IRC Logs

*******************************************************************
[03:05] BBB ([email protected]) joined #gstreamer.
[03:56] thaytoo ([email protected]) joined #gstreamer.
[04:12] jcsston ([email protected]) joined #gstreamer.
[04:20] ChrisHJW ([email protected]) left irc: Remote closed the connection
[04:21] ChrisHJW ([email protected]) joined #gstreamer.
[04:23] <wheels> CompIBook: Hmm, stuff still isn't working here, but I'll continue tomorrow.  I'm getting segfaults and am having trouble getting a useful backtrace...
[04:25] <wheels> All that I'm getting is: #0  0x40001e30 in _start () from /lib/ld-linux.so.2
[04:26] Action: BBB has no clue
[04:27] <BBB> are you debugging the right binary?
[04:28] <wheels> I certainly think so.  :-)
[04:29] <bitshifter> wheels: I haven't been able to get any useful backtraces with gdb either for some time, but valgrind seems to work quite well
[04:29] <wheels> bitshifter: no luck with valgrind either
[04:30] <BBB> wheels: as in, gst-launch-0.8, not gst-launch
[04:30] <BBB> :)
[04:30] <wheels> BBB: "gst-launch filesrc location=/home/scott/test.mp3 ! mad ! osssink" works, but "gst-launch kiosrc location=file:/home/scott/test.mp3 ! mad ! osssink" crashes with the above...
[04:30] <wheels> BBB: They're the same here...
[04:30] <BBB> no
[04:30] <BBB> the one is a script wrapper
[04:30] <BBB> or it's supposed to be
[04:30] jcsston ([email protected]) left irc: Read error: 60 (Operation timed out)
[04:30] Action: wheels tries again
[04:31] <wheels> BBB: Actually I didn't look at the files -- just the install time so I knew that gst-launch was installed with my HEAD install...
[04:31] <BBB> :)
[04:31] <wheels> Ah, now I'm getting somewhere...
[04:32] <wheels> Hmm, I actually get some sound under valgrind.
[04:33] <wheels> Ok, so it's memory corruption somewhere.  I'll hit it with valgrind tomorrow.  There's an invalid free, but the bt doesn't show the kiosrc anywhere...
[04:36] <BBB> :)
[04:36] Action: BBB has a scene detection plugin
[04:36] <BBB> whee :)
[04:36] <BBB> and it's pretty good, too
[04:37] <BBB> and fast
[04:37] <BBB> it takes a 3-5% CPU on my 1GHz, which means it's at least 20x realtime :)
[04:37] <BBB> and only a few false positives, and no false positives :)
[04:37] Action: BBB cheers
[04:38] <BBB> er
[04:38] <BBB> no false negatives
[04:38] Action: wheels throws a parse error
[04:38] <wheels> ah
[04:38] Action: BBB wonders what happened to the video editor that he was going to integrate this into
[04:42] CompIBook ([email protected]) left irc: Read error: 60 (Operation timed out)
[04:42] Action: wheels -> sleep()
[04:42] CompIBook ([email protected]) joined #gstreamer.
[04:42] Nick change: CompIBook -> Company
[04:43] #gstreamer: mode change '+o Company' by ChanServ!ChanServ@services.
[04:43] Topic changed on #gstreamer by [email protected]: if you're trying to debug gst-launch and fail, try gdb gst-launch-0.8
[04:43] #gstreamer: mode change '-o Company' by [email protected]
[04:45] <Company> so there
[04:51] jcsston ([email protected]) joined #gstreamer.
[04:51] <BBB> hm
[04:51] <BBB> one false positive in a minute
[04:51] <BBB> zero false negatives
[04:51] <BBB> and +/- 15 rightly detected
[04:51] Action: BBB very happy :)
[04:52] Action: ds-work attempts to compile gstreamer on windows
[04:53] <Company> ds-work: mingw?
[04:53] <ds-work> yes
[04:53] <Company> good luck
[04:53] <Company> i tried glib and stock glib doesn't compile under mingw
[04:54] <ds-work> so far, other packages seem to compile "randomly"
[04:54] <ds-work> I compiled glib 4 times, and on the 4th time, it worked
[04:54] <Company> and it only build .la files where i wanted dlls
[04:54] <Company> that was back in december when i wanted my tetris to compile
[04:57] <BBB> sweet! all correct this time!
[04:57] Action: BBB tries a different movie for a change
[04:57] <Company> BBB: what do you try?
[04:58] <BBB> automatic scene detection
[04:58] <BBB> integrated in GStreamer :)
[04:59] <BBB> step 1 in opening a movie in an editor
[04:59] <BBB> and I needed it for work, so ... :D
[05:00] <BBB> the goal for now is to make a list of images in HTML and make each of them clickable to start playing a file in a video player
[05:01] <Company> if you got it working, i'll integrate it in the gdk-pixbuf loader
[05:02] <Company> oh shit
[05:02] <Company> after fixing my partition i have 3 gigs more space
[05:07] <BBB> why would you want that?
[05:07] <BBB> in a pixbuf loader?
[05:12] <Company> because it gives you a pretty good idea about which pic you want to use as a thumbnail
[05:12] <BBB> actually
[05:13] <BBB> in nautilus, you should give the option to open a movie as a directory
[05:13] <BBB> and then show each scene as a file
[05:13] <BBB> and allow clicking those and then opening the movie in totem at that specific point
[05:13] <BBB> [/dream]
[05:13] <BBB> like chapters in DVD :)
[05:14] <BBB> if I have two pipelines and I want them both to negotiate to the same size, how do I do that?
[05:14] <BBB> so I want pipeline two to negotiate to the caps of ximagesink of pipeline one
[05:15] <Company> get the caps and use them as filter i'd say
[05:18] Action: BBB tries
[05:18] <BBB> does fakesrc work already?
[05:18] <BBB> from apps?
[05:18] Action: BBB loves such evil hacks
[05:18] <Company> fakesrc?
[05:19] <BBB> fakesrc, handoff...
[05:19] <BBB> guess what I'm trying to do ;)
[05:19] <BBB> and why I need caps
[05:19] <Company> fakesrc or fakesink?
[05:19] <BBB> fakesrc
[05:19] <BBB> I want to *input* data into a pipeline
[05:20] <Company> oh
[05:20] <Company> don't know that works
[05:20] <BBB> fakesrc's only real use ;)
[05:21] <BBB> too bad fakesrc isn't loopbased
[05:21] <BBB> that'd be more useful
[05:21] <Company> huh?
[05:22] <BBB> then *I* decide when to input buffers instead of getting a handoff signal and then waiting on the other pipeline to provide data
[05:22] <BBB> if you know what I mean
[05:22] <Company> you should write your own element
[05:23] <BBB> too much effort
[05:23] Action: BBB lazy
[05:24] Action: BBB is gonna write some evilish code
[05:24] <BBB> 2 threads
[05:24] <BBB> run one
[05:24] <BBB> wait for signal
[05:25] <BBB> thread_wait inside handler
[05:25] <BBB> run other pipeline
[05:25] <BBB> wait for data
[05:25] <BBB> wake up one
[05:25] <BBB> etc.
[05:25] <BBB> [/evil]
[05:25] <BBB> and the good thing is, I'll need to fix jpegenc while I'm at it :)
[05:26] <Company> i need a working x server
[05:27] <Company> everytime i end an x session i can't use the display anymore
[05:27] <Company> whoops
[05:28] <Company> running make start X on the other machine
[05:28] <Company> or running gcc
[05:29] <BBB> ever considered installing something like suse or redhat? :)
[05:29] <Company> debian
[05:29] <Company> but there's 2 reasons for not doing it yet
[05:30] <Company> 1) it doesn't install source by default and i can't make it to do that
[05:30] <Company> 2) it doesn't install debug-enabled libs by default and i can't make it to do that
[05:30] <BBB> 3) it works?
[05:30] <Company> it wouldn't on a broken hd either
[05:37] <ds-work> BBB: what's wrong with jpegenc?
[05:47] <BBB> it segfaults for me...
[05:48] <BBB> dunno why, never cared to fix it
[05:48] <BBB> it's been segfaulting since before 0.6.x
[05:48] <BBB> I'm surprised nobody ever filed a bug
[05:48] <ds-work> at one point, I had it working with videotestsrc ! jpegenc ! jpegdec ! xvimagesink
[05:49] <ds-work> with correct size negotiation
[05:49] <BBB> hm, interesting
[05:49] <BBB> then it crashed for some other reason
[05:49] <BBB> it always crashed when testing with v4lsrc
[05:49] <ds-work> but it doesn't quite work now
[05:50] <ds-work> file a bug with a backtrace
[05:50] <BBB> I'll fix it... I need it right now anywya, so...
[05:50] <ds-work> btw, we need to switch globally to image/jpeg
[05:55] <BBB> go ahead
[05:55] <BBB> MJPEG is not the same as JFIF/JPEG imo, but ohwell...
[05:55] <BBB> I won't care for now
[05:56] <BBB> http://bugzilla.gnome.org/show_bug.cgi?id=138582 <- for you guys
[06:02] <ds-work> right
[06:02] <ds-work> only for true streams of JFIF images
[06:03] <ds-work> video/x-mjpeg for mjpeg
[06:03] <ds-work> which is what it is currently, I think
[06:04] Action: ds-work wishes bugzilla.g.o would switch back to UTF-8
[06:05] <BBB> it's video/x-jpeg
[06:05] <BBB> not x-mjpeg
[06:10] <BBB> why doesn't fakesrc's handoff signal handoff a GstData ** instead of a GstBuffer *?
[06:11] <ds-work> in case you want to send an event?
[06:12] <BBB> yes
[06:12] <BBB> like...
[06:12] <BBB> GST_EVENT_EOS :)
[06:13] <BBB> if fakesrc/fakesink are supposed to be application entry/exit points for a pipeline (like identity is a middle point), then I think it'd be more useful
[06:14] <ds-work> perhaps
[06:14] <ds-work> I don't think fakesrc/fakesink are at all transparent in that capacity, however
[06:15] <BBB> does it matter?... it's supposed to be used for quickhacks imo
[06:16] <ds-work> not everything is going to fit the gstreamer model
[06:17] <ds-work> if we want people to convert their current applications to gstreamer, it would be nice to have decent appsrc/appsink elements
[06:19] <BBB> I consider fakesrc/fakesink those elements
[06:19] <BBB> and appfilter = identity
[06:19] <BBB> right?
[06:19] <ds-work> sure, but they suck for that purpose
[06:19] <ds-work> they work in a hack sense
[06:25] <BBB> that's their whole goal!
[06:25] <BBB> like appsrc wouldn't be a hack
[06:35] spyder482 ([email protected]) left irc: Read error: 110 (Connection timed out)
[06:38] trow ([email protected]) left irc: "Leaving"
[06:51] <BBB> bbl
[06:51] BBB ([email protected]) left irc: "Client exiting"
[06:57] apwbd ([email protected]) joined #gstreamer.
[07:02] <Company> omg
[07:02] <Company> it's 7 in the morning
[07:02] <thaytoo> hi
[07:02] <thaytoo> GOOD MORNING!
[07:02] Action: Company kicks his hd once again and then goes to bed
[07:23] <Misirlou> <Company> it's 7 in the morning
[07:23] <Misirlou> But it's always seven in the morning!
[07:23] <Misirlou> AM/PM IS A LIE TO KEEP THE AMERICANS QUIET
[07:31] Company ([email protected]) left irc: Read error: 60 (Operation timed out)
[07:31] apwbd ([email protected]) left irc: Read error: 60 (Operation timed out)
[07:32] #gstreamer: mode change '+o ds' by ChanServ!ChanServ@services.
[07:32] #gstreamer: mode change '+o ds-work' by ChanServ!ChanServ@services.
[07:32] #gstreamer: mode change '-o ds' by [email protected]
[07:32] #gstreamer: mode change '+o ds' by ChanServ!ChanServ@services.
[07:33] Topic changed on #gstreamer by [email protected]: if you're trying to debug gst-launch and fail, try gdb gst-launch-0.8 | Media playback tracker bug: #138435
[07:33] #gstreamer: mode change '-o ds' by [email protected]
[07:33] #gstreamer: mode change '+o ds' by ChanServ!ChanServ@services.
[07:33] #gstreamer: mode change '-o ds-work' by [email protected]
[07:34] wu-wei ([email protected]) joined #gstreamer.
[08:01] ChriHJW_log ([email protected]) joined #gstreamer.
[08:04] mxpxpod ([email protected]) left irc: "To quit or not to quit.. wait... was that a question?"
[08:07] ChrisHJW_log ([email protected]) left irc: Read error: 60 (Operation timed out)
[08:14] mxpxpod ([email protected]) joined #gstreamer.
[08:20] ChrisHJW ([email protected]) left irc: Read error: 110 (Connection timed out)
[08:54] thomasvs ([email protected]) joined #gstreamer.
[09:07] jcsston` ([email protected]) joined #gstreamer.
[09:11] markey ([email protected]) joined #gstreamer.
[09:13] ChrisHJW ([email protected]) joined #gstreamer.
[09:19] jcsston ([email protected]) left irc: Read error: 110 (Connection timed out)
[09:24] #gstreamer: mode change '-o ds' by [email protected]
[09:27] Nick change: mathrick|sleep -> mathrick
[09:27] <mathrick> mornin
[09:36] kmaraas ([email protected]) left irc: Client Quit
[10:01] wheels ([email protected]) left irc: "using sirc version 2.211+KSIRC/1.3.9"
[10:04] LeRoutier ([email protected]) joined #gstreamer.
[10:04] <LeRoutier> hello
[10:05] <LeRoutier> anyone know the owner/maintainer of control-center ?
[10:05] <LeRoutier> 2.6.0.1 still looks for gst 0.7, not 0.8
[10:13] Nick change: mathrick -> mathrick|away
[10:14] thomasvs ([email protected]) left irc: Read error: 110 (Connection timed out)
[10:20] dolphy ([email protected]) joined #gstreamer.
[10:21] <LeRoutier> hello dolphy
[10:24] <dolphy> hi
[10:25] <LeRoutier> seems control-center people have a not up to date at all gstreamer. already 2 gstreamer related bugs in control-center 2.6.0.1
[10:30] mathrick|away ([email protected]) left irc: Remote closed the connection
[10:31] thomasvs ([email protected]) joined #gstreamer.
[10:33] jdahlin ([email protected]) joined #gstreamer.
[10:37] <dolphy> morning dudes :)
[10:39] <jdahlin> morning dolphy
[10:40] LeRoutier ([email protected]) left irc: Read error: 60 (Operation timed out)
[10:47] LeRoutier ([email protected]) joined #gstreamer.
[10:52] Shoragan ([email protected]) joined #gstreamer.
[10:56] ChrisHJW ([email protected]) left irc: Remote closed the connection
[10:57] ChrisHJW ([email protected]) joined #gstreamer.
[11:06] teuf ([email protected]) joined #gstreamer.
[11:07] <teuf> hi
[11:07] <LeRoutier> lut teuf
[11:08] jonathang ([email protected]) joined #gstreamer.
[11:22] pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) joined #gstreamer.
[11:39] thomasvs ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[11:43] thomasvs ([email protected]) joined #gstreamer.
[11:47] sublett ([email protected]) joined #gstreamer.
[11:56] <LeRoutier> who is gst-recorder main dev ?
[11:57] <jdahlin> Ronald (BBB)
[11:57] markey ([email protected]) left irc: "leaving"
[12:01] <LeRoutier> ok, thx
[12:14] Rotty ([email protected]) joined #gstreamer.
[12:14] harshy ([email protected]) joined #gstreamer.
[12:27] Nick change: LeRoutier -> LeRaway
[12:44] thaytan ([email protected]) joined #gstreamer.
[12:47] <LeRaway> hum, gstream .desktop files are borked
[12:48] <LeRaway> _Name & _Comment in final .desktop files (could have been broken by i18n)
[12:50] <thomasvs> which .desktop files ?
[12:50] <LeRaway> all gst-editor & gst-player ones
[12:51] <LeRaway> they all have _Name & _Comment once installed so they don't show in menu
[12:51] Nick change: LeRaway -> LeRoutier
[12:51] <thomasvs>  here they look fine
[12:51] <thomasvs> sounds like a problem with your intltool
[12:51] <LeRoutier> well, perhaps, looking at it
[12:53] <thaytan> harro!
[12:53] <thomasvs> I find gst-player in my menus, but not gst-editor...
[12:53] <thomasvs> and indeed gst-editor for me also has _
[12:54] <LeRoutier> i rechecked, gst-player is in my menu. only gst-editor parts are missing from Dev
[12:56] <LeRoutier> while you're at it (.desktop files), could you add bug-buddy infos for gst-editor ? (X-GNOME-Bugzilla-*)
[12:56] <thomasvs> LeRoutier: well, if you know how to, please give us a patch
[12:57] <LeRoutier> ok, don't touch them, i'll do patches for gst-editor & -recorder in the following hours
[12:57] <thaytan> rock!
[12:57] <thaytan> I _have_ to check in this DVD stuff tonight
[12:57] <thaytan> although my esdsink changes are, ahem, unorthodox
[12:58] <thomasvs> thaytan: what happened ?
[12:58] <thaytan> I have working still frames, and with Martin's dvddemux, you can select 'English Subtitles' from the menu, and it works!
[12:58] <thaytan> all using a simple gst-launch pipeline
[12:59] <thomasvs> really ? that's cool :)
[12:59] <thaytan> only a couple of problems remaining: it crashes on a seek, and it doesn't figure out aspect ratios
[13:01] <thaytan> oh, and testing more DVDs
[13:01] ensonic ([email protected]) joined #gstreamer.
[13:05] <ensonic> hi hi
[13:06] Nick change: harshy -> harshyZ
[13:26] ploum ([email protected]) joined #gstreamer.
[13:35] kmaraas ([email protected]) joined #gstreamer.
[13:35] jcsston` ([email protected]) left irc: Read error: 110 (Connection timed out)
[13:44] markey ([email protected]) joined #gstreamer.
[13:48] sublett ([email protected]) left irc: "Why are all the really good songs short?"
[13:56] <LeRoutier> thomasvs, i found the gst-editor .desktop problem : they should not be in configure.ac/AC_OUTPUT (end of file) because they are generated by intltool
[13:57] <LeRoutier> i'll do a patch soon and post it to bugzilla
[14:02] <LeRoutier> http://bugzilla.gnome.org/show_bug.cgi?id=138623
[14:09] Company ([email protected]) joined #gstreamer.
[14:10] <LeRoutier> and a similar one for gst-recorder : http://bugzilla.gnome.org/show_bug.cgi?id=138624
[14:10] <LeRoutier> hi Company
[14:10] <Company> good morning angels
[14:11] teuf ([email protected]) left irc: Remote closed the connection
[14:12] <ensonic> thomasvs: I just look at the source of the adder and noticed that I not registers? a debug category (it not appears in the "gst-launch-0.8 --gst-debug-help" list
[14:15] <ensonic> thomasvs: I was going to see if I can get it to tell me what it does
[14:16] <thaytan> adding a debug category is pretty easy
[14:16] <ensonic> should every plugin register with the debug-log system? if yes, can anyone point me to a plugin I should look at to do a patch
[14:16] Nick change: LeRoutier -> LeRaway
[14:18] <thomasvs> http://gstreamer.freedesktop.org/data/doc/gstreamer/head/manual/html/section-debugging-adding.html
[14:19] <thomasvs> ensonic: look there
[14:21] Company ([email protected]) left irc: Read error: 60 (Operation timed out)
[14:22] sublett ([email protected]) joined #gstreamer.
[14:27] thaytoo ([email protected]) left irc: "leaving"
[14:27] Action: ensonic is reading section-debugging-adding.html
[14:35] <ensonic> when runnigngst-register-0.8 I get 
[14:35] <ensonic> (process:20495): GLib-GObject-CRITICAL **: file gparamspecs.c: line 1846 (g_param_spec_float): assertion `default_value >= minimum && default_value <= maximum' failed
[14:35] <ensonic>  
[14:35] <ensonic> (process:20495): GLib-GObject-CRITICAL **: file gobject.c: line 275 (g_object_class_install_property): assertion `G_IS_PARAM_SPEC (pspec)' failed
[14:36] <thomasvs> ensonic: use --gst-fatal-warnings and run from gdb to see what causes it
[14:39] markey ([email protected]) left irc: "leaving"
[14:41] <ensonic> thomasvs: done that, is not immediately helpful : gst_element_register (plugin=0x80b8d58, name=0x80d04e8 "ladspa-alienwah-stereo", rank=0, type=135070848)
[14:41] <ensonic>     at gstelementfactory.c:204
[14:41] <ensonic> klass = GST_ELEMENT_CLASS (g_type_class_ref (type));
[14:42] markey ([email protected]) joined #gstreamer.
[14:42] <ensonic> i'll look at at that later, probably is a ladspa plugin with a bad descriptor
[14:43] <alley_cat> artssink doesn't build with arts 1.2.1 unless i set CPPFLAGS="-I/usr/include/arts"
[14:44] markey ([email protected]) left irc: Client Quit
[14:47] <thomasvs> ensonic: hm, looks like there was a warning before it then
[14:48] <ensonic> thomasvs: should I paste the backtrace (20 lines)?
[14:50] <LeRaway> ++
[14:50] LeRaway ([email protected]) left irc: "Leaving"
[14:50] <thomasvs> ensonic: privately then
[15:12] <thomasvs> ensonic: what I'm saying is, I think you had criticals or warnings before the one about default_value, since the backtrace doesn't mention g_parm_spec_float
[15:15] <ensonic> thomasvs: no there are not others
[15:15] <ensonic> thomasvs: but you are right that is strange
[15:17] <ensonic> thomasvs: ahh no, it is there, I have just skipped the first lines, which are all in glib
[15:20] <thomasvs> ensonic: where is that alienwah ladspa plugin coming from ?
[15:20] Action: thomasvs doesn't have it
[15:23] <ensonic> thomasvs: suse9 (or http://www.tux.org/pub/tux/ladspa/)
[15:24] wheels ([email protected]) joined #gstreamer.
[15:24] <ensonic> thomasvs: anyway most ladspa rpm have it bundled
[15:26] <thomasvs> ensonic: I've never seen it.  is that link the upstream location ?
[15:29] <ensonic> thomasvs: i've just googled for it, and have seen that it is contained in several rpms
[15:30] <thomasvs> ensonic: cool, I should add it then.
[15:31] <thomasvs> ensonic: but since it's only that LADSPA plugin, it's likely that the plugin is doing something wrong
[15:31] <thomasvs> ensonic: so file a bug for us with the dl location of that ladspa plugin so we can add checks to our general ladspa plugin
[15:32] <ensonic> thomasvs: I'll do that
[15:32] <ensonic> thomasvs: btw. that debuging stuff in the adder now works, I try to fix it as a practive in gst_plugins
[15:36] <alley_cat> does artssink build for anyone?
[15:46] iain ([email protected]) joined #gstreamer.
[15:46] <iain> I managed to write a 2000 word essay this morning
[15:46] <iain> pity it had to be in for last thursday
[15:54] KoRnouille ([email protected]) left irc: "Leaving"
[16:08] kmaraas ([email protected]) left irc: "Leaving"
[16:20] smoke ([email protected]) joined #gstreamer.
[16:23] sublett ([email protected]) left irc: "Why are all the really good songs short?"
[16:28] mxpxpod ([email protected]) left irc: "To quit or not to quit.. wait... was that a question?"
[16:50] foser ([email protected]) joined #gstreamer.
[17:07] trow ([email protected]) joined #gstreamer.
[17:08] sublett ([email protected]) joined #gstreamer.
[17:17] smoke ([email protected]) got netsplit.
[17:23] smoke ([email protected]) got lost in the net-split.
[17:35] Company ([email protected]) joined #gstreamer.
[17:36] BBB ([email protected]) joined #gstreamer.
[17:36] <Company> i need a filesystem that can recover gracefully on HD failures
[17:37] <BBB> no, stay here you!
[17:38] <wheels> Company : tmpfs?  :-)
[17:38] <thaytan> anyfs on RAID :)
[17:39] yippi ([email protected]) left irc: Remote closed the connection
[17:45] Action: BBB needs cygwin or so
[17:53] <wheels> Company : How well did the kiosrc actually work for you?
[17:53] Action: wheels has just recently gotten it to where it'll sort of play a couple seconds of stuff before segfaulting...
[17:54] <thomasvs> if pads are float audio, what datatype does the buffer use ? gfloats ?
[17:54] <ensonic> has gstreamer an API to query a list of plugins like filter(mime_class,plugin_type), e.g. filter("audio","source")
[17:56] Company ([email protected]) left irc: Read error: 113 (No route to host)
[17:56] <wheels> thomasvs : GstBuffer?
[17:57] <thomasvs> wheels: no, I mean, what does the buffer contain ? gfloat, gdouble, not defined ?
[17:57] <ensonic> thomasvs: i'll check the ladspa wrapper, as most loadspa plugins use float
[17:58] Company ([email protected]) joined #gstreamer.
[17:58] <ensonic> thomasvs: gfloat
[18:02] BB1 ([email protected]) joined #gstreamer.
[18:02] BBB ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[18:03] <BB1> thomasvs: depends on 'depth'
[18:03] <BB1> if depth is 4, it's gfloat, if depth is 8, it's gdouble
[18:03] <BB1> see the PWG or docs/random/mimetypes
[18:03] <BB1> it's all documented
[18:03] <Company> 32 and 64, not 4 and 8
[18:03] <thomasvs> BB1: then a lot of plugins don't advertise a depth range for float audio
[18:04] <thomasvs> and I hope Company is right, and you didn't choose byte values for depth :)
[18:04] <Company> and it's width not depth
[18:04] <BB1> thomasvs: er, that's fundamentally broken then
[18:04] <BB1> :)
[18:04] <BB1> oh, width?
[18:04] Action: BB1 didn't read the docs himself O:-)
[18:04] <thomasvs> width and depth actually
[18:04] <BB1> no, only one of the two
[18:04] <BB1> there's no padding in float caps
[18:04] <BB1> as in int
[18:04] <thomasvs> right, so they're equal
[18:04] <Company> depth is how many bits are used
[18:04] Uraeus ([email protected]) joined #gstreamer.
[18:05] <Company> since float uses all bits, we don't have it there
[18:05] <Uraeus> good evening
[18:05] <thomasvs> hm, imo to be consistent it should just have both values
[18:05] <BB1> er, and yes, it's in bits
[18:05] <BB1> so it's 32/64
[18:05] Action: BB1 is sleepy...
[18:05] Action: BB1 got up at 4:30am this morning
[18:05] Action: BB1 saw Janet Jackson live in Good Morning America :)
[18:05] Action: thomasvs still not sure now if adder should add width or depth to the float caps
[18:05] <BB1> in the park near wall street :)
[18:06] <BB1> thomasvscheck the PWG!
[18:07] <thomasvs> the mimetypes doc says "width"
[18:07] <thomasvs> I'm just not entirely sure that's the conceptually correct one
[18:07] <Company> width is the number of bits per frame
[18:08] <Company> depth only exists for int and is the number of used bits per frame
[18:08] <wheels> Company : If you want some fun try kiosrc under valgrind.  :-)
[18:08] <thomasvs> so what are we going to do when we have 12bit float audio ?
[18:08] <Company> wheels: i have enough fun installing stuff on my new hd, thank you very much
[18:08] <jonathang> thomasvs, you're looking at gstadder?
[18:09] <Company> thomasvs: what's 12 bit float?
[18:09] <BB1> thomasvs: we'' see when we get there
[18:09] <BB1> don't worry about impossible things until they're possible
[18:09] <ensonic> IEEE basically defines 32bit, 64bit, 80bit floats
[18:09] <BB1> same for 64bit RGB
[18:10] <Company> in case someone invents it he can use audio/x-raw-float-with-12-bits
[18:10] <thomasvs> there are hardware devices that do 12bit audio float
[18:10] <ensonic> Company: ;-)
[18:10] <thomasvs> it's not like as if anyone of us has seen int audio with width != depth :)
[18:10] <BB1> in theory
[18:10] <ensonic> thomasvs: but not floats then, PCM-12bit
[18:10] <thomasvs> so I really don't see why we're making a difference between int and float on this point
[18:10] <Company> thomasvs: i have
[18:10] <BB1> avi/wav supports it
[18:10] <Uraeus> hey, we get a nice mention in the Ars Technica review :)
[18:11] <BB1> yup, noticed
[18:11] <BB1> but the comment is right
[18:11] <BB1> we need totem in 2.8
[18:11] <BB1> so we need to make totem rock really fast
[18:11] Action: BB1 works on that
[18:12] <Company> we've got half a year
[18:12] <BB1> what is dolphy doing?
[18:12] <Company> should be enough to get scheduling, autoplugging and tag writing done
[18:12] <Company> oh, and DVD
[18:12] <Company> maybe even subtitles
[18:12] <Company> Debian or LFS?
[18:13] <BB1> ?
[18:13] <thomasvs> thaytan says he has subtitles 
[18:13] <thomasvs> <thaytan> I have working still frames, and with Martin's dvddemux, you can select 'English Subtitles' from the menu, and it works!
[18:13] <thomasvs> <thaytan> all using a simple gst-launch pipeline
[18:13] <thomasvs> <thomasvs> really ? that's cool :)
[18:13] <Company> i was talking about generic subtitle file support
[18:13] <BB1> *nod*
[18:13] <Company> not just mpeg subtitles
[18:13] <BB1> avi, matroska, ...
[18:13] <Company> ideally with copy/paste btw
[18:14] <BB1> mpeg has no subtitles, that's a DVD extension
[18:14] <BB1> Company: I guess that in a video player, it's just part of the frame
[18:14] <BB1> in a subtitle editor, it'd be a separate textbox below the video widget
[18:14] <Company> we have a framework to allow that
[18:14] <Company> capture mouse events and send selection events
[18:15] <BB1> it'd be cool, but very low priority
[18:15] <Company> yup
[18:15] <BB1> I have moer important things on my mind than copypasting from a videoscreen
[18:15] <Company> hehe
[18:15] <BB1> like
[18:15] <BB1> getting mpeg4 to playback and so on
[18:20] <jonathang> thomasvs, if you're looking at gstadder and i can help you, i'd be glad to :)
[18:20] markey ([email protected]) joined #gstreamer.
[18:21] <ensonic> Company: as we have already found out spider is not using "audioscale ! audioconvert", has this a reason or is just not yet done?
[18:21] <ensonic> jonathang: we have basicly fixed it, we just need to make sure we *properly* fix it
[18:21] <thomasvs> jonathang: it's almost fixed
[18:21] <jonathang> fantastic!
[18:21] <thomasvs> jonathang: I'm going through bits of ensonic's patch and fixing some other stuff I see
[18:22] <BB1> thomasvs: your blog on 29/3 mentions license issues for media, that needs more attention
[18:22] Action: BB1 writes blog entry
[18:22] <thomasvs> more attention ?
[18:22] Action: jonathang is reeeeeeally happy
[18:22] <ensonic> anyone knows, if gstreamer has an API to query a list of plugins like filter(mime_class,plugin_type), e.g. filter("audio","source")
[18:22] <ensonic> jonathang: whatfor do you need the adder?
[18:22] <jonathang> for a "DJ" application
[18:23] smoke ([email protected]) joined #gstreamer.
[18:23] <jonathang> a dual player with mixing in other words
[18:23] <jonathang> so the adder is really needed... :)
[18:23] <ensonic> jonathang: same situation here ;-)
[18:24] <BB1> thomasvs: I want gst-ffmpeg in gnome?
[18:24] <BB1> it's ridiculous that we moved it out (users' point of view)
[18:24] <thomasvs> BB1: it doesn't matter, it can't go in gnome
[18:24] <BB1> I actually asked jeff, and he just mentioned that it "had not been submitted for inclusion"
[18:24] <BB1> where has this discussion taken place?
[18:24] <BB1> I'd love to read it back
[18:25] <thomasvs> BB1: it's been discussed with the foundation
[18:25] <thomasvs> feel free to bring it back up again
[18:25] <BB1> WHERE
[18:25] <thomasvs> the foundation doesn't discuss openly
[18:25] <thomasvs> just mail them
[18:25] <thomasvs> they clearly state GNOME cannot ship code that infringes on patents
[18:25] <BB1> I'm a foundation member, don't I have access to what they do?
[18:25] <thomasvs> they pulled the springload folder too
[18:25] <BB1> and 99% of ffmpeg doesn't infringe on patents
[18:25] <thomasvs> BB1: why are you complaining to the wrong person all the time
[18:25] <BB1> not to you ;)
[18:25] <jonathang> ensonic, hey, that's cool. Have you already got something working? (I began with another framework and had bits working... but nothing satisfying or showable :( )
[18:25] <thomasvs> BB1: it doesn't really matter, it's the 10% that does infringe that matters
[18:26] <BB1> I'm just asking you because you told me it couldn't go in so I assumed you knew where to find info on that
[18:26] <thomasvs> BB1: because I discussed with the board :)
[18:26] <BB1> what's the foundation's email?
[18:26] <thomasvs> BB1: so you should do the same if you really want it in
[18:26] <thomasvs> BB1: it's an ml on gnome, check lists.gnome.org
[18:26] <thomasvs> cc me please
[18:27] <BB1> if it's a foundation ml, then you're on it
[18:27] <thomasvs> no
[18:27] <BB1> you're a foundation member aren't you?
[18:27] <thomasvs> board
[18:27] <BB1> oh
[18:27] <thomasvs> it's the board
[18:27] <BB1> sorry
[18:27] Action: BB1 checks
[18:27] <thomasvs> BB1: but you realize that a) gnome cannot ship patent-troubled code and b) ffmpeg contains some of that ?
[18:28] <thomasvs> BB1: ie, it's pretty clear cut, and you don't accept it because you would really like it to be there and chose to ignore the fact that it can't due to clearcut patent problems ?
[18:28] <BB1> I know that
[18:28] <BB1> but I want gst-ffmpeg in in some way
[18:28] <BB1> there's lots of unpatented code in there
[18:28] <thomasvs> but why ?
[18:28] <thomasvs> everyone can use it already
[18:28] <BB1> because it's 90% of our codecs
[18:28] <thomasvs> what do you care about losing a legal battle ?
[18:29] <thomasvs> BB1: it's already ten times better as a situation than, say, springloaded folders, or AA hinting.
[18:29] <thomasvs> since we already are pluggable
[18:29] <thomasvs> it's just a matter of compiling or getting the package
[18:29] <thomasvs> we should be happy
[18:29] <thomasvs> why do you want to start a discussion that you already lost before beginning ?
[18:29] <BB1> I'd love a default redhat to be able to playback, say, ADPCM
[18:29] <BB1> it doesn't
[18:29] <BB1> I understand that it won't playback MPEG
[18:29] <thomasvs> BB1: then spend your time convincing ffmpeg to split up
[18:29] <BB1> no
[18:30] <thomasvs> that's the right thing to do
[18:30] <BB1> that's useless
[18:30] <thomasvs> no it's not
[18:30] <thomasvs> it should be pluggable :)
[18:30] <BB1> you know that
[18:30] <BB1> it's not, ffmpeg has no interest in legal issues
[18:30] <jdahlin> why not move ADPCM to gst-plugins?
[18:30] <BB1> jdahlin: I'm _NOT_ going to fork any more ffmpeg code
[18:30] <thomasvs> BB1: ok, so there's our problem.
[18:30] <BB1> and ADPCM is an example
[18:30] <thomasvs> BB1: we can't solve it if we can't make ffmpeg care.
[18:30] <BB1> half of the codecs in there are unpatented
[18:30] <BB1> there has to be a way to dist a tarball without the legally problematic code
[18:31] <thomasvs> BB1: it's always easier to convince people that have no reason to do something you want, than convince people that have good enough reasons not to do what you want
[18:31] <BB1> I mean, gst-ffmpeg has --disable-risky
[18:31] <BB1> and if that means I have to publically announce that I know what code is troublesome, then so be it
[18:32] <thomasvs> if what means what ?
[18:32] <thomasvs> you jumped a few steps
[18:32] <BB1> disting two gst-ffmpeg tarballs
[18:32] <ensonic> jonathang, see private message
[18:32] <BB1> one complete, one clean
[18:32] <thomasvs> BB1: since there's no --disable-non-risky, that'll be tricky
[18:32] <thomasvs> BB1: anyway, it's too late for gnome 2.6 anyway
[18:32] <BB1> --disable-risky ;)
[18:32] <BB1> I know
[18:33] <BB1> it's for gnome-2.10 or 2.8
[18:33] <BB1> I don't care much
[18:33] <thomasvs> ok, so don't jump the gun and take some time to think it through
[18:33] <BB1> but do you see my problem?
[18:33] <thomasvs> it's really not as simple as you want it to be
[18:33] <thomasvs> BB1: of COURSE I see the problem.
[18:33] <BB1> I understand
[18:33] <thomasvs> BB1: I just don't think shipping with gnome is the solution
[18:33] <thomasvs> BB1: people are smart enough
[18:33] <BB1> I'm affraid that not shipping with gnome will give many people bad opinions about us
[18:33] <thomasvs> everyone and their dog installed the mp3 plugins for fedora core 1
[18:34] <thomasvs> BB1: that's not true at all.
[18:34] <BB1> especially since we're not very explicit about why we only ship part of our code into gnome
[18:34] <thomasvs> BB1: the bad opinions are because we don't play those formats properly :)
[18:34] <thomasvs> BB1: if you want me to be explicit to gnome about why gst-ffmpeg is not in, I can :)
[18:34] <BB1> we playback anything on my HD, and I'm working on all problematic files I get from people right now
[18:34] <thomasvs> BB1: but people already know why xine and mplayer are not able to be a part of gnome right now.
[18:35] <thomasvs> anyone have suggestions on how to test the float part of adder now ?
[18:35] <BB1> trow, some time ago, told us that "people have the idea that gst is slow, unstable and supports little formats"
[18:35] <BB1> that's not true
[18:35] <BB1> so how do we convince them?
[18:35] <thomasvs> by being fast and stable for more than a week
[18:35] <thomasvs> the little formats we explain people how to solve it
[18:36] <thomasvs> the fedora people already can subscribe to our repo and play back *everything*
[18:36] <thomasvs> that's all you need to do
[18:36] <thomasvs> *companies* will license codecs and get code written to decode it
[18:36] <thomasvs> I already wrote my blog entry :)
[18:36] <BB1> why do people generally use totem-xine then?
[18:36] <thomasvs> BB1: simple - because it's been a lot more stable for a lot longer time
[18:36] <thomasvs> people TRUST xine
[18:36] <thomasvs> and I can't blame them - it's been a long time since I've managed to crash xine
[18:37] <thomasvs> having your machine be able to play back all files doesn't really fix bastien not even being able to play a single file with his totem-gst :)
[18:37] <BB1> I fixed the file he provided ;)
[18:37] Action: thomasvs fixed a few obvious bugs that might have helped in that respect, should ask bastien
[18:37] <thomasvs> currently, I'm even unable to play back anything myself with cvs
[18:37] <thomasvs> while my 0.8.0 rpms work fine, strange
[18:38] <BB1> hm...?
[18:38] <thomasvs> anyway, those are the issues we need to work on
[18:38] Action: BB1 uses CVS only
[18:38] <thomasvs> it needs to work for everyone, not just one of us
[18:38] <thomasvs> anyway, let me fix up adder first, then move on to go back to the player and what other errors are preventing playback
[18:38] <BB1> so I need test files that don't work
[18:38] <BB1> from any of you
[18:38] <BB1> and from bastien
[18:38] <BB1> and from whatever user out there
[18:39] <BB1> so I can sit one day a week, and fix each and every playback issue
[18:39] <BB1> (until someone takes that damn job over from me)
[18:39] <thomasvs> hehe :)
[18:39] <thomasvs> you're doing good, don't worry
[18:39] <thomasvs> I love gst-ffmpeg
[18:39] <thomasvs> just accept like everyone that it's not shippable as-is with gnome
[18:40] <BB1> I know, but it's not perfect yet
[18:40] <BB1> I need more ;)
[18:40] <BB1> especially exotic file formats
[18:40] <BB1> like...
[18:40] <BB1> anyone have doom3 movies>? castle wolfenstein movies?
[18:40] <BB1> those
[18:40] <thomasvs> mplayer has support for tons of those game movie formats
[18:40] <jdahlin> BB1, download the demos?
[18:41] <BB1> jdahlin: probably
[18:41] <BB1> I need a bigger HD
[18:41] Action: BB1 wants to buy an ipod
[18:41] <BB1> thomasvs: via ffmpeg :) we support them too, but I cannot test them
[18:41] <BB1> so there's probably trivial issues
[18:41] <jdahlin> BB1, do you have all from mplayerhq.hu/MPlayer/samples ?
[18:42] <dolphy> BB1: i m quite busy at that time
[18:42] <jdahlin> once you can get a few more of these working we can start add them to the testsuite
[18:43] <BB1> sure
[18:43] Action: BB1 marks that link
[18:45] <BB1> the bad thing is that ffmpeg includes very basic demuxers
[18:45] <BB1> mplayer forks those and adds advanced stuff like seeking and so on
[18:45] <BB1> ffmpeg doesn't have that :(
[18:46] <jdahlin> ftp that is
[18:47] <thomasvs> can someone tell me again what buffer-frames does again, and why the standard audio float caps specify 1-max while vorbisdec specifies 0 ?
[18:47] <thomasvs> BB1: so, let's diff and add mplayer code as patches ?
[18:48] <Company> buffer-frames is the bumber of frames per buffer
[18:48] <Company> 0 means undefined
[18:48] <BB1> thomasvs: GPL...
[18:48] <thomasvs> Company: so why specify it then in the case of 0 ?
[18:49] <thomasvs> Company: because if vorbisdec serves 0 then it cannot connect to caps listing 1-max
[18:49] <thomasvs> hrm, I really can't test float adder it seems
[18:52] <Company> thomasvs: correct and it shouldn't
[18:52] Action: benow wispers CVS to thomasvs
[18:52] kmaraas ([email protected]) joined #gstreamer.
[18:52] <thomasvs> benow: I'm waiting for an fdo admin to approve you
[18:53] <Company> thomasvs: there may be elements which require all buffers being the same size - those specify 1-MAX
[18:53] <benow> sure.  I'm trustworthy... honest.
[18:54] walters ([email protected]) joined #gstreamer.
[18:56] <thomasvs> Company: hm, so those elements probably are wingo-elements, no ? :)
[18:57] <thomasvs> imo in general adder and similar elements should advertise buffer-frames 0 then, correct ?
[18:57] <Company> thomasvs: 0-MAX probably
[18:58] <Company> thomasvs: all elements that can cope with variable input should be able to cope with defined input sizes, too
[18:59] <BB1> it should just check for the property on caps on sink_link()
[18:59] dolphy ([email protected]) left irc: "Network down, IP Packets delivered via UPS"
[18:59] <BB1> but not require it
[18:59] <BB1> so don't put it in the sink_getcaps()
[18:59] <BB1> unless elements further on do require it
[19:03] <ensonic> is it valid to use a bin with two elements as an element in a thread?
[19:04] <thomasvs> right, so how about I just fix the int part then ask for input on the float part :)
[19:04] yippi ([email protected]) joined #gstreamer.
[19:05] <BB1> ensonic: yes
[19:05] <BB1> go right ahead, thomasvs
[19:05] Action: BB1 wants to try out roq video tonight
[19:07] <ensonic> BB1: then I found a bug:  gstoptimalscheduler.c(1423):gst_opt_scheduler_state_transition: element "audio-source" has no group
[19:09] <BB1> are they linked?
[19:13] <ensonic> BB1: the elements in the bin are linked, the ghost-pads are created and then the bin is inserted in the thread and the thread elems are linked
[19:14] <thomasvs> jonathang: if you want, you can try updating cvs and try out adder
[19:15] <jonathang> ok
[19:15] kwm ([email protected]) joined #gstreamer.
[19:16] decadix ([email protected]) joined #gstreamer.
[19:16] <ensonic> BB1: uhm, forgot to *add* the elements to the inner bin
[19:16] sub_pop ([email protected]) left irc: "Client exiting"
[19:16] <decadix> so, what's wrong with GstThread?
[19:18] <ensonic> BB1: now its crashing nicely ;-)
[19:18] <BB1> sweet ;)
[19:18] <decadix> ok, i guess i need to use a queue?
[19:18] <BB1> decadix: yes
[19:19] <ensonic> ... without a backtrace being available in gdb
[19:19] <BB1> use gst-launch-0.8, not gst-launch
[19:19] <BB1> see topic :)
[19:19] <BB1> gst-launch is a wrapper
[19:19] <decadix> back to work
[19:20] pb_ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) left irc: "bbl"
[19:23] apoc ([email protected]) joined #gstreamer.
[19:24] <apoc> hi
[19:26] <jonathang> i always have problems with translation (.po) files after a cvs update. is it normal? should i redo a fresh checkout?
[19:27] <thomasvs> jonathang: it is fairly normal.  if they conflict, delete them
[19:27] mxpxpod ([email protected]) joined #gstreamer.
[19:28] <bitshifter> jonathang: you can do  cvs update -C   - that will replace all locally modified files with fresh copies (don't do that if you work on code of course =)
[19:28] <jonathang> ok thanks!
[19:28] <BB1> hi apoc
[19:29] <BB1> apoc: want to help me working gradually on improving exotic media file support? you've done quite some work on that in the past :)
[19:29] <BB1> apoc: jdahlin pointed me to some sample ovies, so that should be doable
[19:29] <apoc> BB1: ok
[19:29] <BB1> we've got code for all of it, it basically needs checking and small-fixing
[19:29] <BB1> like svq3
[19:30] <apoc> BB1: there are still problems with svq3 ?
[19:30] <BB1> http://mplayerhq.hu/MPlayer/samples/ <- that's your source
[19:30] <BB1> apoc: no, works perfectly
[19:30] <apoc> BB1: ok
[19:30] <BB1> but it didn't work six months ago ;)
[19:32] wheels ([email protected]) left irc: Read error: 60 (Operation timed out)
[19:32] <apoc> BB1: what format/codec should be supported ?
[19:33] <BB1> anything in http://mplayerhq.hu/MPlayer/samples/game-formats/
[19:33] <decadix> ok, helloworld.c gives chunky output, and queue.c gives no output (GStreamer-WARNING **: bin "bin" is not the managing bin, can't be iterated on!)
[19:33] <BB1> other formats can be tested etc.
[19:34] <bitshifter> apoc: btw, is there any chance you could check your replay gain plugins into gst-sandbox or put them up somewhere? :)
[19:34] <BB1> decadix: what are you trying to do?
[19:34] <decadix> i'm trying to program with GStreamer, BB1
[19:34] <decadix> the examples provided in the documentation just don't work
[19:35] <BB1> that sounds sort of odd...
[19:35] Action: BB1 should check those one day
[19:35] <decadix> perhaps the examples are specific to 0.8?
[19:35] <decadix> i'm running 0.6.4, btw
[19:35] <BB1> did you read the application development manual?
[19:35] <decadix> yes
[19:35] <BB1> try 0.8.0, 0.6.4 is deprecated
[19:35] <ensonic> anyone an idea what to do when geting a segmentation fault, but no backtrace in gdb?
[19:35] <decadix> i will
[19:35] <BB1> ensonicdebug gst-launch-0.8, not gst-launch?
[19:35] <BB1> ensonic: or provide info on what you're debugging
[19:36] thomasvs ([email protected]) left irc: Read error: 60 (Operation timed out)
[19:36] <apoc> BB1, thomasvs : will we create a CVS branch for 0.8.x ?
[19:36] <BB1> not yet
[19:36] <BB1> we will, later on
[19:37] <apoc> I've a replaygain plugin ... should I put it into gst-sandox or gst-plugins ?
[19:37] <apoc> no extra libs needed
[19:40] sub_pop ([email protected]) joined #gstreamer.
[19:41] wheels ([email protected]) joined #gstreamer.
[19:43] Action: bitshifter would vote for gst-plugins if he had a vote
[19:45] <ensonic> BB1: own code
[19:47] <ensonic> BB1: the example from above, a thread(sinesrc,bin(audioscale,audioconv),esdsink), I can set i up, but right after starting to play, it seg-faults
[19:49] <ensonic> BB1: when not wrapping audioscale,audioconv it works
[19:50] <bitshifter> ensonic: tried with valgrind, or with gdb and export LD_ASSUME_KERNEL=2.4.1 ?
[19:56] <ensonic> bitshifter: export LD_ASSUME_KERNEL=2.4.1 doesn't makes any difference
[19:58] apoc_ ([email protected]) joined #gstreamer.
[20:00] thomasvs ([email protected]) joined #gstreamer.
[20:01] jonathang ([email protected]) left irc: "Leaving"
[20:01] jdahlin ([email protected]) left irc: "Leaving"
[20:02] Company ([email protected]) left irc: Read error: 113 (No route to host)
[20:02] Company ([email protected]) joined #gstreamer.
[20:02] pb_ (~pb@2002:5160:4550:1:240:95ff:fe30:aa84) joined #gstreamer.
[20:03] <ensonic> bitshifter: for valgrind I at first need to find out how to use it valgrind ./myapp myopts does just start and idle
[20:03] Action: ensonic will continue tomorrow
[20:03] <ensonic> cu'
[20:04] <bitshifter> ensonic: with valgrind things will run A LOT slower
[20:04] <ensonic> a so I will wait
[20:04] <bitshifter> like: factor 50-100x or so (rough guess)
[20:04] ploum ([email protected]) left irc: "Leaving"
[20:05] apoc ([email protected]) left irc: Read error: 60 (Operation timed out)
[20:07] Uraeus ([email protected]) left irc: Remote closed the connection
[20:07] <ensonic> bitshifter: it reported just two Invalid free() / delete / delete[]
[20:08] <BB1> that's a start
[20:08] dolphy ([email protected]) joined #gstreamer.
[20:09] <ensonic> its not as it is in gst_init_check_with_popt_table (gst.c:333), and stuff like that
[20:09] <ensonic> amazingly, my prog is not seg-faulting in valgrind ...
[20:11] <bitshifter> that's normal (ie. that it doesn't segfault in valgrind even if it would otherwise)
[20:11] <ensonic> the lst thing I get when I 'crawl' my app with ---gst-debug="*:5" is gstoptimalscheduler.c(1164):loop_group_schedule_function: calling loopfunc of element audio-source in group 0x81bca90
[20:12] benow ([email protected]) left irc: Read error: 113 (No route to host)
[20:14] <ensonic> okay, let have a look at this tomorrow, many thanks though
[20:14] <ensonic> ciao
[20:14] ensonic ([email protected]) left #gstreamer ("Leaving").
[20:24] alley_cat ([email protected]) left irc: Read error: 60 (Operation timed out)
[20:36] alley_cat ([email protected]) joined #gstreamer.
[20:42] sublett ([email protected]) left irc: "Why are all the really good songs short?"
[20:44] jdahlin ([email protected]) joined #gstreamer.
[20:46] apoc_ ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[20:50] wheels ([email protected]) left irc: "switching clients"
[20:51] wheels ([email protected]) joined #gstreamer.
[20:54] Shoragan ([email protected]) left irc: "Leaving"
[21:00] abo_ ([email protected]) joined #gstreamer.
[21:09] <Company> i so hate the mozilla feature to download into $tmp
[21:10] <Company> especially when downloading isos on windows
[21:12] <jdahlin> trow: nice patch for pygtk
[21:13] teuf ([email protected]) joined #gstreamer.
[21:14] <trow> jdahlin: do you think it is OK to commit?
[21:14] LeRoutier ([email protected]) joined #gstreamer.
[21:14] <LeRoutier> hello
[21:15] <LeRoutier> BB1, ready to roll on test cases to fill bugs
[21:15] <jdahlin> trow, looks pretty safe, but I'd rather have james eye it first
[21:18] <trow> jdahlin: probably a good idea
[21:20] <jdahlin> trow: I find the refcounting and sinking a bit scary
[21:23] <trow> jdahlin: yeah, me too
[21:32] <BB1> LeRoutier: go ahead!
[21:32] <BB1> trow: can you do the videodev2.h patch please?
[21:32] <BB1> trow: it affects mdk10 too
[21:32] <BB1> :(
[21:32] <LeRoutier> BB1, already  2 filled
[21:32] <BB1> good, I'll start working on it sunday or so
[21:33] <LeRoutier> ok, no hurry
[21:33] <BB1> friday and saturday, I'm gone all day
[21:33] <BB1> thursday I need to sleep
[21:33] <LeRoutier> i did a small trivial patch for gst-recorder to have a working .desktop file, saw it ?
[21:33] <trow> BB1: Do you know the bug number, by any chance?
[21:34] Action: trow searches, finds that it is 135919
[21:34] <trow> b.g.o sure is a lot nicer now...
[21:35] <Company> heh
[21:35] <LeRoutier> BB1 : should i also link bugs like "my avi file plays but with dirty colors (compared to mplayer)" to this tracker or should i only link crashers ?
[21:35] <Company> people have different opinions about b.g.o
[21:35] <Company> i liked the old interface better
[21:36] <BB1> LeRoutier: file anything that's not good
[21:36] <BB1> so wrong colors is also part of that
[21:36] <LeRoutier> ok, filling a new one for the george michael clip so
[21:40] <LeRoutier> done, next
[21:41] <Company> anyone knows about debian packages of the freedesktop stuff? project utopia and keith's xserver in particular?
[21:44] <ds-work> Company: they're in experimental, iirc
[21:52] <Company> how does that work? if i add experimental to my sources.list, will it install all eperimental packages or just those where there's no equivalent in unstable?
[21:54] <pb_> by default, it'll still prefer unstable.  you have to use '-t experimental' to get stuff from experimental.
[21:54] <pb_> but you can tweak apt's configuration to prefer the bits from experimental if you're sufficiently hardcore.
[21:56] <Company> nope
[21:56] <Company> i just want stuff from experimental when the package is not available in unstable
[21:59] caro ([email protected]) joined #gstreamer.
[22:03] mathrick ([email protected]) joined #gstreamer.
[22:03] <mathrick> wassup
[22:04] Nick change: harshyZ -> harshy
[22:05] <mathrick> yo harshy, any news on libburn front?
[22:06] <harshy> nothing really
[22:07] Action: mathrick can't wait for dropping xcdroast
[22:22] <markey> uh, my gst engine has a memleak
[22:22] <markey> big one
[22:22] <markey> in the handoff callback, I'm getting a GstBuffer
[22:22] <markey> do I need to free that?
[22:22] <markey> GstBuffer* rather
[22:22] <LeRoutier> BB1: most of my bugs are of 2 types : 1) uncompressed PCM audio not handled kills avidemux  2) thread problem
[22:24] <markey> gst_buffer_default_free(buf) does not work
[22:25] <mathrick> markey: you should probably unref it, if I were to guess
[22:25] <markey> what function to use?
[22:25] <markey> gst_buffer_unref() maybe?
[22:25] <mathrick> lemme check if it's GObject
[22:25] <markey> I'm not sure this is calling the leak, but it looks like it
[22:26] <markey> since the callback gets called very frequently
[22:26] <markey> quickly adds up
[22:26] <mathrick> markey: yep, gst_buffer_unref()
[22:26] <markey> thx!
[22:27] <mathrick> markey: maybe you should consider installing devhelp just for gst docs? :)
[22:28] <markey> devhelp?
[22:28] <mathrick> gnome's api browser
[22:28] <mathrick> very handy
[22:28] <mathrick> extremely even
[22:28] <markey> (process:13091): GStreamer-CRITICAL **: file gstdata.c: line 238 (gst_data_unref): assertion `GST_DATA_REFCOUNT_VALUE (data) > 0' failed
[22:28] <markey> uh
[22:28] <markey> oh, I've got a web browser, konq
[22:29] <markey> docs are nicely on the web, so it's okish
[22:29] <mathrick> markey: do you have incremental search amongst all gst api in konq? :)
[22:30] <mathrick> markey: if so, I'll promptly shut up :>
[22:30] <markey> I bet it's hidden in one of the 5000 options ;)
[22:30] <BB1> mathrick: I noticed, that's ok
[22:30] <BB1> I'll look anyway
[22:31] dolphy ([email protected]) got netsplit.
[22:31] mxpxpod ([email protected]) got netsplit.
[22:31] harshy ([email protected]) got netsplit.
[22:31] Rotty ([email protected]) got netsplit.
[22:31] taaz-gone ([email protected]) got netsplit.
[22:31] danb ([email protected]) got netsplit.
[22:31] Rotund ([email protected]) got netsplit.
[22:31] jdahlin ([email protected]) got netsplit.
[22:31] Misirlou ([email protected]) got netsplit.
[22:31] wtay_ ([email protected]) got netsplit.
[22:31] ds ([email protected]) got netsplit.
[22:31] md` ([email protected]) got netsplit.
[22:31] steve_b_ ([email protected]) got netsplit.
[22:31] jimmy_dean ([email protected]) got netsplit.
[22:33] <mathrick> BB1: huh?
[22:33] <mathrick> BB1: noticed what?
[22:33] dolphy ([email protected]) returned to #gstreamer.
[22:33] mxpxpod ([email protected]) returned to #gstreamer.
[22:33] harshy ([email protected]) returned to #gstreamer.
[22:33] Rotty ([email protected]) returned to #gstreamer.
[22:33] md` ([email protected]) returned to #gstreamer.
[22:33] steve_b_ ([email protected]) returned to #gstreamer.
[22:33] Rotund ([email protected]) returned to #gstreamer.
[22:33] danb ([email protected]) returned to #gstreamer.
[22:33] taaz-gone ([email protected]) returned to #gstreamer.
[22:33] jimmy_dean ([email protected]) returned to #gstreamer.
[22:33] jdahlin ([email protected]) returned to #gstreamer.
[22:33] Misirlou ([email protected]) returned to #gstreamer.
[22:33] wtay_ ([email protected]) returned to #gstreamer.
[22:33] ds ([email protected]) returned to #gstreamer.
[22:34] ChrisHJW ([email protected]) left irc: Read error: 110 (Connection timed out)
[22:34] decadix ([email protected]) left irc: "leaving"
[22:35] md` ([email protected]) left irc: Connection timed out
[22:42] <LeRoutier> BB1: i'm on the last video
[22:43] <BB1> yay
[22:43] <BB1> I noticed that quite a few of your bugs are PCM related
[22:43] <BB1> I'll work on that
[22:45] <LeRoutier> most are either PCM related (crashes avidemux) or thread problem
[22:45] <mathrick> <e1f> faculty member got a document that had a virus that corrupts the system bios by flashing it with garbage
[22:45] <mathrick> yay for brave new world of trustworthy computing ;)
[22:47] kwm ([email protected]) left irc: "Leaving"
[22:49] <LeRoutier> BB1, and the thread problem seems to occur only when the video codec is not wired in gstreamer (iv32/iv50/mrle/msvc)
[22:50] adruab ([email protected]) joined #gstreamer.
[22:51] <adruab> Does anyone know the status of the C++ bindings for gstreamer?
[22:51] <adruab> or perhaps a web location that's valid?
[22:52] <adruab> the main link on the gstreamer.freedesktop.org, is down, so I'm not sure where to look
[22:53] <LeRoutier> it should be in gst-bind in fd.o CVS but i don't know if tarballs are available
[22:54] <LeRoutier> i think python bindings are the only one which are up to date
[22:54] harshy ([email protected]) left irc: "I quit for now"
[22:54] <adruab> ok
[22:54] <adruab> hmm
[22:55] apoc ([email protected]) joined #gstreamer.
[22:56] <mathrick> adruab: they're almost up to date
[22:56] apoc ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[22:56] <mathrick> adruab: but not commited yet, AFAIK
[22:58] <adruab> so the lastest run is still be worked on?
[22:58] <mathrick> kinda, yes
[22:58] Action: mathrick points @ mxpxpod
[22:58] <adruab> :P I see :)
[22:58] <mxpxpod> oh no
[22:58] <mxpxpod> what did I do?
[22:59] <adruab> I'm just asking about the C++ bindings for gstreamer
[22:59] <mxpxpod> oh
[22:59] <mxpxpod> they're a mess
[22:59] <mxpxpod> I should probably update them this weekend for 0.8
[22:59] <adruab> ah :)
[23:00] <mxpxpod> I'm not even the maintainer
[23:00] <adruab> a mess meaning behind, or a mess in general?
[23:00] <adruab> ah
[23:00] <mxpxpod> a mess meaning behind and in general
[23:00] <adruab> it's ok, I was just interested in using them because, the rest of the project I'm working in uses gtkmm and the like
[23:01] <mxpxpod> what project?
[23:01] <adruab> ok, so should I just break down and use the C version
[23:01] <adruab> it's a company project atm
[23:01] <mxpxpod> ah, cool
[23:01] <mxpxpod> adruab: I'd recommend using the C stuff
[23:01] <mxpxpod> the c++ stuff isn't stable enough
[23:02] <adruab> ok, thanks then, I'll check it out
[23:02] <adruab> and I'll probably be back here again :)
[23:02] <mxpxpod> heh
[23:03] <adruab> thanks for the help
[23:04] adruab ([email protected]) left irc: "Leaving"
[23:09] <BB1> LeRoutier: that might be bad...
[23:09] <BB1> LeRoutier: it means we don't have proper pad usability checks anymore
[23:09] <BB1> ohwell
[23:13] smoke ([email protected]) got netsplit.
[23:14] <LeRoutier> BB1: you have enought data to eat or should i continue to report bugs for mpegs ?
[23:14] <LeRoutier> or mov
[23:19] smoke ([email protected]) got lost in the net-split.
[23:27] BB1 ([email protected]) left irc: "Download Gaim: http://gaim.sourceforge.net/"
[23:33] md` ([email protected]) joined #gstreamer.
[23:43] walters ([email protected]) left irc: "out"
[23:50] markey ([email protected]) left irc: "leaving"
[00:00] --- Thu Apr  1 2004
[00:04] dolphy ([email protected]) left irc: "Network down, IP Packets delivered via UPS"
[00:05] Action: thomasvs laughs maniacally at wheels
[00:06] <wheels> thomasvs: I couldn't resist.  :-)
[00:07] <thomasvs> wheels: it had me for all of twenty seconds though
[00:07] spyder482 ([email protected]) joined #gstreamer.
[00:09] <Company> can we get a common format for mentioning people in changelogs?
[00:09] <Company> or is there already one?
[00:11] <Company> taaz-gone: it'd be nice if there were a gstreamer-plugins-dev package
[00:12] Action: Company hopes he'll read this
[00:25] Shoragan ([email protected]) joined #gstreamer.
[00:25] caro ([email protected]) left #gstreamer.
[00:27] <ds-work> Company: if someone sends a patch, I usually create a changelong entry specifically for them
[00:27] <Company> ds-work: it'd be nice if it were script-extractable for release-notes
[00:27] <ds-work> Company: if they just suggested the change, I usually mention them, but not in any way that can be extracted
[00:27] <ds-work> patches of more than a few lines I consider contributions
[00:28] <ds-work> a line or two, I don't bother
[00:28] <Company> i'd like to have it like owen does it with glib or gtk
[00:28] <ds-work> which is?
[00:29] <Company> ds-work: http://lists.gnome.org/archives/gtk-devel-list/2004-March/msg00195.html
[00:29] <Company> (see bottom)
[00:30] <Company> it's heavily motivating to see your name there if your not an open source regular
[00:31] <mathrick> night
[00:31] <wheels> Company: Indeed.  I've actually had a couple of folks very excited that they made it into the JuK about box...
[00:31] Nick change: mathrick -> mathrick|sleep
[00:34] Action: Company wonders if he already made it into the gnome about box
[00:36] harshy ([email protected]) joined #gstreamer.
[00:36] <wheels> Company: You're in the RB one.  :-)
[00:36] <Company> wheels: close, but rb is not gnome ;)
[00:36] Kabuto ([email protected]) joined #gstreamer.
[00:36] <wheels> Company: is there a general "About GNOME" with names?
[00:36] <Company> wheels: library writers have a hard time getting mentioned prominently in about boxes though...
[00:37] <Kabuto> Anyone using gstreamer .8 and .ogg?
[00:37] <teuf> wheels: yep
[00:37] <Company> wheels: gnome-about
[00:37] <Company> wheels: it's its own app with included easter-eggs
[00:37] <Company> Kabuto: me
[00:37] Action: wheels fires it up...
[00:38] <wheels> Company: Well, in KDE it's always kind of been the Konq about box, but I've always been too lazy to add myself...
[00:38] <thaytan> Kabuto: yep, all day long
[00:38] <Kabuto> Company: using rythmbox .6.10 and gstreamer .8.  mp3 play fine but ogg are mono.  Also can't get gst-launch-ext-0.8 to work right for testing.
[00:38] <thaytan> the mono issue is fixed in CVS
[00:38] <Company> yeah
[00:38] <thaytan> what's the problem with gst-launch-ext "?
[00:39] <Company> gst-launch-ext needs to be fixed
[00:39] <Company> we changed the way oggs are handled
[00:39] <Company> and i bet ity wasn't updated
[00:39] <Kabuto> thaytan: Whens the next release and/or what needs to be patched?
[00:40] <Company> audioconvert needs to be patched
[00:40] <Company> no idea about next release but i want one ;)
[00:41] <Kabuto> Company: kinda new to gstreamer.  What dir is audioconvert in?
[00:42] <Company> Kabuto: gst-plugins/gst/audioconvert
[00:42] mort_ ([email protected]) joined #gstreamer.
[00:42] <Kabuto> Company: The #137556 fix?
[00:43] <Kabuto> Company: for fixate?
[00:43] teuf ([email protected]) left irc: "Client exiting"
[00:43] <Company> Kabuto: yeah
[00:43] <Kabuto> Company: thx a bunch.  Will try it.
[00:43] <Company> np
[00:48] <LeRoutier> who's sound-juicer maintainer ?
[00:48] <Company> ross
[00:48] <Company> burton
[00:49] sublett ([email protected]) joined #gstreamer.
[00:49] <LeRoutier> k, i'll try to see on his site if a version supporting gstreame 0.8 natively is soon out
[00:50] <Company> you could just modify the configure script manually
[00:50] <LeRoutier> well, i already did it
[00:52] <LeRoutier> but i always prefer native fixes to patches everywhere in my local portage tree
[00:52] <LeRoutier> (say i'm lazzy, it's true)
[00:56] <Company> then just poke ross
[01:00] <ds-work> "sunfreeware does not have the ars
[01:00] <ds-work> technica rss feed on the wrong side of the large intestine, terminating at
[01:00] <ds-work> the moment."
[01:03] <wheels> thomasvs: woohoo -- a bite  :-)
[01:04] abo_ ([email protected]) left irc: Read error: 113 (No route to host)
[01:04] <Company> wheels: what are you 2 talking about?
[01:04] <wheels> Company: see the list
[01:07] trow ([email protected]) left irc: "Leaving"
[01:07] iain ([email protected]) left irc: "leaving"
[01:08] md` ([email protected]) left irc: Read error: 60 (Operation timed out)
[01:13] Kabuto ([email protected]) left irc: "I like core dumps"
[01:16] thaytan ([email protected]) left irc: Read error: 113 (No route to host)
[01:16] <jdahlin> wheels: I'm soo slow
[01:16] Action: jdahlin realized it first now
[01:19] <wheels> /me will have to add this to his sigs though: "Umm, well, I'm no core developer, but off the top of my head, I'm thinking
[01:19] <wheels> what you're proposing is lunacy.
[01:20] Action: Company officially switched his desktop to KDE
[01:21] <Company> s/KDE?Debian/
[01:21] <Company> s/KDE/Debian/
[01:21] <Company> now i can install KDE there
[01:21] <Company> everything that doesn't exist in Debian doesn't exist for me from now on... ;/
[01:22] <ds-work> what about gstreamer-0.8?
[01:22] Shoragan ([email protected]) left irc: "Leaving"
[01:32] ChrisHJW ([email protected]) joined #gstreamer.
[01:35] <jdahlin> Debian Desktop Environment?
[01:36] <LeRoutier> ah, good, gnome 2.6 is now quite complete in gentoo
[01:37] alley_cat ([email protected]) left irc: "May the Source be with you!"
[01:38] Rotty ([email protected]) left irc: Remote closed the connection
[01:38] alley_cat ([email protected]) joined #gstreamer.
[01:39] <Company> ds-work: i'm busy enough getting all the n-dev packages right now - i want gst-plugins-dev
[01:41] <ds-work> I had a script that would install everything gstreamer wanted, but I don't remember where it is
[01:44] sublett ([email protected]) left irc: "Why are all the really good songs short?"
[01:45] hyriand ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[01:46] hyriand ([email protected]) joined #gstreamer.
[01:47] foser ([email protected]) left irc: "[ I want to believe ]"
[01:48] BBB ([email protected]) joined #gstreamer.
[01:48] hyriand ([email protected]) left #gstreamer.
[01:52] mort_ ([email protected]) left irc: Remote closed the connection
[02:00] BBB ([email protected]) left irc: "Client exiting"
[02:13] thaytoo ([email protected]) joined #gstreamer.
[02:23] ChrisHJW ([email protected]) left irc: Remote closed the connection
[02:24] pb_ (~pb@2002:5160:4550:1:240:95ff:fe30:aa84) left irc: "Client exiting"
[02:24] ChrisHJW ([email protected]) joined #gstreamer.
[02:28] yippi ([email protected]) left irc: "Client exiting"
[02:29] tim__ ([email protected]) joined #gstreamer.
[02:41] <LeRoutier> xvid 1.0 final is out
[02:44] dholmes ([email protected]) joined #gstreamer.
[02:45] bitshifter ([email protected]) left irc: Read error: 110 (Connection timed out)
[02:47] <LeRoutier> good night ppl
[02:47] LeRoutier ([email protected]) left irc: "Leaving"


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