_work ([email protected]) joined #gstreamer.
[18:35] teuf ([email protected]) left irc: "Client exiting"
[18:36] mxpxpod ([email protected]) joined #gstreamer.
[18:38] pb_ (~pb@2002:5160:45ef:0:240:95ff:fe30:aa84) joined #gstreamer.
[18:39] Action: thomasvs hates packaging even more now
[18:40] sub_pop ([email protected]) joined #gstreamer.
[18:41] sxpert_work ([email protected]) left irc: "Leaving"
[18:44] <ensonic> shit, the new libtool stuff does not work :-(
[18:45] Action: ensonic is filing anoter bug
[18:46] Nick change: markey -> markey|afk
[18:47] Nick change: mathrick -> mathrick|afk
[18:47] sxpert_work ([email protected]) joined #gstreamer.
[18:48] <ensonic> uhm, just a make clean did it
[18:48] <ensonic> ciao
[18:48] ensonic ([email protected]) left #gstreamer.
[18:50] <jdahlin> thomasvs: what do I need to have installed to get xmlto working?
[18:50] <thomasvs> jdahlin: xmlto ? :)
[18:59] <taaz> hmm.  freshmeat just send me a bad-link-warning-will-remove-project-in-7-days for gst-python
[18:59] <taaz> need to get that /bindings/ dir active again soon i suppose
[19:00] <taaz> jdahlin: the /bindings/guile/ (?) dir is probably where to find guile info ;)
[19:00] <taaz> or just look at them in cvs i guess
[19:00] <jdahlin> taaz: okay
[19:00] <jdahlin> taaz: got some interface code working today btw
[19:00] <jdahlin> taaz: but it's far from ready
[19:02] sxpert_work ([email protected]) left irc: Client Quit
[19:04] <taaz> jdahlin: why did you write logging.py instead of using standard logging module?
[19:05] <jdahlin> taaz: only available in 2.3
[19:08] <taaz> a standalone package has been available for a long time.  i used it before 2.3 came out.
[19:08] sxpert_work ([email protected]) joined #gstreamer.
[19:09] <jdahlin> I'd like to depend on as little as possible for now
[19:09] <jdahlin> logging.py is only a couple of minutes work anyway
[19:11] <taaz> the 2.3 logging module is 2223 lines.  and it's cool.  i don't see why you have this thnig against using code that's maintained by someone else.
[19:12] <jdahlin> 1) dependency of 2.3 or an external package
[19:12] <jdahlin> it's really quite hard to install non standard stuff on three different build bots
[19:12] <jdahlin> with more to come
[19:12] <jdahlin> 2) It's far too complicated for what I needed
[19:13] <jdahlin> it's also quite difficult to extend if you don't know it perfectly
[19:13] <jdahlin> with my module I have control
[19:14] <jdahlin> but it doesn't matter much, or?
[19:15] <taaz> i'm not even sure how to respond.  apparently we don't think at all alike ;)
[19:15] smoke ([email protected]) joined #gstreamer.
[19:16] <jdahlin> :-)
[19:17] <taaz> 2.3 has been out for 9 months or so?  i've had debian betas installed since 2002 i think.  i don't even consider that an issue.  i'm not going to even think about 2.2 support if i want to use another 2.3 feature.
[19:17] <jdahlin> Well, FC1 still uses python 2.2
[19:17] <jdahlin> As do YD3, RH9
[19:18] <taaz> they don't have 2.3 packages?
[19:18] <jdahlin> not easily installable
[19:18] <jdahlin> I've spent 2 hours today building a separate 2.3 package for FC1
[19:18] <taaz> jeez.  i didn't realize this was even an issue.  more reason to switch to a real distro like debian ;)
[19:19] <jdahlin> Are there 2.3 packages for stable?
[19:19] <taaz> speaking of that... if it's easy to setup a build process on remote machines i can add a debian builder
[19:19] <thomasvs> taaz: it's fairly easy yeah
[19:19] <taaz> stable?  i have no idea.  stable is older than 2.3
[19:20] <jdahlin> so, then there will be issues
[19:22] <taaz> 2) logging module is not hard to use.  it takes 3 lines.  i think there's a default mode which uses less too.
[19:22] <taaz> the only reason it's a large module is it's too easy to extend so they did and have many logging methods
[19:22] Action: jdahlin ignores taaz
[19:25] <taaz> well fine.
[19:25] <taaz> where's the love?
[19:25] Action: taaz out for a bit
[19:25] <dolphy> that's the way love goes....
[19:28] <KoRnouille> jdahlin: in python, how do I setup filtered caps ? video/x-raw-rgb,width=XXX,height=YYY ?
[19:29] <jdahlin> KoRnouille: dunno, how do you do it in c?
[19:29] <KoRnouille> i dunno, I can only tell you how to do it on command line
[19:30] <KoRnouille> { queue ! mpeg2dec ! videoscale ! video/x-raw-rgb,width=XXX,height=YYY ! xvimagesink }
[19:30] <jdahlin> how is that done in C?
[19:30] Action: jdahlin looks at dolphy and thomasvs
[19:30] <KoRnouille> i dunno, thomasvs told me these were called filter caps...
[19:32] <dolphy> jdahlin: gst_element_link_filtered
[19:32] <KoRnouille> what do you put in () ?
[19:32] <jdahlin> hmm, caps is not really wrapped in python
[19:32] <jdahlin> KoRnouille: I know what you can do!
[19:32] <jdahlin> KoRnouille: file a bug
[19:32] <KoRnouille> jdahlin: I don't even know how to do that
[19:33] <KoRnouille> :/
[19:33] <jdahlin> KoRnouille: go to http://bugzilla.gnome.org
[19:33] <jdahlin> then, report a new bug (using advanced interface)
[19:33] <jdahlin> look for the GStreamer product
[19:33] <KoRnouille> gstreamer.Element.link_filtered <-- that's seems available
[19:33] <KoRnouille> gstreamer.Element.link_pads_filtered <-- that's seems available too
[19:34] <jdahlin> KoRnouille: yes, but the caps is not properly implement in python
[19:35] <ds> moo
[19:35] <ds> can we switch to optgthread as the default scheduler?
[19:36] <ds> after a particularly bloody battle with gdb last night, I decided that I'm no longer dealing with cothreads
[19:36] <jdahlin> default opt is not using threads, right?
[19:38] <KoRnouille> jdahlin: ooh, it that why I get GStreamer-CRITICAL **: file gstcaps.c: line 517 (gst_caps_is_fixed): assertion `caps != NULL' failed /// GStreamer-WARNING **: trying to set non fixed caps on pad udpsrc:src, not allowed ???
[19:38] <ds> KoRnouille: that looks like udpsrc is broken
[19:39] <KoRnouille> but that's using python.
[19:40] Nick change: mathrick|afk -> mathrick
[19:40] nturner|work ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[19:41] teuf ([email protected]) joined #gstreamer.
[19:45] walters ([email protected]) joined #gstreamer.
[19:46] <walters> word
[19:46] <walters> so 0.7.5 doesn't have that crash-on-unref bug right?
[19:46] <teuf> walters: I think it does have it
[19:47] <walters> :(
[19:47] <thomasvs> we're releasing soon anyway
[19:47] <thomasvs> walters: I think I'm going to add release-device properties to the audio sinks so rb can pause that way instead of going to NULL
[19:47] <teuf> thomasvs: still on monday ?
[19:48] <walters> thomasvs: this is to avoid the clock-resetting issue?
[19:48] <thomasvs> teuf: well :)
[19:48] <thomasvs> walters: no, to avoid you setting it to NULL :)
[19:48] <thomasvs> walters: setting it to NULL is just wrong, but I understand why you're doing it
[19:48] <walters> thomasvs: hmm.  why is it "just wrong"?
[19:48] <thomasvs> walters: and however way you put it, people blame it on gst anyways
[19:48] <thomasvs> walters: because NULL is for when you want to get rid of something.
[19:49] <thomasvs> walters: the only reason you're setting it to NULL is because gst only frees the device when going to NULL
[19:49] <thomasvs> walters: which I consider to be wrong on our part, but people seem to disagree on that.
[19:50] <thomasvs> walters: gst is doing the right thing when you unpause and it waits for x seconds.
[19:50] <walters> thomasvs: isn't the waiting due to the clocking reset?
[19:51] <teuf> walters: yep, NULL more or less means "reset everything since a new stream will begin"
[19:51] <thomasvs> walters: yeah, exactly.  so gst is doing the right thing.
[19:51] <teuf> and when you go from NULL to playing, it notices the stream doesn't start at time 0, so it waits
[19:51] <walters> thomasvs: ok then, sounds reasonable to me.
[19:51] <thomasvs> walters: the output sink is waiting on the time to catch up to where the rest of the pipeline is.
[19:52] <thomasvs> walters: but I fully realize it's very annoying if you as an app developer wants to release the device :)
[19:52] <thomasvs> walters: I'm just having a really hard time explaining to the others that this is a flaw in our design.
[19:52] <thomasvs> or, maybe they know something I don't that explains why this isn't a flaw in our design :)
[19:52] <walters> thomasvs: so this will be in 0.8.0?
[19:53] <thomasvs> walters: I'd feel really really embarassed if rhythmbox with a stable gst actually has this bug.
[19:53] <thomasvs> walters: so, as far as I'm concerned, it won't be in, whatever it takes.
[19:53] <walters> ah...i mean, your release-device property will be in 0.8.0?
[19:54] <thomasvs> yeah
[19:54] jdahlin ([email protected]) left irc: "Leaving"
[19:54] <thomasvs> unless ds or company come up with something better this weekend at least :)
[19:55] <walters> ok.  it does seem a little dirty though.
[19:55] ChrisHJW ([email protected]) joined #gstreamer.
[19:55] <walters> maybe we need a state between NULL and READY
[19:55] <thomasvs> walters: yeah, that's IMO the right solution.  or, at least, move device grabbing to between READY and PAUSED
[19:55] <walters> right
[19:56] <walters> adding a new state probably isn't something to do right before 0.8 though :)
[19:56] <thomasvs> no kidding :)
[19:56] <thomasvs> walters: well, I need support from apps developers so heads aren't getting stuck in the sand when it gets brought up :)
[19:57] <teuf> btw, did jorn complain about gst0.7 on bugzilla ? he said he had 2 major issues with it today when I asked him about it
[19:57] <thomasvs> teuf: don't think I saw any bugs filed by jorn no
[19:57] <thomasvs> but I could have overlooked them
[19:58] <walters> thomasvs: makes sense.  just ping me or teuf once it's commmitted, we'll switch over
[19:58] <teuf> thomasvs: damn jorn ;)
[20:00] Nick change: markey|afk -> markey
[20:08] walters ([email protected]) left irc: Client Quit
[20:13] <KoRnouille> thomasvs: if I ever wanna bugzilla something about gst-python, which "Component" should I choose ?
[20:14] <thomasvs> KoRnouille: it's not there ?
[20:15] <KoRnouille> I might be blind... just have a look
[20:15] Action: thomasvs keeps forgetting how to debug GST_CAPS from C
[20:17] <KoRnouille> There is "documentation, don't know, gst-editor, gst-ffmpeg, gst-player, gst-plugins, gst-rec, gstmm, gstreamer (core)"... But apparently, nothing about gst-python...
[20:18] <thomasvs> KoRnouille: ok, added
[20:18] <KoRnouille> :p thanks
[20:19] <KoRnouille> Okay, I wanna be the 1st one. ;)
[20:20] <KoRnouille> should I send stuff about not beeing able to use xvimagesink into a socket to an X application ?
[20:20] <KoRnouille> about XOverlay
[20:21] <KoRnouille> Actually, this is more a request than a bug
[20:21] <KoRnouille> so I don't know if I should pollute the bugzilla with these kinda problems
[20:22] <ds-work> what problem?
[20:22] <KoRnouille> XOverlay not working with python
[20:22] <thomasvs> KoRnouille: requests are fine
[20:22] <KoRnouille> thomasvs: okay
[20:23] <KoRnouille> thomasvs: btw, I asked before (but not sure if you were there) :
[20:23] <KoRnouille> in the videowall application, I would use the server to read many files (playlist) and send this on the client.
[20:24] <KoRnouille> but how would I reset the pipeline of the client (STATE_NULL) when the server starts a new file ? For flushing.
[20:24] Company ([email protected]) joined #gstreamer.
[20:24] <thomasvs> KoRnouille: why do you need to reset on the client ?
[20:25] <ds-work> thomasvs: I think the release-device idea is ok
[20:25] <KoRnouille> because when I read multiple mpeg1 files (mpeg2 doesn't work for me yet), the client reads the 1st file, everything's fine, but on the next file, the video get's a little crappy, and after a while, I get mpeg2dec errors, and finally a seg fault
[20:25] <ds-work> thomasvs: /me is suddenly interested in any hack that will fix the bug
[20:26] <thomasvs> ds-work: I agree, it's embarassing
[20:26] <thomasvs> and I want to spend most of my time this weekend on media-info and the bloody mpeg playback bug
[20:27] <KoRnouille> thomasvs: do you get my point ?
[20:28] <ds-work> thomasvs: try using a different scheduler
[20:28] <Company> mpeg playback bug?
[20:28] <ds-work> thomasvs: I started using basicgthread, and the problem disappeared
[20:28] m_wheels ([email protected]) joined #gstreamer.
[20:29] Nick change: m_wheels -> wheels
[20:31] Shoragan ([email protected]) left irc: "Leaving"
[20:34] <ds-work> Company: what do you think the best scheduler is currently?
[20:35] <Company> ds-work: they're all borked one way or another
[20:35] <thomasvs> ds-work: I'd rather fix the scheduler.  it looks to me like osssink just isn't getting shceduled to run, so the clock doesn't get updated
[20:35] dolphy ([email protected]) left irc: "Network down, IP Packets delivered via UPS"
[20:36] <ds-work> a first step is to get make check to pass with the various schedulers
[20:43] <Company> who is Matthias Saou?
[20:43] <Company> irc nick?
[20:47] <thomasvs> Company: thias
[20:47] <thomasvs> Company: he's not on irc much
[20:48] <Company> hm
[20:48] <thomasvs> he runs freshrpms
[20:48] <thomasvs> and we have a really hard time having anything gst-related work on his machines :)
[20:48] <Company> yeah
[20:49] <Company> that's why i need him
[20:49] <thomasvs> he just left :)
[20:49] <Company> alsa got a bugtracker ;)
[20:49] <thomasvs> Company: how angry would you be if I add hw:0 as the default device to alsasink ?
[20:50] <Company> thomasvs: i'd be quite annoyed
[20:50] <Company> ha
[20:50] Last message repeated 1 time(s).
[20:50] <Company> jdahlin
[20:50] <Company> get him in herer!
[20:50] <thomasvs> he just left too :)
[20:50] <thomasvs> Company: well, any other suggestions to make alsa work out of the box for the huge amount of people that will be sending us bugs next month on it ?
[20:51] <Company> yes
[20:51] <Company> alsa got a bugtracker
[20:51] <Company> get the bug fixed upstream :)
[20:51] <thomasvs> that only works *before* it ends up in a very widely deployed distro :)
[20:51] <Company> yes
[20:51] <thomasvs> right now we'll have to punt somehow so this doesn't bury us
[20:52] <Company> i'm afraid of starting that early applying duct tape everywhere
[20:52] <Company> loads of plugins are still full of that crap from 0.6 punting
[20:53] <Company> in fact there's a hell of a lot of spider workarounds everywhere :/
[20:54] <thomasvs> yeah, I know
[20:54] <thomasvs> but, there's nothing to be done about it.
[20:54] <thomasvs> we're supposed to release a 0.8 :)
[20:54] <Company> i'd like to have a list where all those things are documented
[20:54] <thomasvs> the best we can do is write down all the hacks we have to do this weekend, then discuss the right way of fixing them for the next cycle
[20:54] <thomasvs> our mistake was probably not to document the hacks we were adding
[20:54] <thomasvs> I've been thinking how we can filter out all those hacks and discuss them
[20:54] <Company> gst-plugins/HACKS
[20:55] <Company> CrapLog
[20:56] <ds-work> or use the comment /* HACK */
[20:56] kmaraas ([email protected]) joined #gstreamer.
[20:56] <Company> no
[20:56] <Company> i'd prefer to have all of them in one place with a good explanation of why it was done
[20:56] somex1 ([email protected]) joined #gstreamer.
[20:56] <Company> so we can easily figure out what's missing in 0.9
[20:57] <thomasvs> Company: agreed
[20:57] <Company> i'd like the ChangeLog format
[20:57] <Company> with longer explanations
[20:57] <thomasvs> let's not call it CrapLog though
[20:58] <Company> yeah
[20:58] <sxpert> thomasvs: cr@plog would be fine ;D
[20:58] <Company> --enable-segfault and CrapLog are not that nice...
[21:00] Action: ds-work thinks --enable-segfault is cute
[21:00] <ds-work> --enable-bugs
[21:05] ChrisHJW ([email protected]) left irc: Client Quit
[21:05] ChriHJW ([email protected]) joined #gstreamer.
[21:06] <Company> hm
[21:06] <Company> osssink needs a fixate function
[21:07] <thomasvs> it so does
[21:07] <ds-work> doesn't it have one?
[21:07] <thomasvs> one that prefers 44.1/stereo
[21:08] <Company> hrm
[21:08] <Company> audioconvert has one, too?
[21:09] <thomasvs> I think so, but it prefers changing the channel count
[21:09] <thomasvs> I lost track of all changes to audioconvert :)
[21:09] <thomasvs> all right !
[21:09] <Company> it doesn't here
[21:09] Action: thomasvs runs apt-get upgrade and gets 25 new packages installed, both 0.6 and 0.7
[21:10] <thomasvs> Company: what version of openjade do you have compiled ?
[21:10] <Company> openjade:I: "openjade" version "1.3.2"
[21:10] <Company> openjade:I: "OpenSP" version "1.5.1"
[21:11] <thomasvs> Company: do the docs say they won't build, or does the build fail somewhere ?
[21:11] <Company> thomasvs: haven't tried in a while
[21:11] <Company> hm
[21:11] <Company> spideridentity has no fixate function
[21:11] <ds-work> spideridentity should not have one
[21:12] <Company> is there no proxy_fixate or so?
[21:12] <ds-work> no
[21:13] <Company> because spider ! osssink (which translates to ... ! audioconvert ! spideridentity ! osssink ) fixates to audio/x-raw-int, rate=(int)44100, channels=(int)1, endianness=(int)1234, width=(int)8, depth=(int)8, signed=(boolean)true
[21:13] <Company> which is not good
[21:13] <thomasvs> Company: probably related to what that gentoo guy is having
[21:13] <thomasvs> I had something similar going wrong at first too, which is what causes his static, "but I can still hear the song somewhat"
[21:14] <ds-work> that sounds like audioconvert and spideridentity are doing the fixate
[21:14] <ds-work> instead of osssink
[21:15] <Company> yeah
[21:15] <Company> probably
[21:15] Action: thomasvs checks how buildbot is doing
[21:16] sub_pop ([email protected]) left irc: Read error: 60 (Operation timed out)
[21:21] wheels ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[21:22] <thomasvs> gst_ogg_get_pad_by_pad (ogg=0x96081d0, pad=0x95fb0cc) at gstoggdemux.c:276
[21:22] <thomasvs> 276         cur = (GstOggPad *) walk->data;
[21:22] <thomasvs> (gdb) print walk
[21:22] <thomasvs> $1 = (GSList *) 0x19
[21:22] <thomasvs> crap :)
[21:23] <ds-work> hrm
[21:24] wheels ([email protected]) joined #gstreamer.
[21:24] <ds-work> one way around the bad fixation problem is for the app to renegotiate some link near the sink, and make sure to fixate it properly
[21:25] sub_pop ([email protected]) joined #gstreamer.
[21:25] <KoRnouille> i have posted my 1st bug
[21:25] <KoRnouille> # 136334
[21:25] <thomasvs> ds-work: for simple use cases, the core should be able to do it though
[21:26] <KoRnouille> I hop it's really a bug and not a problem because the user is a newby and doesn't know how to use gstreamer
[21:27] <ds-work> thomasvs: not really.  We specifically don't handle the problem well
[21:27] Action: KoRnouille is gone
[21:27] <thomasvs> damn, my oggdemux is bogus
[21:28] mariano ([email protected]) joined #gstreamer.
[21:28] <mariano> trivial question: what's the g in gstreamer?
[21:28] <teuf> is that normal to always get cvs conflicts in the po files ?
[21:28] <ds-work> Good
[21:29] Action: ds-work wonders if it's obvious that I'm making shit up
[21:31] <thomasvs> teuf: sort of
[21:33] <Company> mariano: i think it was named GnomeStreamer once (when it still required gtk)
[21:34] <thomasvs> no, I think it was glib from the start
[21:34] <thomasvs> gnome streamer is some project that wanted to use gstreamer to do some radio apps
[21:35] sub_pop ([email protected]) left irc: Read error: 60 (Operation timed out)
[21:35] Action: ds-work amuses himself with www.schröder.de
[21:36] <mariano> ds-work: firefox does not like the ö in the url
[21:36] <mariano> (nor does the rfc, iirc...)
[21:36] <ds-work> mariano: you haven't read the news lately, have you
[21:37] <mariano> it depends on which news you are referring to, I guess
[21:37] <Company> use utf8 in irc, damnit
[21:37] <Company> it's ?
[21:37] <Misirlou> Company: That *is* UTF-8.
[21:37] <mariano> yeah ;)
[21:37] <ds-work> i18nized domain names went live recently
[21:38] <mariano> firefox did not get that memo, I guess
[21:38] <Misirlou> aew@duende:~$ host www.schröder.de
[21:38] <Misirlou>  *** invalid answer name www.schr\195\182der.de after A query for www.schröder.de
[21:38] <Misirlou> www.schröder.de does not exist, try again
[21:38] <Misirlou> ds-work: Linky! Mariano and I have bugs to file. :)
[21:39] <mariano> haha
[21:40] <Misirlou> mariano: My Firefox doesn't mind it, by the way.
[21:41] <Misirlou> ds-work: Does it lead to www.shuttle.de for you?
[21:41] <ds-work> yes
[21:42] <Misirlou> 'K, just wanted to make sure it wasn't Googling things behind my back.
[21:45] sub_pop ([email protected]) joined #gstreamer.
[21:53] sjoerd ([email protected]) got netsplit.
[21:53] markey ([email protected]) got netsplit.
[21:53] foser ([email protected]) got netsplit.
[21:53] sublett ([email protected]) got netsplit.
[21:53] ds ([email protected]) got netsplit.
[21:53] alley_cat ([email protected]) got netsplit.
[21:53] hyriand ([email protected]) got netsplit.
[21:53] rvalles ([email protected]) got netsplit.
[21:53] kwm ([email protected]) got netsplit.
[21:53] pb_ (~pb@2002:5160:45ef:0:240:95ff:fe30:aa84) got netsplit.
[21:53] wheels ([email protected]) got netsplit.
[21:53] bitshifter ([email protected]) got netsplit.
[21:53] Jara[zZ] ([email protected]) got netsplit.
[21:53] md` ([email protected]) got netsplit.
[21:53] kh_zZz ([email protected]) got netsplit.
[21:53] sxpert_work ([email protected]) got netsplit.
[21:53] mathrick ([email protected]) got netsplit.
[21:53] yippi ([email protected]) got netsplit.
[21:53] danb ([email protected]) got netsplit.
[21:53] teuf ([email protected]) got netsplit.
[21:53] sub_pop ([email protected]) got netsplit.
[21:53] bilboed_ ([email protected]) got netsplit.
[21:53] bluejay ([email protected]) got netsplit.
[21:53] Company ([email protected]) got netsplit.
[21:53] tromey ([email protected]) got netsplit.
[21:53] smoke ([email protected]) got netsplit.
[21:53] ChrisHJW_log ([email protected]) got netsplit.
[21:53] wtay-zZz ([email protected]) got netsplit.
[21:53] KoRnouille ([email protected]) got netsplit.
[21:53] aldug ([email protected]) got netsplit.
[21:53] Misirlou ([email protected]) got netsplit.
[21:53] shawarma ([email protected]) got netsplit.
[21:53] ChriHJW ([email protected]) got netsplit.
[21:53] mxpxpod ([email protected]) got netsplit.
[21:53] somex1 ([email protected]) got netsplit.
[21:53] iain ([email protected]) got netsplit.
[21:53] camh ([email protected]) got netsplit.
[21:53] bilboed__ ([email protected]) got netsplit.
[21:53] ds-work ([email protected]) got netsplit.
[21:53] taaz ([email protected]) got netsplit.
[21:53] dilinger ([email protected]) got netsplit.
[21:53] kmaraas ([email protected]) got netsplit.
[21:53] mariano ([email protected]) got netsplit.
[21:53] sxpert ([email protected]) got netsplit.
[21:53] jimmy_dean ([email protected]) got netsplit.
[21:53] lilo ([email protected]) got netsplit.
[21:54] sub_pop ([email protected]) returned to #gstreamer.
[21:54] mariano ([email protected]) returned to #gstreamer.
[21:54] wheels ([email protected]) returned to #gstreamer.
[21:54] ChriHJW ([email protected]) returned to #gstreamer.
[21:54] somex1 ([email protected]) returned to #gstreamer.
[21:54] kmaraas ([email protected]) returned to #gstreamer.
[21:54] Company ([email protected]) returned to #gstreamer.
[21:54] teuf ([email protected]) returned to #gstreamer.
[21:54] smoke ([email protected]) returned to #gstreamer.
[21:54] sxpert_work ([email protected]) returned to #gstreamer.
[21:54] pb_ (~pb@2002:5160:45ef:0:240:95ff:fe30:aa84) returned to #gstreamer.
[21:54] mxpxpod ([email protected]) returned to #gstreamer.
[21:54] markey ([email protected]) returned to #gstreamer.
[21:54] kwm ([email protected]) returned to #gstreamer.
[21:54] mathrick ([email protected]) returned to #gstreamer.
[21:54] foser ([email protected]) returned to #gstreamer.
[21:54] iain ([email protected]) returned to #gstreamer.
[21:54] sjoerd ([email protected]) returned to #gstreamer.
[21:54] bilboed_ ([email protected]) returned to #gstreamer.
[21:54] sublett ([email protected]) returned to #gstreamer.
[21:54] alley_cat ([email protected]) returned to #gstreamer.
[21:54] sxpert ([email protected]) returned to #gstreamer.
[21:54] jimmy_dean ([email protected]) returned to #gstreamer.
[21:54] ds ([email protected]) returned to #gstreamer.
[21:54] bitshifter ([email protected]) returned to #gstreamer.
[21:54] camh ([email protected]) returned to #gstreamer.
[21:54] ChrisHJW_log ([email protected]) returned to #gstreamer.
[21:54] hyriand ([email protected]) returned to #gstreamer.
[21:54] wtay-zZz ([email protected]) returned to #gstreamer.
[21:54] rvalles ([email protected]) returned to #gstreamer.
[21:54] KoRnouille ([email protected]) returned to #gstreamer.
[21:54] lilo ([email protected]) returned to #gstreamer.
[21:54] Jara[zZ] ([email protected]) returned to #gstreamer.
[21:54] yippi ([email protected]) returned to #gstreamer.
[21:54] md` ([email protected]) returned to #gstreamer.
[21:54] kh_zZz ([email protected]) returned to #gstreamer.
[21:54] bilboed__ ([email protected]) returned to #gstreamer.
[21:54] aldug ([email protected]) returned to #gstreamer.
[21:54] Misirlou ([email protected]) returned to #gstreamer.
[21:54] tromey ([email protected]) returned to #gstreamer.
[21:54] bluejay ([email protected]) returned to #gstreamer.
[21:54] danb ([email protected]) returned to #gstreamer.
[21:54] ds-work ([email protected]) returned to #gstreamer.
[21:54] dilinger ([email protected]) returned to #gstreamer.
[21:54] taaz ([email protected]) returned to #gstreamer.
[21:54] shawarma ([email protected]) returned to #gstreamer.
[22:00] <Company> ok, audioconvert works now
[22:02] <Company> i'm now gonna test cicl
[22:02] jcsston ([email protected]) joined #gstreamer.
[22:15] thomasvs ([email protected]) left irc: Read error: 60 (Operation timed out)
[22:20] sublett ([email protected]) left irc: "Nice of you to put me on the money"
[22:37] <Company> # gst-launch-0.7 filesrc location=/mnt/windows-alt/Temp/New/Avril\ Lavigne\ -\ Sk8er\ Boi\ \(Skater\ Boy\).ogg ! oggdemux ! vorbisdec ! osssink device=/dev/dsp1
[22:37] <Company> WARNING: erroneous pipeline: could not link vorbisdec0 to osssink0
[22:37] <Company>          Trying to run anyway.
[22:37] <Company> RUNNING pipeline ...
[22:37] <Company> ERROR: pipeline doesn't want to play.
[22:40] sxpert ([email protected]) left irc: "Client exiting"
[22:41] sxpert ([email protected]) joined #gstreamer.
[22:43] <Misirlou> Avril Lavigne?????
[22:43] Action: Misirlou vomits repeatedly
[22:43] <Company> it's the only ogg i have :p
[22:44] <Misirlou> Let me give you a decent one, sir.
[22:44] <Company> i like that
[22:44] Action: Company has "Pop" as his favorite music
[22:45] <Company> out of curiosity: what song would you have given me?
[22:47] <ds-work> The_White_Stripes/The_White_Stripes/01_Jimmy_the_Exploder.ogg
[22:47] <Misirlou> DJ Danger Mouse - Grey Album - 11 - Interlude.ogg :)
[22:47] <teuf> not sure any of those is more listenable than avril lavigne ;)
[22:48] <Company> i prefer avril
[22:49] <Company> though both songs don't tell me much
[22:50] thomasvs ([email protected]) joined #gstreamer.
[22:50] <Company> the song i dig most atm is Sophie Ellis-Bextor - I Won't Change You anyway - it's just that that is an mp3
[22:51] Action: Company kicks the syncmail script
[22:51] <Company> cicl is cool btw
[22:51] <ds-work> yeah
[22:52] <ds-work> using pcl and cicl is easier than whatever I did before
[22:52] <thomasvs> Company: feel free to revert that recursive patch
[22:52] <thomasvs> are we having a weekendlong hackaton btw ?
[22:52] <Company> Misirlou: dcc doesn't work here
[22:52] <ds-work> yeah, the recursive state changing thing is evil
[22:52] <Misirlou> Company: I was never sending anything ;)
[22:52] <Company> Misirlou: yeah, it was just fyi :p
[22:54] Action: Company wished wingo patched opt
[22:55] <ds-work> we need to get wingo a decent net connection
[22:55] <Company> we need to get me a job
[22:55] <ds-work> like, say, fedexing him a CD every day
[22:55] <Company> i currently hate working at Lufthansa
[22:55] <ds-work> Company: willing to move?
[22:56] <Company> a student job
[22:56] <ds-work> ah
[22:57] <ds-work> right
[22:57] <Company> well, i'd take a fulltime job, too
[22:57] <Company> if someone paid enough and i could work from here ;)
[22:58] <Company> actually the "work from here/home" thing is pretty much my goal after university anyway
[22:59] <ds-work> ARGH!
[22:59] <Company> ?
[23:00] Action: ds-work desperately needs a sound daemon
[23:00] <teuf> do you have anything against reusing a function from xine relicensed to lgpl in gstmad ? (for the parsing of the xing header)
[23:00] <ds-work> requirement: 0% suckage
[23:01] <ds-work> teuf: as long as it's documented, it's cool
[23:01] <Company> teuf is evil
[23:02] <Company> he should write an xingtag plugin for that :p
[23:02] <Company> i'll bug teuf as long as it takes to get him do it :)
[23:02] <teuf> Company: the id3 code needs to be removed from mad first for that ;)
[23:03] <Company> teuf: don't say that too loud or i'll do that right now :p
[23:03] <Company> (that would include seperating mad and id3tag into seperate directories, which is a good thing anyway)
[23:03] <teuf> doesn't sound like something to do 2 days before 0.8.0 ;)
[23:04] <Company> i've fixed 4 bugs today
[23:04] <Company> i may now introduce at least one now :p
[23:05] <ds-work> 0.8.0 is going to hurt real bad
[23:06] <Company> probably
[23:07] sublett ([email protected]) joined #gstreamer.
[23:08] <Company> somehow it's nice to be in "don't care" mode ;/
[23:08] <Misirlou> Just from my relatively outside perspective, it sounds like you guys rewrite the entire project every few days.
[23:08] <ds-work> yeah, really.  otherwise I'd be stressing and wouldn't actually get anything done
[23:08] walters ([email protected]) joined #gstreamer.
[23:09] <ds-work> Misirlou: we've, at most, rewritten 20% of gstreamer since 0.6
[23:09] <Company> we'll have rewritten 80% of the core with 0.10
[23:10] <Company> i'd say we've rewritten 40% of the core code in 0.6=>0.8
[23:10] <Misirlou> ds-work: Wow, that's quite a bit. Are you happy with the changes?
[23:11] <Company> i am (but i'm not ds-work)
[23:11] <ds-work> we've touched 40%
[23:11] <ds-work> but rewrites were gstdebug, gstcaps, parts of gstpad
[23:11] <Company> pipeline parsing
[23:12] <ds-work> scheduling
[23:12] <Company> nope
[23:12] <Company> not realkly :/
[23:12] <ds-work> opt wasn't in 0.6
[23:12] <Company> it was
[23:12] <Company> it's the default 0.6 scheduler even ;)
[23:12] Action: ds-work doesn't pay attention
[23:12] <Company> 0.6 is really long ago
[23:13] <Company> we haven't touched the interesting things yet - replugging playing pipelines
[23:13] <Company> that'll blow up left and right
[23:15] <ds-work> we need more developers :)
[23:15] <Company> the cool thing is that the .tar.bz2 of the core grew only by 2kB from 0.6.5 to 0.7.5
[23:16] <Company> we need someone to do a video editor
[23:16] <ds-work> that's because GstCaps and GstProps was huge and useless
[23:16] <Company> someone rela good
[23:16] thaytan ([email protected]) joined #gstreamer.
[23:17] <ds-work> I'd work on a video editor, but you'd have to take away my CVS access
[23:17] <ds-work> so I don't get distracted
[23:17] <Company> nope
[23:17] <Company> i need you for core hacking
[23:17] <Company> we need someone _else_ for that ;)
[23:17] <Misirlou> Wasn't that college team working on a non-linear editor?
[23:17] <Company> dunno
[23:17] <tromey> yeah, they were
[23:18] <Company> but college students != someone real good most of the time ;)
[23:18] thomasvs ([email protected]) left irc: Read error: 110 (Connection timed out)
[23:18] <tromey> I started working on one last year, but I've been distracted by other projects
[23:18] <Company> gst is a lot more stable though
[23:19] <Company> it's still far from perfect or even release quality
[23:19] Action: bitshifter wonders where company fits into his equation :P
[23:19] thomasvs ([email protected]) joined #gstreamer.
[23:19] <ds-work> Company: do you have time for a meeting in the next month or 2?
[23:19] joeforker (dholth@3ffe:bc0:8000:0:0:0:0:262b) joined #gstreamer.
[23:19] <joeforker> hi
[23:19] <joeforker> how's gstreamer doing with theora?
[23:19] <Company> ds-work: sure
[23:20] <Company> joeforker: the only theora video i have is nicely decoded
[23:20] <Company> joeforker: i haven't worked on encoders
[23:20] <ds-work> Company: we need to plan some 0.9 projects
[23:21] <ds-work> we should just show up at thomasvs's door
[23:21] <Company> heh
[23:21] <Company> we need a working 0.8 (incl. new autopluggers) first
[23:23] <ds-work> depends if we want 0.10 for gnome-2.8
[23:24] <thaytan> ds-work: do it the 21-23rd april, and I'll be there too
[23:25] <joeforker> according to the webpage, the release is 0.6.5, but according to the topic it's 7.  which?
[23:25] <ds-work> thaytan: where are you going to be?
[23:26] <ds-work> btw, I was thinking of a meeting solely focussed on core issues
[23:26] <thaytan> ds-work: at thomasvs'
[23:27] <thaytan> that's the 3 days we're there
[23:27] Action: ds-work notes that thomasvs only has one couch
[23:30] <thaytan> we should talk to the other Barcelona hackers too then, eh?
[23:32] <thomasvs> joeforker: both - one is stable, the other is unstable
[23:33] <thomasvs> I'd be hesitant to target 0.10 for gnome-2.8
[23:33] <thomasvs> I have a feeling six months is too short for us
[23:34] <ds-work> we'd just have to choose our goals carefully
[23:34] <ds-work> and not slip
[23:34] <thomasvs> well, ok, yeah
[23:35] <thomasvs> if we were to focus on a few welldefined items
[23:35] <thomasvs> but I have a feeling we need some designs this time around
[23:35] <ds-work> perhaps we should only focus on goals that don't affect the API too much
[23:35] <thomasvs> and it would be nice if we would choose our points based on the user apps and their expectations
[23:35] <thomasvs> and, it would be nice if we could focus on fixing bugs in 0.8 for at least six weeks
[23:35] <ds-work> we need feedback from 0.8 for that
[23:35] <teuf> I also think 6 months will be too short for you
[23:35] <teuf> and also agree on the bug fixing period suggested by thomasvs ;)
[23:36] <thomasvs> how about we start writing some designs based on use scenarios
[23:36] ChrisHJW ([email protected]) joined #gstreamer.
[23:36] <thomasvs> like, proof our state definitions with actual use cases
[23:36] <ds-work> our state definitions suck
[23:36] <ds-work> anyway, this is why we need a meeting
[23:37] ChriHJW ([email protected]) left irc: Read error: 110 (Connection timed out)
[23:37] steve_b ([email protected]) joined #gstreamer.
[23:39] jcsston` ([email protected]) joined #gstreamer.
[23:39] <thomasvs> ds-work: and you don't mind flying over for that again ?
[23:40] <thomasvs> or, inviting us to sf :) ?
[23:40] <ds-work> it's easier for me to go there
[23:40] <ds-work> of course, you're all welcome to come here
[23:41] <Company> imo we should target 0.10 for gnome 3
[23:41] <Company> (or earlier)
[23:41] <Company> and target rb and gst-totem on 0.8 for gnome 2.8
[23:42] <Misirlou> Company: Is the Theora decoder in HEAD?
[23:42] <Company> Misirlou: yes
[23:42] <Company> plugins/ext/theora
[23:42] <Misirlou> Awesome.
[23:43] joeforker (dholth@3ffe:bc0:8000:0:0:0:0:262b) left irc: Remote closed the connection
[23:43] <ds-work> we certainly have the option of making 0.10 almost (or entirely) compatible with 0.8
[23:43] <ds-work> i.e., not change much
[23:44] <Company> yeah
[23:44] <Company> that's an option if there are some limitations we wanna get around
[23:44] <Company> we can do a 0.10 and then do the threading stuff for 0.12
[23:45] <ds-work> imo, we're about 2 months later than we want to be in the release cycle
[23:46] <ds-work> so we can only do 1/3-2/3 as much as we'd like to
[23:47] <ds-work> one thing I definitely would want to do for a fall release is do fixating correctly
[23:50] ChrisHJW ([email protected]) left irc: Read error: 60 (Operation timed out)
[23:50] <Company> +i wanna have autoplugging, clock/sync/time stuff and src ! queue ! queue ! queue ! sink working
[23:53] <Company> i'm wondering if gnome is going to break API again some day
[23:53] jcsston ([email protected]) left irc: Read error: 110 (Connection timed out)
[23:54] <Company> and if they need a longer time between releases then
[23:55] <ds-work> probably
[23:55] <ds-work> Gtk looks like it could use an ABI bump
[23:56] <ds-work> for gtk, it's different though.  Going to 3.0 just means spending a lot of time getting rid of the cruft that has been sitting around just to keep abi stability
[23:56] kwm ([email protected]) left irc: "Leaving"
[23:57] <Company> dunno
[23:57] <Company> someone will want to do transparent windows
[23:57] <Company> with xdamage and friends
[23:57] <ds-work> actually, /me wants transparent windows
[23:57] <thomasvs> gtk-directfb does it fine
[23:58] <thomasvs> probably with incompatible api changes though
[23:58] <ds-work> I'd like to put buttons on top of an ximagesink
[23:58] <ds-work> that's nto really transparent windoes, though
[23:59] <bitshifter> ds-work: I thought Gtk wanted to move to Cairo with 3.0, which to me sounds like a bit more work than just getting rid of cruft in the API 
[23:59] <Company> thomasvs: i was talking about the stuff keithp does with those bugs, not saying "the whole window is x% translucent"
[00:00] --- Sat Mar  6 2004
[00:00] <Company> and i want animations when opening menus or moving bubbles in my aqua-like scrollbars
[00:01] <Company> and i want gtkrc written in CSS
[00:02] Action: bitshifter wants to be able to turn around windows and scribble notes on their back
[00:03] Action: mathrick saw demo doing something like that
[00:03] <bitshifter> mathrick: yeah, Sun 'Looking Glass' desktop thingy
[00:04] <mathrick> bitshifter: ah, yeah, that was the name, I remembered it had something to do with Sun in it :)
[00:05] markey ([email protected]) left irc: "nite"
[00:06] <Company> isn't the current stable core 0.6.5?
[00:07] <Company> it is
[00:07] <Company> thomasvs: why doesn't ftp.gnome.org have gstreamer-0.6.5 packages?
[00:08] <thomasvs> Company: I've been trying to get jeff's attention to ask him if it's the right thing to do or not
[00:08] jcsston` ([email protected]) left irc: Read error: 110 (Connection timed out)
[00:09] <Company> thomasvs: "right thing"?
[00:09] <Company> ?
[00:10] <thomasvs> Company: if we should do that, if it won't mess up our latest release, and so on
[00:10] <Company> why would it mess up our latest release?
[00:10] <thomasvs> I don't know :) I haven't checked hte code for his install-module stuff :)
[00:11] <Company> ah, that stuff
[00:11] hallibaby ([email protected]) joined #gstreamer.
[00:11] Action: Company understands what thomasvs has been talking about
[00:12] <thomasvs> Company: I want to make sure first it's a) expected that we do so and b) it won't clobber our "latest release" symlinks and so on
[00:13] <Company> yeah, understood
[00:14] <Misirlou> ds-work: Do you have any information about those internationalized domains? I'm trying to find some information with which I can file bug reports.
[00:15] lilo ([email protected]) left irc: "Lost terminal"
[00:15] lilo ([email protected]) joined #gstreamer.
[00:15] jcsston ([email protected]) joined #gstreamer.
[00:16] <Misirlou> I am having the hardest time bringing up Google. :(((
[00:16] <bitshifter> Misirlou: RFC 3490
[00:17] <Misirlou> bitshifter: Thank you!
[00:17] <Misirlou> ds-work: Never mind, I can take it from here.
[00:22] <mathrick> g'night
[00:22] Nick change: mathrick -> mathrick|sleep
[00:25] smoke ([email protected]) got netsplit.
[00:31] smoke ([email protected]) got lost in the net-split.
[00:37] yippi ([email protected]) left irc: "Client exiting"
[00:38] Company ([email protected]) left irc: Remote closed the connection
[00:38] bitshifter ([email protected]) left irc: "bye"
[00:39] Nick change: camh -> camh|away
[00:52] kmaraas ([email protected]) left irc: Read error: 110 (Connection timed out)
[00:54] bitshifter ([email protected]) joined #gstreamer.
[00:57] hallibaby ([email protected]) left irc: "Bye bye (but visit us at http://www.froevel.de)"
[00:59] jcsston ([email protected]) left irc: Read error: 60 (Operation timed out)
[01:00] Nick change: kh_zZz -> khalek
[01:03] wtay_ ([email protected]) joined #gstreamer.
[01:03] wtay-zZz ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[01:29] <thomasvs> wtay_: ping ?
[01:31] <ds-work> what happened with the discussions with wtay?
[01:31] <thomasvs> I have some notes from that
[01:32] <thomasvs> I can write them up, but I don't know if it's useful to have random notes by themselves ?
[01:32] <ds-work> maybe
[01:37] sub_pop ([email protected]) left irc: "Client exiting"
[01:47] iain ([email protected]) left irc: Read error: 110 (Connection timed out)
[01:48] teuf ([email protected]) left irc: Remote closed the connection
[01:51] ChrisHJW_log ([email protected]) left irc: Read error: 110 (Connection timed out)
[01:52] foser ([email protected]) left irc: "[ I want to believe ]"
[02:11] pb_ (~pb@2002:5160:45ef:0:240:95ff:fe30:aa84) left irc: "Client exiting"
[02:20] jcsston ([email protected]) joined #gstreamer.
[02:28] alley_cat ([email protected]) left irc: "May the Source be with you!"
[02:28] sublett ([email protected]) left irc: "Nice of you to put me on the money"
[02:30] kmaraas ([email protected]) joined #gstreamer.
[02:33] sub_pop ([email protected]) joined #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