IRC Logs

*******************************************************************
[03:03] BBB ([email protected]) joined #gstreamer.
[03:03] iain ([email protected]) left irc: "leaving"
[03:17] smoke ([email protected]) got netsplit.
[03:23] smoke ([email protected]) got lost in the net-split.
[03:39] <ds-work> wheels: still around?
[03:57] BBB ([email protected]) left irc: Read error: 110 (Connection timed out)
[04:01] trow ([email protected]) left irc: "Leaving"
[04:03] BBB ([email protected]) joined #gstreamer.
[04:09] jcsston ([email protected]) joined #gstreamer.
[04:25] BBB_ ([email protected]) joined #gstreamer.
[04:26] BBB__ ([email protected]) joined #gstreamer.
[04:26] walters ([email protected]) joined #gstreamer.
[04:26] BBB ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[04:32] walters ([email protected]) left irc: Client Quit
[04:32] BBB__ ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[04:33] BBB_ ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[04:48] joem ([email protected]) left irc: Remote closed the connection
[04:51] Company ([email protected]) left irc: Remote closed the connection
[04:58] BBB ([email protected]) joined #gstreamer.
[04:59] BBB_ ([email protected]) joined #gstreamer.
[05:00] BBB ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[05:00] BBB__ ([email protected]) joined #gstreamer.
[05:01] walters ([email protected]) joined #gstreamer.
[05:06] Nick change: BBB__ -> BBB
[05:08] BBB_ ([email protected]) left irc: Read error: 60 (Operation timed out)
[05:10] BBB_ ([email protected]) joined #gstreamer.
[05:11] BBB ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[05:11] BBB__ ([email protected]) joined #gstreamer.
[05:13] harshy ([email protected]) left irc: "I quit for now"
[05:17] BBB_ ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[05:18] BBB__ ([email protected]) left irc: Read error: 60 (Operation timed out)
[05:29] BBB ([email protected]) joined #gstreamer.
[05:47] BBB_ ([email protected]) joined #gstreamer.
[06:01] BBB_ ([email protected]) left irc: "Client exiting"
[06:04] BBB ([email protected]) left irc: Read error: 110 (Connection timed out)
[06:04] walters ([email protected]) left irc: "Off to build shareholder value"
[06:15] jimmy_dean ([email protected]) left irc: Read error: 110 (Connection timed out)
[06:50] thomasvs ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[06:55] Nick change: kh_out -> khalek
[07:06] ChrisHJW ([email protected]) joined #gstreamer.
[07:08] thomasvs ([email protected]) joined #gstreamer.
[07:13] Marsupilami23 ([email protected]) left irc: Remote closed the connection
[07:18] joem ([email protected]) joined #gstreamer.
[07:21] Action: mathrick laughs at his own stupidity
[07:21] <thaytoo> yo mathrick 
[07:21] <mathrick> yo aussie
[07:23] <mathrick> heh, I tried to save time by using online translation service. I thought I'd do slight corrections and I'd be done ;)
[07:23] <mathrick> obviously, this didn't work as expected :)
[07:24] <thaytoo> oops :)
[07:25] thaytoo ([email protected]) left irc: "hometime!"
[07:31] Nick change: mx|gone -> mxpxpod
[07:42] <mathrick> mxpxpod: I see you're half asleep ;)
[07:42] <mxpxpod> lol
[07:42] <mxpxpod> had to talk to some guys in #breakmygentoo
[07:43] Nick change: mxpxpod -> mx|sleep
[07:43] <mathrick> nite :)
[07:43] <mx|sleep> now I'm completely asleep :)
[07:43] <mx|sleep> nite
[08:17] Shoragan ([email protected]) joined #gstreamer.
[08:20] Action: sxpert is back (gone 07:48:02)
[08:21] ChrisHJW ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[08:24] jimmy_dean ([email protected]) joined #gstreamer.
[08:29] joem ([email protected]) left irc: Remote closed the connection
[08:37] jcsston ([email protected]) left irc: Read error: 110 (Connection timed out)
[08:45] joem ([email protected]) joined #gstreamer.
[09:18] apoc__ ([email protected]) left irc: "Leaving"
[09:42] ChriHJW_log ([email protected]) joined #gstreamer.
[09:51] thaytan ([email protected]) joined #gstreamer.
[09:56] ChrisHJW_log ([email protected]) left irc: Read error: 110 (Connection timed out)
[10:04] markey ([email protected]) joined #gstreamer.
[10:32] jdahlin ([email protected]) joined #gstreamer.
[10:34] sublett ([email protected]) joined #gstreamer.
[10:35] somex1 ([email protected]) left #gstreamer.
[10:40] <jdahlin> danb?
[10:40] dolphy ([email protected]) joined #gstreamer.
[10:41] <danb> jdahlin: yeah
[10:43] <jdahlin> danb: I'd recommend you not to use python threads in a gst program
[10:43] <danb> jdahlin: hum
[10:43] <jdahlin> danb: it'll take quite a bit of hacking to get it to work correctly
[10:43] <danb> jdahlin: why not?
[10:44] <jdahlin> danb: I've looked a bit at it, and it's completely broken.
[10:44] <danb> hmm...
[10:46] <danb> jdahlin: is there a short description of the problem?
[10:46] <danb> jdahlin: or a long one that you're willing to give? ;)
[10:47] <jdahlin> danb: basically, there is one big lock for python 2.2 or earlier
[10:48] <jdahlin> danb: to be able to have more than one thread accessing python code we need to require 2.3
[10:48] <jdahlin> danb: on top of that gobject isn't threadsafe, nor are gstreamer itself
[10:48] <danb> jdahlin: hehe. that's fine with me. aren't we already requiring gst 0.8?
[10:49] <danb> jdahlin: and i can hack around lack of thread-safetly in gobject and gst. i already know where i need to do it
[10:51] <danb> jdahlin: are those the only problems that you know of?
[10:54] <jdahlin> danb: integrating it with python and pygobject is the main problem right now
[10:54] <jdahlin> which I have no idea how to do, to be honest
[10:54] <danb> ic
[10:54] <danb> if i make my own uses of pygobject stuff synchronized, will i be alright?
[10:55] <jdahlin> no, you need support in the gst bindings first
[10:55] <danb> is this a safety or liveness issue?
[10:56] <danb> is this a deadlock issue or just lack of synchronization?
[10:56] <jdahlin> something similar to GDK_THREADS_ENTER/LEAVE
[10:57] <danb> hmm
[10:57] <danb> i'm not familiar with that
[10:58] <jdahlin> then you have to protect all gst calls with gst.threads_enter()
[10:58] <danb> interesting
[10:59] <danb> is there an index on the gstreamer reference docs?
[10:59] <jdahlin> not that I am aware of
[11:00] <danb> that seems slightly braindead...
[11:00] <danb> do you know if threads_enter is in the reference, and if so, where?
[11:02] <danb> and if i wrap all gst calls, will i be alright?
[11:02] Nick change: markey -> markey|afk
[11:02] <jdahlin> no, there is no threads_enter in gst
[11:02] <jdahlin> but there is one is gdk..
[11:02] <danb> oh...
[11:03] <danb> k, found a reference on that, but there are no words to describe it. hum
[11:03] <danb> ah, here is one with descriptions
[11:09] <danb> jdahlin: so you mean gdk.threads_enter() (instead of gst.)?
[11:10] Rotty ([email protected]) joined #gstreamer.
[11:10] <jdahlin> danb, I want to write gst.threads_enter()/leave(), mimicing the gdk ones
[11:12] <thaytan> yo danb, jdahlin
[11:13] <jdahlin> lo thaytan
[11:16] markey|afk ([email protected]) left irc: "leaving"
[11:24] <danb> ho
[11:24] jonathang ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[11:24] <danb> jdahlin: ic
[11:28] pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) joined #gstreamer.
[11:50] bilboed_ ([email protected]) got netsplit.
[11:50] aldug ([email protected]) got netsplit.
[11:50] hyriand ([email protected]) got netsplit.
[11:50] bitshifter ([email protected]) got netsplit.
[11:50] sxpert_work ([email protected]) got netsplit.
[11:50] sublett ([email protected]) got netsplit.
[11:50] sub_pop ([email protected]) got netsplit.
[11:50] snikkt ([email protected]) got netsplit.
[11:50] Shoragan ([email protected]) got netsplit.
[11:50] pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) got netsplit.
[11:50] alley_cat ([email protected]) got netsplit.
[11:50] sri ([[email protected]) got netsplit.
[11:50] camh ([email protected]) got netsplit.
[11:50] Jara[zZ] ([email protected]) got netsplit.
[11:50] grub_booter ([email protected]) got netsplit.
[11:50] sjoerd ([email protected]) got netsplit.
[11:50] md` ([email protected]) got netsplit.
[11:50] thomasvs ([email protected]) got netsplit.
[11:50] mx|sleep ([email protected]) got netsplit.
[11:50] thaytan ([email protected]) got netsplit.
[11:50] tromey ([email protected]) got netsplit.
[11:51] bilboed_ ([email protected]) returned to #gstreamer.
[11:56] snikkt ([email protected]) got lost in the net-split.
[11:56] grub_booter ([email protected]) got lost in the net-split.
[11:56] Jara[zZ] ([email protected]) got lost in the net-split.
[11:56] camh ([email protected]) got lost in the net-split.
[11:56] bitshifter ([email protected]) got lost in the net-split.
[11:56] sxpert_work ([email protected]) got lost in the net-split.
[11:56] sjoerd ([email protected]) got lost in the net-split.
[11:56] sri ([[email protected]) got lost in the net-split.
[11:56] md` ([email protected]) got lost in the net-split.
[11:56] tromey ([email protected]) got lost in the net-split.
[11:56] alley_cat ([email protected]) got lost in the net-split.
[11:56] hyriand ([email protected]) got lost in the net-split.
[11:56] mx|sleep ([email protected]) got lost in the net-split.
[11:56] aldug ([email protected]) got lost in the net-split.
[11:56] sub_pop ([email protected]) got lost in the net-split.
[11:56] thomasvs ([email protected]) got lost in the net-split.
[11:56] Shoragan ([email protected]) got lost in the net-split.
[11:56] thaytan ([email protected]) got lost in the net-split.
[11:56] sublett ([email protected]) got lost in the net-split.
[11:56] pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) got lost in the net-split.
[11:56] aldug ([email protected]) joined #gstreamer.
[11:56] hyriand ([email protected]) joined #gstreamer.
[11:56] sxpert_work ([email protected]) joined #gstreamer.
[11:56] bitshifter ([email protected]) joined #gstreamer.
[11:59] Shoragan ([email protected]) joined #gstreamer.
[12:01] thomasvs ([email protected]) joined #gstreamer.
[12:01] mx|sleep ([email protected]) joined #gstreamer.
[12:01] thaytan ([email protected]) joined #gstreamer.
[12:01] tromey ([email protected]) joined #gstreamer.
[12:01] pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) joined #gstreamer.
[12:01] alley_cat ([email protected]) joined #gstreamer.
[12:01] md` ([email protected]) joined #gstreamer.
[12:01] sri ([[email protected]) joined #gstreamer.
[12:01] sjoerd ([email protected]) joined #gstreamer.
[12:01] camh ([email protected]) joined #gstreamer.
[12:01] grub_booter ([email protected]) joined #gstreamer.
[12:01] Jara[zZ] ([email protected]) joined #gstreamer.
[12:01] sublett ([email protected]) joined #gstreamer.
[12:01] sub_pop ([email protected]) joined #gstreamer.
[12:01] snikkt ([email protected]) joined #gstreamer.
[12:01] wheels ([email protected]) left irc: "using sirc version 2.211+KSIRC/1.3.9"
[12:01] sub_pop ([email protected]) left irc: Excess Flood
[12:01] sub_pop ([email protected]) joined #gstreamer.
[12:10] danb ([email protected]) left irc: "hrm"
[12:11] danb ([email protected]) joined #gstreamer.
[12:51] iain ([email protected]) joined #gstreamer.
[13:00] smoke ([email protected]) joined #gstreamer.
[13:14] <iain> hmm, deinterleave is breaking for multiple channels because it always thinks its getting audio/x-raw-int data
[13:14] <iain> even when its getting float
[13:19] <iain> ohhh, dumb bug
[13:20] <iain> thomasvs: how do I get emacs to do the desired gstreamer formating?
[13:21] <iain> yey, marlin works again
[13:21] <thaytan> iain: rock! 
[13:27] <thomasvs> iain: jdahlin was going to put the emacs mode line together, I was going to do the vim one
[13:27] <thomasvs> jdahlin: poke ?
[13:28] <jdahlin> thomasvs: peek
[13:28] <jdahlin> thomasvs: I started to look into it, but it appers to be a lot of work.
[13:29] <jdahlin> thomasvs: eg, finding out exactly what the indentation options are called in emacs
[13:29] markey ([email protected]) joined #gstreamer.
[13:30] <iain> so how do I reformat my code so I can commit it?
[13:30] <jdahlin> you don't need to
[13:30] <jdahlin> it'll be indented on commit
[13:30] Action: jdahlin shrugs
[13:31] <iain> so when I recheck out, it'll be different and I'll need to recompile?
[13:31] <thomasvs> iain: or run gst-indent on it before you commit
[13:32] <jdahlin> or just put /* INDENT-OFF */ in the beginning of the file ;)
[13:32] <jdahlin> /* *INDENT-OFF* */
[13:33] Action: iain still dislikes the indent style
[13:33] <iain> seeing as it doesn't match ANY other projects indent style
[13:33] <jdahlin> at least one more
[13:36] <jdahlin>        When indent is run twice on a file, with the same  profile,  it  should
[13:36] <jdahlin>        never  change  that  file  the second time.  With the current design of
[13:36] <jdahlin>        indent, this can not be guaranteed, and it  has  not  been  extensively
[13:36] <jdahlin>        tested.
[13:44] <iain> committed :)
[13:44] Action: iain thinks marlin is destined to always require CVS HEAD
[13:50] <jdahlin> iain: would be great to have some tests of deinterleave, so we can make sure it doesn't break in the future
[13:50] <jdahlin> iain: do you have a small test application, that can be run to see if it works properly?
[13:52] jdahlin ([email protected]) left irc: "Leaving"
[13:58] sgf ([email protected]) left irc: "Off on holiday for a fortnight."
[14:13] jonathang ([email protected]) joined #gstreamer.
[14:14] pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) left irc: Read error: 110 (Connection timed out)
[14:15] foser ([email protected]) joined #gstreamer.
[14:19] apoc ([email protected]) joined #gstreamer.
[14:19] <apoc> hi
[14:39] markey ([email protected]) left irc: "leaving"
[14:41] md`` ([email protected]) joined #gstreamer.
[14:41] adiabatic ([email protected]) joined #gstreamer.
[14:42] md` ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[14:43] jonathang_ ([email protected]) joined #gstreamer.
[14:43] Nick change: md`` -> md`
[14:54] <thaytan> hi apoc
[14:55] jonathang ([email protected]) left irc: Read error: 110 (Connection timed out)
[14:55] mathrick ([email protected]) left irc: Remote closed the connection
[14:56] Nick change: jonathang_ -> jonathang
[14:56] <adiabatic> I have a FLAC decoding bug in rhythmbox that I've been able to reproduce with `gst-launch gnomevfssrc location=/Volumes/executrix/Audio/flac/Jim\'s\ Big\ Ego/01.\ They\'re\ Everywhere.flac ! spider ! volume ! $(gconftool-2 -g /system/gstreamer/default/audiosink)` on OS X with fink's gstreamer version 0.6.4-5; has anyone else noticed that playing FLAC files causes the CPU to spike to 100% and delays the start of play by several secon
[14:56] <adiabatic> ds?
[14:57] <thaytan> I have no FLACs
[14:57] <adiabatic> Got broadband? I can give you that one; it's Creative Commons licensed. Oh, and I have smaller FLACs somewhere.
[14:58] Action: adiabatic digs for tiny flacs
[14:58] <thaytan> can you put it up somewhere?
[14:58] <thaytan> I can check if it still appears in CVS
[14:59] md` ([email protected]) left irc: Read error: 60 (Operation timed out)
[15:00] <adiabatic> ...?
[15:01] <adiabatic> WARNING: sound not available
[15:01] <adiabatic> This is new.
[15:02] <adiabatic> http://splat.cyclooctane.com/ 
[15:03] <adiabatic> the biggun is everywhere.flac
[15:03] <adiabatic> (reload)
[15:04] <thaytan> does delta.flac also break?
[15:04] <adiabatic> It breaks differently.
[15:04] <adiabatic> For unexplained reasons.
[15:04] <adiabatic> Let me try it again after playing this ogg vorbis file...
[15:06] <adiabatic> yeah, delta.flac breaks in the same way.
[15:06] md` ([email protected]) joined #gstreamer.
[15:06] <thaytan> it seems to play fine here
[15:06] <thaytan> playback starts immediately
[15:06] <adiabatic> Are you on OS X?
[15:07] <thaytan> ahr, no
[15:07] <thaytan> missed that detail
[15:08] <adiabatic> still a useful datum...
[15:09] <thaytan> you might like to try the 0.8 packages available on the website :)
[15:10] <sublett> hmm I've noticed some problems with rb/flac, but only with really big files. After seeking cpu load remains around 100%, and if they're big enough (have a couple of 200mb flacs) they don't make any sound at all (but rb still looks like it's playing)
[15:11] Action: adiabatic prays for no patches for 0.6.whatever
[15:12] <adiabatic> eep.
[15:14] <adiabatic> No, that won't do.
[15:14] <adiabatic> Or at least it won't be done by me first.
[15:16] <thaytan> wha?
[15:17] <adiabatic> Well, it might do.
[15:17] <adiabatic> all that stuff might have gone upstream already.
[15:18] <adiabatic> Where might I download the thing, since http://gstreamer.freedesktop.org/download/ doesn't even provide a link to the source?
[15:19] <thaytan> http://gstreamer.freedesktop.org/news/
[15:19] <thaytan> http://www.freedesktop.org/~gstreamer/modules/ (better)
[15:21] <adiabatic> do I need anything other than gstreamer for flac playback?
[15:21] <adiabatic> (with that one command-line thing)
[15:21] <thaytan> gst-plugins also
[15:26] pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) joined #gstreamer.
[15:26] <adiabatic> that's the messiest ipv6 address I've seen yet.
[15:27] <iain> how do you set the event handler function?
[15:27] <iain> on a pad
[15:28] <thaytan> gst_pad_set_event_function
[15:28] <iain> I'm using gst_pad_set_event_function but that only sets the GST_RPAD_EVENTFUNC
[15:28] <thaytan> what did you want to set?
[15:28] <iain> but when you do a gst_pad_send_event it looks for the GST_RPAD_EVENTHANDLER
[15:29] <iain> and I can't find anything that sets EVENTHANDLER
[15:30] <iain> is one of them a typo?
[15:30] <thaytan> I don't know what's up there at all
[15:30] Action: iain shrugs
[15:31] <thaytan> event handlers definitely work
[15:31] <iain> I'm getting GStreamer-WARNING **: pad src:src1 has no event handler
[15:31] <iain> when I try to send an event, because I only set the EVENTFUNC and not EVENTHANDLER
[15:32] <iain> (c.f.  gstpad.c line 3785 and line 541)
[15:35] <thaytan> it looks wrong to me too
[15:35] jcsston ([email protected]) joined #gstreamer.
[15:35] <thaytan> I'm trying to figure out why my upstream events are working
[15:37] pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) got netsplit.
[15:39] <adiabatic> gar
[15:40] pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) returned to #gstreamer.
[15:46] kwm ([email protected]) joined #gstreamer.
[15:47] <thaytan> iain: ahr, it's the schedulers that set up EVENTHANDLER from the EVENTFUNC
[15:53] Uraeus ([email protected]) joined #gstreamer.
[15:55] <iain> oh bugger...opt breakage?
[15:56] <thaytan> have you also set an event mask?
[15:56] <iain> yes
[15:56] <iain> shouldn't I?
[15:56] <thaytan> yeah, you should
[15:57] <thaytan> yeah, I guess it's opt fault
[15:57] <iain> yey
[16:08] Nick change: khalek -> kh_zZz
[16:09] alley_cat ([email protected]) left irc: Read error: 110 (Connection timed out)
[16:09] <thaytan> g'night
[16:10] <iain> night
[16:11] alley_cat ([email protected]) joined #gstreamer.
[16:14] jdahlin ([email protected]) joined #gstreamer.
[16:20] ChrisHJW ([email protected]) joined #gstreamer.
[16:25] markey ([email protected]) joined #gstreamer.
[16:25] ChriHJW_log ([email protected]) left irc: Read error: 60 (Operation timed out)
[16:26] markey ([email protected]) left irc: Client Quit
[16:31] ChrisHJW_log ([email protected]) joined #gstreamer.
[16:33] mathrick ([email protected]) joined #gstreamer.
[16:33] <mathrick> 'sup
[16:42] md` ([email protected]) left irc: Read error: 110 (Connection timed out)
[16:46] Nick change: mx|sleep -> mxpxpod
[16:53] walters ([email protected]) joined #gstreamer.
[16:53] tromey ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[16:56] md` ([email protected]) joined #gstreamer.
[16:57] walters ([email protected]) left irc: Client Quit
[17:01] kmaraas ([email protected]) joined #gstreamer.
[17:04] yippi ([email protected]) joined #gstreamer.
[17:11] thomasvs ([email protected]) left irc: Read error: 60 (Operation timed out)
[17:22] hyriand ([email protected]) left irc: "BOOT!"
[17:22] adiabatic ([email protected]) left #gstreamer ("Leaving").
[17:26] bitshifter ([email protected]) left irc: "bye"
[17:27] pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) left irc: "bbl"
[17:31] bitshifter ([email protected]) joined #gstreamer.
[17:40] walters ([email protected]) joined #gstreamer.
[17:48] trow ([email protected]) joined #gstreamer.
[17:54] tromey ([email protected]) joined #gstreamer.
[17:54] jonathang ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[18:01] apoc_ ([email protected]) joined #gstreamer.
[18:09] jcsston ([email protected]) left irc: Read error: 110 (Connection timed out)
[18:09] apoc ([email protected]) left irc: Read error: 60 (Operation timed out)
[18:10] markey ([email protected]) joined #gstreamer.
[18:18] Nick change: Uraeus -> Ura_gym
[18:34] jcsston ([email protected]) joined #gstreamer.
[18:37] foser ([email protected]) left irc: "[ I want to believe ]"
[18:38] kwm ([email protected]) left irc: "Leaving"
[18:39] kwm ([email protected]) joined #gstreamer.
[19:03] dolphy ([email protected]) left irc: "Network down, IP Packets delivered via UPS"
[19:09] jcsston ([email protected]) left irc: Read error: 110 (Connection timed out)
[19:14] jdahlin ([email protected]) left irc: "Leaving"
[19:23] sublett ([email protected]) left irc: "You got to kill your babies sometimes"
[19:38] spyder482 ([email protected]) joined #gstreamer.
[19:41] Nick change: Ura_gym -> Uraeus
[19:41] <Uraeus> spyder482: so you are the guy getting thanked for not releasing your nut muxer :)
[19:42] <spyder482> yes
[19:43] <Uraeus> spyder482: what is NUT? I never heard of it before
[19:43] yippi ([email protected]) left irc: "Client exiting"
[19:43] <spyder482> it's a sort of conflict of interests obviously ;)
[19:43] <spyder482> it's a container
[19:43] <spyder482> developed by the mplayer people
[19:43] <spyder482> originally called MPCF
[19:44] <Uraeus> and why is people happy you didn't release the muxer? because it would have made NUT a matroska competitor?
[19:44] <spyder482> well, it's not even finished...only half working
[19:44] <spyder482> but yes that's why i don't release it
[19:45] <spyder482> not really because it would be a competitor...it lacks some features that many like in matroska
[19:45] <spyder482> but more because i am a matroska dev...no need to go supporting yet another new format :)
[19:45] <spyder482> right now it's no better than the ffmpeg one anyway
[19:51] yippi ([email protected]) joined #gstreamer.
[20:02] <ChrisHJW> spyder482 : how is NUT overhead compared to matroska in your tests ?
[20:02] jcsston ([email protected]) joined #gstreamer.
[20:03] <spyder482> well, it's smaller...but no index yet
[20:03] <spyder482> and my files are still b0rked somehow
[20:03] <spyder482> i need to rewrite the spec for myself
[20:03] <spyder482> it's kinda scattered
[20:15] Uraeus ([email protected]) left irc: "Client exiting"
[20:16] dolphy ([email protected]) joined #gstreamer.
[20:18] Company ([email protected]) joined #gstreamer.
[20:31] Uraeus ([email protected]) joined #gstreamer.
[20:34] md`` ([email protected]) joined #gstreamer.
[20:34] <Uraeus> walters: around?
[20:35] md` ([email protected]) left irc: Nick collision from services.
[20:35] Nick change: md`` -> md`
[20:37] sub_pop ([email protected]) left irc: "Client exiting"
[20:38] Rotty ([email protected]) left irc: Remote closed the connection
[20:57] jcsston ([email protected]) left irc: ""Matroska Not Included.""
[20:59] md`` ([email protected]) joined #gstreamer.
[21:00] trasher ([email protected]) joined #gstreamer.
[21:00] <trasher> debian packages of gstreamer 0.8 anyone that knows if someone's working on that?
[21:04] <Company> taaz knows
[21:04] md`` ([email protected]) left irc: Remote closed the connection
[21:04] <Company> since he's the maintainer of the Debian ackages ;)
[21:04] <mathrick> because he's working on them :)
[21:04] <Company> i'll leave claiming that to him :p
[21:05] <mathrick> Company: wanna backlog? ;P
[21:06] jcsston ([email protected]) joined #gstreamer.
[21:07] <trasher> ok
[21:08] jcsston ([email protected]) left irc: Client Quit
[21:09] <Company> man, being knowledgable is bad when reading slashdot
[21:09] <Uraeus> yes, quite saddening
[21:09] <Company> i haven't read that much crap in one story before
[21:09] <Company> _everything_ was missing at least one point
[21:09] <Company> (havocs c# / java thingy)
[21:10] <mathrick> Company: you mean, havoc's writeup made it to /. ?
[21:10] <mathrick> Company: didn't see it
[21:11] <Company> mathrick: 2 days ago or so
[21:11] md`` ([email protected]) joined #gstreamer.
[21:11] <Company> mathrick: http://developers.slashdot.org/article.pl?sid=04/03/17/1335210&mode=thread&tid=126&tid=131&tid=156&tid=189
[21:12] md` ([email protected]) left irc: Connection timed out
[21:12] <mathrick> haven't been reading /. that regularly recently, so it may be
[21:13] Nick change: md`` -> md`
[21:20] hyriand ([email protected]) joined #gstreamer.
[21:20] trasher ([email protected]) left #gstreamer ("Klienten avslutas").
[21:21] Zeenix ([email protected]) joined #gstreamer.
[21:21] <Zeenix> hello
[21:22] <mathrick> lo Zeenix
[21:25] kmaraas ([email protected]) left irc: Read error: 110 (Connection timed out)
[21:26] <mathrick> Company: "story" as in "that short piece of text shown on /. main page", "comments section", or "havoc's paper"? Ie. which one is that shitty?
[21:27] yippi ([email protected]) left irc: Remote closed the connection
[21:28] <Company> mathrick: comments
[21:28] <Company> mathrick: the impressive thing is that all of them are shitty, not just most
[21:29] <mathrick> Company: hmm, I don't see how they are worse than usual :>
[21:29] <Company> they all miss what havoc wrote about
[21:29] <Company> every single one of them
[21:29] <mathrick> Company: yea, good point
[21:30] <mathrick> although, I suppose you didn't read all 684 of them? :)
[21:30] <Company> i assumed the lower rated ones were as bad as the higher rated ones
[21:31] <mathrick> heh, probably true
[21:31] <mathrick> anyway, Havoc's raising some interesting issue there
[21:31] <mathrick> s/issue/\&s/
[21:31] <Company> yeah
[21:31] <Company> but it'll naturally sort itself out
[21:32] <Company> and i'd bet 70% that mono wins
[21:32] <mathrick> and honestly, I don't think that Java == .NET, but from different vendor
[21:32] Action: ds-work still can't figure out what the hell he's ranting about
[21:33] <Company> ds-work: he wants the simple apps inside gnome written in !C
[21:33] <Company> that's all
[21:33] <tromey> and parts of the non-simple apps
[21:33] <Uraeus> actually I think it is more about positioning 'the plattform' in regards to commerical vendors
[21:33] <mathrick> Java is language, .NET is platform. I don't know if I like .NET, but it definitely doesn't seem like simple equivalent to me
[21:33] <Company> the big apps aren't the real problem
[21:33] <Company> the platform itself is what he talks about
[21:34] <Company> if someone writes a gnome presentation app that rocks the boat and it's in ruby, well all install ruby
[21:35] yippi ([email protected]) joined #gstreamer.
[21:35] <Company> the problem is if you want to market some gnome libs/simple apps that are written in ruby it's not so nice to do
[21:35] <Company> if gcalctool required mono, we'd all live without it
[21:36] <mathrick> wrt providing FLOSS independent platform, cloning (or rather, "staying compatible with") MS classes isn't necessarily that bad thing, it's even rather good thing
[21:36] jcsston ([email protected]) joined #gstreamer.
[21:36] <mathrick> I guess I'm biased by reading too much Joel recently ;), but being compatible is good
[21:37] <Company> being compatible isn't easy unfortunately
[21:37] <Company> they haven't even cloned MS DOS right
[21:37] sub_pop ([email protected]) joined #gstreamer.
[21:37] <mathrick> hmm, DOS isn't exactly on the same level as .NET :)
[21:38] <Company> no, it works with 640k
[21:38] <Company> .NET requires around 640M ;)
[21:38] <mathrick> no, I mean DOS is emulated, but .NET is implemented
[21:38] <mathrick> that's not the same thing
[21:39] <Company> it is
[21:39] <Company> c# and the JIT are basically a computer
[21:39] <mathrick> yep
[21:39] <Company> so they have to emulate a computer in both cases
[21:40] <mathrick> both cases == mono and net? or mono and dos?
[21:40] <Company> mono and dos
[21:41] m_wheels ([email protected]) joined #gstreamer.
[21:41] <mathrick> well, but MS .NET isn't any different here, it also does its own computer, doesn't it
[21:41] <mathrick> lo m_wheels
[21:42] <Company> well, in DOS you have to emulate an OS from the 80s
[21:42] <Company> that could be installed from 6 floppies
[21:42] <mathrick> yes, and that's very much of difference
[21:42] <mathrick> .NET is (at least meant to) kind of api supposed to be enough to support any kind of devel work
[21:43] <Company> it should be far easier to emulate stuff from 6 floppies than WInXP + .NET
[21:43] <Company> well, you can support it half-assed (without windows.forms)
[21:43] <mathrick> DOS is half-assed, quick and dirty hack bought for $150 from guy who was going on vacation and had no time to advertise it himself ;)
[21:43] <Company> yeah, but it's only 6 floppies
[21:44] <Company> Windows.Forms is a hack going deep into 40 million lines of code that is a hack, too
[21:44] <mathrick> yeah, but 6 floppies full of worst desing (or rather lack thereof) mankind has ever seen :)
[21:44] <Company> it's still easier than WinXP
[21:45] <m_wheels> hi mathrick 
[21:45] Nick change: m_wheels -> wheels
[21:46] <Company> our eurovision nationals all suck :/
[21:46] <mathrick> Company: I don't think emulating some braindead piece of code, making any possible assumptions about hw it's running on, is any easier than implementing well designed API
[21:46] <mathrick> not to say WinForms is good design, btw
[21:47] <Company> no
[21:47] <Company> 6 floppies of bad design is far easier than 40 million lines of code
[21:47] <mathrick> no, it's just smaller
[21:47] <Company> which makes it easier
[21:48] <Company> doing SVG + CSS is harder than MS HTML for exactly that reason ;)
[21:49] <mathrick> uh-huh, considering how deep in gross x86 shit DOS was diving, I doubt .NET could ever beat it :P
[21:49] <mathrick> ahh, no. You're confusing things now
[21:49] <mathrick> CSS + SVG != HTML, for one
[21:49] Action: wheels loves the concept of an open IRC meeting in a closed channel.  ;-)
[21:50] <Company> mathrick: it's both about interpreting a file
[21:50] <mathrick> but anyway, MS HTML is easy b/c it just ignored standards in hard parts, and did things hackishly
[21:50] <Company> and DOS != .NET ;)
[21:50] <wheels> Cheese != Monkeys
[21:50] <wheels> Oh, wait...
[21:50] <Company> well, s/MS HTML/MS Word/
[21:51] <mathrick> now, implementing FULL SVG + CSS standard is hard, precisely b/c it's full impl
[21:51] <Company> yeah
[21:51] <Company> because it's a lot
[21:51] <Company> just lkike WinXP
[21:51] <mathrick> implementing 95% of spec is ~10x easier than 100%
[21:52] <mathrick> but we're digressing too much :)
[21:52] <Company> implementing 95% of SVG+CSS is ok, but 95% of WIndows.Forms just isn't enough ;)
[21:52] <mathrick> how so?
[21:53] <Company> because programs will use the other 5%, too
[21:53] <mathrick> anyway, .NET is meant to be that "good spec", so we're over that part :)
[21:53] <mathrick> just like webpages / SVG files 
[21:53] <Company> no, i gave you that point for free :p
[21:53] <mathrick> heh :)
[21:54] <mathrick> and MS is moving from WinForms, too
[21:55] <mathrick> not that it makes implementing it any easier ;)
[21:56] <mathrick> but think about all these other funky, TLA out there - ADO, ASP and lotsa things I have no clue about
[21:57] <Company> yeah, those might work some day
[21:57] <mathrick> AFAIK, they're supposed to work today, too
[21:57] <Company> but we're talking desktop here, the server is already Linux territory
[21:57] <mathrick> at least ASP is pretty much complete now, AFAIR
[21:58] <Company> nice, but doesn't matter - windows will never get the server
[21:58] <Company> hey, somneone declared Xouvert dead - the project I claimed to be dead from the start :)
[21:59] <mathrick> not really. There are many devs that use MS platform now, and they don't do desktop only
[21:59] <mathrick> now, if they have opportunity to run their code on Linux for free / very cheaply, many of them will promptly check it out
[22:00] <Company> yeah
[22:00] <Company> but that's only the remaining 20% still running their servers on IIS :p
[22:01] <mathrick> that's exactly the thing Miguel is saying - there are many folks on the other side, who just can't afford changing their environment completely in one step
[22:01] <mathrick> well, AFAIK, ADO is DB stuff
[22:01] <Company> yeah, it's .NET ODBC
[22:02] <mathrick> and surprisingly many ppl actually use MS SQLish crap
[22:02] <Company> it's nice to have, but not needed
[22:02] <Company> what's needed is a way to conquer the desktop :)
[22:03] <mathrick> ok, but that's pretty different from "providing base classes as a completely independent solution"
[22:03] <Company> yeah, ok
[22:04] <mathrick> as I understand it, for .NET specific desktop stuff is only cherry on top
[22:04] <Company> someone should just run mono bytecode on Parrot and we're set ;)
[22:04] <mathrick> so, implementing everything except winforms is much already
[22:05] <mathrick> does parrot make it possible to interleave Perl and Python code freely for example?
[22:05] NFusi0n ([email protected]) joined #gstreamer.
[22:05] <mathrick> that's supposedely the biggest strength of .NET
[22:05] <tromey> I think that's largely a myth
[22:06] <tromey> it lets you interoperate provided both languages have similar type systems
[22:06] <mathrick> dunno, that's what they say
[22:06] <walters> Uraeus: i am now..
[22:06] <tromey> which usually isn't the case
[22:06] <Uraeus> walters: congrats with the Slashdotting :)
[22:07] <Company> well, you might at least get easier bindings for C stuff
[22:07] <walters> Uraeus: heh thanks
[22:07] <mathrick> Company: with net or parrot or both?
[22:07] <Uraeus> walters: btw, what are your plans for enabling tag editing in RB ?
[22:07] <Company> mathrick: parrot - .NET gives you that anyway
[22:07] <walters> Uraeus: once it works ;)
[22:08] <Uraeus> walters: I though it worked now? any specific bugs hindering it?
[22:08] <Company> Uraeus: autoplugging
[22:08] <Uraeus> hmm
[22:09] Action: Uraeus looks pleadingly at Company
[22:09] <mathrick> Company: that's the point - if net gave you method to invoke arbitrary language, instead of creating bindings, we're already lightyears ahead in terms of interoperability
[22:09] <tromey> there's no free lunch here
[22:10] <Company> mathrick: and we're lightyears back into windows-.only land
[22:10] <tromey> type systems must align or you forfeit some interoperability
[22:10] <mathrick> Company: why?
[22:10] <tromey> stubs must be generated either dynamically or AOT, either way you pay
[22:10] <mathrick> Company: I speak of .NET platform now, not MS NET runtime
[22:11] <Company> mathrick: because if you can easily link to other libs, you will be tempted to do so
[22:11] <Company> mathrick: and that makes people link into windows libs
[22:11] <mathrick> sure I pay, but I pay and I use, not wait indefinitely until someone creates bindings for exact combination of languages / libs I'm looking for
[22:12] <mathrick> Company: I don't see how Linux devs would probably link into windows libs
[22:12] <tromey> even with C# you have to write the bindings -- the differences are just syntactical
[22:12] <tromey> pinvoke is easier than JNI, but you're still writing declarations and you still have the maintenance problems
[22:13] <mathrick> so, it's impossible to write code in C#, and call it from VB (bleeh) instantly?
[22:13] <Company> afaik it works if you compile the code into .NET bytecode
[22:13] <Company> what's the acronym for that btw? CLR?
[22:13] <Uraeus> hmm, CVS totem just segfaults for me on startup
[22:14] <mathrick> Company: i think so
[22:14] <mathrick> or IL?
[22:14] <dolphy> Uraeus: you might want to join the matroska meeting
[22:14] <Uraeus> dolphy: where?
[22:14] <Uraeus> dolphy: channel is invite only
[22:14] <Company> boo
[22:15] <tromey> mathrick: no, you're right, if it all compiles to .NET bytecode, then you don't need stubs
[22:15] <tromey> I thought you meant something else
[22:15] <Company> tromey: does that work for parrot bytecode, too?
[22:15] <ChrisHJW> Uraeus : different server
[22:15] <mathrick> tromey: so, basically, we're free from bindings hassle, right?
[22:15] <ChrisHJW> Uraeus : irc.corecodec.com
[22:15] <dolphy> irc.corecodec.com
[22:15] <Company> mathrick: only if you compile everything to bytecode ;)
[22:15] <mathrick> that'd be already enormous leap forward
[22:16] <tromey> if it is all bytecode.  but then the type systems must align
[22:16] <mathrick> Company: yeah, but you're supposed to anyway, right?
[22:16] <tromey> but if you're using the C#-ish type system, why bother writing in anything but C#?
[22:16] <Company> if you compile GSTreamer to .NET bytecode, it's probably only half as fast and then some ;)
[22:16] <mathrick> tromey: how is it determined if they align?
[22:16] <tromey> it's a language issue
[22:16] <tromey> eg, AIUI, .NET doesn't have the notion of multiple inheritance
[22:17] <mathrick> Company: but that means you write one binding - GST.NET, and all net languages get it for free
[22:17] <tromey> So C++ code using this feature can't be compiled to .NET bytecodes
[22:17] <Company> people should compile languages into other languages that are human readable
[22:17] <mathrick> tromey: AIUI?
[22:18] <tromey> as I understand it
[22:18] <mathrick> ah, k
[22:19] <mathrick> tromey: so, does it mean that C# has some type system, or it's .NET which has it?
[22:19] <walters> Uraeus: there are several problems.  first the mode in gstid3tags.c always seems to get set to tags-only, so it doesn't pass the buffers through, meaning your end file will be empty
[22:19] <walters> Uraeus: secondly there's no ogg muxer
[22:19] <mathrick> tromey: ie, can C# and INTERCAL# have different type systems, being both managed by .NET?
[22:19] <walters> Uraeus: and thirdly there's no autoplugging
[22:20] <tromey> mathrick: my understanding is that .NET has an implicit type system, which C# exposes most of to the user
[22:20] <Company> autoplugging tag writing is hard
[22:20] <tromey> so there is some room for non-C# languages
[22:21] <Company> because tagging format1 will not result in format1 files in every case
[22:21] <mathrick> tromey: k, so it would mean that any .NET language is inherently type aligned with any other .NET one?
[22:21] <tromey> right
[22:22] <mathrick> that would be great help, IMO
[22:22] <tromey> yes, but it means that interoperability is a myth
[22:22] <tromey> because different languages in fact have different type systems
[22:22] <tromey> there may be some useful subset of course
[22:23] <tromey> but in general you have to rewrite
[22:23] <mathrick> rewrite what?
[22:23] <tromey> the code you want to reuse
[22:24] <mathrick> hmm, I don't get it
[22:24] <mathrick> is "code I want to reuse" already managed by .NET or not?
[22:25] <Company> not managed by .NET
[22:25] <tromey> well, why is language interoperability interesting?  Becuase you have code in some other language you want to reuse
[22:25] <tromey> so suppose you have a big chunk of C++ code
[22:25] <tromey> you try to copmile it to .NET
[22:25] <tromey> but the type systems differ, so you can't.  Instead you have to rewrite parts of it
[22:25] <tromey> But then this isn't interoperability, it is rewriting
[22:25] <tromey> or you can expose parts via some API.  But then you need stubs to manage the type system differences
[22:26] <mathrick> huh? I thought there's C++.NET, isn't there?
[22:26] <tromey> there is "Managed C++", which I believe is a subset of C++
[22:26] <tromey> meaning, there are valid C++ programs that won't work with it
[22:26] <mathrick> hmm, that would truly suck
[22:26] <tromey> maybe you can get lucky and it turns out your program doesn't need rewriting, or not much
[22:29] <wheels> mathrick: well, there are valid C++ programs that won't compile under MSVC++ anyway...
[22:30] <wheels> (IIRC it still doesn't support partial template specialization.)
[22:32] <mathrick> wheels: could you remind me what it is exactly?
[22:32] <Company> #include <windows.h>
[22:32] <Company> MyObject x = MAX(new MyObject (5), new MyObject (4));
[22:32] <Company> that should already fail iirc
[22:33] <wheels> mathrick: it's where you can specify different template behavior for different classes of types.  i.e. what a list should do when it contains pointers.
[22:34] <mathrick> Company: what is it supposed to do?
[22:35] <mathrick> wheels: hmm, how do you specify it? Ie. have a snippet handy? :)
[22:35] <mathrick> Company: and, AFAICT, that would leak
[22:36] <wheels> mathrick: http://webcvs.kde.org/cgi-bin/cvsweb.cgi/kdesupport/taglib/toolkit/tlist.tcc?rev=1.18&content-type=text/x-cvsweb-markup
[22:37] <wheels> mathrick: note the first two template class declarations -- it's used there to automatically delete items in a list (if that's enabled) when the list is cleared or destroyed.
[22:38] BBB ([email protected]) joined #gstreamer.
[22:39] Action: BBB waves
[22:40] <Uraeus> hi BBB
[22:40] <Uraeus> BBB: you just missed the vote
[22:40] <mathrick> yo BBB
[22:40] <BBB> vote?
[22:40] <BBB> what did we vote fot?
[22:40] <Uraeus> BBB: you where voted to fix gstreamer so that Walters can enable tag editing in RB :)
[22:40] <bitshifter> BBB: yeah, gstreamer is gonna be rewritten in c#
[22:40] <BBB> rrrrrrright
[22:41] Action: BBB likes bitshifter's idea
[22:41] <BBB> actaully
[22:41] <BBB> I fell more for perl
[22:41] <BBB> that will be truely readable
[22:41] <mathrick> wheels: ah, that thing. thx
[22:41] <ds-work> Uraeus: where's the bug report?
[22:41] <bitshifter> does perl support macros? 
[22:41] <walters> BBB: heh
[22:42] <bitshifter> ah well, I guess one could just capitalise everything anyway 
[22:42] <Uraeus> ds-work: I think walters should write the bug reports as he knows exactly where the shoe hurts
[22:42] <mathrick> bitshifter: only :)
[22:42] <Uraeus> ds-work: btw, your perstistent bug work is about to take us off the top 15 bug list, 2 more closed and we are off I think :)
[22:43] <ds-work> Uraeus: I'd have to start fixing bugs to decrease the list more
[22:43] Action: BBB opens quicktime bugs
[22:44] <Uraeus> walters: will you file a bug report for the tagging issues please? would help increase their visibility 
[22:45] <mathrick> wheels: funny part about MSVC is that it doesn't really support C either. ie., I don't think you can get pure C binary w/o going through lotsa hoops with MSVC linker
[22:47] <dolphy> BBB: irc.corecodec.com #matroska
[22:47] <BBB> ?
[22:47] <BBB> now?
[22:49] hyriand ([email protected]) left irc: "Lost terminal"
[22:52] <walters> Uraeus: i'm not totally convinced the id3tag issue isn't an rb bug yet
[22:52] <walters> Uraeus: ogg and autoplugging are pretty well known i think
[22:54] <Uraeus> walters: bugs not in bugzilla tend to be forgotten or go to the bottom of the todo list, and it might not be well known by the developer who ends up actually fixin it. Doesn't help if Company or ds-work knows about it if they don't have time to fix it
[22:58] Action: BBB supports that statement
[22:58] Action: BBB wants bugs to fix
[22:58] <BBB> and I want gst-rec to work ;)
[23:00] <bitshifter> BBB: you could apply http://bugzilla.gnome.org/show_bug.cgi?id=137448 (mpegaudioparse crash) ;)
[23:02] md` ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[23:02] md` ([email protected]) joined #gstreamer.
[23:02] <BBB> bitshifter: will do tonight
[23:02] <BBB> I'm not at home
[23:02] <bitshifter> thanks
[23:03] wingo ([email protected]) joined #gstreamer.
[23:03] <BBB> hi andy
[23:04] <Uraeus> yo wingo
[23:04] hyriand ([email protected]) joined #gstreamer.
[23:05] <wingo> hey hey hey!
[23:06] <wingo> hey Uraeus, i read you're coming :)
[23:06] <Uraeus> wingo: yup :)
[23:06] Action: wingo happy
[23:06] <Uraeus> wingo: I am really looking forward to it (well who wouldn't mind traveling the world for 2 months :)
[23:06] <wingo> and also a bit drunk -- been a while since i've been drunk on irc :-}
[23:07] <wingo> sweet. i'm really looking forward to you coming.
[23:07] <dolphy> being drunk on irc is fun
[23:07] <wheels> Uraeus: You could just do the austrailian thing and never go back.  :-)
[23:08] <dolphy> wheels: nah we need him :)
[23:08] <wheels> dolphy: There are many places in the world with internet connections.  :-)
[23:08] <Uraeus> wheels: what can I say, the world is so big and there is just one of me, I need to spread out as much as possible to be fair
[23:08] <BBB> oh my god, an epidemic
[23:08] Action: wheels is hoping to hit 4 continents this year.  Up to 3 so far...
[23:08] Action: ds-work needs a drink
[23:09] <Uraeus> heh :)
[23:09] <wheels> Err, 5.
[23:09] <Uraeus> this will be my first ever visit to both Oceania and Africa
[23:09] <Zeenix> wheels: would that include pakistan?
[23:09] <wheels> Austrailia and Antarctica will have to wait.  :-)
[23:10] <wheels> Zeenix: close -- I'll be on both sides of it -- I'm going to be in Saudi Arabia, UAE and India.  :-)
[23:10] <Uraeus> doubt I ever will try to go to Antartica, living in Norway has cured me from any need of visiting cold places
[23:11] <wingo> ds: why can't the fixate function work on audioconvert's src? i don't see any technical reason. you just fixate, by default, to the number of channels already negotiated on the sink.
[23:11] <ds-work> wingo: because in the fixate function, the pads' caps are not set
[23:11] <Zeenix> wheels: a near hit is a miss :(
[23:12] <wheels> Uraeus: I've heard that there's a ferry from Argentina just for the people that want to have covered all of the continents.  :-)
[23:12] <ds-work> cruises to antarctica are popular
[23:13] Action: wheels will be pretty close there too -- but at the wrong time of the year.  I'm planning on being in Chile in August or so...
[23:14] <wingo> ds-work: what's the general solution to that kind of problem? i am unconvinced multiple elements is the solution. and do you consider changing audioconvert's behavior to be an incompatible change that should be avoided?
[23:14] Action: ds-work knows several people who have been to the south pole
[23:14] <Uraeus> wheels: well ok go to Antartica then, but remember to bring your laptop so you can update the Qt GStreamer bindings while there :)
[23:14] <wheels> Uraeus: I'm hoping to get to that over the easter weekend.  But I just copied them to my laptop since I'm heading out again tomorrow...  :-)
[23:14] <ds-work> wingo: multiple elements.  audioconvert can stay the same, since it's useful for autoplugging
[23:15] david_ ([email protected]) joined #gstreamer.
[23:17] <david_> looking for gst.h to compile some examples on web
[23:17] <ds-work> david_: install gstreamer
[23:18] <david_> mandrake 10.0 rpm's as far as I could
[23:18] <Zeenix> Uraeus: what new you would find in americas? come to pk, i'll show you plenty of new things (includnig death if you want) :)
[23:18] <david_> ?
[23:18] <Uraeus> Zeenix: yeah, finding my own death is one of the reasons stopping in .pk is not to tempting :)
[23:19] <ds-work> david_: then add $(pkgconfig --cflags gstreamer-0.6) to your CFLAGS
[23:19] <Uraeus> david_: maybe you are missing the gstreamer-devel package?
[23:19] Shoragan ([email protected]) left irc: "Leaving"
[23:19] markey ([email protected]) left irc: "leaving"
[23:19] <ds-work> and $(pkgconfig --libs gstreamer-0.6) to your LIBS
[23:19] <Zeenix> Uraeus: shortest possible stay would quite effectivelly decrease the chances
[23:20] <Zeenix> Uraeus: btw, why would someone like to kill a norwegian for, do the bush care for norwegians 
[23:21] <david_> I didn't expect absolute bollocks!
[23:21] <Uraeus> Zeenix: there are norwegian soldiers in both Afghanistan and Pakistan, and besides who care if I am norwegian or american when they are looking for a infidel to kill for islam
[23:21] <Uraeus> david_: your question wasn't very detailed, sorry if our answers where to generic
[23:22] <david_> this is about gstreamer?
[23:22] <Uraeus> david_: no, this about some of us here knowing eachother quite will over the years
[23:23] <david_> o.k.
[23:23] <Zeenix> Uraeus: i'll teach you 'qalma' and you have the proof to be a muslim :)
[23:24] <wheels> Zeenix: Well, it's not really clear if you see a westerner whether they're Norwegian or American...
[23:24] <Uraeus> david_: the gst.h file should be included in the gstreamer-devel rpm and be placed in /usr/include/gst/gst.h I think
[23:24] Action: wheels got an invitation to speak at a Linux event in Algeria recently...
[23:25] <Uraeus> david_: actually this is my path to the file: /usr/include/gstreamer-0.8/gst 
[23:25] <Uraeus> wheels: I will be doing a GStreamer and SVG talk in Sydney actually :)
[23:26] <wheels> Uraeus: bah, Sydney is easy.  I'm doing a Linux talk (this one work related) in Riyadh on Wednesday.  :-)
[23:26] <Uraeus> and maybe I and wingo can do a talk for a LUG in Windhoek :)
[23:26] <Uraeus> wheels: why is that hard? you have to wear a burka? ;)
[23:26] <wheels> Uraeus: Well, uhm, there's a bit less of a culture gap.  :-)
[23:27] <wheels> Uraeus: Or language gap.
[23:27] <bitshifter> david_: look if there are some -devel RPMs as well
[23:27] <Zeenix> so they (arabs) give some of the money they get from the pilgrim on Linux too, maybe i need to change my idea regarding 'haj' then :)
[23:27] Action: wheels only *thinks* that he actually has a hotel reservation.  :-)
[23:27] <david_> el others gave bad signatures
[23:27] <wheels> Zeenix: heh
[23:27] <wheels> Zeenix: I think it's more like some of the money from oil.  ;-)
[23:28] <david_> bitshifter  just downloaded them thanks others had bad signatures which put me off
[23:33] <bitshifter> david_: you could uninstall the gstreamer package(s) you just installed and see if the fedora RPMs work on Mandrake (mind you, I don't know if they do or don't)
[23:34] <ds-work> BBB: is there a reason we can't use metadata for esds atoms?
[23:35] <BBB> metadata?
[23:35] <BBB> er?
[23:35] <BBB> ?
[23:35] <BBB> caps
[23:35] <BBB> not metadata
[23:35] <ds-work> stream metadata
[23:35] <david_> bitshifter trying libgstreamer6
[23:35] Action: BBB doesn't understand
[23:36] <bitshifter> david_: hmm, you should look for 0.8 packages (unless you need 0.6 for some specific app)
[23:36] <david_> bitshifter: I must go for source no easy way I can see
[23:38] Misirlou ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[23:38] Misirlou ([email protected]) joined #gstreamer.
[23:38] Action: ds-work wonders why he didn't make the structure name just a field called "media_type"
[23:39] <BBB> why stream metadata ds?
[23:39] <BBB> why not caps
[23:40] <ds-work> BBB: because metadata is where we currently put informational data about a stream
[23:40] <ds-work> BBB: but arguably, it's not just informational
[23:42] <BBB> also, that's stream data
[23:42] <BBB> this needs to be known before codec setup
[23:42] <Company> i think the docs even say that metadata is purely optional and the stream must work without
[23:43] <Company> besides, using metadata for buffers with a different type is kinda lame
[23:43] Action: mathrick decides to go to sleep
[23:43] <mathrick> nite
[23:43] Action: BBB agrees
[23:43] <BBB> just put it in the caps
[23:43] <BBB> we can add a GstBuffer to caps now
[23:43] <BBB> let's use that
[23:43] wingo ([email protected]) left irc: Read error: 110 (Connection timed out)
[23:43] Nick change: mathrick -> mathrick|sleep
[23:43] <ds-work> yeah
[23:43] <ds-work> it was one of my goals while rewriting caps
[23:44] Action: md` waves at Uraeus and dolphy
[23:44] <Uraeus> heh :)
[23:44] <BBB> ds-work: so you achieved the goal and now you don't want to use it? ;)
[23:44] <ds-work> BBB: I just want to make sure it's the right way
[23:44] <Company> does it work with to/from_string?
[23:44] <ds-work> and Company said exactly the right words to convince me
[23:45] <ds-work> Company: no
[23:45] <Company> it should
[23:45] <Company> even if it just converts to/from hex values
[23:45] <ds-work> Company: file a bug :)
[23:46] <BBB> Company: I will for GstBuffer
[23:46] <BBB> just give me some time dudes ;)
[23:46] wingo ([email protected]) joined #gstreamer.
[23:46] <BBB> esds_data=0x0000aadde457840313e4f5bc87a543
[23:46] <BBB> something like that?
[23:46] <Company> will be funny when people start complaining about 300 line long caps :)
[23:46] <ds-work> Company: it should print something uninteresting for GstBuffer, but not anything that would let you reconstruct the buffer from
[23:46] <Company> BBB: yeah
[23:47] <BBB> it contains data and size: (strlen(x)-2)/2
[23:47] <BBB> :)
[23:47] <ds-work> probably does (GstBuffer)0xdeadbeef, with the pointer
[23:47] <Company> ds-work: why not? i might want that for filtered caps
[23:47] <BBB> filtered caps doesn't work :(
[23:47] Action: BBB is having issues with it
[23:48] <Company> what doesn'T work there?
[23:49] <BBB> I tried videotestsrc ! video/x-raw-yuv,framerate=25 ! videodrop ! video/x-raw-yuv,framerate=5 ! ximagesink
[23:49] <BBB> but it gave me 25-fps video output
[23:49] <BBB> somewhat more difficult testcase, but you get the idea
[23:49] <ds-work> BBB: framerate=25.0
[23:49] <ds-work> it's double
[23:49] <BBB> I'll make a bugreport one day
[23:49] <BBB> oh
[23:49] <BBB> thn why doesn't it error out?
[23:50] <ds-work> because gst-launch is stupid
[23:50] Action: BBB files bug
[23:50] <Uraeus> ok, time for me to sleep, see ya
[23:50] Uraeus ([email protected]) left irc: "Client exiting"
[23:50] <Zeenix> bye
[23:50] <ds-work> BBB: actually, it doesn't work here
[23:50] <Company> gst-launch should give an error but try to run anyway, no?
[23:51] <ds-work> it works fine with 25.0 and 5.0
[23:55] <BBB> hm
[23:55] <BBB> odd
[23:55] Action: BBB must be doing something wrong
[23:55] <BBB> ohwell
[23:56] <BBB> laterstuff
[23:56] Action: BBB needs to fix gst-rec first
[23:58] kwm ([email protected]) left irc: "sleeping"
[23:59] BBB ([email protected]) left #gstreamer.
[00:00] --- Sat Mar 20 2004
[00:00] benow ([email protected]) joined #gstreamer.
[00:01] <benow> heyas... is there a Benjamin Otte loitering in here?
[00:01] Action: ds-work points at Company
[00:01] <ds-work> benow: how's it going?
[00:02] Action: Company looks up
[00:02] <ds-work> benow: we still have your radio clip in our media test collection :)
[00:02] <benow> heya ds-work, not bad.  hoping to squish some gst-java bugs this wkend.
[00:04] <benow> nice.  yeah, I'd like to get back to using a gstreamer backend.  am using a java wrapped libshout for raw streaming now, but would be able to do much more with gstreamer
[00:04] <wingo> hey, it's a benow!
[00:04] <benow> heya wingo, oh, and yeah, i'm calling gst-init ;)
[00:04] <wingo> ds-work: i might have missed a reply earlier..
[00:05] <wingo> ah, that was you ;)
[00:05] Nick change: bilboed_ -> bilboed
[00:05] jonathang ([email protected]) joined #gstreamer.
[00:05] <ds-work> wingo: ?
[00:06] Action: benow is thinking there must be a better way to glue that JNI
[00:07] <ds-work> JMF?
[00:07] Action: ds-work ducks
[00:08] <benow> a small wrapper around gst would blow jmf out of the water.
[00:13] ChrisHJW ([email protected]) left irc: Connection timed out
[00:16] <ds-work> wingo: "This is very easily solved by having simpler elements that mess
[00:16] <ds-work> with individual features of the stream, and not everything at once
[00:16] <ds-work> like audioconvert.  I have no problems with audioconvert _itself_,
[00:16] <ds-work> and audioconvert and the simpler elements could probable share
[00:16] <ds-work> up to 90% of their code."
[00:17] Action: ds-work dislikes /msg
[00:18] <wingo> as do i.
[00:18] <wingo> two issues, really.
[00:19] <wingo> 1) the proper grouping of functionality with respect to element boundaries. (i.e. what functionality in which / how many elements).
[00:19] <wingo> 2) can we change the advertised behavior of the named element "audioconvert" after 0.8.0?
[00:20] <ds-work> wingo: 1) number of channels is fundamentally different than just "format"
[00:20] <ds-work> 2) no
[00:21] <ds-work> I think of it similarly to videoscale and colorspace.  colorspace changes format, videoscale changes something akin to number of channels
[00:21] <wingo> ds-work: wrt 1), i agree
[00:21] <wingo> but 2) is a problem
[00:22] <ds-work> why?
[00:22] Action: wingo wonders if we are doomed to the int2float (an example) hell
[00:22] <ds-work> audioconvert, as it is currently, is necessary for spider to work properly
[00:22] <wingo> no,
[00:23] <wingo> i agree that you can't change the advertised behavior of audioconvert
[00:23] <wingo> i.e. it is, in 0.8, an element that can convert caps
[00:23] <wingo> for audio
[00:23] <wingo> so to fix it is to change the problem space that it operates on
[00:23] <ds-work> but we need a separate channelconvert (or something)
[00:23] <wingo> and as such is an incompatible change
[00:24] <wingo> i see the need, but what do we do now that 0.8 is stable?
[00:24] <wingo> audioconvert2?
[00:24] <ds-work> "fixing the problem" involves creating new elements
[00:24] <wingo> that's an ugly solution if i've ever heard of one
[00:25] Action: wingo was referring to audioconvert2 in the last line
[00:25] <wingo> ds-work: yes, but at the same time, changing the functionality of audioconvert is an incompatible change.
[00:25] <ds-work> "audioformat" and "audiospace", perhaps
[00:26] <ds-work> who is talking about changing audioconvert?
[00:26] Action: ds-work wonders if wingo is drunk :)
[00:26] <wingo> you, by removing channel conversion?
[00:26] <wingo> oh yes, i'm getting there ;)
[00:27] <ds-work> _changing_ audioconvert is not an option, for the main reason that it will no longer work with spider
[00:29] <wingo> how do you propose to split out channel conversion while not removing it from audioconvert?
[00:29] Action: wingo struggles to understand
[00:29] <wingo> i mean, i trust you, but i don't understand, and i don't like that situation :-/
[00:30] <ds-work> add "audioformatconvert" and "audiochannelconvert" elements
[00:31] <wingo> and leave audioconvert as it is
[00:33] <ds-work> yes
[00:33] <wingo> ah, ok.
[00:34] <wingo> this whole issue of being stable is a bitch sometimes ;)
[00:34] <ds-work> it keeps us honets
[00:34] <ds-work> honest, too
[00:36] <taaz> hey wingo 
[00:37] <wingo> hey taaz
[00:39] <taaz> not used to seeing you here at this time of day
[00:40] camh ([email protected]) left irc: Read error: 111 (Connection refused)
[00:40] <taaz> isn't it nice to see work done on the python bindings? ;)
[00:40] <ds-work> ok, gst-plugins compiles now without X
[00:41] walters ([email protected]) left irc: "out"
[00:51] Nick change: mxpxpod -> mx|gone
[00:52] camh ([email protected]) joined #gstreamer.
[00:54] <wingo> taaz: it is nice, yes. and i hope to beat you soon :)
[00:58] <NFusi0n> where's the gapless love? :(
[00:58] <wingo> aaag... cheap red wine...
[00:58] <iain> yo
[00:59] <iain> wingo: fixed a silly bug in deinterleave :)
[00:59] benow ([email protected]) left irc: Remote closed the connection
[00:59] wheels ([email protected]) left irc: Read error: 60 (Operation timed out)
[01:09] <wingo> iain: um, i'm sure i put it there intentionally just to check your skills...
[01:09] <wingo> er.
[01:09] <wingo> good :)
[01:09] Nick change: kh_zZz -> khalek
[01:09] <iain> it confused me for like 10 seconds when I looked at the code
[01:10] <iain> but it's been annoying me for ages :)
[01:12] wheels ([email protected]) joined #gstreamer.
[01:13] jdahlin ([email protected]) joined #gstreamer.
[01:22] bilboed_ ([email protected]) joined #gstreamer.
[01:23] Action: wingo trying to set up a blog for african hackers
[01:25] <Misirlou> There are African hackers? AWESOME
[01:25] <spyder482> lol
[01:25] <spyder482> that's weird
[01:25] <ds-work> er, wingo is an African hacker
[01:26] <jdahlin> so gstreamer is ported to win32
[01:26] bilboed ([email protected]) left irc: Read error: 110 (Connection timed out)
[01:26] <spyder482> cool
[01:26] <Misirlou> ds-work: I know, and that's AWESOME
[01:26] Action: ds-work finds it amusing/disgusting that every African he knows is white
[01:26] Action: Misirlou ate too much sugar
[01:26] <spyder482> lol
[01:27] <spyder482> so you shout?
[01:27] <spyder482> :P
[01:32] jcsston` ([email protected]) joined #gstreamer.
[01:32] kmaraas ([email protected]) joined #gstreamer.
[01:37] thaytan ([email protected]) left irc: Connection timed out
[01:39] <wingo> ds-work: I've met a few quality black african hackers this week. it's been enlightening.
[01:40] BBB ([email protected]) joined #gstreamer.
[01:40] <wingo> hey BBB
[01:40] <BBB> howdy
[01:40] Action: BBB tries to remember what to fix tonight
[01:41] <Company> gst-rec :p
[01:41] <BBB> no doubt
[01:41] <BBB> but which part? ;)
[01:44] thaytan ([email protected]) joined #gstreamer.
[01:46] ds-work ([email protected]) left irc: Remote closed the connection
[01:47] jcsston ([email protected]) left irc: Read error: 110 (Connection timed out)
[01:47] <wingo> wordpress rocks!
[01:47] <wingo> hey thaytan
[01:49] <spyder482> <nanoflower> Syskin: you have to go for some pansy OS like LInux to find a system that can only stay up for six days. :)
[01:49] <spyder482> funny ;)
[01:50] <spyder482> <spyder482> nanoflower: linux isn't a pansy system...maybe you're just a pansy user ;)
[01:53] <spyder482> night
[01:54] spyder482 ([email protected]) left irc: "Leaving"
[01:55] <Company> ds: didn't you claim to have fixed ffmpeg?
[01:58] lala768 ([email protected]) joined #gstreamer.
[01:59] lala768 ([email protected]) left #gstreamer ("Lämnar").
[02:10] Action: wingo officially drunk.
[02:11] <wingo> and still hacking on blogs.fossfa.org.
[02:11] smoke ([email protected]) got netsplit.
[02:11] <jdahlin> wingo: so you're not up for some opt discussion then? :)
[02:11] <iain> jdahlin: now is the perfect time to get him to discuss it
[02:12] <jdahlin> iain: yay!
[02:12] <iain> in fact, he might even be able to fix it too!
[02:12] <jdahlin> wicked!
[02:13] iain ([email protected]) left irc: "To the bedcave robin!"
[02:14] <wingo> is there a problem?
[02:14] <BBB> omfg
[02:14] <BBB> what the fuck happened to gst-rec's coding style?
[02:14] <BBB> this is beyond hideous
[02:14] Action: wingo has lubricated his though rotors
[02:14] <wingo> *thought
[02:14] <jdahlin> wingo: the last thing I heard was smp *issues*
[02:14] <wingo> ah.
[02:14] <Company> opt can't have smp issues
[02:15] <wingo> that's a bold statement ;)
[02:15] <Company> because schedulers only run in one thread
[02:15] <Company> at least they're supposed to
[02:15] <wingo> i would like bugs.
[02:16] <wingo> for the record, i have some problems with dynamic parallel pipelines. but i know what the solution is. just for the record ;)
[02:17] smoke ([email protected]) got lost in the net-split.
[02:17] <wingo> BBB: just chill, bro. find you a pabst blue ribbon, and chill.
[02:17] Action: wingo is at approximately 5 kelvin.
[02:18] <bitshifter> wingo: http://bugzilla.gnome.org/show_bug.cgi?id=137711  some of the stuff in the second post might be an opt issue (I don't really know, but it looks like it to me)
[02:19] <BBB> wingo: there's vodka & beer coming up
[02:20] <jdahlin> BBB, blame me!
[02:20] <jdahlin> BBB, (for the coding style)
[02:21] Action: BBB kicks jdahlin
[02:21] <BBB> what did you do?!?
[02:21] Action: BBB cries
[02:21] <BBB> it was so beautiful
[02:21] <jdahlin> BBB, I just like gnu style
[02:21] <BBB> it was so perfect
[02:21] <BBB> I don't
[02:21] <jdahlin> So, we have a conflict :)
[02:21] <BBB> I like GNOME's coding style
[02:21] Action: BBB gets a gun and shoots
[02:22] <BBB> (solution to the conflict)
[02:22] <jdahlin> What's wrong with gnu?
[02:22] <jdahlin> Okay, the gnome coding standard is nice too
[02:22] <jdahlin> but we don't need to do something like running indent on all the code like thomas did just on gstreamer
[02:22] <jdahlin> do we
[02:22] <jdahlin> ?
[02:24] <wingo> 5 kelvin, guys.
[02:24] ds-work ([email protected]) joined #gstreamer.
[02:25] <jdahlin> Company/wingo, some one (eg andersca) told me that after switching from opt to basicgthread jamboree worked perfectly on his smp system
[02:25] <jdahlin> might be related to esdsink too, maybe it triggers something only when ran under smp
[02:26] <wingo> look, let's put it this way.
[02:26] <Company> i'm not gonna look at smp issues
[02:26] <wingo> the opt scheduler followed a broken refcounting strategy before 1.61 (cvs version).
[02:26] <jdahlin> me neither, I just wanted to point it out
[02:26] <Company> until i know the code is smp clean
[02:26] <Company> and it's not
[02:27] <Company> smp is much too random
[02:27] <jdahlin> random??
[02:27] <wingo> following a proper refcounting strategy that actually allows for groups / chains to be destroyed exposed a lot of bugs.
[02:27] <jdahlin> concurrent I'd say
[02:27] <wingo> so we just have to fix these bugs now and not be sloppy.
[02:27] <ds-work> anyone remember the last message to me?  I missed it right as X crashed
[02:27] Action: jdahlin will push it to someone who understands it well enough to be able to fix it
[02:27] pb_ ([email protected]) joined #gstreamer.
[02:28] <Company> ds-work: didn't you claim to have fixed ffmpeg?
[02:28] <Company> that was the last thing i said to you at least :)
[02:29] <ds-work> ah
[02:29] <ds-work> Company: not really.  I rely on BBB to tell me what to do
[02:29] <Company> dsputil_mmx.c: In function `h263_h_loop_filter_mmx':
[02:29] <Company> dsputil_mmx.c:614: error: can't find a register in class `GENERAL_REGS' while reloading `asm'
[02:30] <BBB> -fomit-frame-pointer, and don't use -ggdb
[02:30] <BBB> (I've been told)
[02:30] <ds-work> -ggdb == -g
[02:31] <Company> i do use -g on everything i compile, especially gstreamer plugins ;)
[02:32] pb_ ([email protected]) left irc: Client Quit
[02:34] <BBB> hm
[02:34] <BBB> -g should compile
[02:34] <BBB> it does here
[02:34] <BBB> hm
[02:34] <BBB> what else could eat registers?
[02:34] <BBB> do you use anything else than default?
[02:34] <BBB> like -O0?
[02:35] <Company> -g -march=athlon-xp
[02:35] <ds-work> -fPIC and -fomit-frame-pointer
[02:35] <ds-work> are the only flags I know that will change register allocation
[02:35] <BBB> remove the -fPIC ;)
[02:35] <ds-work> fix the asm
[02:35] <ds-work> it's broken anyway
[02:36] Action: BBB is bad at asm
[02:36] <ds-work> I fixed all but one chunk
[02:36] <BBB> if it's broken, go tell michael... I know less than nothing about asm so I cannot explain him
[02:36] <ds-work> which by its algorithm, actually uses 8 registers
[02:36] <Company> i know that -O2 has helped me quite often
[02:36] <Company> with asm - especially from ffmpeg
[02:36] Nick change: mx|gone -> mxpxpod
[02:40] sub_pop ([email protected]) left irc: "Client exiting"
[02:43] Company ([email protected]) left irc: Remote closed the connection
[02:49] <wingo> jesus, it's about four in the morning. i'm going to sleep.
[02:50] Nick change: wingo -> wingo-z
[02:50] <BBB> bye andy
[02:53] jdahlin ([email protected]) left irc: "Leaving"
[02:54] BBB ([email protected]) left irc: "Client exiting"


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