IRC Logs

*******************************************************************
[03:00] <ds-work> we need a queue subclass that has adaptive feedback to determine a good queue depth
[03:01] <deadchip> is decoder simply interchangable with spider when constructing a pipeline or do i need to take care of something specific?
[03:03] <deadchip> ok i see it is but when i do so, i get the above error
[03:03] <ds-work> deadchip: what pipeline?
[03:04] <Company> ds-work: how big for timebased? 3 seconds? 5?
[03:05] <Company> 2?
[03:05] <ds-work> Company: default, probably about 0.5 seconds
[03:05] <deadchip> ds-work: filesrc ! spider ! osssink
[03:05] <deadchip> testing with gst-launch actually now
[03:05] <ds-work> otherwise you end up with a pretty heavy queue for v4lsrc ! queue ! xvimagesink
[03:05] <deadchip> but the app works when using a specific decoder
[03:06] <ds-work> deadchip: what kind of file?
[03:06] <mathrick|code> ds-work: how about 0.5s of raw RGB?
[03:06] <deadchip> ds-work: tried with vorbis and mp3
[03:06] <deadchip> hmm gst-launch shows more
[03:07] <deadchip> ERROR: /pipeline0/spider0/sink_ident: Could not find media type
[03:07] <deadchip> ERROR       scheduler( 9433) gstoptimalscheduler.c(2080):gst_opt_scheduler_iterate: [GstOptScheduler@0x8099080] in error state
[03:07] <deadchip> execution ended after 1 iterations (sum 3080716000 ns, average 3080716000 ns, min 3080716000 ns, max 3080716000 ns)
[03:07] <deadchip> the plugin is installed in any case
[03:07] <deadchip> for the media type i mean (it's an mp3)
[03:07] <deadchip> oh sorry it's ogg
[03:07] <deadchip> but that's installed as well
[03:08] <deadchip> same error w/ mp3
[03:09] <Company> ds-work: 0.5 seconds is not good - I have AVIs here that are 2 seconds out of A/V sync
[03:09] <ds-work> Company: then it needs to be solved in a different way
[03:10] <ds-work> Company: in those cases, the video queue should normally be short.  the audio queue can be as long as necessary
[03:10] <Company> that won't work
[03:10] <Company> at least not if the video is ahead
[03:11] <ds-work> ok, right
[03:11] <ds-work> audio is usually "late" when properly muxed
[03:11] <Company> the problem is that we put raw audio/video in our queue
[03:11] <ds-work> because it's assumed that video pipelines are higher latency
[03:11] <Company> those files aren'T properly muxed
[03:12] <Company> they have 500k audio chunks
[03:13] <ds-work> does avidemux push buffers in order of presentation timestamp?
[03:13] <Company> no
[03:13] <ds-work> that's what it *should* do
[03:13] <Company> that'd require buffering in avidemux...
[03:13] <ds-work> so
[03:13] Action: ds-work doesn't care
[03:14] <ds-work> qtdemux does it
[03:14] <Company> qtdemux seeks like mad
[03:14] <ds-work> that works, too
[03:14] <Company> not with gnomevfssrc location=http://
[03:15] <ds-work> put a cache element in front
[03:15] <Company> in that case, it works
[03:15] <ds-work> besides, avi's aren't always streamable
[03:15] <ds-work> arguably, http:// should support seeking
[03:15] <Company> it does
[03:15] <Company> it resyncs on the other end
[03:15] <ds-work> in gnomevfssrc?
[03:16] <Company> i hope so
[03:16] <Company> seeking should be rare
[03:16] <ds-work> maybe over http
[03:16] <Company> events should be rare
[03:16] <ds-work> but not in gstreamer
[03:16] <Company> yes, in gstreamer
[03:17] <Company> because you can't assume events are handled fast
[03:17] <ds-work> not according to wim and erik
[03:17] <ds-work> seeking is supposed to be cheap
[03:17] <Company> oh
[03:18] <Company> not according to me though
[03:18] <Company> seeking normally is cheap because you work on local files
[03:19] <ds-work> or, you use a cache element
[03:19] <Company> yes, that works for seeking backwards
[03:19] <Company> and mybe forwards, too
[03:20] <Company> but it kills you for big amounts of data (which AVIs are) because you copy the whole file
[03:21] <Company> it kills you for streams and video conferences (which at least qt are) because you need to save the whole stream
[03:21] <ds-work> not necessarily
[03:22] <Company> yes, you can try to use an intelligent cache that discards data and only saves the last $x bytes
[03:22] <Company> but that way you only push the problem to a different element ;)
[03:23] <mathrick|code> nite
[03:23] <ds-work> well, you want to store 5 seconds of raw video
[03:23] <Company> no, i don't
[03:23] Nick change: mathrick|code -> mathrick|sleep
[03:23] <Company> i want a solution to the problem 
[03:23] <ds-work> the solution is to push buffers in presentation order
[03:24] <Company> yeah
[03:24] <Company> pretty much
[03:24] <Company> i'll write an object for that
[03:24] <ds-work> I wouldn't be surprised if there are AVIs out there that have audio in a completely different area of the file from video
[03:25] <Company> yeah, in that case you have to seek
[03:25] <ds-work> mplex used to mux MPEG like that :)
[03:25] <Company> *rimshot*
[03:26] <Company> you want an object that does that buffering for you
[03:26] <Company> gst_order_cache_new ()
[03:26] <Company> gst_order_cache_add_pad (cache, pad);
[03:27] <ds-work> well, either way, really
[03:27] <ds-work> I think qtdemux could decrease its reliance on seeking, really
[03:27] <Company> gboolean gst_order_cache_push_next ()
[03:27] <thaytan> Company: (diff < 0 && element->base_time < abs (diff)) reads weird to me :)
[03:28] <thaytan> I like the clocking changes from reading over them
[03:28] <thaytan> compiling now
[03:28] <Company> gboolean gst_order_cache_push (cache, pad, buffer);
[03:29] <ds-work> Company: qtdemux and avidemux could probably subclass a more generic cache element
[03:29] <Company> gst_order_cache_clear ()
[03:29] <ds-work> something like bytestream on the sink pad that says "I may need this range of bytes to decode the next second of data"
[03:29] <Company> what about mpeg?
[03:29] <ds-work> mpeg is supposed to muxed in presenation order, roughly
[03:30] <Company> does avi know the range it needs?
[03:30] <Company> ok, same goes for ogg
[03:30] <ds-work> maybe
[03:30] <ds-work> it could be calculated
[03:31] <ds-work> I'm thinking of it more like "definately drop this segment.  if I need it, I'll seek"
[03:31] <Company> i thought you just push in buffers and the cache tells you when it knows where to push
[03:31] <ds-work> Company: ah, that's thinking of it from the srcpad side
[03:31] <ds-work> Company: that's cool, too
[03:32] <Company> you only need to know if you're full there
[03:32] <Company> i'll leave that to you to figure out what's better
[03:33] <Company> and to figure out a good default for the queue values
[03:33] <ds-work> I think I like the srcpad side, but not with a subclass
[03:33] <Company> yeah, a generic element somewhere in gst-libs that does that for you
[03:33] <ds-work> if the video queue is empty, you just pull buffers until you can fill it
[03:34] <ds-work> perhaps a GObject that you add to an element, not a GstElement
[03:34] ChrisHJW ([email protected]) joined #gstreamer.
[03:34] <Company> yeah, i meant object, not element
[03:34] <Company> ChrisHJW: awake or autoconnect?
[03:35] ChrisHJW ([email protected]) left irc: Remote closed the connection
[03:35] ChrisHJW ([email protected]) joined #gstreamer.
[03:36] <Company> something seeks incorrectly in gstplay
[03:37] <Company> i need to fix x(v)imagesink to use adjust_time instead of set_time anyway
[03:37] hadley ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[03:37] <Company> but not today
[03:37] <Company> s/today/tonight/ ;)
[03:38] <Company> gnight
[03:38] ChrisHJW ([email protected]) left irc: Remote closed the connection
[03:39] ChriHJW ([email protected]) joined #gstreamer.
[03:39] <thaytan> g'night Company
[03:44] n6ugp ([email protected]) joined #gstreamer.
[03:46] n6ugp ([email protected]) left #gstreamer.
[03:53] <taaz> i think caca code is broken
[03:54] mxpxpod ([email protected]) joined #gstreamer.
[03:54] <taaz> silly clock patches messing things up ;)
[03:55] <taaz> anyone know the fix?
[03:56] <walters> so, Company rocks
[03:56] Action: walters is very very happy.
[03:56] <taaz> how so?
[03:56] <walters> taaz: gst metadata works again
[03:56] <taaz> i mean, how so specifically today ;)
[03:56] <walters> taaz: for both ogg and mp3
[03:57] <walters> it's pretty slow
[03:57] <walters> but it works
[03:57] <mxpxpod> walters: what did Company do for you? :)
[03:57] <ds-work> it was a spider caps negotiation bug :)
[04:00] <walters> ** ERROR **: file gailtreeview.c: line 4189 (get_tree_path_from_row_index): should not be reached
[04:00] <walters> aborting...
[04:00] <walters> bah
[04:00] Action: walters disables accessibility
[04:01] <walters> [0x8276498] [rb_metadata_save] rb-metadata-gst.c:411 (22:00:57): saving metadata for uri: file:///audio-video/music/Tool/Lateralus/01-tool-the_grudge-ksi.mp3
[04:01] <walters> mmmm
[04:01] <walters> but did it work?
[04:01] Company ([email protected]) left irc: Read error: 110 (Connection timed out)
[04:02] <walters> nope, it stripped the file.
[04:02] <walters> oh well, getting closer anyways i guess...
[04:02] ofer ([email protected]) joined #gstreamer.
[04:02] <ofer> heya
[04:02] <ofer> are there any mythtv / freevo like projects involving gstreamer?
[04:02] <walters> and emptied it even
[04:02] <walters> ofer: dave/dina maybe?
[04:03] <ofer> walters: isnt that just an rpm based distribution that includes mythtv?
[04:03] <ofer> you dont need gstreamer for mythtv
[04:04] <ofer> although it seems like gstreamer is pretty suited for it
[04:04] <ds-work> we're missing a few key plugins, that's all
[04:06] <taaz> like the DiskQueue one ;)
[04:06] <walters> ofer: i guess i thought it would since a gstreamer hacker is involved in the project.
[04:07] <mxpxpod> hmm, esdsink is being dumb....
[04:07] <mxpxpod> RUNNING pipeline ...
[04:07] <mxpxpod> ERROR: /pipeline0/esdsink0: not negotiated
[04:07] <mxpxpod> ERROR       scheduler( 1347) gstoptimalscheduler.c(2074):gst_opt_scheduler_iterate: [GstOptScheduler@0x10020518] in error state
[04:07] <mxpxpod> Execution ended after 3 iterations (sum 40549000 ns, average 13516333 ns, min 107000 ns, max 37638000 ns).
[04:09] <mxpxpod> anyone know what that means?
[04:09] <ds-work> yeah
[04:09] <ofer> i have a problem with the gstreamer gnomevfs input
[04:09] <ds-work> your pipeline couldn't be negotiated
[04:09] ChriHJW ([email protected]) left irc: Remote closed the connection
[04:10] <ofer> it seems to request a whole lot out of samba and eventually the samba server says "no more processes"
[04:10] <mxpxpod> ds-work: esd is running, every other app that uses it works
[04:10] <walters> ok ok so
[04:10] <walters> now that metadata sorta works
[04:10] <walters> how about you guys think about a release? :)
[04:10] <walters> so i can release
[04:10] <ds-work> mxpxpod: that's irrelevant
[04:10] <mxpxpod> ds-work: ok... so how do I get it to negotiate?
[04:11] ChriHJW ([email protected]) joined #gstreamer.
[04:11] <ds-work> mxpxpod: esdsink won't take audio formats it doesn't handle
[04:11] <ds-work> mxpxpod: what is your pipeline?
[04:11] <ds-work> walters: probably a good idea
[04:11] <ds-work> walters: we'll do it right after the ABI freeze
[04:11] <mxpxpod> ds-work: gst-launch-0.7 filesrc location="Rush - Tears.flac" ! flacdec ! "esdsink"
[04:11] <ds-work> walters: which should happen any day now
[04:12] <ds-work> mxpxpod: try adding an audioconvert or audioscale
[04:12] <mxpxpod> ds-work: how am I supposed to tell rhythmbox to do that?
[04:13] <ds-work> mxpxpod: change your output gconf key
[04:14] <taaz> can somone try something for me?  -inspect cacasink and do you have non-alpha chars at the beginning of the output?
[04:14] <ds-work> taaz: what I'd like to know is why it starts up a window
[04:14] <ds-work> Zeenix really needs to improve his code quality
[04:15] <taaz> oh... maybe that's it.  i'm doing it over ssh and it just fubars my screen ;)
[04:15] <taaz> and use the ChangeLog ;)
[04:17] <mxpxpod> ds-work: thanks for the help... there's another strange thing... I don't need the audioscale for vorbis, but I do for flac
[04:17] <taaz> hmm.  i don't think cacasink is working for basic mpeg2 es decode
[04:18] <taaz> works for aasink.  just some random blobs of color for cacsink
[04:18] <taaz> maybe it's my random guess at removing the ->clock bug line ;)
[04:27] ChrHJW ([email protected]) joined #gstreamer.
[04:29] thaytan ([email protected]) left irc: Read error: 110 (Connection timed out)
[04:40] ChrHJW ([email protected]) left irc: 
[04:40] ChrHJW ([email protected]) joined #gstreamer.
[04:47] steveb_ ([email protected]) joined #gstreamer.
[04:48] ChrHJW ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[04:57] steveb_ ([email protected]) left irc: Remote closed the connection
[05:00] A_ ([email protected]) joined #gstreamer.
[05:16] sub_pop ([email protected]) joined #gstreamer.
[05:43] zane ([email protected]) joined #gstreamer.
[05:43] zane ([email protected]) left irc: Client Quit
[05:44] zane ([email protected]) joined #gstreamer.
[05:44] zane ([email protected]) left irc: Client Quit
[05:45] zane ([email protected]) joined #gstreamer.
[05:45] zane ([email protected]) left irc: Client Quit
[06:06] hadley ([email protected]) joined #gstreamer.
[06:20] thaytan ([email protected]) joined #gstreamer.
[06:25] NFusi0n ([email protected]) joined #gstreamer.
[06:35] A_ ([email protected]) left irc: 
[06:44] thaytan ([email protected]) left irc: Read error: 110 (Connection timed out)
[06:52] steveb ([email protected]) joined #gstreamer.
[07:03] thaytan ([email protected]) joined #gstreamer.
[07:05] <thaytan> yo, all
[07:09] NFusi0n ([email protected]) left irc: Read error: 60 (Operation timed out)
[07:21] steveb ([email protected]) left irc: Read error: 110 (Connection timed out)
[07:42] wahoo ([email protected]) joined #gstreamer.
[07:47] <wahoo> I was just reading /. article and I have to say gstreamer looks really cool.  guess I will have to work on getting my sound card working.
[07:50] wahoo ([email protected]) left #gstreamer ("Client Exiting").
[07:56] hadley ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[08:03] Rayban_ ([email protected]) joined #gstreamer.
[08:03] Rayban ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[08:11] ChriHJW ([email protected]) left irc: Ping timeout: 14400 seconds
[08:27] jdahlin_ ([email protected]) joined #gstreamer.
[08:27] jdahlin ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[08:38] Mithrandir_ ([email protected]) left irc: "ChatZilla 0.9.35 [Mozilla rv:1.5/1]"
[08:39] jdahlin_ ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[08:39] natyaht ([email protected]) joined #gstreamer.
[08:40] Miko5881 ([email protected]) left irc: "ChatZilla 0.8.31 [Mozilla rv:1.4/20030928]"
[08:40] thaytan ([email protected]) left irc: Nick collision from services.
[08:41] natyaht ([email protected]) left irc: Client Quit
[08:41] natyaht ([email protected]) joined #gstreamer.
[08:41] Nick change: natyaht -> thaytan
[08:47] jdahlin_ ([email protected]) joined #gstreamer.
[08:52] somex1 ([email protected]) left #gstreamer.
[08:58] matrixise ([email protected]) left irc: Read error: 113 (No route to host)
[09:01] Uraeus ([email protected]) joined #gstreamer.
[09:01] <Uraeus> good morning :)
[09:04] mathrick|sleep ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[09:04] mathrick ([email protected]) joined #gstreamer.
[09:05] <mathrick> mornin
[09:06] <Uraeus> the commentary on slashdot seemed to be relativly mild mannered, which I guess is a good sign :)
[09:12] thaytan ([email protected]) left irc: Read error: 110 (Connection timed out)
[09:15] Action: mathrick wonders if it's so low KDE rants ratio, or does he have some filters applied to comments :)
[09:16] bitserf ([email protected]) joined #gstreamer.
[09:22] swentel ([email protected]) joined #gstreamer.
[09:22] <wheels> Uraeus: Do you still want me to post something on the dot?
[09:22] <Uraeus> wheels: would be nice if you did :)
[09:22] <wheels> Uraeus: Ok, I'll do so during a low point at work today.  :-)
[09:22] Action: wheels has a meeting first...
[09:23] <wheels> (which would qualify as a low point, but...)
[09:23] <Uraeus> only worry I have is a rehash of the 'glib is part of gtk+/no its not/yes it is/no its not' threads on slashdot :)
[09:23] <bitserf> i was wondering if anyone had any bright ideas for converting from one pixel aspect ratio into another so it "looks right" ;)
[09:23] <wheels> Uraeus: Yeah, it probably will, but I might try to head some of that off in the summary.
[09:23] <bitserf> struggling with a scaling routine at the moment
[09:24] <Uraeus> bitserf: sorry, I have no idea, and I am not sure how many of our video experts are awake yet/still
[09:25] <wheels> Uraeus: I don't know if you ever look at the headlines, but the big theme lately has been Freedesktop-style integration and so I'll probably use some verbage to make it fit in with that theme.
[09:25] <Uraeus> wheels: I do, I tend to check the dot at least a couple of times a week to see if there is some multimedia or SVG related stories :)
[09:27] ofer ([email protected]) left irc: Read error: 60 (Operation timed out)
[09:27] <mathrick> bitserf: you may look at videoscale element code
[09:28] kmaraas ([email protected]) left irc: "Leaving"
[09:29] rk ([email protected]) joined #gstreamer.
[09:29] Nick change: rk -> rk|away
[09:30] Nick change: rk|away -> rk
[09:30] <Uraeus> rk: that was quick :)
[09:31] <rk> Uraeus: Yeah, I'm very quick today :)
[09:32] <Uraeus> rk: so are you hacking on a gstreamer using project?
[09:33] <rk> Uraeus: Actually just starting
[09:33] wheels ([email protected]) left irc: "work"
[09:34] <Uraeus> rk: ah, anything specific planed?
[09:34] wheels ([email protected]) joined #gstreamer.
[09:35] <bitserf> mathriek: thanks
[09:35] <rk> Uraeus: Yes. I want to use gstreamer for nonlinear video editing somehow
[09:37] <Uraeus> rk: you probably want to talk to bilboed when he comes in then, he and some fellow students are making a nonlin editor as their final year project using gstreamer, afaik so will they base their work upon -> http://gnonlin.sf.net
[09:38] <rk> Uraeus: Thanks! I saw this project but not yet got into details.
[09:39] <wheels> Uraeus : Hmm, looking at their webcvs it doesn't seem that any of the files have been committed to in over a year...
[09:39] BBB ([email protected]) joined #gstreamer.
[09:40] <mathrick> yo BBB
[09:40] <rk> Uraeus: btw I'm writing my dimploma this year and it's topic is nonlinear video editor too :)
[09:40] <BBB> howdy
[09:40] <BBB> NLE is getting popular
[09:40] <BBB> so where's the code? ;)
[09:40] <Uraeus> wheels: I know, it was a project of one of the previous gst devs, but he abandoned it, but the french guys plan on picking it up again afaik
[09:40] <mathrick> maybe we should charge $2 for every gst NLE project out there ;)
[09:40] <BBB> Uraeus: I didn't see any code yet...
[09:41] <mathrick> it would easily make enough funds for needed HW and such :)
[09:41] <Uraeus> BBB: me neither, but I expect we will see some or they flunk their class :)
[09:41] <BBB> *g*
[09:41] <BBB> what a choice
[09:41] <wheels> Uraeus : You'd be surprised what passes someone as a final project.  I remember one guy at my Uni that just did a very simple drum machine in Java and that was acceptable...
[09:42] <Uraeus> BBB: I saw that you planed on working on some playback today. I have some troublesome wma files if you want (some drm crap in them)
[09:42] <Uraeus> wheels: he even got a good grade?
[09:42] <BBB> drm? I don't do drm ;)
[09:42] <BBB> actually one of the persons in my lab is a terrible student
[09:42] <Uraeus> BBB: well the hope was that you find a way for us to circumvent it :)
[09:42] <walters> hmm
[09:42] <wheels> (It was nothing but 4/4 time and the most fine grained stuff that you could notate was a 16th note.)
[09:42] <BBB> his introduction to the group was: "you'll pass, no matter what you do"
[09:42] <BBB> they just want to get rid of him ;)
[09:42] <walters> should mp3s really be detected as application/x-id3?
[09:43] <BBB> walters: if it's got an id3 tag, yes
[09:43] <wheels> Uraeus : yes -- and actually that wasn't the worst of them, it's just the one that I remember...
[09:43] <walters> BBB: hmm.  ok.
[09:43] <BBB> run typefind on the srcpad of id3tag to find the actual media type
[09:43] <mathrick> BBB: http://anime.swat.pl/~mathrick/[ax]arc_the_lad_-_episode_10_[BCF58409].mkv
[09:45] Action: walters is really starting to think he should push some/most of rb-metadata-gst.c into a gst lib
[09:49] teuf ([email protected]) joined #gstreamer.
[09:49] <teuf> Uraeus: your gstreamer article rocks ;)
[09:50] <BBB> nice stuff on slashdot... :)
[09:51] <BBB> mathrick: ty!
[09:51] <mathrick> BBB: you on decent connection now?
[09:51] <BBB> yup
[09:51] <BBB> is this mpeg4aac?
[09:51] <Uraeus> teuf: thanks :)
[09:51] <mathrick> BBB: how fast does it go?
[09:51] <mathrick> BBB: mkvinfo says so, typedfind too
[09:52] <BBB> ok
[09:52] <BBB> I'll check and make it work
[09:52] <mathrick> BBB: and no probs with resolving hostname?
[09:52] Action: BBB seriously needs to fix matroska typefinding
[09:52] <BBB> mathrick: no, downloads fine
[09:52] <BBB> ETA 9:36
[09:53] <mathrick> hmm, ok, so it was only Misirlou who had broken DNS then
[09:53] <mathrick> BBB: I wish it was my home connection ;)
[09:56] <BBB> well, it's not even that fast :p
[09:57] <BBB> and then again, I always wish my home connection was fast, too
[09:58] <mathrick> BBB: well. you get 100KB/s don't ya?
[09:59] <mathrick> BBB: max throughput noted was 1000+, it's 10BaseT direct link ;)
[09:59] matrixise ([email protected]) joined #gstreamer.
[10:03] <BBB> this works well
[10:03] <BBB> ok, I got it
[10:03] Action: BBB searches iain for the m4a file
[10:07] <teuf> BBB: you need any m4a file, or a specific one which causes problems ?
[10:08] <BBB> just one
[10:08] <BBB> for fixing autoplugging, which appears broken for some people
[10:08] <teuf> I can upload one
[10:08] <BBB> please :)
[10:08] <BBB> dcc is fine, too
[10:09] <teuf> not sure it works from here
[10:09] <BBB> I simply need some example files for each media type that we want to play back
[10:09] <BBB> then I'll make sure each works
[10:09] <teuf> BBB: it's being uploaded at http://cfergeau.free.fr/15%20Coda.m4a
[10:09] <BBB> can I start dl'ing it?
[10:09] <teuf> (it's copyrighted though, so don't include it in a distributed test suite ;)
[10:10] <teuf> BBB: it's not finished, but if only the beginning of the file matters, it should be enough
[10:10] <teuf> will be finished in a bunch of sconds
[10:10] <BBB> I'd rather have a full file
[10:10] <teuf> that's done
[10:10] <BBB> for proper EOS et all
[10:10] <BBB> it's not meant for a testsuite, it's really just for internal use here
[10:11] <BBB> if something's broken, one of you guys kicks me, benjamin, ds, dolphy, thomas or whoever, and we then have an example file locally so we can fix it immediately instead of having to ask for a sample file
[10:11] <BBB> and if we canot reproduce it, we can always ask for the specific problem-file
[10:11] <BBB> hm, 404...
[10:12] borgu ([email protected]) joined #gstreamer.
[10:12] <BBB> is the link correct?
[10:12] <borgu> morning
[10:12] <mathrick> hi borgu
[10:13] Action: BBB pokes teuf ;)
[10:13] <teuf> BBB: http://cfergeau.free.fr/15%20Coda.m4a
[10:13] <teuf> doesn't work ?
[10:13] <BBB> no
[10:13] <teuf> oh
[10:13] <BBB> 10:13:52 ERROR 404: Not Found
[10:13] <teuf> it went away
[10:13] <mathrick> BBB: btw, my file should be copyright-free outside od Japan ;)
[10:13] Markey- ([email protected]) joined #gstreamer.
[10:13] <Markey-> moin
[10:13] <BBB> mathrick: then it's all fine, no? ;)
[10:13] <teuf> BBB: oh, I still have http://cfergeau.free.fr/01%20Civili.m4a
[10:13] <BBB> mornin' markey-
[10:13] <mathrick> BBB: or at least, semi-legal ;)
[10:14] <BBB> that's good enough for me
[10:14] <mathrick> hi Markey-
[10:14] <BBB> as long as I don't download full movies, I guess I don't mind
[10:14] <Markey-> BBB: getting no signal out of my identity :/
[10:14] <BBB> trailers etc. are fine
[10:14] <Markey-> ;)
[10:14] <mathrick> BBB: hmm, it's not trailer :)
[10:14] <Markey-> I'm a person without identity
[10:14] <BBB> Markey-: hm, can you provide some sample code? you can email it privately
[10:14] <BBB> mathrick: even then :p
[10:15] <Markey-> BBB: ok, but I bet it'll make you cry, since there will be _stupid_ errors
[10:15] <mathrick> BBB: just make sure not to travel to Japan too soon ;P
[10:15] <Markey-> remember, /me is gst n00b
[10:15] <BBB> no worries ;)
[10:15] <BBB> mathrick: hey, I'm going to the US in a month from now
[10:15] <BBB> poor me
[10:15] <mathrick> BBB: yeah, I pity ya, d00d
[10:17] <Markey-> BBB: email address?
[10:17] <BBB> [email protected]
[10:17] <Markey-> thx
[10:22] <Markey-> sent.
[10:23] <Uraeus> BBB: http://www.linuxrising.com/files/ <- that mp3 here crash gstreamer/rhythmbox
[10:24] <Uraeus> BBB: that wma files fails in mplayer too, so I guess you should forget about it :)
[10:25] <BBB> DRM WMA is a bit hard
[10:25] <Uraeus> BBB: but the mp3 file needs gstreamer love :)
[10:25] <BBB> will look
[10:25] <BBB> company might need to help with mp3 sicne he wrote the new id3 stuff
[10:25] <BBB> (and someone needs to remove gst/id3/)
[10:25] <Uraeus> gstcacasink.c: In function `gst_cacasink_chain':
[10:25] <Uraeus> gstcacasink.c:338: error: structure has no member named `clock'
[10:25] <Uraeus> is this Company's new sync stuff not yet implemented in cacasink?
[10:27] <BBB> don't know... I'm recompiling right now so I'll get on that if I have it too
[10:27] <BBB> btw, your mp3 is a riff file, that might be a problem
[10:27] <BBB> wavparse is slightly broken, iirc
[10:27] <Uraeus> well you only get it if you have libcaca installed :)
[10:27] <BBB> I don't :p
[10:28] <Uraeus> well ok, so it is a wav test file then :)
[10:28] <BBB> I like your theme
[10:28] <BBB> (metanics.jpg)
[10:29] <Uraeus> hmm, personally I hate it and haven't worked on it for many months
[10:30] <BBB> I think it's cute
[10:30] <BBB> the icons and so
[10:30] <Markey-> can gst plugins provide a config GUI?
[10:30] Nick change: jdahlin_ -> jdahlin
[10:30] <mathrick> urk! they're soooo big
[10:31] <BBB> Markey-: slight error
[10:31] <BBB> Markey-: gst_element_link_many( m_pFilesrc, spider, m_pAudiosink, pIdentity, NULL ); <- note how you're linking and not checking return value
[10:31] <BBB> the right way to link is this:
[10:31] <BBB> gst_element_link_many( m_pFilesrc, spider, pIdentity, m_pAudiosink, NULL );
[10:31] <BBB> (i.e., swap identity and audiosink)
[10:32] <Markey-> oh!
[10:32] <Markey-> thanks.. had no clue
[10:32] <Markey-> let's try
[10:32] <Markey-> BBB: also, seeking still br0ken
[10:32] <Markey-> only works "sometimes"
[10:32] <Markey-> seemingly random
[10:32] <BBB> does that go for all file types?
[10:33] <Markey-> hmm I think so.. can try
[10:33] <BBB> it'd help if you could use --gst-debug-level=5 and catch errors related to seeking
[10:33] <BBB> if something goes wrong, it'll show up there
[10:33] <Markey-> ok
[10:34] <Uraeus> morning jdahlin
[10:34] <BBB> is he awake?
[10:34] <BBB> ;)
[10:34] <Uraeus> he changed his nick so I guess so :)
[10:34] <jdahlin> Uraeus: morning :)
[10:35] <BBB> oh, god, I'm sleepy, didn't even notice
[10:35] <BBB> :o
[10:35] <jdahlin> Uraeus: nice article btw
[10:35] <Uraeus> jdahlin: takkar :)
[10:38] <mathrick> Uraeus: ok, smartass, any pointers to online norwegian courses? :)
[10:39] <Uraeus> mathrick: actually that was swedish :)
[10:39] <jdahlin> psuedo-swedish :)
[10:39] <jdahlin> it's tackar :-D
[10:40] <Uraeus> ah right, I forgot that the swedish can't spell :)
[10:40] <jdahlin> bah!
[10:41] <mathrick> Uraeus: but anyway, do you know any?
[10:41] <mathrick> swedish is ok too, I think any scandinavian language is fine :)
[10:41] <Uraeus> not really, if Mayam comes in here soon she might not as she learned norwegian not long ago
[10:41] <Uraeus> s/not/know/
[10:42] Action: mathrick sets notice on maYam
[10:42] <mathrick> s/notice/notify/
[10:43] <BBB> she's coming to you?
[10:43] <BBB> we could really interpret that very wrongly, Uraeus... ;)
[10:43] <Uraeus> heh :) well since and Wim broke up I guess there no problem no matter who you interpret it ;)
[10:44] <BBB> well, if you interpret it wrongly, you could say that you're indeirectly keeping wim away from gstreamer :p :o
[10:44] Action: BBB is really very evil
[10:45] <BBB> gr, why does compiling take s long?
[10:45] <BBB> anyone wanna give me a quad xeon 2GHz?
[10:45] <mathrick> BBB: walters'd say: ccache :)
[10:45] <walters> and distcc :)
[10:45] <teuf> BBB: k, where should I ship it ?
[10:45] <teuf> ;)
[10:45] <BBB> teuf: to me of course ;)
[10:45] <Markey-> compiling c++ takes even longer.. my beard's already that long: <----------------------->
[10:45] <BBB> dcc it :p
[10:46] <teuf> hehe
[10:46] <BBB> walters is probably right
[10:46] <BBB> but I'm too lazy and scared to install anything that I dont' have redhat RPMs of
[10:46] <walters> BBB: you run FC1?
[10:46] <BBB> RH9, actually
[10:46] dolphy ([email protected]) joined #gstreamer.
[10:46] <BBB> didn't get to instaling FC1 yet
[10:46] <BBB> hi dolphy
[10:46] <walters> BBB: i used dag's packages
[10:47] <Uraeus> morning dolphy
[10:47] <dolphy> mornig :)
[10:47] <Uraeus> walters: I thought you where a sworn in Debian user :)
[10:47] <walters> Uraeus: you didn't know?  I got hired by Red Hat
[10:47] <BBB> wow
[10:47] <Uraeus> walters: didn't know, congrats!
[10:47] <BBB> for developing RB in there?
[10:47] <BBB> congats indeed!
[10:47] <walters> BBB: GNOME development
[10:48] <BBB> cool, sweet, superb!
[10:48] <walters> thanks :)
[10:48] <walters> i'm pretty excited
[10:48] <Markey-> BBB: now getting: ** (process:17172): CRITICAL **: file gstsearchfuncs.c: line 341 (gst_autoplug_sp): assertion `sinkcaps != NULL' failed
[10:48] <Uraeus> walters: that is great, now we have a gstreamer man on the inside of Red Hat :)
[10:48] <BBB> Markey-: I saw that error before, let me try
[10:48] <dolphy> yeah :)
[10:48] <dolphy> congrats
[10:49] swentel ([email protected]) left irc: Connection timed out
[10:49] <BBB> with what file type, Markey-?
[10:49] <BBB> I don't get that error
[10:49] <Markey-> mp3
[10:49] <BBB> hm
[10:49] <Markey-> I can try ogg
[10:49] <BBB> id3 header or not?
[10:50] <Markey-> umm guess so
[10:50] <Markey-> I'll try another
[10:50] jeroen ([email protected]) joined #gstreamer.
[10:50] <jeroen> hi
[10:50] <jeroen> any alsa users around/
[10:50] <jeroen> ?
[10:50] <BBB> Markey-: try with hexedit, I need to know so I know what pipeline it's using
[10:51] <jeroen> i'm running kernel 2.6 and i'm new to alsa, so i don't know if it's configured properly
[10:52] <jeroen> cat /proc/asound/devices gives me this:
[10:52] <jeroen>   0: [0- 0]: ctl
[10:52] <jeroen>  16: [0- 0]: digital audio playback
[10:52] <jeroen>  24: [0- 0]: digital audio capture
[10:52] <jeroen>  33:       : timer
[10:52] <Markey-> BBB: yep, id3
[10:52] <jeroen> shouldn't there be a mixer in there?
[10:52] <Uraeus> jeroen: only person I know for sure use alsa is Company since he maintains the alsa plugin, but as you can see he is not here atm
[10:52] <jeroen> Uraeus: hmm, ok
[10:52] <jeroen> Uraeus: kick-ass article btw! (it even made /. )
[10:52] <teuf> jeroen: I have 1 sequencer
[10:52] <teuf> jeroen: otherwise I have the same stuff as yours
[10:52] <BBB> damn
[10:52] <BBB> I can't find an id3 file locally
[10:53] <rk> jeroen: isn't that 'ctl' a mixer?
[10:53] <teuf> BBB: http://cfergeau.free.fr/01%20-%20Airbag.mp3 prolly have id3 in it
[10:53] <Uraeus> BBB: what do you need? a mp3 with a id3v1 tag?
[10:53] <rk> jeroen: I have no mixer in there too
[10:53] <jeroen> teuf: after inserting module "snd-seq", i know have a sequencer as well. What does it do? :)
[10:53] <teuf> jeroen: dunno ;)
[10:54] <teuf> jeroen: let's say it does some really cool stuff
[10:54] <jeroen> heh
[10:54] <jeroen> is gst-mixer part of jhbuild yet btw?
[10:54] <BBB> hm
[10:54] <BBB> works for me too
[10:54] <BBB> Markey-: can't really reproduce that...
[10:55] <BBB> jeroen: yes, in gnome-media
[10:55] <jeroen> cool
[10:55] <BBB> Markey-: id3v2, right?
[10:55] <teuf> is gnome-volume-control gst-based in gnome head ?
[10:55] <teuf> it segfaults here ;)
[10:55] <Markey-> BBB: yap.. I tried all sorts of mp3
[10:55] <Markey-> always the same
[10:56] <Uraeus> teuf: it works fine for me, but I use OSS, maybe it has alsa problems
[10:56] <jeroen> teuf: older version i had (rebuilding gnome 2.5 now) just didn't support alsa (red line through the speaker icon)
[10:56] <BBB> Markey-: do you have 'id3tag'?
[10:56] <Markey-> plugin?
[10:56] <teuf> jeroen: I wouldn't mind a red line through the speaker icon, but I'd rather if it didn't segfault on me ;)
[10:57] <jeroen> heh
[10:57] <BBB> Markey-: yes
[10:57] <Markey-> Long name:    id3 tag extractor
[10:57] <Markey-> got it
[10:57] <BBB> hm
[10:57] <BBB> odd...
[10:57] <BBB> could you send me the file?
[10:57] <BBB> id3 files play fine locally
[10:57] <Markey-> maybe a different reason, not related to mp3?
[10:57] <Markey-> something else b0rked in my code
[10:58] <BBB> could be
[10:58] <BBB> but I first want the file
[10:58] <Markey-> I guess so
[10:58] <BBB> I'll check after that
[10:58] <Markey-> ok
[10:58] <Markey-> dcc ok?
[10:58] <jeroen> does anyone what the status with totem+gstreamer is btw?
[10:58] <jeroen> it hasn't been proposed for addition to gnome 2.6
[10:58] <BBB> jeroen: we're working on it, but probably only for gnome-2.8
[10:58] thomasvz ([email protected]) left irc: Read error: 110 (Connection timed out)
[10:58] <BBB> Markey-: yes
[10:59] <teuf> ok, it no longer segfaults but can't find any mixer eleemnt ;)
[10:59] <dolphy> Company's modifications to clocking look awesome
[11:01] <Markey-> BBB: /dcc get markey
[11:02] <Uraeus> dolphy: except he didn't update libcaca plugin ;)
[11:03] <dolphy> shouldn't be hard to fix
[11:03] <dolphy> the api is quite straight :)
[11:03] lypanov ([email protected]) joined #gstreamer.
[11:03] <Uraeus> yeah, well I guss zeenix has quite some more work to do on it anyways to implement it the right way
[11:03] <lypanov> did noone submit the dot story?
[11:03] <lypanov> that sucks :(
[11:03] <Uraeus> lypanov: wheels promised to do so sometime today
[11:03] <teuf> hi lypanov
[11:03] <lypanov> aah. okay
[11:03] <lypanov> hey teuf 
[11:04] <Markey-> oyoy lypanov 
[11:04] <lypanov> long time no see :)
[11:04] <lypanov> moin Markey- 
[11:04] Nick change: Markey- -> Markov
[11:04] alley_cat ([email protected]) left irc: "May the Source be with you!"
[11:08] <Markov> getting same error with .sid tunes
[11:08] <Markov> seems not mp3 related
[11:08] <BBB> hm
[11:08] <BBB> I don't get them
[11:08] <BBB> but I can't really get wavparse to work at all...
[11:08] <Markov> not that bad, either
[11:09] <Markov> with the gst_bin_add_many() call, the argument order does not matter, right?
[11:10] <BBB> doesn't really matter, no
[11:10] <Markov> k
[11:10] <BBB> Uraeus: your file doesn't seem to playback at all...
[11:10] <Uraeus> BBB: it used to playback fine before
[11:10] <BBB> hrm
[11:10] <BBB> wavparse doesn't like it
[11:11] <Uraeus> BBB: xmms play it fine
[11:11] <BBB> the file looks fine
[11:16] Nick change: Markov -> Markey-
[11:17] Markey- ([email protected]) left irc: "bbl"
[11:18] deadchip_ ([email protected]) joined #gstreamer.
[11:19] <BBB> ah, found it
[11:19] <Uraeus> yay :)
[11:19] <BBB> wavparse should use my new rifflib and this wouldn't happen ;)
[11:19] <BBB> it doesn't handle unknown elements correct
[11:19] <Uraeus> port it over then :)
[11:19] thomasvs ([email protected]) joined #gstreamer.
[11:19] <BBB> and it also doesn't handle elements with 'non-normal' size correctly
[11:19] <BBB> i.e. it constantly loses sync
[11:19] <Uraeus> morning thomasvs
[11:20] <BBB> port it is not that easy...
[11:20] <thomasvs> morning
[11:20] Action: thomasvs feels energized to finish error stuff today
[11:20] Action: Uraeus has a cramp in his finger from scrolling the #gstreamer channel member list :)
[11:20] <dolphy> thomasvs: hehe :)
[11:21] deadchip ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[11:21] <mathrick> Uraeus: use mouse wheel :)
[11:21] <thomasvs> I have a cramp in my arm and some hurt in my back from riding into matthias
[11:22] <thomasvs> Is ds still around ?
[11:22] <Uraeus> thomasvs: please keep your gay details private ;)
[11:22] <dolphy> did you really write "into" ?
[11:22] <thomasvs> Uraeus: why are you giving me the cold shoulder ? You were a lot more open to my affection at new year
[11:22] <thomasvs> dolphy: I crashed into him, sort of
[11:22] <Uraeus> I was drunk
[11:23] <thomasvs> Uraeus: that's no excuse.  It was your true nature coming out
[11:23] <thomasvs> ds: ping ?
[11:23] <BBB> er..
[11:23] <Uraeus> ok, well Lotte gave me the cold shoulder so I had to go somewhere for comfort :)
[11:23] <thomasvs> haha :)
[11:24] <dolphy> Uraeus: you tried Lotte ?
[11:24] <dolphy> Uraeus: you viking ! :)
[11:27] <BBB> lotte?
[11:27] <BBB> one of thomas' belgian friends?
[11:27] bitserf ([email protected]) left irc: "Leaving"
[11:27] <dolphy> BBB: yeah
[11:28] <thomasvs> Uraeus: did you actually try something ?
[11:28] <thomasvs> you sneaky bastard :)
[11:29] Action: BBB smiles
[11:30] <BBB> well, Uraeus, I gotta acknowledge, belgian girls are definately better-looking thatn dutch ones imo
[11:31] Action: lypanov opens his eyes
[11:31] <lypanov> amsterdam is good
[11:31] <lypanov> but. overall. i agree
[11:31] teuf ([email protected]) left irc: Remote closed the connection
[11:31] Action: thomasvs goes to amsterdam on friday
[11:32] <lypanov> cool
[11:32] <lypanov> what for?
[11:32] <thomasvs> hey, suppose you can't fork, and you'd have to describe that as a resource error - what sort of error would it be ?
[11:32] <thomasvs> lypanov: concert
[11:32] <lypanov> cool. who?
[11:32] <lypanov> :)
[11:32] <lypanov> NotFuckableError
[11:32] jdahlin_ ([email protected]) joined #gstreamer.
[11:32] <lypanov> i mean. Fork
[11:33] <thomasvs> lypanov: Twilight Singers, a little known but fantastic rock band with a soul twist
[11:33] <lypanov> :)
[11:33] Action: lypanov ain't heard of 'em
[11:33] <lypanov> what sort of resource error?
[11:33] <lypanov> well, its just a "can't fork" resource error right?
[11:33] <deadchip_> can anyone think of any reason why libgstmad.so isn't seen by gstreamer while all other plugins are?
[11:33] <lypanov> it can be for various reasons...
[11:34] <BBB> hm, something weird is going on here
[11:34] Nick change: deadchip_ -> deadchip
[11:34] <thomasvs> yeah.  the man page says "no memory" basically.  ie, the kernel couldn't copy the task struct, or could not allocate memory for it
[11:34] <lypanov> right
[11:34] <thomasvs> deadchip: what do you mean, isn't seen ?
[11:34] <lypanov> but also if pids ran out
[11:34] <thomasvs> deadchip: does it error on registering ?
[11:34] <lypanov> or process quotas were on
[11:34] <deadchip> thomasvs: gst-register simply doesn't register it
[11:34] <lypanov> i imagine it would produce the same error
[11:34] <thomasvs> lypanov: hm, yeah, but it doesn't mention it in the man page
[11:34] <deadchip> thomasvs: not really an error, it just isn't 'seen'
[11:34] <lypanov> thomasvs: strange
[11:35] <deadchip> gst-register simply doesn't pick it up
[11:35] <thomasvs> deadchip: in which case it probably does have an error
[11:35] <thomasvs> deadchip: can you try ldd on the .so file and paste the result ?
[11:35] <deadchip> thomasvs: ok i'll try with verbosity
[11:35] <deadchip> ok
[11:35] <deadchip> Fire@x510dl-r:~$ ldd /usr/lib/gstreamer-0.7/libgstmad.so 
[11:35] <deadchip>                 libmad.so.0 => /usr/lib/libmad.so.0 (0x4001c000)
[11:35] <deadchip>         libc.so.6 => /lib/tls/libc.so.6 (0x40032000)
[11:35] <deadchip>         /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x80000000)
[11:35] <thomasvs> hm, looks ok
[11:35] <deadchip> yeah
[11:35] <thomasvs> then I can't think of a reason, you have to check the verbose log yeah
[11:35] <BBB> oh fsck
[11:35] Action: BBB hates wavparse
[11:36] <deadchip> ok
[11:36] <BBB> ds will have to fix the last bit
[11:36] <BBB> I've fixed wavparse, but capsnego still fails
[11:37] <deadchip> hey BBB
[11:37] <thomasvs> ok, I'll mark it as a TOO_LAZY error for now and think about it some more
[11:37] <BBB> hi
[11:37] <deadchip> your mixer work is great :)
[11:37] <BBB> ty :)
[11:37] <BBB> I pretty much like it, too
[11:37] <deadchip> i'm still fighting with gst in bmp, hope we'll have it by 0.9.8
[11:38] <Uraeus> deadchip: bmp?
[11:38] <deadchip> Uraeus: our gtk2 version/fork of xmms
[11:38] <Uraeus> ah right :)
[11:38] <deadchip> Uraeus: beepmp.sf.net.. i think we talked about it
[11:38] <deadchip> yeah :)
[11:38] <deadchip> we release 0.9.7 end of the week, and 0.9.8 is meant to be libglade-ized and gst-ized
[11:39] <BBB> I've tracked some bugs for them already :)
[11:39] <thomasvs> deadchip: it looks nice, but will you keep theming ?
[11:39] <thomasvs> anyone have a pipeline for me using pipefilter ?
[11:39] <deadchip> thomasvs: we will keep theming with SVG skins in the future, but we'll also be having a GTK-only-UI mode
[11:39] <thomasvs> I'm not quite sure how to use it
[11:40] <BBB> Uraeus: it works when explicitely specifying elements, but it fails horribly with spuder
[11:41] jdahlin ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[11:41] <BBB> ds needs to fix that, it's a capsnego error
[11:41] <Uraeus> ok :)
[11:42] Action: Uraeus looks outside, yesterday was 'world is a slurpy' day, today is snow pooring down again day
[11:42] <deadchip> here it's stormy and raining
[11:42] <deadchip> perfect day to stay at home and code :)
[11:43] <dolphy> today is shiny day with 17 degrees
[11:43] Action: dolphy loves barcelona
[11:43] <Uraeus> skipping work today too, but that is because I don't really have anything to do there combined with a slight cold
[11:43] Action: BBB is actually working right now
[11:44] <deadchip> is there any other plugin that can play mp3's except for mad and mpegaudio? (mpegaudio has an FPE when starting playback)
[11:46] <sxpert_work> Uraeus, hope it will be gone before june...
[11:47] <BBB> Uraeus: fixed in CVS
[11:47] <deadchip> interesting
[11:47] Action: BBB continues to the matroska file
[11:47] <deadchip> stat64("/usr/lib/gstreamer-0.7/libgstmad.so", {st_mode=S_IFREG|0755, st_size=235403, ...}) = 0
[11:47] <deadchip> rt_sigaction(SIGSEGV, {0x4005d87c, [], 0}, NULL, 8) = 0
[11:47] <deadchip> rt_sigaction(SIGSEGV, {SIG_DFL}, NULL, 8) = 0
[11:47] <Uraeus> deadchip: I think ds-work started on a new mp3 library in swfdec CVS (LGPL one), but it is not done yet
[11:47] <BBB> deadchip: ffdec_mp3?
[11:47] <deadchip> BBB: no that's libgstmad
[11:47] <BBB> is it?
[11:47] <BBB> I don't think it is
[11:48] <deadchip> BBB: uhmm what do you mean?
[11:48] <deadchip> the strace?
[11:48] <BBB> that you could use ffdec_mp3 for decoding
[11:48] <deadchip> ahh sorry
[11:48] <deadchip> you recommended it
[11:48] <BBB> instead of mad or mpegaudio
[11:48] <BBB> yes
[11:48] <deadchip> the question mark irritated me
[11:48] <deadchip> :)
[11:48] <BBB> it's LGPL
[11:48] <deadchip> ok
[11:49] <thomasvs> deadchip: hm, so let's figure out why mad segfaults
[11:49] <deadchip> yeah
[11:49] <deadchip> i'll run it trough gdb
[11:49] <thomasvs> deadchip: also, why is it segfaulting on stat ??? that makes no sense
[11:49] <deadchip> yeah it really makes no sense
[11:50] <BBB> (process:22073): GStreamer-CRITICAL **: file gstelement.c: line 1051 (gst_element_add_pad): assertion `gst_object_check_uniqueness (element->pads, GST_PAD_NAME(pad)) == TRUE' failed, whoops, I suck
[11:50] <deadchip> thomasvs: gdb is of no use it just runs trough
[11:50] <deadchip> wth
[11:50] <deadchip> yeah why would it SEGV on stat
[11:51] <deadchip> this here is even more weird imho
[11:51] <deadchip> can i paste it? it's not really a flood but it's more than the lines above
[11:52] <deadchip> or into some other channel?
[11:56] <thomasvs> yeah, just dialog me
[11:59] <thomasvs> [gst-cvs] [thomas@otto po]$ gst-launch fakesrc ! fakesink state-error=3
[11:59] <thomasvs> EXECUTION du tube en cours ...
[11:59] <thomasvs> DEBUG: thomas: emitting error_2 signal
[11:59] <thomasvs> DEBUG: thomas: source 0x9a26a28, error 0x9a28530, debug (nil)
[11:59] <thomasvs> ERROR: from element fakesink0: Echoué a changer d'état.
[11:59] <thomasvs> ERREUR: le tube refuse de s'executer.
[11:59] <thomasvs> [gst-cvs] [thomas@otto po]$
[11:59] <thomasvs> this error stuff is starting to get fun :)
[12:00] metric ([email protected]) joined #gstreamer.
[12:00] <Uraeus> heh, guess I should tranlate the error stuff to norwegian too then :)
[12:02] <thomasvs> yep yep yep
[12:02] <thomasvs> I mean - bork bork bork
[12:04] <thomasvs>  *                    2001 Dominic Ludlam <[email protected]>
[12:04] <thomasvs> never heard of this guy
[12:05] bilboed ([email protected]) joined #gstreamer.
[12:05] <bilboed> hi all
[12:06] <Uraeus> hi bilboed
[12:06] <bilboed> Uraeus : just read the article, sweet :)
[12:06] <Uraeus> bilboed: hope you didn't mind me giving you guys some PR :)
[12:06] <bilboed> not at all :)
[12:06] <bilboed> except there are two very small mistakes
[12:06] <mathrick> rk: hint: bilboed is the guy from NLE team :)
[12:07] <bilboed> we're 9 and it's a multi-media (not just audio+video) nonlinear editor :)
[12:07] <rk> mathrick: thanks
[12:07] <bilboed> so midi, once it's here, should be included in it...
[12:07] <mathrick> doesn't midi count as audio? :)
[12:07] <Uraeus> bilboed: well then by saying you where 7 I have set expections so low that 9 people should be able to easily ace it :)
[12:08] <bilboed> Uraeus: sure will
[12:08] <Uraeus> bilboed: leif was in here yesterday talking about his incoming midi plugins :)
[12:08] <rk> bilboed: What is the current status of your project?
[12:09] <bilboed> rk: 3 days away from starting dev, we've been looking through all the documentation and doing some planning/scheduling
[12:09] <BBB> matroska typefinding is done too...
[12:09] foser ([email protected]) joined #gstreamer.
[12:09] foser ([email protected]) left irc: Connection reset by peer
[12:09] <rk> bilboed: Are you going to use gnonlin for video editing?
[12:09] <BBB> audio is still broken, ds can fix that
[12:09] <BBB> :)
[12:09] Action: BBB continues to mp4
[12:09] <bilboed> rk: we're taking it over
[12:10] <bilboed> rk: i.e. we're continuing it's dev within our project
[12:10] <bilboed> making it media-agnostic too
[12:10] <Uraeus> BBB: be sure to update the TODO file for ds :)
[12:11] <rk> bilboed: Cool!
[12:11] <BBB> I'll kick him when he wakes up ;)
[12:11] <BBB> mp4 is capsnego too
[12:11] Action: BBB leaves all that for ds
[12:11] <bilboed> rk: which means first updating all the existing code to the new api
[12:11] <BBB> which is very boring ;)
[12:11] <BBB> when did you start, bilboed?
[12:11] <BBB> right after newyear?
[12:12] <rk> bilboed: I'm preparing to start writing some kind of NLE too
[12:12] jdahlin_ ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[12:12] <bilboed> BBB: beginning of december
[12:12] <bilboed> but there was a 2 week break for Xmas+NY
[12:12] <rk> bilboed: It will be my diploma project
[12:12] <bilboed> rk: us too :)
[12:13] <bilboed> rk: got more details of your project ?
[12:13] <BBB> rk: please cooperate with bilboed's team
[12:13] <BBB> we don't want ten NLE GStreamer backends
[12:13] <bilboed> :)
[12:13] <rk> bilboed: Yes, I know. Guys from here alredy introduced you :)
[12:13] <BBB> cooperation is a "good thing"(tm) in the opensource world
[12:14] <rk> bilboed: what do you think about cooperation? :)
[12:14] <bilboed> rk: cooperation ROCKS
[12:15] <BBB> that's the spirit!
[12:16] <Uraeus> BBB: btw, the mixer gui is weird atm. It starts really small so that all the sliders is not seeable unless I scroll
[12:16] <rk> bilboed: btw what is your university project deadline?
[12:16] <bilboed> but, (there always is a but), we need to check out what each of us want to do, and maybe we can have 2 projects with there own speciality but with a 75%+ common base
[12:16] <bilboed> rk: final deadline is end of jan 2005, with an intermediary one end of july 2004
[12:16] <bilboed> rk: but we can extend the dev another year
[12:17] <bilboed> and whatever, that's only the school's deadline, I haven't got the intention of stopping the dev then :)
[12:17] <rk> bilboed: My deadline is july 2004 but I'm going to continue development
[12:17] <bilboed> rk: that's the spirit
[12:19] <rk> bilboed: Actually I thought I'll be writing it alone (at least on the first stage) and planned to implement only essential NLE features
[12:20] <bilboed> rk: featurs like those in gnonlin ?
[12:20] <rk> bilboed: Yes, and some usable gui for it.
[12:21] <bilboed> well then we do have some common grounds, and we intended to have a minimal editor by july
[12:21] <bilboed> s/and/as
[12:22] ChrisHJW ([email protected]) joined #gstreamer.
[12:22] <rk> bilboed: Will it be C language?
[12:23] <bilboed> rk: yep, with glib2.0, and gtk2.0 for the GUI
[12:23] <rk> bilboed: good :)
[12:23] fsfman ([email protected]) joined #gstreamer.
[12:23] <bilboed> we'll follow the GObject model, so we can do bindings in other languages (like C++, python, ...)
[12:23] <mathrick> ChrisHJW: if i have some mkv that opens correctly, but sfv computed is other than expected, it that file good then?
[12:24] <thomasvs> can someone advise me on
[12:24] <thomasvs> #define QUARK(string) G_STMT_START{                                     \
[12:24] <thomasvs> GQuark gst_core_ ## string ## _quark (void) {                           \
[12:24] <thomasvs>   static GQuark quark;                                                  \
[12:24] <thomasvs>   if (!quark)                                                           \
[12:24] <thomasvs>     quark = g_quark_from_static_string ("gst-core-" ## string ## "-quark"); \
[12:24] <thomasvs>   return quark;                                                         \
[12:24] <thomasvs> }G_STMT_END
[12:25] <thomasvs> and how to do this right ?
[12:25] <rk> bilboed: When did you plan to put something about the project on the web?
[12:25] <thomasvs> is it at all possible like this ?
[12:25] <rk> bilboed: and where?
[12:26] <mathrick> thomasvs: looks ok, does it work?
[12:26] <thomasvs> nope
[12:26] <thomasvs> it fails when actually calling it
[12:26] <bilboed> we've got some webspace somewhere, we're finishing all the paperwork and then we'll send infos to the gstreamer team and set up a website
[12:26] <mathrick> what does gcc -E output?
[12:26] <thomasvs> it cannot create a preprocessor thingy for the from_static_string call
[12:28] <rk> bilboed: As you said it won't be NLVE only, what else will it be able to do?
[12:28] <thomasvs> gsterror.c:92:14: pasting ""gst-core-"" and "stream" does not give a valid preprocessing token
[12:28] <thomasvs> gsterror.c:92:14: pasting "stream" and ""-quark"" does not give a valid preprocessing token
[12:28] <thomasvs> these are the errors I get
[12:28] <bilboed> rk: audio and video for a start (so nothing very different), but then midi when it's there, and maybe text (subtitles, info, etc...)
[12:29] <ChrisHJW> <mathrick> ChrisHJW: if i have some mkv that opens correctly, but sfv computed is other than expected, it that file good then?  <--- how did you check the sfv, and to what value did you compare it to ?
[12:30] <mathrick> ChrisHJW: cksfv filename[SFV_SUM_HERE].mkv
[12:30] <bilboed> rk: the gui itself will be divided in two parts : the main part is media-agnostic, and there are plugins for handling specifif GLOBAL types (audio/*, video/*, midi/*, text/*)
[12:30] <BBB> Uraeus: I know, I don't exactly know how to fix it
[12:30] <BBB> I hope one of the gnome UI gurus will come with a one-liner patch
[12:31] <bilboed> rk: so the main gui part is in fact only the graphical representation of the gnonlin and gstreamer layers
[12:31] <rk> bilboed: and what about timeline?
[12:31] <Uraeus> BBB: try asking trow to take a look at it
[12:32] <bilboed> rk: timeline logic is in gnonlin, the main aspect is in the main gui part, and the graphical representation of the sequences will be handled by plugins
[12:32] <BBB> plugins for UI?
[12:33] <BBB> isn't that a bad idea?
[12:33] <thomasvs> n/m, the problem is I can't use a bare string from QUARK(domain) to use both for the function prototype and the string to be quoted
[12:33] <bilboed> BBB: only for globals types (audio/video/midi/text)
[12:34] <rk> bilboed: what these UI plugins responsible for?
[12:34] <bilboed> you don't "represent" video the same way you do audio (one is frames, the other is frequency or FFT) in a NLE
[12:35] jdahlin_ ([email protected]) joined #gstreamer.
[12:35] <BBB> hm...
[12:35] <BBB> sounds acceptable
[12:35] <mathrick> thomasvs: #define QUARK(string) quark = g_quark_from_static_string ("gst-core-" #string "-quark")
[12:35] <bilboed> if you look at a timeline in a NLE, the different sequences (audio or video) have got a graphical representation... This seems to be the only media-specific issue in the GUI
[12:35] <BBB> make sure you don't spread everything over new plugins
[12:35] <BBB> plugins generally suck
[12:36] <BBB> I try to avoid plugins everywhere except the gstreamer plugin system
[12:36] <mathrick> thomasvs: QUARK(foo) the yields: quark = g_quark_from_static_string ("gst-core-" "foo" "-quark");
[12:36] <BBB> the more plugin systems, the worse
[12:36] <bilboed> it enables easy integration of new general types, but I do understand your fear about plugins
[12:37] <mathrick> s/the/then/
[12:37] <bilboed> BBB: or maybe we'll integrate them as libraries (but then we're just playing on words really)
[12:37] <BBB> plugins sounds ok for this, I guess...
[12:38] <BBB> just make sure you don't make anything like "plugins required to make gstreamer plugins work in gnonlin"
[12:38] <BBB> if you do that, you kill the whole thing
[12:38] <BBB> nobody will use/do that
[12:38] <BBB> I made that mistake with mixers... in the end, I had to backport everything to gst-plugins from gst-rec... and now, everyone's happy ;)
[12:38] <mathrick> anyone familiar with TTL circuits?
[12:39] iain ([email protected]) joined #gstreamer.
[12:39] <iain> oh, not that many more people in here after our /.ing
[12:40] <rk> bilboed: Do you have some kind of design description written already?
[12:40] foser ([email protected]) joined #gstreamer.
[12:40] matrixise ([email protected]) left irc: Read error: 113 (No route to host)
[12:40] <thomasvs> mathrick: thanks !
[12:41] <mathrick> thomasvs: np, you will help me with auto* magic later ;)
[12:41] <thomasvs> hehe, I will
[12:42] <thomasvs> [gst-cvs] [thomas@otto po]$ gst-launch filesrc location=/tmp/kernel-source#2.4.22-1.2140.nptl_2.4.22-1.2140.nptl_i386.rpm ! typefind
[12:42] <thomasvs> EXECUTION du tube en cours ...
[12:42] <thomasvs>  
[12:42] <thomasvs> (process:3091): GStreamer-WARNING **: No error messages for domain gst-stream-error-quark
[12:42] <thomasvs> man, developing uninstalled ROCKS SO HARD
[12:44] <bilboed> rk: I'm in the middle of finishing it, it'll be in french first :(
[12:44] <bilboed> rk: but I'll translate before I leave friday for the snow
[12:46] <rk> bilboed: I started thinking about using french-english translator :)
[12:48] <bilboed> bon, well folks I've got to get back to finishing these specifications, or it'll be finished in time
[12:48] <bilboed> never be finished...
[12:50] bilboed ([email protected]) left #gstreamer.
[12:51] <thomasvs> [gst-cvs] [thomas@otto po]$ gst-launch filesrc location=/tmp/kernel-source#2.4.22-1.2140.nptl_2.4.22-1.2140.nptl_i386.rpm ! typefind
[12:51] <thomasvs> EXECUTION du tube en cours ...
[12:51] <thomasvs> ERROR: from element typefindelement0: Echoué a déterminer le type du flux.
[12:52] <BBB> qu'est-ce que c'est ca?
[12:52] <thomasvs> un erreur en francais venant du element typefind
[12:52] <thomasvs> but typefind isn't too good yet at figuring out it can't play files
[12:52] <ChrisHJW> <mathrick> ChrisHJW: cksfv filename[SFV_SUM_HERE].mkv  <--- yes, but who gave you the 2nd svf number, the one you compare it to, and how was that done ?
[12:52] <thomasvs> so we need to test it on all files
[12:53] <BBB> go ahead...
[12:53] <BBB> ;)
[12:53] <mathrick> ChrisHJW: sum is embedded in file name, ie. [ax]arc_the_lad_-_episode_01_[c1d6690b].mkv
[12:54] <ChrisHJW> mathrick : this is probably the md5 hash on the block headers, the matroska specific file ID
[12:54] <mathrick> ChrisHJW: and I compare it to sum computed by cksfv
[12:54] <ChrisHJW> mkvmerge does it
[12:54] <mathrick> ChrisHJW: for all but two, it is identical
[12:54] <ChrisHJW> yes, but you do it on the whole file, not only on the block headers
[12:54] <ChrisHJW> oh ...
[12:55] <ChrisHJW> [ax] = anime extreme group ?
[12:55] <mathrick> ChrisHJW: probably
[12:55] <iain> hmm, ardour has a new website design
[12:55] <iain> http://ardour.sourceforge.net/
[12:55] <iain> does it look familiar?
[12:55] <iain> http://marlin.sf.net
[12:56] <mathrick> ChrisHJW: I don't really know fansubber's responsible for that one ;)
[12:56] <mathrick> s/'//
[12:56] <ChrisHJW> Hmmm ..... havent talked to them in a while, no idea what hash they use .... but if it opens fine, what do you want :) ...
[12:56] <mathrick> ChrisHJW: so, you say that if mkv opens fine, it's definitely good?
[12:56] <BBB> isn't ardour the media thing from that gst-bahser on the gtk list?
[12:57] <iain> its the media thing (multitrack-HDR) from Paul Davis
[12:57] <iain> who probably is the gst-basher but I dunno
[12:58] <ChrisHJW> mathrick : not 100% sure, but as long as it plays fine, and the track headers are ok, who cares if one pixel in one frame is incorrect
[12:58] <BBB> iain: your roadmap rocks :p
[12:58] <BBB> davis is the gst-basher, yes
[12:59] <iain> okay
[12:59] <mathrick> ChrisHJW: how do i check track headers?
[12:59] <iain> he's an everything basher
[12:59] <BBB> that too
[12:59] <ChrisHJW> mathrick : if mkvinfo reports them all, they are fine
[12:59] <iain> gnome and kde are just frameworks for word processors
[12:59] <iain> blahblahblah
[12:59] <BBB> yup
[12:59] <ChrisHJW> you can also use mkvverify from alexnoe
[13:00] <mathrick> ChrisHJW: thx
[13:01] <iain> BBB: from what I remember, things only don't suck if he does them. I believe he's planning a GTK replacement that is "geared towards audio use"
[13:01] <iain> but if the screenshots for ardour are anything to go by, it'll be bloody hideous :)
[13:03] <mathrick> damn, this guy surely bashes every HIG available out there ;)
[13:03] <BBB> ugh, cinelerra like
[13:04] <iain> apparently audio engineers can only use audio software if its the same colours as their hardware
[13:04] <iain> they're a very fashion conscious bunch of people
[13:05] <foser> hehe, you guys know a bit about bashing too ;)
[13:06] <ChrisHJW> mathrick : links are http://mkvtoolnix.matroska.org ( you have to compile it for Linux i guess ) and http://alexnoe.matroska.org
[13:06] <BBB> of course, part of the jo
[13:07] <ChrisHJW> mathrick : the latter will only work on win32 AFAIK
[13:07] <iain> foser: yup:) you suck
[13:07] <mathrick> ChrisHJW: too much work for now, will investigate later ;)
[13:08] <ChrisHJW> lol
[13:09] <foser> iain: don't start with me, you don't know what evil you are unleashing
[13:09] teuf ([email protected]) joined #gstreamer.
[13:09] <mathrick> iain: you should be sentenced at least 2yr for roadmap like that ;P
[13:10] <iain> mathrick: 2 years doing what?
[13:10] <mathrick> iain: in prison
[13:10] Action: teuf loves iain roadmap
[13:11] <mathrick> it causes me to drop everything i'm doing and laugh :)
[13:11] <BBB> it should be updated
[13:12] <BBB> the blue dot is before oct2003
[13:12] <BBB> and it's really 2004 by now
[13:12] <iain> yeah, I have a new version somewhere
[13:12] <iain> but I might have lost it
[13:13] <BBB> :(
[13:13] Nick change: fsfman -> matrixise
[13:15] Tomat ([email protected]) joined #gstreamer.
[13:16] <Uraeus> iain: the marlin website looks nice now, much better than the last time I looked :)
[13:19] <iain> Uraeus: wow really? its been like that for months
[13:19] Company ([email protected]) joined #gstreamer.
[13:21] Action: Company goes looking how famous he is now
[13:21] <Tomat> hiya
[13:21] <Uraeus> hi Tomat
[13:21] <Tomat> i came here via slashdot - congrats for getting on the front page
[13:21] <Uraeus> Tomat: thanks :)
[13:22] <Tomat> i have a rather specialised videoediting need
[13:22] <Tomat> i'm a community leader for a group of deaf students
[13:22] <Company> i assume it's some evil slashdot/osnews conspiracy that gets stories from there to the front page of slashdot ;)
[13:23] <Uraeus> Tomat: sounds interesting, so what are you looking for?
[13:23] <Tomat> we campaigned for and won the right to present our papers in sign lanuage not english
[13:23] <Tomat> but now we are looking for video editors that will let us edit our sign lanuage paers
[13:23] <Tomat> papers
[13:23] <Tomat> premiere et al are way too complicated
[13:24] <Tomat> i can use premiere but other deaf students can't.
[13:24] <Uraeus> Tomat: hmm, sorry to say that I don't think we have some ready for you just yet :(
[13:24] <BBB> we don't have a usable video editor yet...
[13:24] <Tomat> yup i know that :)
[13:24] <Uraeus> Tomat: but the nonlin guys just said they planned on having a basic editor ready by june/july
[13:24] <Tomat> just wondering if u have any ideas
[13:24] <Uraeus> could be that this basic solution will suit your needs
[13:25] <Uraeus> the GUI will be gtk2 so it will also support all the SUN accessibility stuff
[13:25] <Tomat> also, cos its so specialised, if there is nothing out there, i can start pushing for fundraising to create an open source app that meets out needs
[13:25] <thomasvs> Tomat: that's an awesome application
[13:25] <thomasvs> Tomat: it should definately allow you to raise funds for it
[13:25] <thomasvs> Tomat: and maybe you could find someone here willing to do it too if you have the funds raised
[13:25] <mathrick> maybe creating subtiltles type plugin to process english -> signs is a good idea?
[13:26] <Tomat> we need things like embedded links to other papers (whether video or text) and references , cos rememebr this is for academic use
[13:26] <thomasvs> mathrick: man, that would be awesome too
[13:26] <Uraeus> SMIL should help with that
[13:26] <Tomat> i've not seen any video editor capable of doing that
[13:26] <thomasvs> Tomat: is there anything sign-language-like as ap roject out there ?
[13:26] <thomasvs> Tomat: for example, standard graphics used to do sign language ?
[13:27] <Tomat> mmm I'm in the UK, i'm aware of a project in Holland, but thats all
[13:27] <iain> http://www.osnews.com/comment.php?news_id=5648&offset=30&rows=40#187300
[13:27] <iain> thats the best comment on osnews
[13:27] <iain> paraphrasing..."So, ummm, its another sound server?"
[13:27] <Tomat> erm, yes, but graphics to represent SL are mainly devloped by hearing people, and are not really accpetable to deaf people 
[13:27] <BBB> *g*
[13:27] walters ([email protected]) left irc: Remote closed the connection
[13:27] <Tomat> SMIL?
[13:28] <thomasvs> Tomat: hehe :) well, it would be sweet to have a sign language overlay plugin, but we would need the input graphics for that to work
[13:28] <thomasvs> Uraeus: btw, this would be a perfect SVG application too !
[13:29] <Tomat> what do u mean by overlay ? do u mean as in translation of the audio? that's not what i'm looking for at all.
[13:29] <Uraeus> Tomat: I thank about SMIL in the article, it is a w3c standard for adding a lot of extra info to media
[13:29] <thomasvs> Tomat: yeah, I know - I'm just running with what you gave me.
[13:29] <thomasvs> Tomat: sorry, I'll get back on topic :)
[13:30] <dolphy> mxpxpod: ping
[13:30] <dolphy> mxpxpod: can you test again your flac file with latest CVS head of flacdec
[13:30] <dolphy> mxpxpod: Company did some fixes for tagged flac files
[13:30] <iain> BBB: http://62.188.60.41/~iain/01_Civili.m4a
[13:30] <Tomat> SL interpretion is usually added as an boxed window in the corner, with a stream of someone signing.
[13:31] <thomasvs> [gst-cvs] [thomas@otto cvs]$ wc patches/gsterror.patch
[13:31] <thomasvs>    1595    5978   52616 patches/gsterror.patch
[13:31] <iain> BBB: and also gst-mixer and gnome-volume-control crash
[13:31] <thomasvs> and core is done.
[13:31] <Tomat> :)
[13:31] <thomasvs> can someone please test this patch for me ?
[13:31] <dolphy> mxpxpod: ah you did already it seems :)
[13:31] <thomasvs> I want to make sure it compiles properly an a clean one
[13:31] <Company> dolphy: he already did that - i did the fixes together with him :)
[13:31] <dolphy> mxpxpod: the bug is closed
[13:31] <dolphy> Company: yeah just seen the closed bug
[13:32] <Tomat> mmm reading up on SMIL on the W3.org page..  must look at some of these apps
[13:33] <Uraeus> Tomat: you know of any freely available sources with sign language images?
[13:33] <thomasvs> Company: GST_ELEMENT_ERROR is used in the element state enum, does it matter if I change it so I can reuse GST_ELEMENT_ERROR ?
[13:33] <Tomat> Uraeus, what do u mean by images, do u mean photos, or videos, or drawings?
[13:33] <Uraeus> Tomat: drawings
[13:34] <BBB> iain: how do they crash? please provide backtraces (or rebuild or so)
[13:34] <BBB> works fine here
[13:35] <iain> gst-mixer crashes
[13:35] <iain> #3  0x0804a907 in create_mixer_collection (notebook=0x1) at mixer.c:311
[13:35] <iain> #4  0x0804ac2b in main (argc=1, argv=0xbffffa44) at mixer.c:457
[13:35] <iain> which isn't a good sign
[13:35] <Tomat> just2let u know, the hands only carry about 50% of the info, the rest is in face, body postion, subtle details of face expression, and grammar is embedded in how you flow from one sign to another, so any format that is not based on viewing a human signer will loose a lot of information
[13:36] <Tomat> will have a look anyway :)
[13:36] <Company> thomasvs: i think it'd be better if GST_ELEMENT_ERROR stayed as it is
[13:36] <Uraeus> Tomat: oh ok, didn't know that. Cause I was considering to offer to get such a thing started, in a similar vein to the flag gallery I spearheaded
[13:36] <thomasvs> Company: why ?
[13:36] <Company> thomasvs: what do you want to use it for and how do you propose we change this enum name?
[13:36] <iain> BBB: and gnome-volume-control is
[13:36] <iain> 0x0804ae09 in create_mixer_collection (notebook=0x822bf28) at mixer.c:344
[13:36] <iain> 344         for (n = 0; n < array->n_values; n++) {
[13:36] <iain> (gdb) bt
[13:36] <iain> #0  0x0804ae09 in create_mixer_collection (notebook=0x822bf28) at mixer.c:344
[13:36] <iain> #1  0x0804b166 in main (argc=1, argv=0xbffffa24) at mixer.c:503
[13:37] <teuf> oh, I've got the same one
[13:37] <deadchip> i'm confused as how to make volume control work... how do i attach a GstDParamManager to an output sink? simply link into the pipeline?
[13:37] <teuf> I added a if (array == NULL) goto next; right before this line
[13:37] <BBB> iain: 'print array'?
[13:37] <thomasvs> Company: if the enum is GstElementFlag then IMO the enum values should be GST_ELEMENT_FLAG_*
[13:37] <Tomat> Uraeus: yeah its a subtle area - a lot of people fall over on that - graphic representation of SL is something that looks easy but is bloody hard, much like AI :)
[13:37] Action: BBB guesses array is NULL... oops
[13:38] <thomasvs> Company: btw, in     if (! gst_alsa_start_audio (this)) {
[13:38] <thomasvs>       gst_element_error (GST_ELEMENT (this), "Probed format doesn't work");
[13:38] <thomasvs>       return GST_PAD_LINK_REFUSED;
[13:38] <BBB> and what element delivers the array?
[13:38] <thomasvs> what was the format probed from ?
[13:38] <Tomat> But thanks for yr offer
[13:38] <iain> BBB: (gdb) print array
[13:38] <iain> No symbol "array" in current context.
[13:38] <iain> grrr
[13:38] <BBB> :(
[13:38] <BBB> frame X before that
[13:38] <teuf> BBB: it was NULL here (I added a printf)
[13:38] <thomasvs> deadchip: I think so yeah - but there should be sample apps
[13:38] ChrisHJW ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[13:39] <thomasvs> Company: as for why, because I want to use GST_ELEMENT_ERROR as the macro
[13:39] ChrisHJW ([email protected]) joined #gstreamer.
[13:39] <deadchip> ok i'll try to find some code
[13:39] <iain> BBB: still same
[13:39] <BBB> uhm
[13:39] <BBB> ok
[13:39] <BBB> anyway, it's a bug, yes
[13:39] <BBB> sorry
[13:39] <BBB> ...
[13:40] <Tomat> OK umm we have students trying to do thair papers right now, can u reccomend anything? People have said Mac imovie, so thats kinda the bar, but windows/linux suggestions r welcome too. 
[13:40] <Company> thomasvs: hm, in that case you could rename the enum value to GST_ELEMENT_IN_ERROR - but I'd prefer to make the macro lowercase
[13:40] <BBB> iain: I'll commit a fix
[13:40] <Uraeus> Tomat: looked at Kino? not sure where they are exactly at, but it might be worth a shot
[13:40] <BBB> weird I never saw this before
[13:40] <Tomat> And thnks for giving me some ideas re SMIL - will keep an eye on that 
[13:40] <Company> thomasvs: because it should be treated just like a function
[13:41] <thomasvs> Company: hm - we have lots of function macros in uppercase no ?
[13:41] <Uraeus> Tomat: http://kino.schirmacher.de/article/static/61/
[13:41] <iain> BBB: cool
[13:42] <Uraeus> Tomat: http://kino.schirmacher.de actually :)
[13:42] <Company> thomasvs: i don't think we have that many - only the debugging statements, which are in uppercase because debugging is always uppercase...
[13:42] <BBB> iain: fixed...
[13:43] <BBB> iain: who maintains gst-mixer is dead, btw
[13:43] <BBB> I'll remove it
[13:43] <BBB> it's not gnome-volume-cotnrol
[13:43] <thomasvs> Company: ok, then I'll change to lowercase
[13:43] <thomasvs> *sigh* :)
[13:43] <Tomat> Uraeus: looking at it now ...
[13:43] <thomasvs> Company: if alsa cannot sync it's params after for example an xrun, how would you describe that error in terms of resources ?
[13:43] <thomasvs> a failed open ?
[13:43] <thomasvs> or, wrong settings ?
[13:43] sublett ([email protected]) joined #gstreamer.
[13:44] <Company> ugh
[13:44] <iain> BBB: oh, I know what the problem was...
[13:44] <thomasvs> ugh would work too
[13:44] <Company> if it cannot use the same params after an xrun that's an intternal error ;)
[13:44] <thomasvs> right
[13:44] <iain> BBB: I had my soundcard turned off
[13:44] <BBB> iain: NULL is still a valid reply
[13:44] <BBB> so...
[13:44] <BBB> I'll add the check anyway
[13:44] <BBB> actually
[13:44] <BBB> I already added it
[13:44] <iain> but now I just get 
[13:44] <Company> "internal error, please file a bug report against alsalib" :)
[13:44] <iain> ERROR            alsa( 1096) gstalsamixer.c(158):gst_alsa_mixer_init: [GstAlsaMixer@0x8237ee8] Cannot load mixer settings.
[13:45] <BBB> have fun fixing ALSA ;)
[13:45] <iain> company can do it
[13:45] Action: iain runs
[13:45] <BBB> it works for leif...
[13:45] <BBB> Company: does the mixer work for you?
[13:45] <iain> seeing as the alsa devels have me on their ignore lists
[13:45] <Uraeus> I noticed people on osnews commenting on the alsa mixer not working too
[13:45] <iain> BBB: I bet none of them have a card as fucked up as the extigy though
[13:46] <Company> iain: let's swap: you do the new autoplugger, i'll fix alsa
[13:46] <iain> Uraeus: the alsa devels seem far more interested in doing stupid things like implementing lisp interpreters than actually getting alsa to do basic things like work on the cards they claim it does
[13:46] <teuf> Here I get "Sorry, no mixer elements and/or devices found" with gnome-volume-control 
[13:47] <iain> Company: hmmmm....will the autoplugger I write have to work better than spider does?
[13:47] jdahlin__ ([email protected]) joined #gstreamer.
[13:47] <Company> BBB: no, it just says "no mixer devices found"
[13:47] <BBB> Company: and that's not correct, is it? ;)
[13:47] <Company> iain: it will have to work for every test people throw at you, including retagging and transparently handling overlay of subtitles :)
[13:47] <BBB> teuf: well, that's a normal error if it finds nothing
[13:48] <Company> BBB: no, it's not - but since it didn't crash, i never cared
[13:48] <BBB> I think it worked for Leif, though
[13:48] <teuf> BBB: yeah, except that I have a mixer somewhere ;)
[13:48] <teuf> acme used to work fine
[13:48] <BBB> iain: well, do you think you can try fixing it somewhere?
[13:48] <BBB> I don't have alsa
[13:48] <BBB> so I cannot fix it...
[13:48] <Tomat> Uraeus: kino - nice interface, but without SMIL support (ability to embedd references) it wont meet academic requirements.  mmm SMIL deffo way to go
[13:49] <Company> teuf: you use alsa or oss?
[13:49] ChriHJW ([email protected]) joined #gstreamer.
[13:49] <teuf> Company: alsa with oss emulation
[13:49] <Company> hm
[13:49] <Company> that finds all my devices with oss fine
[13:49] <Company> in gst-mixer that is
[13:50] <Tomat> Uraeus: so i have a target, how much do u think is needed to pay to have whatever is out there tweaked to meet our shopping list? of course it would then be released to benefit deaf students globally.
[13:50] Action: thomasvs tries building alsa for fedora.us
[13:50] <iain> teuf: I think most people use alsa with oss emulation
[13:50] <Company> go thomasvs go
[13:50] <iain> BBB: I'll "take a look"
[13:51] <iain> BBB: "sometime"
[13:51] <Company> alsa without oss emulation is a pain still with non-normal soundcards, isn't it?
[13:51] <BBB> iain: *g*
[13:51] <iain> Company: its a pain for normal cards too I think
[13:51] <iain> Company: (and no, I think I'll pass on the autoplugger)
[13:52] <BBB> oss works pretty well
[13:52] <Tomat> *simplified interface, *help pages would also be in sign language, *tools for layout and paper construction embedded
[13:52] <BBB> then again, oss is easy
[13:52] <Company> then again, oss is old and bugfixed
[13:52] <Company> alsa will be like oss in 2 years
[13:52] <Uraeus> Tomat: very hard to give an estimate on, I would say betwee4n 5000 USD and 40 000 USD :)
[13:52] Action: BBB will wait 2 years before usign alsa
[13:52] <Company> when Linu 2.8 comes out :)
[13:52] <Tomat> Uraeus: mm about what i thought
[13:53] Action: Company is seriously thinking about reverting his clocking "fix"
[13:53] <BBB> because...?
[13:54] <Tomat> Uraeus: it's doable, possibly by this summer or end of the year. Once the money is raised, how long to implement? another year?
[13:54] <Uraeus> Tomat: there is a GStreamer company starting up, you can mail [email protected] with a specification and ask for an offer
[13:54] jdahlin_ ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[13:54] <thomasvs> Tomat: also, mail the -devel list too
[13:55] <Tomat> thanks all. will contact julien / -dev when our project is more advanced
[13:56] ChrisHJW ([email protected]) left irc: Operation timed out
[13:57] <Tomat> Gotta go now. thanx for your help :)
[13:57] <Tomat> (waves)
[13:57] <Uraeus> Tomat: bye :)
[13:57] Tomat ([email protected]) left #gstreamer.
[14:00] <Uraeus> sales department here I come :)
[14:00] <BBB> you're bad :p
[14:00] <thomasvs> hahaha :)
[14:01] <Uraeus> Company: you don't like your own fix? btw, the libcaca plugin needs to be updated with the fix
[14:01] Action: thomasvs has lots of unproductive fun changing GST_ELEMENT_ERROR_2 to gst_element_error_2
[14:02] <Company> thomasvs: sed ?
[14:02] <Company> Uraeus: I thought I did the libcaca plugin. Didn't I?
[14:02] <Company> Uraeus: it has a serious issue with seeking...
[14:03] <Uraeus> Company: I got a message about not knowing about clock or something when trying to compile
[14:03] <Company> Uraeus: exact error?
[14:03] Action: Company should probably install libcaca
[14:04] <Uraeus> Company: I am re-installing libcaca now so I will get the error again
[14:09] <Uraeus> Company: ugh, so the fix breaks seeking or just makes is less accurate?
[14:09] <iain> hmmm, wonder if I should get Marlin to use the profiles thing for saving...
[14:10] <Company> Uraeus: less accurate - currently in the range of 3 seconds a/v diff
[14:20] Company ([email protected]) left irc: Remote closed the connection
[14:20] <Uraeus> Company: so when we seek we risk having audio and video going out of sync with 3 seconds?
[14:23] <BBB> iain: it'd be nice, yes
[14:28] mxpxpod ([email protected]) left irc: Read error: 60 (Operation timed out)
[14:38] mxpxpod ([email protected]) joined #gstreamer.
[14:46] <mathrick> do we support JACK?
[14:46] <Uraeus> mathrick: yes
[14:46] <mathrick> nice, but w/o preserving RT i suppose?
[14:47] <Uraeus> mathrick: I think Wingo aimed at it still working for RT
[14:48] <mathrick> Uraeus: and how succesful he was? Would using only jack elements now mean RT?
[14:49] joh ([email protected]) joined #gstreamer.
[14:49] <Uraeus> mathrick: well afaik so did noone ever test it, but in theory it was supposed to work
[14:49] <Uraeus> mathrick: you would need to mail Wingo to get the gory details :)
[14:50] <mathrick> hehe
[14:50] <sublett> does the JACK sink work now?
[14:50] <Uraeus> have never tried it so I have no idea
[14:50] <Uraeus> it compiles ;)
[14:50] <sublett> hehe
[14:51] <Uraeus> hi joh
[14:53] sublett ([email protected]) left irc: "I like food, food is good!"
[15:00] <mxpxpod> dolphy: yeah, I tested it... and she works like a charm :)
[15:03] Nick change: jdahlin__ -> jdahlin
[15:23] metric ([email protected]) left irc: "User abortion with 5 coathooks"
[15:25] mxpxpod ([email protected]) left irc: "Guns kill people as much as spoons made Rosie O'Donnell fat.."
[15:39] gheet ([email protected]) joined #gstreamer.
[15:40] <Uraeus> hi gheet
[15:40] <gheet> Yo Uraeus!
[15:41] <Uraeus> gheet: what kind of soundsystem do the Sunray's use?
[15:41] trow ([email protected]) joined #gstreamer.
[15:42] <Uraeus> hi trow
[15:42] <gheet> It pretty much redirect from th epesudo device to the sunray device
[15:42] <Uraeus> BBB: did you get that mixer gui issue fixed?
[15:42] <Uraeus> gheet: solaris audio then?
[15:42] <trow> hi Uraeus
[15:43] <gheet> Uraeus: You made it to osnews :)
[15:43] <Uraeus> gheet: yup and slashdot :)
[15:44] <Uraeus> trow: is your mixer applet patch merged upstream?
[15:45] <trow> Uraeus: yes
[15:45] <Uraeus> great :)
[15:45] <gheet> Uraeus: SUnray architecture, everything runs on the server, so the SW has to intercept the audio and redirect to the Sunray device
[15:45] <BBB> Uraeus: not yet
[15:46] <BBB> trow: maybe you can help?
[15:47] <trow> BBB: Well, I don't know anything about  Solaris and I don't have access to any Sun hardware.  But other than that...
[15:48] joh ([email protected]) left irc: Read error: 60 (Operation timed out)
[15:48] jeroen ([email protected]) left irc: Nick collision from services.
[15:49] <BBB> trow: no, no, no, not sun/solaris
[15:49] <BBB> ;)
[15:49] <BBB> trow: it's really simple
[15:49] <BBB> trow: two things, actually
[15:49] <BBB> trow: first, I noticed a small bug in gst-mixer... you probably want to update your acme/applet patches accordingly
[15:50] <BBB> trow: the gst_property_probe_probe_and_get_values() function may return NULL, so you need to check for that
[15:50] <trow> Oh, I misread gheet's comment as being from you.
[15:50] Action: trow needs new glasses.
[15:50] <BBB> that patch was just applied to gnome-media
[15:51] <BBB> you need the same patch in acme
[15:51] <BBB> I can apply it myself too if you want
[15:51] <BBB> (and applets)
[15:51] <trow> BBB: OK.  If it returns NULL, is there anything that we can conclude about what has gone wrong?  (i.e. is there a helpful error message we could display?)
[15:51] <BBB> just, no device
[15:51] <BBB> the error will be displayed elsewhere
[15:51] <BBB> it's on detection of elements and devices
[15:52] <BBB> if no element/device was found, another handler will display a message already
[15:52] <BBB> _("Sorry, no mixer elements and/or devices found")); <- that one in gnome-media
[15:53] <BBB> array = gst_property_probe_probe_and_get_values (probe, devspec); -> if (!(array = gst_property_probe_probe_and_get_values (probe, devspec))) goto next;
[15:53] <BBB> that was the patch in gnome-media
[15:53] jeroen_ ([email protected]) joined #gstreamer.
[15:54] <BBB> trow: that's 1... now the other... do you know something about gtk?
[15:58] <trow> BBB: You should probably just go ahead and send patches to the acme and gnome-applet maintainers (hadess and kevinv, respectively).
[15:59] <trow> BBB: And yes, I guess I know a little bit about gtk.  What is the problem?
[15:59] <Uraeus> ..a little bit...understatement of the year :)
[16:01] <BBB> trow: gnome-media's gst-mixer statrs up far too small! how do I get the requested size of the widget in the viewport so that I can enlarge it if the screen/display isn't too small to hodl the whole window?
[16:01] Action: BBB is a gtk idiot
[16:02] <jeroen_> is nautilus-media supposed to build?
[16:02] <Uraeus> jeroen_: yes
[16:02] <jeroen_> i'm getting "GstProps" undeclared errors
[16:02] <jeroen_> audio-properties-view.c: In function `apv_get_format_int':
[16:02] <Uraeus> jeroen_: but I am not sure if it is supposed to due much more than build atm
[16:02] <jeroen_> audio-properties-view.c:72: `GstProps' undeclared (first use in this function)
[16:02] Zahir ([email protected]) joined #gstreamer.
[16:02] <Zahir> hi all
[16:02] <BBB> hi
[16:02] <jeroen_> Uraeus: well, it doesn't even do that :)
[16:03] <Uraeus> hi Zahir
[16:03] <mathrick> BBB: my guess is that size is set somewhere. It shouldn't be, and then window is automatically sized
[16:03] <BBB> I'm just too downright bad at gtk to understand it at all
[16:03] <BBB> I'll stick to playing with glade for now ;)
[16:03] <mathrick> BBB: throw me a link to its code
[16:04] <BBB> cvs.gnome.org -> gnome-media/gst-mixer/src/mixer.c
[16:05] <mathrick> BBB: was it laid out by glade?
[16:05] alley_cat ([email protected]) joined #gstreamer.
[16:05] <trow> BBB: Well, the easiest thing to do is to call gtk_widget_set_size_request on the viewport and make it bigger by default.   Viewports have a habit of being much smaller than you want them to be.
[16:05] <BBB> http://cvs.gnome.org/bonsai/cvsblame.cgi?file=gnome-media%2Fgst-mixer%2Fsrc/mixer.c&rev=&root=/cvs/gnome
[16:06] <trow> BBB: The alternative would be to listen to various size_request signals and try to intervene, but that sort of thing can get very messy.
[16:06] <BBB> trow: but how do I get the right size?
[16:06] <mathrick> trow: too hacky
[16:06] <BBB> I mean, for OSS, it should just 'not be' a viewport
[16:06] <BBB> i.e., the viewport should do nothing
[16:06] <BBB> for alsa, it should, but only in width
[16:06] <BBB> the height should be 'unviewported'
[16:06] <jeroen_> i'm also getting the "no mixer elements and/or devices found" when i run g-v-c
[16:06] <trow> BBB: Well, you can set the policy for the scrollbars.  That isn't hard.
[16:06] <jeroen_> how do i get it to work?
[16:07] <Uraeus> jeroen_: are you using alsa?
[16:07] <jeroen_> Uraeus: yep
[16:07] <BBB> jeroen_: alsa seems slightly broken for now...
[16:07] <BBB> please use oss emulation
[16:07] <trow> BBB: gtk_scrolled_window_set_policy.  You can make it always or never use a scroll bar for height or width.
[16:07] <jeroen_> hmm, ok
[16:07] <BBB> trow: ok, so I set 'never' for height
[16:08] <hyriand> BBB: gtk_widget_size_request
[16:08] <trow> BBB: And you could do something like set never for width if the # of sliders is < some small number.
[16:09] <Uraeus> gheet: any chance we could get you to make a Solaris Audio plugin for GStreamer?
[16:11] <mathrick> BBB: try adding gtk_window_set_default_size(window, -1. -1); in mixer.c:484
[16:12] <mathrick> BBB: something may be "smart" enough to try setting it's size manually
[16:13] Action: BBB tries
[16:13] <mathrick> BBB: (-1,-1) means "right" size, i.e. the one that comes from children size request
[16:13] <iain> gtk_window_set_resizable to FALSE?
[16:14] <mathrick> iain: huh?
[16:14] <mathrick> iain: what good would that do?
[16:14] <iain> if you set resizable to FALSE the window will use whatever space the children require
[16:14] matrixise ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[16:14] <Uraeus> iain: yes, but as you know the ALSA mixer gets crazy then
[16:15] <iain> ah, I thought it was for vertical size we needed it, not horizontal :)
[16:15] <mathrick> iain: no, it'd just prevent user from resizing
[16:16] <iain> mathrick: it also makes it take whatever space is needed
[16:16] <mathrick> iain: no
[16:16] <trow> dolphy: Did you ever get a chance to look into why gst_play_seek_to_time wasn't working?
[16:17] Nick change: rk -> rk|away
[16:17] <dolphy> trow:  not yet
[16:17] <iain> mathrick: yes
[16:17] <mathrick> gtk_widget_set_size_request(GTK_WIDGET(window), -1, -1) is for that
[16:17] <dolphy> trow: i m busy on finding why i get a BadAccess X error with ximagesink and not xvimagesink in a special case :)
[16:17] <iain> mathrick: that may be another way to do it
[16:18] <iain> mathrick: but setting it to not be resizable does the same thing
[16:18] <mathrick> iain: _set_resizable just controls boolean property - if user can resize window
[16:18] <iain> yes
[16:18] <iain> and if its not resizable then the size it takes is the size that is required to fit all the widgets
[16:18] <mathrick> iain: it's just side effect, not obvious one, btw
[16:18] <mathrick> void        gtk_window_set_resizable        (GtkWindow *window,
[16:18] <iain> I think its obvious
[16:18] <mathrick>                                              gboolean resizable);Sets whether the user can resize a window. Windows are user resizable by default.
[16:19] <iain> if your window isn't resizable and it doesn't take the required size then you can't get to some widgets
[16:19] <iain> but its how I've made all dialogs be "the right size" on marlin
[16:19] <mathrick> iain: setting window unresizable is impolite anyway
[16:19] <iain> just make them unresizable
[16:20] harshy ([email protected]) left irc: Read error: 60 (Operation timed out)
[16:20] <iain> no-one's ever complained
[16:20] <mathrick> iain: it's because i don't use marlin :)
[16:20] Action: mathrick hates unresizable windows
[16:20] <iain> no-one ever complained for the old mixer either
[16:21] <mathrick> iain: anyway, it's not the right way to do it
[16:21] <iain> if its a window that has no real need to be resized I don't see that it should be
[16:21] <mathrick> iain: if there is even single label inside, there usually is need for resizing
[16:22] <iain> mathrick: why?
[16:22] <iain> mathrick: if its unresizable then the label gets all the space it needs
[16:22] <mathrick> iain: b/c too often i have windows that somehow didn't get sized correctly, and I can't resize
[16:23] <iain> then thats a bug
[16:23] <iain> not a reason that they should be resizable
[16:23] <BBB> ok, I fixed that too
[16:23] <iain> on a semi related note
[16:23] <mathrick> iain: maybe, i didn't see marlin's dialogs. But it is *wrong* way to get correct size
[16:24] <mathrick> BBB: so it's sized correctly now?
[16:24] <iain> the inkscape guys are adding "Default width and height" to their preferences dialog so you can specify the default width
[16:24] <iain> AND they're also adding if that width and height is in pixels or percentage of screen dimensions
[16:24] Action: iain rolls his eyes
[16:24] <mathrick> iain: but only for image windows I presume?
[16:24] <iain> mathrick: for the main window
[16:25] <iain> mathrick: but its still 3 gratuitious preferences that aren't needed
[16:25] Action: trow recoils in horror
[16:25] <mathrick> iain: if they ever add such a strange pref, it should be settable in %
[16:25] <mathrick> iain: and for graphics app, it *may* be usable pref
[16:26] <iain> mathrick: then you just resize the window and have it remember that for later :)
[16:26] <mathrick> iain: but when you create another drawing, it's another window
[16:27] <iain> mathrick: pretty much all windows will be the same size
[16:27] <mathrick> iain: not necessarily, as not all drawings are the same size
[16:27] <mathrick> and pretty often you have one "master" drawing in big window, and many small temporary ones
[16:28] <mathrick> so for image editing app, there may be point in having such pref. It all depends on how often users request it :)
[16:29] <iain> mathrick: I still think all that can be done without needing a preference
[16:29] <iain> mathrick: by working out which are the "temporary" windows
[16:29] <iain> mathrick: and basing from the master
[16:29] <mathrick> iain: that would require embedding lisp machine for AI to decide ;P
[16:30] <iain> get the alsa guys to do that then
[16:30] <iain> they like adding lisp to places its not needed
[16:30] <mathrick> huh? really?
[16:30] <BBB> mathrick: "basically", yes
[16:30] <teuf> what did the alsa guys do with lisp ?
[16:30] <iain> yeah, .asoundrc has been redone so that its in lisp
[16:30] <mathrick> iain: scheme probably?
[16:31] matrixise ([email protected]) joined #gstreamer.
[16:31] <iain> no...well, I dunno, theyjust keep talking about lisp on the mailing list
[16:31] <mathrick> iain: gtk menu dumps are in written scheme too ;)
[16:31] <mathrick> s/\(in\)\(written\)/\2\1/
[16:31] <iain> mathrick: there's an alsalisp dir in alsa-lib
[16:32] <teuf> maybe their ultimate plan is to compete with emacs
[16:32] <mathrick> wow, maybe they'll reimplement emacs in alsalisp too? ;)
[16:32] <mathrick> hehe :)
[16:32] <Uraeus> me remembers the gnome snooker game the alsa guys wrote with scheeme or lisp
[16:32] <iain> and alsalisp is their interpreter
[16:32] <iain> lovely
[16:32] <mathrick> it's nothing wrong about using scheme in alsa, it's implementing it that's bad :)
[16:33] <mathrick> s/scheme/lisp/
[16:33] <mathrick> BBB: "basically"?
[16:34] <mathrick> BBB: you mean it basically works? or there is basically lisp in alsa?
[16:34] <iain> (princ "  name    : " (Aresult (Acall 'card_get_name card)) "\n")
[16:34] <iain> mmm, yummy
[16:35] <mathrick> not very lispish lisp it is, i must say
[16:35] <iain> its got lots of ( and ) so I count it as lisp
[16:36] <mathrick> (princ (concat "  name    : " (Aresult (Acall 'card_get_name card)) "\n")) is more lispish
[16:36] Company ([email protected]) joined #gstreamer.
[16:36] <mathrick> yo Company
[16:36] <Company> oy
[16:36] <BBB> mathrick: it basically resizes the UI correctly
[16:37] <mathrick> BBB: and what about "advanced"? :)
[16:37] <mathrick> BBB: btw, you have devhelp installed, don't ya
[16:39] jeroen__ ([email protected]) joined #gstreamer.
[16:40] <BBB> mathrick: no
[16:40] <BBB> what's that?
[16:41] <mathrick> BBB: it's the thing that makes you happy, not to say it's necessary for GTK development ;). Gtk-doc reader, w/ indexing and xreferences
[16:41] <thomasvs> BBB: YOU DON'T HAVE DEVHELP ??????
[16:41] <thomasvs> BBB: DUDE !
[16:42] <thomasvs> BBB: want a screenshot of it with gst ?
[16:42] <BBB> what is devhelp?
[16:42] <iain> the developers version of yelp
[16:42] <mathrick> BBB: it's absolutely necessary for development, I told ya already
[16:42] <BBB> nope, don't have that
[16:42] <iain> I've not used it for ages
[16:43] Action: BBB ususally digs around in header files for that purpose
[16:43] <mathrick> BBB: Gtk-doc reader, w/ indexing and xreferences
[16:43] <iain> BBB: yeah, me too
[16:43] <Uraeus> have you guys seen this : http://developers.slashdot.org/comments.pl?sid=92710&cid=7970111
[16:43] <mathrick> BBB: you just throw gtk-doc generated books at it, and it takes care of everything else
[16:43] <iain> yey paul davis :)
[16:44] <thomasvs> BBB: http://thomas.apestaart.org/download/tmp/devhelp-gst.png
[16:44] <thomasvs> BBB: I don't know how you can program gst without it :)
[16:45] <mathrick> thomasvs: true
[16:45] <BBB> thomasvs: I know the API ;)
[16:46] <thomasvs> Uraeus: seems like that comments is completely right, no ?
[16:46] <BBB> at least paul is not flaming gst any longer
[16:46] <BBB> he's growing uop
[16:46] <BBB> wow
[16:46] <BBB> thomasvs: looks quite funny, but do you actually use it?
[16:46] Action: BBB is an old-fashioned vim programmer
[16:47] <Uraeus> thomasvs: no idea, I just though I make people aware considering PBD trackrecord
[16:47] <jdahlin> thomasvs: you use devhelp quite a bit don't you?
[16:47] jeroen_ ([email protected]) left irc: Read error: 60 (Operation timed out)
[16:47] <jdahlin> in spanish too! :)
[16:47] <iain> thomasvs: I used to have a macro for emacs that fired up devhelp when you pressed F1 at the word the cursor was on
[16:48] <iain> thomasvs: we have other ways of routing data between processes
[16:48] <iain> thomasvs: not just though jack
[16:48] <mathrick> BBB: it's impossible to program w/o devhelp :)
[16:48] <BBB> I proove the opposite
[16:49] <iain> yeah me too
[16:49] Action: iain wants to be able to revise
[16:51] <mathrick> iain: do you have that macro around? I'm still firing it up manually
[16:52] <iain> mathrick: no, it was years ago, sorry
[16:52] <iain> and I'm not an elisp person
[16:52] <mathrick> iain: ok, i'll write it myself then
[16:52] <iain> I forget who it was gave me it
[16:52] <iain> fejj maybe
[16:53] <Company> i don't grok why jack is better than gst?
[16:53] <Company> just because nobody has written callback based elements yet?
[16:54] <iain>       (gdb (concat "gdb -x /home/matt/bin/gdb_commands " (getenv "EXE_TO_GDB")))
[16:54] Action: iain wonders who matt is and how he got into iain's init.el file
[16:55] <Company> actually i want callback based elements
[16:55] <iain> Company: I think jack is only better because of the whole RT thing
[16:55] <Company> currently only sources are callback based
[16:55] <Company> RT based?
[16:55] <Company> RT thing?
[16:56] <jdahlin> realtime perhaps
[16:56] <Company> yeah, but why does that make it better? ;)
[16:56] <iain> because for recording 16 tracks of audio
[16:56] <iain> they all come in smoothly
[16:56] <iain> and don't have pops and clicks
[16:57] <Uraeus> Company: decided what to do in regards to your fix and the seeking issue? still considering a revert?
[16:57] <iain> Company: but only complicated things like ardour really need it essentially I think in practise
[16:57] <iain> practice
[16:58] Nick change: Rayban_ -> Rayban
[16:58] <Company> ardour is an advanced sample editor with an ugly GUI
[16:58] <Company> where does that something that gst can't do?
[16:59] <iain> can gst do 32 track recording with no pops and clicks?
[16:59] <Company> if you want it to, sure
[16:59] jeroen__ ([email protected]) left irc: "Client exiting"
[16:59] <iain> no pops and clicks?
[17:00] <iain> (recording 32 tracks to disk not into memory)
[17:00] <Company> at least not because of the design
[17:00] <iain> I know not because of the design
[17:00] <iain> but in practice I dont think it could handle it
[17:00] <Company> yes
[17:00] <iain> (but if you give me a couple of RME hammerfalls, I'll be happy to check for you)
[17:00] <Company> that's because all devs have a focus on multimedia
[17:01] <Company> but i think if someone with a focus on pro-audio shows up and does the low latency schedulers and stuff, that'd not be a problem
[17:02] <Company> in fact, video/audio editing apps will require this some day
[17:03] ChrisHJW_log ([email protected]) left irc: Read error: 60 (Operation timed out)
[17:03] <Company> but i think it's better to go the current way from common multimedia to specialised
[17:03] <Company> instead of the other way
[17:03] <Company> because that gives you a huge number of testers for free
[17:04] Rayban ([email protected]) left irc: "ChatZilla 0.9.52B [Mozilla rv:1.7a/20040101]"
[17:05] <Company> gstreamer is currently in the unfortunate business of doing nothing really well
[17:05] <Company> but everything a bit
[17:05] <iain> yey for halfassed
[17:06] <Company> :(
[17:06] <Company> but gstreamer is the only framework that has a sample editor, a music manager and a video player
[17:07] <teuf> Company: I'm pretty confident that it'll get stable enough in a few months from now ;)
[17:07] <Company> i guess so
[17:08] <Company> if people would only do ardour with gst i wouldn't need to write an autoplugger :/
[17:09] <Company> ok, flacenc/flacdec transcode correctly wrt tags now
[17:10] <Company> and on BE, too
[17:10] <BBB> Company: and a video recorder ;)
[17:10] <BBB> and a mixer
[17:11] <Company> and a TV viewer ;/
[17:11] <iain> and a pipeline editor
[17:11] <BBB> most of which are unreleased
[17:11] <Company> yup
[17:11] <BBB> in short: company's right, we suck at everything ;)
[17:11] <Company> i remember someone saying halfassed
[17:12] Action: BBB will try to work on integrating profiles in gst-rec tonight
[17:12] Action: thomasvs is BORED with fixing errors
[17:12] <thomasvs> we have crap errors :)
[17:12] <BBB> no kidding doc
[17:12] <BBB> actually, we rock
[17:12] <BBB> we'll be i18n'ed very soon
[17:12] <BBB> yay @ thomas
[17:12] <thomasvs> well, the error stuff works really well
[17:13] <thomasvs> but it is BORING to go over *all* of the plugins
[17:13] <Company> and we'll have a dvd player, too
[17:13] <Company> yay!
[17:13] <BBB> I know, I've done it several times
[17:13] <Company> firstr ever mixer to have a dvd player!
[17:13] Action: thomasvs created _TOO_LAZY enums in each of the domains too
[17:13] <thomasvs> for the hard ones that still need to be decided
[17:13] Action: BBB kicks thomasvs for doing that
[17:14] <BBB> at least call it 'UNKNOWN'
[17:14] <BBB> that sounds more professional
[17:14] <Company> i like thomasvs for inventing _TOO_LAZY or --enable-segfault :)
[17:15] <Company> it's too bad people don't have enough humour
[17:18] <Company> ** (gst-player:15802): WARNING **: can't find any mixer element, no volume.
[17:18] <Company> that one is not a programming error, so it shouldn't be a g_warning
[17:19] Action: BBB agrees
[17:19] <BBB> make a bugreport please ;)
[17:19] <thomasvs> BBB: no, the idea is there never is a TOO_LAZY error unless it's under discussion on how to categorize it
[17:19] <BBB> oh, ok
[17:19] <thomasvs> BBB: so there wouldn't be TOO_LAZY at all
[17:19] <BBB> doesn't that go for UNKNOWN too?
[17:19] <thomasvs> BBB: but otherwise we don't keep track of the "bad" messages
[17:19] <thomasvs> BBB: oh give me my pleasures will you :) this is boring enough as it is :)
[17:19] LeRoutier ([email protected]) joined #gstreamer.
[17:19] <BBB> true, true
[17:19] <LeRoutier> hello
[17:20] <mathrick> hi
[17:20] <Company> grmbl
[17:20] <Company> a gstplay pipeline is in which state by default?
[17:20] <Company> dolphy ?
[17:21] <dolphy> Company: hmm
[17:21] <dolphy> Company: you mean just after creation ?
[17:21] <Company> dolphy: yeah
[17:22] <dolphy> Company: hmm i guess it's in NULL
[17:22] <Company> dolphy: when i start gst-player /some/file and then mouse over the video window i get warnings
[17:22] <Company> (gst-player:16013): GStreamer-WARNING **: pad video_scaler:src has no event handler
[17:22] <Company> dolphy: actually gst-player /some/file --gst-scheduler=basicomega
[17:22] <dolphy> Company: well gstplay is just building a pipeline
[17:22] <Company> because opt just stops after 5 seconds
[17:22] Action: Company looks at iain
[17:22] <dolphy> Company: it s not changing the pipeline state
[17:22] <Company> hm
[17:23] <Company> then why is the core allowing it to send events?
[17:23] <dolphy> Company: so creating a bin and adding elements to it and linking them
[17:23] <dolphy> Company: make them READ ?
[17:23] <dolphy> READY ?
[17:24] <Company> well, it's just nav events, but anyway
[17:24] <Company> there shouldn't be any events
[17:24] <dolphy> i will take a look
[17:24] <dolphy> Company: how can i check inside ximagesink if the caps have been negotiated ?
[17:24] <dolphy> Company: get_peer_pad on sinkpad ?
[17:25] <Company> if gst_pad_is_negotiated (pad)
[17:25] <dolphy> Company: because with ximagesink when one calls the xoverlay interface and the pad is not connected yet
[17:25] <dolphy> Company: it tries to set caps on the pad and to create an ximage
[17:25] <dolphy> Company: and that make very funny stuff
[17:26] <Company> understandable :)
[17:26] <Company> gst_pad_is_negotiated is your friend there
[17:28] <dolphy> ok
[17:29] <dolphy> thx
[17:29] Action: iain looks at Company 
[17:29] <BBB> alsa is borked!
[17:30] lypanov ([email protected]) left #gstreamer.
[17:31] <Company> BBB: tell something new please? ;)
[17:31] <BBB> you will fix it!
[17:32] <LeRoutier>  
[17:32] <Company> BBB: tell something new and correct please? ;)
[17:33] <BBB> I just did
[17:33] <BBB> :)
[17:33] <LeRoutier> seems my wav gst-player crash is gone. (got a nego. failed pop-up now in gst-play)
[17:33] <Uraeus> LeRoutier: yeah, ronald fixed wavparse earlier today
[17:33] <LeRoutier> now have some "element work_thread claimed state-change success,but state didn't change to PLAYING. State is PAUSED (NONE_PENDING pending), fix the element"
[17:33] <Uraeus> but we still need ds-work to look over the wav caps-nego
[17:34] <dolphy> Company: that fixes it just nicely
[17:35] <LeRoutier> much of my crashes are when using the opt sched
[17:35] <BBB> LeRoutier: ds needs to fix that last bit
[17:35] <BBB> I fixed the rest
[17:35] <LeRoutier> k
[17:35] BBB ([email protected]) left irc: "Client exiting"
[17:36] <LeRoutier> another usable sheduler than opt ?
[17:37] <Company> basicomega
[17:37] <LeRoutier> ok, so with  --gst-scheduler=basicomega
[17:37] <Company> yeah, that's how i run all my apps these days ;)
[17:38] <Uraeus> Company: but does basicomage work on i686 glibc? I mean it segfaulting on that was the reason we switched to opt in the first place
[17:41] <Company> Uraeus: no idea, but it works here
[17:41] <LeRoutier> got the warning with the "... fix the element" when using basicomega
[17:41] <Company> Uraeus: basicgthread should work everywhere btw
[17:41] <Company> yeah, that warning is the "threading is broken, wait for 0.10" warning
[17:41] <dolphy> Company: A/V sync is completely broken for me with current CVS head
[17:42] <LeRoutier> 0.10 ? miam
[17:43] <LeRoutier> bbl
[17:43] LeRoutier ([email protected]) left irc: "Leaving"
[17:43] <Company> dolphy: huh?
[17:44] <Company> dolphy: you don't use lots of debugging output, do you?
[17:44] <alley_cat> --gst-scheduler=basicomega segfaults here
[17:45] <alley_cat> nptl glibc compiled for athlon
[17:45] <Uraeus> alley_cat: yup, that is why we switched to opt :)
[17:45] <Company> alley_cat: does basicgthread work?
[17:45] <dolphy> Company: i don't
[17:45] <dolphy> Company: i will try with gst-plugins/examples/gstplay
[17:45] <dolphy> Company: as gst-totem might be a bit heavier
[17:46] <Company> dolphy: what audiosink do you use?
[17:47] Phiend ([email protected]) joined #gstreamer.
[17:47] <alley_cat> basicgthread doesn't work
[17:48] <Company> that's too bad for all the poor NPTL users who must now find someone that fixes opt :)
[17:49] <dolphy> Company: osssink
[17:49] <dolphy> Company: hmm with examples/gstplay it works
[17:49] <dolphy> Company: maybe the videobalance element i added in my gconf key
[17:49] <Company> dolphy: i'm finetuning atm
[17:49] <Company> dolphy: don'T think so
[17:49] <dolphy> i have videobalance ! ffcolorspace ! ximagesink
[17:49] <dolphy> as a videosink
[17:49] <Company> that should work
[17:49] <dolphy> it works
[17:50] <dolphy> but A/V sync isn't ther
[17:50] sub_pop ([email protected]) left irc: "Client exiting"
[17:50] <dolphy> and that's the only difference between this and examples/gstplay
[17:50] <Uraeus> Company: need to find someone? I thought we had choosen iain as our designated opt fixer?
[17:51] <Company> dolphy: there are issues when the state change to playing in audio and videosink are too far apart
[17:52] <dolphy> Company: trying with only ximagesink in my gconf key
[17:55] <Company> dolphy: you have updated both corre and plugins?
[17:55] <Company> (that was the obvious question)
[17:56] Action: foser is away: things.. important.. well it's me you know
[17:57] <iain> Uraeus: I'm the "designated one line opt fixer"
[17:57] <iain> Uraeus: if its more than one line...dunno
[17:57] <Uraeus> iain: one line at a time is ok if you do it often enough
[17:58] <Company> iain: you can probably get it all into one line
[17:58] <iain> Company: for what its worth, opt with nptl works fine here
[17:58] <teuf> hmm, have you seen http://advogato.org/person/malcolm/
[17:58] <Company> iain: start with tr "\n" " " gstoptimalscheduler.c
[17:58] <teuf> Gstreamer guys: I pushed somebody in your direction who is interested in writing developer-level documentation. Don't frighten him away!
[17:58] <Company> iain: with alsa?
[17:58] <iain> at least marlin runs and it uses threads for everything
[17:58] <iain> Company: oh, not with alsa
[17:59] <iain> but alsa just has dumb issues for me anyway
[18:00] <Company> you someohw need a loopbased sink to trigger the opt problems
[18:00] <dolphy> Company: yeah
[18:00] <iain> marlinsamplesink is loop based
[18:00] <iain> and it works
[18:01] <Company> well, i can't even run rb with alsa and opt
[18:01] <Company> it plays and stops after ~15-30 seconds
[18:01] <Company> 100% CPU, loops forever
[18:02] <dolphy> Company: so after a few tests
[18:02] <Company> i think filesrc ! spider ! alsasink has the same problem
[18:02] <dolphy> Company: here is what i see
[18:02] <dolphy> Company: if by bad luck i move the mouse over video window
[18:03] <dolphy> Company: some navigation events are going upstream slowing down a bit the video sink
[18:03] <dolphy> Company: then sync is broken
[18:03] <dolphy> Company: and it tries to get back
[18:03] <dolphy> Company: hardly :)
[18:03] <dolphy> Company: 100% cpu
[18:03] <dolphy> Company: and after 30 seconds it almost gets there 
[18:03] <dolphy> Company: going back to 60 % cpu
[18:04] <Company> dolphy: yeah, that's an issue with your machine not having enough power
[18:04] <Company> that's not a clocking issue
[18:04] <Company> it's probably related to software scaling, too
[18:05] <dolphy> Company: hm that test was with xv
[18:05] <Company> that all eats CPU and we don'T discard late buffers yet
[18:05] <Company> hm
[18:05] <dolphy> Company: that's something i could do :)
[18:05] <Company> in that case it's not realted to software scaling ;)
[18:05] <dolphy> Company: discard late buffers
[18:05] <dolphy> Company: should not be too hard isn't it ?
[18:05] <Company> yes it is
[18:06] <Company> because you need a clock-aware element early enough in the stream
[18:06] <Company> and you don't want to discard important buffers (keyframes) either
[18:06] <dolphy> ah you mean discard late buffers in the demuxer
[18:06] <dolphy> i was proposing to discard late buffers in ximagesink
[18:06] <Company> well, discarding late stuff after the stuff is already decoded and has eaten the CPU time is no use ;)
[18:07] <Company> i don't think ximagesink shouldn'T discard late buffers
[18:07] <dolphy> but my cpu was eating 100 % cpu
[18:07] <dolphy> i guess it was processing a pile of buffers in the video queue
[18:07] <dolphy> XPuting them
[18:08] <dolphy> ignoring too late buffers might help a little bit to get back to sync
[18:08] <dolphy> no ?
[18:08] harshy ([email protected]) joined #gstreamer.
[18:09] <alley_cat> hmm, opt scheduler on nptl with alsasink works fine, much better than osssink
[18:10] <alley_cat> only for files with 48khz sampling rate though
[18:11] Nick change: harshy -> harshyWork
[18:12] ack ([email protected]) joined #gstreamer.
[18:12] Nick change: ack -> sack
[18:13] iain ([email protected]) left irc: "restart"
[18:13] <Company> alley_cat: alsasink as sink with rhythmbox?
[18:13] <alley_cat> no, as sink with gst-launch
[18:13] <Company> alley_cat: or do you insert an audioscale in between?
[18:14] <alley_cat> but only if the audio is 48khz, otherwise i get an error
[18:14] Phiend ([email protected]) left #gstreamer.
[18:14] <alley_cat> with audioscale it works for everything
[18:16] Markey- ([email protected]) joined #gstreamer.
[18:16] Action: dolphy added colorbalance interface in videobalance element
[18:16] <dolphy> so now gst totem can set colorbalance also on ximagesink :)
[18:16] <alley_cat> how do i tell audioscale to which rate it should convert?
[18:17] <Company> alley_cat: audioscale ! audio/x-raw-int,rate=48000 ! alsasink
[18:17] <Company> alley_cat: that sets filtered caps with a fixed rate between audioscale and alsasink
[18:21] mxpxpod ([email protected]) joined #gstreamer.
[18:22] <alley_cat> ok, with audioscale and alsa i get perfect a/v sync, with osssink it gets out of sync
[18:23] Action: Company will try osssink
[18:24] borgu ([email protected]) left irc: "will try to boot 2.6.1"
[18:25] <Company> hm, osssink totally craps out nowadays
[18:27] <Company> hrm?
[18:27] Action: Company wonders why it only craps out on mpegs
[18:27] <alley_cat> hmm, and the one file i have that still goes out of sync seems to be a bad file, remuxing it to matroska fixes it
[18:28] <Company> seeking should be broken
[18:28] <Company> but i just fixed that locally
[18:29] <alley_cat> i only have gst-launch for 0.7 currently, so no seeking
[18:30] <Company> you know what sucks?
[18:30] <Company> debugging clocking only works via g_print
[18:31] <Company> stopping the pipeline totally screws everything
[18:31] <Company> so gdb is not an option
[18:42] Action: Company grabs a flowr
[18:42] sub_pop ([email protected]) joined #gstreamer.
[18:42] <Company> i am evil now, i am not evil now, i am evil now, ...
[18:44] gheet ([email protected]) left irc: "Download Gaim [http://gaim.sourceforge.net/]"
[18:46] <alley_cat> is gst-player supposed to show a black image with xvideosink? ;)
[18:46] matrixise ([email protected]) left irc: "Leaving"
[18:48] iain ([email protected]) joined #gstreamer.
[18:49] LeRoutier ([email protected]) joined #gstreamer.
[18:49] <LeRoutier> re
[18:50] matrixise ([email protected]) joined #gstreamer.
[18:50] fsfman ([email protected]) joined #gstreamer.
[18:51] fsfman ([email protected]) left irc: Client Quit
[18:52] jdahlin_ ([email protected]) joined #gstreamer.
[18:52] jdahlin ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[18:53] <Company> is oss documented somewhere?
[18:53] <Company> or more precisely. is there a way to make oss automatically insert $count silent samples?
[18:53] <Company> ?
[18:56] teuf ([email protected]) left irc: "Client exiting"
[18:59] ChrisHJW_log ([email protected]) joined #gstreamer.
[19:00] <deadchip> Company: maybe 4Front will sell you some docs on OSS
[19:00] <deadchip> as open as they are ;)
[19:00] ofer ([email protected]) joined #gstreamer.
[19:03] Misirlou ([email protected]) joined #gstreamer.
[19:07] Nick change: jdahlin_ -> jdahlin
[19:13] jdahlin ([email protected]) left irc: "Klienten avslutas"
[19:16] Nick change: Uraeus -> Ura_out
[19:17] kmaraas ([email protected]) joined #gstreamer.
[19:18] <Company> hrm
[19:34] md`` ([email protected]) joined #gstreamer.
[19:35] md` ([email protected]) left irc: Nick collision from services.
[19:36] Nick change: md`` -> md`
[19:38] kmaraas ([email protected]) left irc: "Leaving"
[19:38] <thomasvs> Company: yeah, tere's a pretty good oss manual
[19:39] <thomasvs> I wish I still had it somewher on my disk
[19:39] <thomasvs> http://www.opensound.com/pguide/oss.pdf
[19:39] <thomasvs> this was it I think
[19:41] pb_ ([email protected]) joined #gstreamer.
[19:42] <Company> i got that
[19:42] <Company> but that's not what I call pretty good, it's mostly blabla
[19:44] kmaraas ([email protected]) joined #gstreamer.
[19:47] mxpxpod ([email protected]) left irc: Read error: 60 (Operation timed out)
[19:47] Markey- ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[19:48] Markey- ([email protected]) joined #gstreamer.
[19:49] mxpxpod ([email protected]) joined #gstreamer.
[19:51] <thomasvs> hm, I thought it explained oss pretty well
[19:52] Markey- ([email protected]) left irc: Client Quit
[19:54] Nick change: Zahir -> ZahirDev
[19:54] minio ([email protected]) joined #gstreamer.
[20:01] minio ([email protected]) left #gstreamer ("Leaving").
[20:01] minio ([email protected]) joined #gstreamer.
[20:02] NotHere ([email protected]) joined #gstreamer.
[20:02] minio ([email protected]) left #gstreamer ("Leaving").
[20:05] minio ([email protected]) joined #gstreamer.
[20:07] minio ([email protected]) left #gstreamer ("Leaving").
[20:08] ofer ([email protected]) left #gstreamer ("Leaving").
[20:14] Action: Company is officially stupid
[20:14] dolphy ([email protected]) left irc: "Network down, IP Packets delivered via UPS"
[20:14] <Company> at least for today
[20:15] NotHere ([email protected]) left #gstreamer ("If I go insane, please don't put your words in my mouth").
[20:15] kmaraas ([email protected]) left irc: "Leaving"
[20:16] <ds-work> #gstreamer needs a Rupert
[20:16] <ds-work> for Company's quotes
[20:18] <Company> you could put the quote into the real Rupert (tm)
[20:19] <mathrick> er, who's Rupert(tm)? :)
[20:20] <Company> a bot with a quote collection
[20:20] <Company> without the (tm)
[20:20] <mathrick> and with?
[20:23] <ds-work> and with pokey quotes
[20:24] <ds-work> <ds-work> Rupert: pokey
[20:24] <ds-work> <Rupert> YE BE MINE SWORN ENEMY!!!
[20:24] thomasvs ([email protected]) left irc: Read error: 113 (No route to host)
[20:29] sack ([email protected]) left irc: "Client Exiting"
[20:29] <Company> avidemux doesn't do seeking?
[20:36] <mathrick> gst-launch-0.7 -v filesrc location="/mnt/win/Misc/Filmy/Clipy/Nyo is War (Di Gi Charat).mpeg" ! spider ! videoscale ! video/x-raw-rgb,height=100,width=100 ! colorspace ! xvimagesink
[20:36] <mathrick> gives funky colors, should I add some format property?
[20:37] <mathrick> hmm, s/rgb/yuv/, it doesn't work at all with rgb
[20:37] aroedl ([email protected]) joined #gstreamer.
[20:37] <ds-work> don't use colorspace
[20:38] <ds-work> for at least a few more days
[20:39] kmaraas ([email protected]) joined #gstreamer.
[20:41] <mathrick> ahh, right, colorspace doesn't work, funky colors are w/ ffcolorspace :)
[20:42] <ds-work> why are you converting to RGB and back to YUV?
[20:42] <mathrick> huh?
[20:42] <mathrick> no, i just pasted wrong pipeline
[20:43] <mathrick> gst-launch-0.7 -v filesrc location="/mnt/win/Misc/Filmy/Clipy/Nyo is War (Di Gi Charat).mpeg" ! spider ! videoscale ! video/x-raw-yuv,height=100,width=100 ! ffcolorspace ! xvimagesink
[20:43] <mathrick> w/o ffcolorspace it's the same
[20:44] <ds-work> use width=104
[20:44] <ds-work> divisible by 8
[20:44] <mathrick> k, it works :)
[20:44] <mathrick> ds-work: and why height=100 works?
[20:45] <ds-work> because the memory layout of I420 depends on the width, not the height
[20:46] <ds-work> I'd bet height=101 doesn't work
[20:46] thomasvs ([email protected]) joined #gstreamer.
[20:47] <mathrick> ds-work: heh, right. It's funny :)
[20:47] Marsupilami23 ([email protected]) joined #gstreamer.
[20:47] <ds-work> for now, just use sizes that are divisible by 8.  x[v]imagesink and videotestsrc are fixed, but nothing else
[20:48] <thomasvs> ds-work: so, after I commit my error stuff, should we think about releasing ?
[20:48] <thomasvs> or is there something blocking us ?
[20:48] <ds-work> thomasvs: yeah, walters is itching
[20:48] <thomasvs> tagging works for him now ?
[20:48] <ds-work> we need to make sure Company is happy with the A/V sync
[20:48] <thomasvs> yeah
[20:49] <thomasvs> we need to make sure we are happy with it too :)
[20:49] <ds-work> he was complaing about it earlier, but I don't know if that was serious or not
[20:49] <thomasvs> I feel completely metal
[20:49] <thomasvs> yah, he wanted to revert no ?
[20:49] <thomasvs> I was so down for a whole month, but things seem to fall into place for us at all levels
[20:50] sublett ([email protected]) joined #gstreamer.
[20:52] Shoragan ([email protected]) left irc: "Leaving"
[21:00] <mathrick> what is the name of xine wrapper plugin?
[21:03] <ds-work> whoever was writing it gave up
[21:04] <Company> there was an ugly bug that i wasn't aware
[21:04] <Company> of
[21:04] <Company> it works good enough now
[21:05] <mathrick> ic, i thought it was completed
[21:06] BBB ([email protected]) joined #gstreamer.
[21:06] <BBB> does anyone know the signal if a textview's contents is changed?
[21:08] Action: mathrick opens Devhelp
[21:09] <BBB> :p
[21:11] <mathrick> BBB: you mean TextView or Buffer?
[21:11] <BBB> view
[21:11] <BBB> if view is buffer, buffer is fine too
[21:11] <mathrick> BBB: View doesn't change it's contents
[21:11] <BBB> textarea then?
[21:12] <mathrick> BBB: Buffer :)
[21:12] <mathrick> GtkTextBuffer::changed
[21:13] <mathrick> BBB: and many others, more specific ones
[21:13] <BBB> changed is good enough
[21:13] Action: BBB has profile editor finished
[21:13] <BBB> it's extremely cool
[21:13] <BBB> I'll commit as soon as 0.8 is branched off
[21:13] <BBB> I don't want this in 0.8 for now
[21:14] <ds-work> we're branching 0.8?
[21:14] <mathrick> BBB: you should really install Devhelp :)
[21:20] <BBB> ds-work: not yet, but we will one day
[21:21] Nick change: matrixise -> _neo_
[21:22] <BBB> yay!
[21:22] Action: BBB has an editor
[21:22] mxpxpod ([email protected]) left irc: "Guns kill people as much as spoons made Rosie O'Donnell fat.."
[21:22] <BBB> I only have to implement XML loading and saving
[21:22] <BBB> :)
[21:27] Action: BBB waves @ company
[21:27] <BBB> Company: still @ media-api?
[21:27] <BBB> they're currently discussing how they'll recreate set/get_property() in their own object system
[21:27] <BBB> and how that will make plugins uniform and UI independent
[21:27] kleppari ([email protected]) joined #gstreamer.
[21:28] <kleppari> can I make a request or contribute a small file?
[21:28] <BBB> sure
[21:28] <kleppari> gstreamer.pc for pkg-config..
[21:28] <BBB> we call it gstreamer-version
[21:28] <BBB> so gstreamer-0.6.pc
[21:29] <BBB> or -0.7.pc
[21:29] <BBB> (depending on your version)
[21:29] <kleppari> ah, ok.
[21:29] <BBB> (try pkg-config --modversion gstreamer-0.6)
[21:29] <Company> we'Re branching 0.8?
[21:29] <BBB> no
[21:30] <BBB> I said "when"
[21:30] <Company> ah
[21:30] <BBB> so long-term future
[21:30] <BBB> ;)
[21:30] <Company> so like in november?
[21:30] <BBB> no, just before gnome-2.6
[21:30] <Company> i didn't intend to branch that early
[21:30] <kleppari> ah, ok, nevermind then :)
[21:30] <BBB> kleppari: heh, you wanted to contribute a .pc file? ;)
[21:30] <Company> i hope we can stay longer with the branch the users use
[21:30] <BBB> kleppari: well, thanks for the idea ;)
[21:31] <kleppari> BBB: well, I thought it wasn't there, faulty locate database
[21:31] <BBB> Company: you mean, you want to continue development?
[21:31] <Company> BBB: i tend to only scan the posts on media-api nowadays
[21:31] <BBB> oh
[21:31] <BBB> well, anyway, they're funny
[21:31] <BBB> but what's your idea for 0.8 then?
[21:31] <Company> BBB: yeah, i thought we would extend the plugins and interfaces for a while without breaking compat
[21:31] <BBB> I mean, we'll have to release 0.8.0 by gnome-2.6
[21:31] <Company> yeah
[21:31] <kleppari> but i'd like to compliment you guys for a great job done so far.
[21:32] <BBB> ok, so you just want to let it stay HEAD for a while?
[21:32] <BBB> that's fine, I guess...
[21:32] <Company> yeah
[21:32] Nick change: ZahirDev -> Zahir
[21:32] <BBB> when will I commit my profile stuff then?
[21:32] <BBB> kleppari: thanks :)
[21:32] <Company> focus on bugfixing, finishing apps and stuf
[21:32] <BBB> kleppari: we hope it's starting to get useful now we're getting closer to 0.8.0
[21:33] <kleppari> yeah, but I really like the idea.
[21:33] <kleppari> people were always reimplenting the same code over and over...
[21:33] <Company> BBB: doesn't avidemux do seeking?
[21:34] <BBB> Company: it does
[21:34] <BBB> or did I break that?
[21:34] Action: BBB thinks
[21:35] <Company> it says it does, but nothing happens
[21:35] <BBB> uhm
[21:35] <BBB> oops
[21:35] <thomasvs> Company: sounds great
[21:35] <thomasvs> Company: I was also hoping we could focus on end user bug fixing a little
[21:35] <BBB> can you file a bug report so I can work on that tonight and tomorrow?
[21:36] <Company> sure
[21:36] <BBB> ty
[21:36] <BBB> (or fix it yourself if you want)
[21:38] <Company> i'm doing clocking and testing seeking...
[21:38] <Company> i just noticed it
[21:38] <BBB> I didn't touch seeking when rifflib'ifying avidemux
[21:38] <BBB> but I might accidently have broken something
[21:44] <thomasvs> can someone tell me what to do so I get the same macro expansion in c++ as in c ?
[21:44] <thomasvs> the preprocessor gives me, for c++, this
[21:44] <thomasvs>       do gst_element_error_2_extended (((((GstElement*) g_type_check_instance_cast ((GTypeInstance*) ((modplug)), ((_gst_element_type)))))), gst_core_error_quark (), GST_CORE_ERROR_NEGOTIATION, g_strdup_printf ("test"), g_strdup_printf ("test"), "gstmodplug.cc", __PRETTY_FUNCTION__, 578) while (0);
[21:44] <thomasvs> I don't get the do/while in C
[21:45] <mathrick> thomasvs: it's G_STMT_START/_END probably
[21:45] <thomasvs> mathrick: no, I don't use that
[21:45] <thomasvs> mathrick: since that adds the do/while around it in C
[21:46] <mathrick> thomasvs: give me that macro definition
[21:46] <thomasvs> oh, hum
[21:47] <thomasvs> the G_STMT stuff did creep back in
[21:47] <thomasvs> #define gst_element_error_2(el, domain, code, message, debug) G_STMT_START \
[21:47] <thomasvs>   gst_element_error_2_extended (GST_ELEMENT(el), \
[21:47] <thomasvs>   GST_ ## domain ## _ERROR, GST_ ## domain ## _ERROR_ ## code, \
[21:47] <thomasvs>   g_strdup_printf message, \
[21:47] <thomasvs>   g_strdup_printf debug, \
[21:47] <thomasvs>   __FILE__, GST_FUNCTION, __LINE__) \
[21:47] <thomasvs> G_STMT_END
[21:47] <thomasvs> d'oh
[21:47] kmaraas ([email protected]) left irc: "Leaving"
[21:47] <mathrick> thomasvs: you need it anyway
[21:47] <thomasvs> removing it fixed it
[21:47] <thomasvs> why do I need it ?
[21:47] <mathrick> thomasvs: otherwise, you break non gcc compat
[21:48] <Company> not in this case
[21:48] <mathrick> hmm, right
[21:48] <thomasvs> hm, seems to work fine without it for both c and c++
[21:48] <mathrick> but generally, it is the case
[21:48] <thomasvs> mathrick: ah, why is that ?
[21:48] <mathrick> thomasvs: it matters for multiline macros
[21:48] <Company> it's only required if one macro expands to multiple statements
[21:49] <Company> (or whatever those are called)
[21:49] <thomasvs> yeah, that's what I thought
[21:49] <mathrick> thomasvs: in gcc, you can enclose it in ({ / }), but in other ones, it's not valid statement
[21:49] ChriHJW ([email protected]) left irc: Read error: 60 (Operation timed out)
[21:50] hadley ([email protected]) joined #gstreamer.
[21:50] dolphy ([email protected]) joined #gstreamer.
[21:50] Action: BBB starts profile loading/saving
[21:50] <BBB> bbl!
[21:50] BBB ([email protected]) left irc: "Client exiting"
[21:53] kleppari ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[21:56] kleppari ([email protected]) joined #gstreamer.
[21:57] kmaraas ([email protected]) joined #gstreamer.
[22:02] Action: Company watches the beginning of Days of Thunder the 200th time today
[22:02] <Company> and since AVI can't seek...
[22:03] <dolphy> ah that's probably why gst-player is not seeking :)
[22:03] <Company> mpeg works
[22:03] <Company> i dunno why avi doesn't
[22:03] <dolphy> any improvement has been done on dvd playback ?
[22:06] wheels ([email protected]) left irc: "back in a minute"
[22:07] wheels ([email protected]) joined #gstreamer.
[22:08] TD ([email protected]) joined #gstreamer.
[22:08] <TD> 'lo guys
[22:16] <dolphy> yo TD
[22:19] Zahir ([email protected]) left irc: "Deu+"
[22:33] <thomasvs> gstv4lmjpegsink.c: In function `gst_v4lmjpegsink_chain':
[22:33] <thomasvs> gstv4lmjpegsink.c:257: warning: implicit declaration of function `gst_element_wait'
[22:33] <thomasvs> make: *** [libgstvideo4linux_la-gstv4lmjpegsink.lo] Error 1
[22:36] sjoerd ([email protected]) left irc: Remote closed the connection
[22:39] <dolphy> thomasvs: update the core probably ?
[22:41] sjoerd ([email protected]) joined #gstreamer.
[22:41] Action: thomasvs is scared about that
[22:45] <dolphy> weee videobalance is now in gstplay
[22:45] <dolphy> but it does not seem to work that well
[22:46] <dolphy> at least it does not product the same effects than xv
[22:48] <thomasvs> in what way ?
[22:49] <dolphy> hmm well as soon as i touch the settings the picture is modified and it s hard to get it back to normal
[22:52] <dolphy> ok got it
[22:53] <dolphy> my range conversion are completely fucke up
[22:53] <dolphy> i ll stop using G_MININT and G_MAXINT
[22:53] <dolphy> that's too much risky
[22:54] <ds-work> why isn't brightness/contrast an xvimagesink property?
[22:56] <dolphy> ds-work: i implemented the interface
[22:56] <dolphy> ds-work: no properties yet
[22:57] <ds-work> dolphy: how am I supposed to play with it in gst-editor then? :)
[22:58] Phiend ([email protected]) joined #gstreamer.
[22:59] <dolphy> ds-work: hey dude my days are only 24 hours :)
[22:59] <dolphy> weee videobalance now works correctly
[22:59] <dolphy> NICCEEEE !! :)
[22:59] <ds-work> dolphy: don't you command a team of high-powered programmers?
[22:59] <dolphy> software color balance is in totem :)
[23:00] <dolphy> ds-work: my only high power programmer is busy rewriting all error messages of all plugins of gstreamer
[23:00] <dolphy> ds-work: so that they can be translated :)
[23:01] <dolphy> ds-work: i ll implement properties right now
[23:01] <dolphy> ds-work: can you fix switch please ?
[23:03] <thomasvs> dolphy: your only high-powered programmer has backstage passes to the best band ever in brussels and amsterdam
[23:03] somex1 ([email protected]) joined #gstreamer.
[23:04] <TD> dolphy: hmm, are you sure they should be translated?
[23:04] <TD> i thought debug messages generally weren't supposed to be translated
[23:04] <ds-work> switch is cool
[23:05] <ds-work> does it pull from both sink pads?
[23:06] _neo_ ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[23:06] mxpxpod ([email protected]) joined #gstreamer.
[23:06] <dolphy> ds-work: it polls from every sink pad
[23:06] <dolphy> ds-work: putting data in a structure
[23:07] <dolphy> ds-work: and only push the active sinks one
[23:07] Action: thomasvs hopes he has network on the plane so he can commit
[23:07] <dolphy> thomasvs: just dream :)
[23:07] Action: ds-work ha's at thomasvs
[23:07] <dolphy> thomasvs: i m implementing properties in x[v]imagesink
[23:07] matrixise ([email protected]) joined #gstreamer.
[23:07] <dolphy> thomasvs: you wanted DISPLAY there ?
[23:08] <thomasvs> dolphy: yeah, but people disagreed
[23:09] <ds-work> eh?
[23:09] <ds-work> there should be a display property in x[v]imagesink
[23:10] <dolphy> who disagreed ?
[23:10] Action: dolphy looks at Company
[23:10] <dolphy> :)
[23:10] <thomasvs> I don't remember :)
[23:12] <dolphy> so what do you want there ?
[23:12] <dolphy> a string
[23:12] <dolphy> when that property is set i cleanup everything and open a connection to that display string
[23:12] <dolphy> ?
[23:12] <ds-work> yes
[23:13] <dolphy> ds-work: when i plug switch in gst-play pipeline i have a green picture of 100 * 100
[23:13] <ds-work> well, no
[23:13] <thomasvs> hm
[23:13] <ds-work> don't change the display until going to NULL
[23:13] <thomasvs> would that allow to transplant a screen to another display on a running pipeline ?
[23:13] <dolphy> ah ok
[23:13] <dolphy> so it's just a CONSTRUCT property
[23:13] <ds-work> thomasvs: almost :)
[23:14] <dolphy> if it s before going to READY i use it
[23:14] <dolphy> otherwise nope
[23:14] <dolphy> right ?
[23:14] <ds-work> CONSTRUCT properties can't be changed after creating the object, correct?
[23:14] <ds-work> that's not exactly right
[23:15] <ds-work> since gst-launch, for example, doesn't do it that way
[23:15] <dolphy> well that was just to describe
[23:15] <dolphy> i won't make it a real construct only property
[23:15] <ds-work> but basically, yes
[23:15] <dolphy> that's what you want ?
[23:15] <dolphy> not live transplantation to another screen ?
[23:15] <dolphy> :)
[23:15] <ds-work> ignore it during READY, PAUSED, PLAYING
[23:16] <dolphy> it's even easier for me
[23:16] <ds-work> you can transplant, it just requires going through NULL first
[23:16] <dolphy> xcontext_get will look at the display name which will be null by default
[23:16] <ds-work> er, not _ignore_, but don't force a display change
[23:16] <dolphy> and state changes will clear xcontext when going to NULL
[23:16] Nick change: Ura_out -> Uraeus
[23:17] <dolphy> so going back from NULL to READY will get the xcontext with the new display
[23:17] <dolphy> deal ?
[23:17] <Uraeus> ds-work: we need you to look at the wavparse caps, BBB fixed the plugin in regards to media handling, but its caps are still not correct
[23:17] <ds-work> Uraeus: ok
[23:18] <Uraeus> good, maybe RB can start playing my wav called an mp3 again then ;)
[23:19] <ds-work> Uraeus: er, what's wrong with it?
[23:19] <Uraeus> ds-work: Ronald said it works when you specify the pipeline, but it is not autoplugged correctly due to bad caps
[23:19] thaytan ([email protected]) joined #gstreamer.
[23:20] <Uraeus> ds-work: could of course be that Ronald eventually fixed it 100% after I talked to him
[23:20] <Uraeus> hi thaytan
[23:20] matrixise ([email protected]) left irc: No route to host
[23:21] <ds-work> Uraeus: could you point me to a mp3-in-a-wav?
[23:22] <Uraeus> ds-work: http://www.linuxrising.com/files/
[23:22] <Uraeus> ds-work: the Spandau Ballet song
[23:23] Action: ds-work fears
[23:23] <LeRoutier> mp3 in a wav ? someone wanted a RIFF compliant mp3 ?
[23:23] matrixise ([email protected]) joined #gstreamer.
[23:24] <ds-work> LeRoutier: it's a common non-id3 way to tag mp3s
[23:24] <ds-work> and actually, much better than id3, imo
[23:25] <LeRoutier> ok, so pure mp3 audio is in an audio stream. and tags are in LIST_INFO or something like that?
[23:26] kmaraas ([email protected]) left irc: "Leaving"
[23:28] minio ([email protected]) joined #gstreamer.
[23:28] <dolphy> ds-work: apart from xhost + what do i need to do on a machine so that external machine can send a window ?
[23:28] <thaytan> good morning
[23:28] <ds-work> dolphy: that should be all
[23:28] <thaytan> maddog is about to do a keynote 
[23:28] <dolphy> ds-work: hmm
[23:28] <dolphy> ds-work: DISPLAY=192.168.2.20:0 gedit 
[23:28] <dolphy> ds-work: reports it can not open display
[23:28] <Uraeus> thaytan: if you see malcomn there ask him to get his act togheter about getting his SMIL library into CVS :)
[23:29] <thaytan> I will bug him
[23:29] <thaytan> he's around in this hall somewhere
[23:29] <thaytan> and I met up with Steve
[23:29] <ds-work> dolphy: you may also need to enable tcp connections in you X server.  many have it turned off
[23:29] <Uraeus> thaytan: ok, make sure steve is more active on the hacking front from here on too :)
[23:30] <ds-work> dolphy: you can test it by using Xnest
[23:30] <thaytan> he's trying to - pressured for time like most of us, of course
[23:30] <ds-work> or Xvnc
[23:30] <ds-work> run Xnest :1
[23:31] <minio> good night I have a question. Would be possible to wirte plugins for eg for .html and for .doc files and use Gstreamer to convert one to other?
[23:31] <thomasvs> thaytan: get jdub to fix the planet too
[23:31] <thomasvs> minio: yeah, it would
[23:31] <thaytan> thomasvs: what's wrong with planet?
[23:32] <thomasvs> minio: gstreamer is in essence extensible to all sort of stream formats
[23:32] <thomasvs> thaytan: advo entries get dropped off after an hour or so
[23:32] <ds-work> minio: yes, but it would be dumb
[23:32] <thomasvs> thaytan: that's why the planet's so sparse the latest few days
[23:32] <thaytan> I'll let him know
[23:32] <dolphy> ds-work: installing xnest
[23:33] <minio> ds-work: why?
[23:33] <ds-work> minio: because .doc and html are not stream formats
[23:34] <Uraeus> ds-work: the binary doc format is isn't it?
[23:34] <ds-work> Uraeus: no
[23:34] <dolphy> ds-work: ok and then ?
[23:34] <dolphy> ds-work: i have the window open
[23:34] <dolphy> ds-work: black one
[23:35] <ds-work> Uraeus: describe how to get to time 7.0s in a .doc :)
[23:35] <ds-work> dolphy: huh?
[23:35] <minio> ds-work: it makes sense :)
[23:35] Phiend ([email protected]) left #gstreamer.
[23:36] <dolphy> ds-work: running Xnest :1 opens a black window :)
[23:36] <dolphy> ds-work: with the X cursor
[23:36] <ds-work> dolphy: ah
[23:36] <ds-work> dolphy: put that in the background, and start a terminal with 'DISPLAY=:1 xterm' or whatever
[23:36] <ds-work> dolphy: or 'DISPLAY=:1 gst-launch videotestsrc ! ximagesink'
[23:37] <ds-work> Uraeus: I think mp3-in-wav is going to be regressed until we get a spider replacement
[23:38] <dolphy> ds-work: Xlib: connection to ":1.0" refused by server
[23:39] hallibaby ([email protected]) joined #gstreamer.
[23:40] hallibaby ([email protected]) left irc: Client Quit
[23:40] bilboe1 ([email protected]) joined #gstreamer.
[23:40] hallibaby ([email protected]) joined #gstreamer.
[23:40] <bilboe1> hi all
[23:40] <ds-work> dolphy: this is all on one machine, right?
[23:40] <Uraeus> hi bilboe1
[23:40] <bilboe1> Uraeus: thx for forwarding the mail
[23:40] Action: ds-work needs to wander off
[23:40] <Uraeus> ds-work: ok, hope we have the spider replacement in before 0.8.0 then :)
[23:40] <dolphy> ds-work: yeah
[23:40] <dolphy> ds-work: it says refused localhost
[23:41] <dolphy> ds-work: probably some security
[23:41] <Uraeus> bilboe1: np, seems we might have found you 2 more developers through my article then :)
[23:41] minio ([email protected]) left #gstreamer ("Leaving").
[23:41] <bilboe1> Uraeus: so it seems, thx a lot
[23:41] <thomasvs> dolphy: xhost + in the terminal you created ?
[23:41] <bilboe1> Uraeus: more brains = better soft
[23:41] <Uraeus> bilboe1: I guess external participants makes your administrator job a little harder though
[23:41] <thomasvs> dolphy: ie, inside the terminal inside the nest session
[23:41] <dolphy> thomasvs: i haven't created any terminal in that session
[23:42] <bilboe1> Uraeus: I'm so close of going crazy... and I'm leaving shortly for a week skiing... followed by 14months of intense coding :)
[23:42] <thomasvs> dolphy: start Xnest with -ac
[23:43] <dolphy> lovely
[23:43] <dolphy> thx
[23:43] <dolphy> k so the display property works fine
[23:43] <thomasvs> dolphy: add it to my christmas bonus, thanks
[23:44] <Uraeus> bilboe1: well I hope you have a plan for how to actually do the coding, with 11 people involved and starting almost from scratch I guess it can easily be a lot of arms and legs flying around without much direction
[23:44] Action: dolphy just think about the case where people might try to set colorbalance on a NULL xvimagesink
[23:44] <dolphy> hmmm
[23:45] <dolphy> i m wondering if i should store the colorbalance value to set them on NULL_TO_READY
[23:45] <dolphy> ds-work: what's your opinion ?
[23:45] <bilboe1> Uraeus: you could picture it like that :(
[23:45] <thomasvs> [gst-cvs] [thomas@otto videofilter]$ gst-inspect videobalance
[23:45] <thomasvs> /home/thomas/gst/cvs/gstreamer/tools/.libs/lt-gst-inspect: relocation error: /home/thomas/gst/cvs/gst-plugins/gst/videofilter/.libs/libgstvideobalance.so: undefined symbol: gst_color_balance_get_type
[23:45] <thomasvs> shite
[23:45] <dolphy> ?
[23:45] <dolphy> it works fine here
[23:46] <bilboe1> Uraeus: but we'll start with a nice division into subgroups so it's a bit easier to maintain
[23:46] <dolphy> fresh checkout completely rebuilt
[23:46] <bilboe1> in fact I was here to see about, maybe, if possible, an account (cvs/wiki) on fdo ?
[23:46] <Uraeus> bilboe1: yeah, I guess you have 2-3 design the core of the app, while the rest do general gstreamer bugfixing, gnonlin updates, new plugins etc.
[23:46] <bilboe1> for the global editor project
[23:47] <bilboe1> Uraeus: and a lot of work on gui and user-friendlyness
[23:47] <Uraeus> bilboe1: thomasvs will surely be able to set you up with that
[23:47] <Uraeus> bilboe1: also you can add new modules to gnonlin CVS if you want, you have full access to that project now
[23:47] <bilboe1> Uraeus: it would mostly be to have a wiki asap, so we can start exchanging ideas easily with a lot of outside people
[23:48] <Uraeus> bilboe1: well you should feel free to use the gstreamer.net wiki
[23:48] herzi ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[23:49] <bilboe1> Uraeus: It'll be very non-linear editing oriented
[23:49] <Uraeus> bilboe1: yeah, but just add it as a new section. Not sure freedesktop has a wiki system set up
[23:49] <bilboe1> I thought there was a wiki and a cvs...
[23:50] <Uraeus> actually freedesktop seems to be a giant wiki :)
[23:50] <bilboe1> lol
[23:50] walters ([email protected]) joined #gstreamer.
[23:50] <Uraeus> ok, time for me to sleep
[23:50] herzi ([email protected]) joined #gstreamer.
[23:50] <Uraeus> hi and goodnight walters
[23:50] Uraeus ([email protected]) left irc: "Client exiting"
[23:50] <hallibaby> sorry for this question @ your irc: OFFTOPIC: some pango specialists in here?
[23:51] minio ([email protected]) joined #gstreamer.
[23:52] minio ([email protected]) left irc: Client Quit
[23:52] <bilboe1> thomasvs: is there wiki on fdo ?
[23:52] <thomasvs> bilboe1: yeah, there is
[23:53] <dolphy> thomasvs: i m wondering if the DISPLAY property is usefull for xvimagesink
[23:53] <dolphy> thomasvs: remote display does not provide XV support
[23:54] <dolphy> thomasvs: hmm still that let you specify the display if you are in a ssh with X11 forwarding
[23:54] <dolphy> thomasvs: hmm might be  usefull
[00:00] --- Thu Jan 15 2004
[00:01] <dolphy> thomasvs: display properties commited
[00:01] <dolphy> thomasvs: the colorbalance one in xvimagesink are not usable yet
[00:01] <dolphy> thomasvs: i ll finnish that tomorrow
[00:03] kmaraas ([email protected]) joined #gstreamer.
[00:04] TD ([email protected]) left irc: "zzzz"
[00:04] dolphy ([email protected]) left irc: "Network down, IP Packets delivered via UPS"
[00:14] <LeRoutier>  
[00:14] <LeRoutier> got big trouble playing a wma file
[00:15] <LeRoutier> asfdemux and ffmpeg shouting
[00:15] <LeRoutier> WARNING **: asfdemux: Audio header contains 10 bytes of surplus data   x1
[00:15] <LeRoutier> WARNING **: ffdec_wmav2: decoding error x ... (a lot)
[00:15] <hallibaby> i'm having trouble in demuxing a mpeg file and muxing again in my own file format. any hints?
[00:18] <mathrick> hallibaby: pipeline?
[00:21] <hallibaby> @mathrick: It was a long ago (ca. 2-3 month) the last time I did some work on my demuxer/muxing stuff. I need some time to remember how to explain my pipeline. Is it okay to use some lines? -> That means I cannot explain in a single line
[00:22] <mathrick> hallibaby: cannot you just describe it like filesrc ! mpegdemux ! ... ?
[00:23] hallibaby ([email protected]) left irc: "Bye bye"
[00:23] hallibaby ([email protected]) joined #gstreamer.
[00:23] <hallibaby> sorry for disconnect. wrong key on keyboard
[00:23] <mathrick> hallibaby: it happens :)
[00:23] <hallibaby> @mathrick: Input: file
[00:24] <hallibaby> @mathrick: demuxer: mpeg2demux
[00:24] <LeRoutier> argh. asf demux only supports ASF v1, not v2. and the file i have is v22
[00:24] <LeRoutier> v2
[00:25] <hallibaby> @mathrick: video0|1 -> queue -> videoscale -> colorspace -> (my_video_convertor) -> queue
[00:25] <hallibaby> @mathrick: audio0|1 -> queue -> audioscale -> (my_audio_converter) -> queue
[00:26] <hallibaby> @mathrick: both (video & audio) queues -> (my_muxer) -> Output: file
[00:26] <hallibaby> @mathrick: Does this describe any pipeline?
[00:26] <hallibaby> @mathrick: Badly I think not! 
[00:26] <hallibaby> @all: correct my english it is okay :)
[00:27] <walters> Company: here?
[00:27] <mathrick> hallibaby: ugh, I wish you just used standard gst-launch syntax ...
[00:28] <hallibaby> @mathrick: and that's the problem :(
[00:32] <mathrick> hallibaby: filesrc ! mpeg2demux name=demux .video_00 ! { queue ! videoscale ! colorspace ! <your video conv> ! queue name=qvid } demux .audio_00 ! { queue ! audioscale ! <your audio conv> ! queue name=qaud } qvid ! <your mux> name=mux ! filesink qaud ! mux ! filesink
[00:32] <mathrick> ?
[00:33] <thaytan> almost
[00:33] <mathrick> right, forgot some threads
[00:33] <hallibaby> @mathrick: I think this could be right????? 
[00:33] <thaytan> demux.audio_00 has no space 
[00:33] <mathrick> thaytan: yes it has
[00:33] <hallibaby> @mathrick: Guessing the syntax!
[00:34] <thaytan> not in 0.7
[00:34] <mathrick> thaytan: yes in 0.7
[00:35] <thaytan> it will look for an element 'demux' to create
[00:35] <thaytan> also, don't need the second mux ! filesink - that should just be 'mux'
[00:36] <mathrick> thaytan: hmm, right
[00:36] <mathrick> thaytan: second filesink is redundant
[00:36] <thaytan> gst-launch-0.7 dvdnavsrc ! tee name=tee ! mpegdemux name=demux .video_00 ! mpeg2dec ! xvimagesink demux .audio_00 ! a52dec ! esdsink tee.src01 ! filesink location=~/test.mpg  < -dvd ripper
[00:36] <thaytan> urkl
[00:36] <thaytan> minus that space :P
[00:37] <thaytan> gst-launch-0.7 dvdnavsrc ! tee name=tee ! mpegdemux name=demux .video_00 ! mpeg2dec ! xvimagesink demux.audio_00 ! a52dec ! esdsink tee.src01 ! filesink location=~/test.mpg
[00:37] <hallibaby> @thaytan: ??????? (does not understand a word :((((
[00:37] <thaytan> hallibaby: it's an example pipeline
[00:37] <thaytan> it plays mpeg off a DVD, and also sends it to a file 
[00:38] <hallibaby> @thaytan: Then it my problem to understand the syntay... By the way how do you hightlight your name in xchat if running?
[00:38] <mathrick> thaytan: ok, you won ;)
[00:38] <thaytan> hallibaby: the highlight is because I mentioned your name
[00:39] <thaytan> breakfast time!
[00:39] <thaytan> see you later :)
[00:40] <hallibaby> @all: How to mention a (pseudo)-name in xchat???? Just for better visibility!
[00:40] thaytan ([email protected]) left irc: "See y'all"
[00:41] <mathrick> hallibaby: gstreamer/docs/random/company/gstparse <-- it's the best info for now, and I will revise manpage in ~2 weeks
[00:41] <mathrick> hallibaby: just write it anywhere on the line
[00:41] <mathrick> like that: hallibaby
[00:42] <hallibaby> mathrick: without writing @mathrick it will work? cool :)===)
[00:42] <hallibaby> mathrick: Are you responsible for the documentation?
[00:43] <mathrick> hallibaby: not really. I'm just looking into manpages to make them more newb-friendly
[00:44] <mathrick> hallibaby: so, it should look like hallibaby: filesrc ! mpeg2demux name=demux .video_00 ! { queue ! videoscale ! colorspace ! <your video conv> ! queue name=qvid } demux.audio_00 ! { queue ! audioscale ! <your audio conv> ! queue name=qaud } qvid ! { <your mux> name=mux ! filesink qaud ! mux } i think
[00:45] <mathrick> hallibaby: sorry, but I don't have any more time to help you now, drop by later :)
[00:45] Action: mathrick is busy (study)
[00:45] <hallibaby> @mathrick: see you
[00:46] <hallibaby> @mathrick: Thinking of I understood the syntax of pipelining: YES!
[00:48] steve_b ([email protected]) joined #gstreamer.
[00:48] Action: foser is back (gone 06:52:26)
[00:49] iain ([email protected]) left irc: "bog off"
[00:49] <LeRoutier>  
[00:50] <LeRoutier> anyone have asfdemux working ?
[00:52] <hallibaby> @all: what's your time? my: 00:51 am or 00:51:00
[00:52] <LeRoutier> hallibaby: same for me 
[00:53] <hallibaby> @LeRoutier: So you should be in the same continent! ;)
[00:55] <bilboe1> GMT+1 rocks :)
[00:56] <LeRoutier> FRance
[00:57] <hallibaby> @LeRoutier: thought of it because of you (pseudo-)name ;)
[00:58] Nick change: mxpxpod -> mx|gone
[00:59] <LeRoutier>  
[00:59] Action: ds-work shows hallibaby /ctcp time
[01:00] <LeRoutier> as most gstreamer devs did not know what to do tomorrow because their TODO list are empty, i've got a new bug for them : http://bugzilla.gnome.org/show_bug.cgi?id=131502
[01:01] <hallibaby> @ds-work: ermmmm???? sorry..... /ctcp did nothing
[01:02] <ds-work> do /ctcp ds-work time
[01:03] kleppari ([email protected]) left irc: Remote closed the connection
[01:03] <hallibaby> @ds-work: oops it's really early in the evening :(
[01:03] Nick change: mx|gone -> mxpxpod
[01:07] bilboe1 ([email protected]) left #gstreamer.
[01:08] <hallibaby> bye... and good night at 01:08 am ;)
[01:09] hallibaby ([email protected]) left irc: "Bye bye"
[01:16] <mathrick> ok, gotta sleep a bit
[01:16] hadley ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[01:16] <mathrick> nite
[01:17] Nick change: mathrick -> mathrick|sleep
[01:20] Miko5881 ([email protected]) joined #gstreamer.
[01:20] <walters> hmmm.
[01:27] Nick change: kmaraas -> km_sleep
[01:28] <taaz> hmmm?
[01:30] <walters> taaz: trying to track down an infloop in opt :/
[01:30] <taaz> good luck
[01:31] <walters> the interesting thing is it only happens on some of my music
[01:32] <walters> so somehow the file content makes a difference
[01:32] <taaz> is it deterministic?
[01:32] <walters> yep
[01:32] <taaz> then you have hope ;)
[01:33] <walters> oh wait a second
[01:33] <walters> this file is empty. :)
[01:33] Action: walters just noticed that his scp returned suspiciously fast...
[01:34] thaytan ([email protected]) joined #gstreamer.
[01:34] <taaz> whoa!  gst-plugins made it into testing!
[01:35] <taaz> as did rb!
[01:35] <thaytan> toot sweet!
[01:35] steve_b ([email protected]) left irc: Read error: 60 (Operation timed out)
[01:35] Action: walters is glad he hard-link-backuped his music
[01:36] Action: walters has apparently already erased several files testing rb metadata writing :/
[01:36] mxpxpod ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[01:39] <Company> walters: pong
[01:39] <thaytan> hi Company
[01:39] Action: Company needs to sleep more
[01:39] <thaytan> clocking changes seem good here
[01:40] mxpxpod ([email protected]) joined #gstreamer.
[01:40] <Company> thaytan: i need to submit a thing and after that seeking should work, too
[01:40] <walters> Company: if you want to look at a fun bug: gst-launch-0.7 -t gnomevfssrc location=/dev/null ! typefind ! spider ! application/x-gst-tags ! fakesink
[01:40] <thaytan> my cute pipeline of the day: gst-launch-0.7 dvdnavsrc ! tee name=tee ! mpegdemux name=demux .video_00 ! mpeg2dec ! xvimagesink demux.audio_00 ! a52dec ! esdsink tee.src01 ! filesink location=~/test.mpg
[01:40] <thaytan> instant dvd ripper :)
[01:41] <sxpert> hahahah
[01:41] sublett ([email protected]) left irc: "I like food, food is good!"
[01:41] <deadchip> :)
[01:41] <sxpert> thaytan: you could throw ffmpeg in there to compress to mp4, for good measure ;D
[01:41] <thaytan> that is just saving the mpeg though, not transcoding at all
[01:42] <thaytan> but yes :)
[01:42] <sxpert> thaytan: possible to do multiple passes ?
[01:44] <thaytan> I guess so - haven't looked into it
[01:44] <ds-work> taaz: have you been following the libjack topic on debian-multimedia?
[01:44] alley_cat ([email protected]) left irc: "May the Source be with you!"
[01:46] <taaz> ds-work: there's a debian-multimedia? ;)
[01:46] <taaz> ie, no.  anything i should check out?
[01:47] <ds-work> taaz: there's talk about doing a libjack transition to 0.91.1 before sarge
[01:47] <ds-work> which will be, er, interesting
[01:48] <taaz> does gst support that?
[01:48] <ds-work> it's api compatible, iirc
[01:48] <taaz> rebuild packages, wait.  what's the problem? ;)
[01:49] <ds-work> because not everything builds all the time
[01:49] <Company> walters: gst-launch-0.7 -t gnomevfssrc location=/dev/null ! typefind ! spider ! application/x-gst-tags ! fakesink --gst-scheduler=basicomega
[01:49] pb_ ([email protected]) left irc: "later"
[01:49] Action: taaz subs to d-m newsgroup on gmane
[01:50] <walters> Company: yeah
[01:50] <walters> Company: it's definitely an opt bu
[01:50] <walters> g
[01:51] <Company> walters: i suspect it's the same one that makes me not use opt anymore
[01:51] <Company> walters: do you use alsasink or osssink?
[01:51] <walters> Company: oss
[01:52] <Company> walters: because I use alsasink and that triggers this inf loop for me during playing
[01:52] <walters> Company: ah.
[01:52] <walters> Company: well, i've tracked down that the scheduler state is successfully set to _STOPPED...i think opt just needs to check that somewhere
[01:52] <Company> walters: but it's great you found that, i'll give iain that pipeline so he can pull out his 1-liners :)
[01:53] <walters> i think i see it
[01:54] <ds-work> why doesn't gst-launch stop anymore when I hit ctrl-c?
[01:54] Action: ds-work wonders if it's an opt regression
[01:55] <Company> ds-work: i fear that's because some race in the signal handlers in gst-launch
[01:55] <walters> ds-work: apparently you have to do it twice
[01:55] <Company> it still doesn't work sometimes
[01:56] <ds-work> Company: unlikely, since it stopped working about a week ago
[01:56] <ds-work> Company: there haven't been any relevant changes to gst-launch.c
[01:56] <Company> hm, i had to kill it earlier
[01:56] <Company> i'll just do a gdb attach when this happens again
[01:57] Action: ds-work loves the '1 week ago' tag
[01:57] <Company> it only happens when I debug other things so i never look at it
[01:57] <Company> i just bitch in here
[01:59] Action: Company hates songs which don't center the drums
[01:59] <Company> at least when listening with headphones
[02:02] <Company> walters: found it?
[02:02] <walters> Company: nope :/
[02:04] thaytan ([email protected]) left irc: "See y'all"
[02:05] Action: Company wants gst_bin_print
[02:08] thaytan ([email protected]) joined #gstreamer.
[02:14] Marsupilami23 ([email protected]) left irc: "I don't wanna grow up, I'm a Toys R Us Kid. There's a million toys at Toys R Us that I can play with!"
[02:17] bforbes ([email protected]) joined #gstreamer.
[02:18] mxpxpod ([email protected]) left irc: Nick collision from services.
[02:18] Nick change: bforbes -> mxpxpod
[02:18] <Company> when are we gonna bite the bullet and disable scheduling paused elements in _pull/_push
[02:19] <Company> walters: found it
[02:19] <Company> walters: spider ;)
[02:19] <walters> Company: ah, cool.
[02:20] <ds-work> Company: we need a --gst-enable-evil
[02:21] <LeRoutier> well, spider + optscheduler is enougth to enable this command line switch
[02:25] <Company> ds-work: "Preferences are bad solutions to performance problems." --hp, 20-nov-0
[02:26] <ds-work> heh, true
[02:26] NFusi0n ([email protected]) joined #gstreamer.
[02:26] <ds-work> but it would be in order to get closer to a solution
[02:27] <Company> enable it by default in cvs builds
[02:27] <walters> Company: going to commit this fix or what? :)
[02:27] <Company> walters: i'm at it
[02:28] <NFusi0n> Has there been a sudden influx of traffic into #gstreamer since the /. and OSNews news?
[02:28] <Company> i needed to poke Rupert for a while until i got that quote ;)
[02:28] steve_b ([email protected]) joined #gstreamer.
[02:29] <ds-work> NFusi0n: no, just you :)
[02:29] <NFusi0n> :)
[02:30] <NFusi0n> i checked gstreamer.sf.net several times a week, never knew there was an irc channel
[02:30] <ds-work> NFusi0n: there have been a few old faces that decided to stop by
[02:30] foser ([email protected]) left irc: "[ I want to believe ]"
[02:31] <Company> NFusi0n: yes, but more because the devs are proud of themselves now and talk more than because of people getting interested ;)
[02:31] <Company> ds-work: there were others earlier today
[02:31] <Company> weeeee, no more --gst-scheduler=basicomega :)
[02:32] <LeRoutier> Company, it is the default now or is opt repaired ?
[02:32] <ds-work> my install now checks for pushing or pulling to/from paused elements
[02:33] <ds-work> and now it segfaults
[02:33] <ds-work> :)
[02:34] <Company> LeRoutier: opt works now
[02:34] <Company> ds-work: great, something to fix for you :p
[02:35] <LeRoutier> good, thanks
[02:35] <Company> ds-work: it's probably a bit problematic wrt threads, because they start running before all elements are in PLAYING
[02:35] <LeRoutier> playing with it so
[02:35] <Company> walters: fix is in
[02:37] matrixise ([email protected]) left irc: "Leaving"
[02:37] thaytan ([email protected]) left irc: Read error: 110 (Connection timed out)
[02:38] <Company> LeRoutier: re #131502 - could you br on g_log to get a bt for the "unimplemented varargs field" thing?
[02:39] <Company> LeRoutier: i suspect it's a varargs function that's not properly terminated inside asfdemux somewhere and it's easier to look at the stacktrace
[02:39] <ds-work> Company: 99% of the time, that's a missing NULL
[02:39] thaytan ([email protected]) joined #gstreamer.
[02:39] <ds-work> or use G_DEBUG=fatal_warnings
[02:39] <Company> ds-work: i know, but i wanna know where :)
[02:39] <LeRoutier> well, i suppose all my asf problems are because asfdemux was not updated to support ASF v2 specs
[02:41] <LeRoutier> gstreamer (with latest optsched fix) compiled and installed. trying now
[02:43] <LeRoutier> with G_DEBUG, it fails before this error message, at the "** WARNING **: asfdemux: Audio header contains 10 bytes of surplus data" level
[02:43] <LeRoutier> G_DEBUG=fatal_warnings gst-launch-0.7 filesrc location=/mnt/windows/R/12\ Tout\ chez\ moi\ l\'habite.wma ! asfdemux ! spider ! osssink
[02:44] <LeRoutier> or is my pipeline bad ?
[02:44] <Company> yeah, that should not be a g_warning
[02:44] <Company> it's normal with wma, but noone removed that yet...
[02:44] Action: ds-work grumbles
[02:44] <ds-work> it's a spider bug
[02:45] <ds-work> spider pulls for typefinding before negotiating the sink pad
[02:46] Action: ds-work goes back to real work
[02:47] <LeRoutier> Company, if you want me to run this pipeline via gdb, i'm ready for your instructions
[02:48] <walters> Company: cool.  now, next bug :)
[02:48] <walters> walters@nexus> examples/retag/retag ~/tmp/NIN\ -\ Wish.mp3                                                         /src/cvs/cvs.freedesktop.org/gstreamer
[02:48] <walters>                                                                                                                                                           
[02:48] <walters> (process:15482): GStreamer-CRITICAL **: file gstpad.c: line 1275 (gst_pad_renegotiate): assertion `GST_PAD_LINK_SINK (pad)' failed
[02:48] <walters>                                                                                                                                                           
[02:48] <walters> ** ERROR **: file gstid3tag.c: line 758 (gst_id3_tag_do_caps_nego): assertion failed: (tag->found_caps == NULL)
[02:49] <walters> aborting...
[02:49] <walters> zsh: abort      examples/retag/retag ~/tmp/NIN\ -\ Wish.mp3
[02:50] matrixise ([email protected]) joined #gstreamer.
[02:53] LeRoutier ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[02:54] LeRoutier ([email protected]) joined #gstreamer.
[02:54] <LeRoutier> re
[02:54] <LeRoutier> hard-lockup. happy 2.4.24


-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html