IRC Logs

*******************************************************************
[03:09] <BBB> dammit
[03:09] <BBB> I have tens of uncommitted fixes and enhancements
[03:09] <BBB> what to do?
[03:11] Action: BBB images thomasvs killing him because he broke the release
[03:11] Action: BBB images all those fancy users that will scream and shout because their favourite media backend doesn't support MS RLE
[03:11] <Company> huh? mplayer doesn't support RLE?
[03:11] <Company> ;)
[03:12] <BBB> xine does :p
[03:14] <BBB> Program received signal SIGTRAP, Trace/breakpoint trap.
[03:14] <BBB> 0x40000c20 in _start () from /lib/ld-linux.so.2
[03:14] <BBB> (gdb) bt
[03:14] <BBB> nice backtrace
[03:14] Action: BBB shoots himself
[03:15] <BBB> _gst_cpu_initialize_i386 (flags=0xfff58ed7, featurelist=0x8055ed8)
[03:15] <BBB>     at gstcpu.c:97
[03:15] <BBB> 97          _gst_cpu_flags |= GST_CPU_FLAG_MMX;
[03:15] Action: BBB looks @ ds
[03:18] Action: BBB also looks @ ds-work
[03:18] <BBB> [ping!]
[03:25] iain ([email protected]) left irc: "leaving"
[03:25] sub_pop ([email protected]) left irc: "Client exiting"
[03:27] sub_pop ([email protected]) joined #gstreamer.
[03:30] robUx4 ([email protected]) left irc: Read error: 110 (Connection timed out)
[03:31] <BBB> [rbultje@shrek gstreamer]$ tools/gst-register-0.8
[03:31] <BBB> Segmentation fault
[03:31] <BBB> [rbultje@shrek gstreamer]$
[03:31] Action: BBB freaks out
[03:31] <BBB> ds, ds-work! HELP!
[03:37] md` ([email protected]) joined #gstreamer.
[03:37] Action: BBB cries
[03:40] <Company> why the HECK is this now reordering functions?
[03:40] <Company> i thought -O2 doesn't do that
[03:56] md` ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[04:02] md` ([email protected]) joined #gstreamer.
[04:06] Action: BBB has several fixes that he wants to test and CVS doesn't work now :/
[04:19] <ds-work> moo
[04:22] <BBB> ah
[04:22] <BBB> there's the evilry
[04:22] <BBB> :p
[04:22] <ds-work> crap.  I don't want to know about breakage to cpu detection
[04:23] <BBB> you broke it! it's all your fault! and your future girlfriends', too
[04:23] <BBB> thomasvs is not gonna be happy ;)
[04:23] <ds-work> what cpu?
[04:24] <ds-work> er, actually, it shouldn't matter
[04:25] <BBB> p3
[04:29] walters ([email protected]) left irc: Read error: 60 (Operation timed out)
[04:40] <Company> i like the humor of the guy in #140046
[04:41] <ds-work> new major stable release, and stuff starts working?  imagine
[04:43] <Company> well, i have a lot of stuff that stopped working with Linux 2.6 ;)
[04:43] <ds-work> yes, but Linux sux
[04:43] <ds-work> BBB: could you email me a copy of libgstreamer_0.8_la-gstcpu.o?
[04:44] <BBB> I just recompiled the old version
[04:44] <BBB> hm
[04:44] Action: BBB recompiles new...
[04:45] Nick change: mx|gone -> mxpxpod
[04:45] <ds-work> actually, .libs/libgstreamer_0.8_la-gstcpu.o
[04:46] <BBB> yeah, I know
[04:46] <BBB> coming...
[04:48] Action: BBB crashes evo more and more since he bought connector
[04:48] <BBB> bah :/;
[04:49] <ds-work> model name      : Pentium 75 - 200
[04:49] <ds-work> cpu MHz         : 149.080
[04:49] <ds-work> poor machine, maybe I'll log off
[04:49] Action: ds-work searches for a PIII
[04:54] <ds-work> BBB: ok, I figured it out
[04:54] Action: BBB notes how ADPCM in .wav is badly screwed
[04:55] <ds-work> BBB: what version of gcc are you using?
[04:55] <BBB> 3.2, provided with RH90
[04:55] <BBB> 3.2.2
[04:55] <ds-work> gcc compiled the code wrong
[04:55] <ds-work> no problem, I asked it to do something tricky
[04:56] <BBB> hm, ADPCM expects data in blocks of 2048 bytes
[04:56] <BBB> but wavparse spits out in 1024 byte blocks
[04:56] <BBB> should I workaround in wavparse or in ffdec?
[04:56] Action: BBB guesses in ffdec
[04:56] <ds-work> probably
[04:58] <Company> file a bug
[04:58] Action: Company hates bugs in compilers, especially because he doesn't speak asm
[04:58] <ds-work> Company: it may just be a feature that I've forgotten was added to 3.3
[04:59] <BBB> the problem is that if I do that... then I might seek to in between two chunks that apparently belong together
[04:59] <Company> that would be fine with me, too
[05:00] <BBB> oh, nm
[05:00] <BBB> wavparse bug
[05:00] <BBB> how surprising ;)
[05:01] Action: BBB wonders why gst-player doesn't pick up the mimetype
[05:04] <BBB> ERROR        GST_PADS(  487) gstpad.c(2403):gst_pad_set_explicit_caps:<wavparse0> (null)
[05:04] <BBB> ERROR        GST_PADS(  487) gstpad.c(2403):gst_pad_set_explicit_caps:<wavparse0> failed to negotiate (try_set_caps with "audio/x-adpcm, layout=(string)microsoft, block_align=(int)2048, rate=(int)22050, channels=(int)2" returned REFUSED)
[05:04] <BBB> what do I conclude from that?
[05:05] <BBB> does spider connect wavparse directly to osssink?
[05:06] <ds-work> impossible to tell from there
[05:06] <ds-work> but likely
[05:06] <BBB> so it's a spider bug?
[05:06] <BBB> hmk...
[05:07] Action: BBB piles up bugfixes until thomasvs awakes
[05:09] <Company> yes, it's a spider bug
[05:09] <Company> spider checks template caps
[05:11] <BBB> bleh
[05:11] Action: BBB leaves it at the current fix then
[05:13] <ds-work> nice, gcc-2.95 generates the wrong code, too
[05:20] <ds-work> BBB: gstcpu.c fix checked in
[05:23] Action: BBB tries
[05:24] <ds-work> ok, I need an opinion.  Should I put the 258-line gstinfo-inline.h into gstinfo.h, or leave it separate?
[05:25] <BBB> k, seems to work
[05:25] <BBB> what's in it?
[05:25] <ds-work> all the variadic macros converted to inline functions
[05:27] <ds-work> I'm going to combine it.  it's silly to split it
[05:28] jcsston` ([email protected]) joined #gstreamer.
[05:28] jcsston ([email protected]) left irc: Nick collision from services.
[05:29] <Company> gstinfo is a mess already, just put it in there
[05:29] <Company> "mess" as in noone's gonna look at it anyway
[05:30] <Company> and it's nicer to not look at one mess instead of needing to not look at 2 messes
[05:31] <BBB> you have a weird reasoning
[05:32] <Company> gnight
[05:32] Company ([email protected]) left irc: Remote closed the connection
[05:49] Nick change: jcsston` -> jcsston
[05:51] <Phython> BBB: sometimes that is called "Justifiable Ugliness"
[05:52] <ds-work> if it were only a bit more complicated, it would be smart to autogenerate it
[05:52] <ds-work> but it seems correct right now, so nobody wants to touch it
[05:53] <ds-work> oh yeah, one of the advantages of macros is that you can use __LINE__ and __FILE__
[06:00] <BBB> that you can concatenate strings
[06:00] <BBB> (via the compiler)
[06:02] <BBB> oh, btw, guys, please don't add bugs to the tracker if they don't have a sample file attached, or at least ask to provide one... debugging those bugs is extremely hard if not simply impossible
[06:05] steve_b_ ([email protected]) joined #gstreamer.
[06:05] steve_b ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[06:06] <BBB> bbl, bedtime
[06:06] BBB ([email protected]) left irc: "Client exiting"
[06:11] steve_b_ ([email protected]) got netsplit.
[06:11] steve_b_ ([email protected]) returned to #gstreamer.
[06:15] steve_b_ ([email protected]) got netsplit.
[06:16] steve_b_ ([email protected]) returned to #gstreamer.
[06:41] sub_pop ([email protected]) left irc: "Client exiting"
[06:44] sub_pop ([email protected]) joined #gstreamer.
[06:48] wheels ([email protected]) left irc: Read error: 113 (No route to host)
[07:01] spider-man ([email protected]) joined #gstreamer.
[07:20] <taaz> where's walters
[07:26] ChrisHJW ([email protected]) joined #gstreamer.
[07:29] sxpert_work ([email protected]) left irc: Read error: 113 (No route to host)
[07:31] sxpert_work ([email protected]) joined #gstreamer.
[07:56] spider-man ([email protected]) left irc: "Client exiting"
[08:13] grub_booter ([email protected]) joined #gstreamer.
[08:17] <grub_booter> slowly getting my head round gstreamer and the demuxer i need - i hope this is a sensible question - do i need to specify a filesrc to load an xml document? we use libxml2 to parse our xml, and this can be a file, url or whatever, so the gstreamer filesrc seems overkill to me...
[08:22] <grub_booter> i guess i just need the demuxer as the first element and this will take an argument of the xml doc and determine from that which pads are applicable... is that valid? are there any similar plugins?
[08:29] ChrisHJW ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[08:42] <ds> grub_booter: what are you trying to do?
[08:42] <ds> what application?
[08:42] #gstreamer: mode change '-o ds' by [email protected]
[08:45] <lilo_booter> ds: i'm trying to provide a mechanism for loading mlt westley projects in gstreamer - basically, westley is just an xml doc which desribes playlists/multiple track arrangements and fx which are all dealt with by the mlt handling - so, to you i would proved an image and audio pads
[08:45] <lilo_booter> proved = provide
[08:45] <ds> lilo_booter: ok, that's what I though
[08:45] <ds> t
[08:46] <ds> no, I don't think it's reasonable to load such files via filesrc
[08:46] <lilo_booter> good :-)
[08:47] <ds> ideally, we'd have a file selection and/or loading interface
[08:47] jcsston` ([email protected]) joined #gstreamer.
[08:47] <ds> so that an app could choose between making the element load the file itself
[08:47] <ds> or the app giving it a bunch of bits that it loaded from the file
[08:48] <ds> or even stream it in from a pad
[08:49] <ds> But for now, I'd suggest just loading a file in from a "filename" property
[08:49] <ds> er, "location"
[08:49] <lilo_booter> ok - and how would that look on the gst-launch command line?
[08:51] <lilo_booter> and how would it interface with gst-player and so on? would changes be required or is there a way to indicate this behaviour in a generic way?
[08:51] LeRoutier ([email protected]) joined #gstreamer.
[08:51] <ds> gst-launch yoursrc location=it.xml ! stuff ! xvimagesink
[08:52] <LeRoutier> good morning
[08:52] <lilo_booter> ok
[08:52] <ds> LeRoutier: yo
[08:52] <ds> gst-player would probably need special code, but that's true for any source
[08:53] <ds> gst-player pretty much only knows gnomevfssrc
[08:53] <lilo_booter> hmm - k
[08:53] <ds> but it can handle fdsrc as well
[08:54] <ds> so the code is there, it just needs another switch
[08:54] <ds> hmmm, but that would be a good reason to support pulling the xml file from a pad
[08:54] <lilo_booter> ok
[08:54] <ds> anyway, no reason to rush -- do the hard stuff, and we can work out the details later
[08:55] <LeRoutier> hum, someone is going to love it, one for AVI extension for avidemux to handle (seems it can't handle OpenDML avis)
[08:55] jcsston` ([email protected]) left #gstreamer.
[08:55] <LeRoutier> one for => one more
[08:56] <lilo_booter> well, i don't think this should be that hard... anyway - np - i'll use your filesrc, read contents into a buffer and add a sepcial case constructor to my xml parser
[08:56] <LeRoutier> yesterday night, i generated some avi from dvgrab and  now, gstreamer can't play it because of unknow stream IDs (ix00 & ix01)
[08:57] <lilo_booter> hmm
[08:57] <lilo_booter> LeRoutier: those are opendml extensions indexes
[08:57] <lilo_booter> i presume you used avi type 2?
[08:57] <LeRoutier> yep
[08:58] <LeRoutier> DV type 2 + OpenDML
[08:58] <LeRoutier> so, gstreamer/avidemux doesn't support OpenDML avis for now
[08:58] <LeRoutier> BBB added rec-list support a few days ago. is that a related extension ?
[08:59] <lilo_booter> not sure
[08:59] <lilo_booter> avi is a pita and is especially useless when it comes to DV ;-)
[09:00] <lilo_booter> really, unless you're using windows editors, stick to raw dv...
[09:00] <lilo_booter> you can unpack those avis using kino and change your preferences to raw dv 
[09:00] <LeRoutier> avi is old, so there were tons of "extensions" for it. what is not handled in the standard becomes an ugly hack
[09:00] <lilo_booter> avi is old and broken to begin with ;-)
[09:01] <LeRoutier> lilo_booter, well, if lots of such avi exists, better add support for it in avidemux
[09:02] <lilo_booter> well, sure, but if you want things to work now and without restriction, raw dv is the way forward
[09:02] <lilo_booter> (well, other than the fact that there aren't any windows apps which tend to understand raw dv that i know of...)
[09:03] <LeRoutier> my problem is that gstreamer breaks at trying to read a sample PAL dv i did yesterday, because of bad sound caps nego
[09:03] <LeRoutier> so, i tried avi
[09:04] <lilo_booter> dv frames are fixed size (144000 for PAL, 120000 for NTSC) - headers and indexes are simply not needed - you just need to know which normalisation the file contains (can be identified by looking at the first few bytes of any frame) - audio information can be similarly extracted
[09:05] <LeRoutier> seems mplayer fails to have sound with it : Cannot find codec for audio format 0x56444152.
[09:06] <lilo_booter> hmm
[09:06] <LeRoutier> direct dv capture from device to file, using cat
[09:07] <lilo_booter> Opening audio decoder: [libdv] Raw DV Audio Decoder
[09:07] <lilo_booter> AUDIO: 48000 Hz, 2 ch, 16 bit (0x10), ratio: 3600000->192000 (28800.0 kbit)
[09:07] <lilo_booter> Selected audio codec: [libdv] afm:libdv (raw DV audio decoder (libdv))
[09:07] <lilo_booter> seems to work for me
[09:07] <lilo_booter> ah - wait... lemme guess...
[09:07] <lilo_booter> your cam is supplying 32khz 4 channel?
[09:08] <ds> XDAT?
[09:08] <LeRoutier> i really don't know which sound format it provides
[09:08] <lilo_booter> ds: ?
[09:08] <ds> 0x56444152 == XDAT
[09:08] <lilo_booter> ah
[09:09] <lilo_booter> LeRoutier: load it in kino, and click on properties in the bottom right
[09:09] <LeRoutier> ok
[09:09] <LeRoutier> 12bits / 32kHz / 1280 samples
[09:10] <lilo_booter> hmm - annoying defaults of all cams... 
[09:10] <lilo_booter> probably why mplayer's having a hard time with it
[09:13] <lilo_booter> hmm - nope - mplayer's handling it correctly on my system (raw dv with a 32khz soundtrack - handled by libdv too...)
[09:13] <lilo_booter> curiouser and curiouser - wonder if there's something funny in the way your cam is encoding the audio...
[09:15] <lilo_booter> my suggestion would be that you change the setting of your cam to 48khz anyway :-) - it won't 'fix' your current recordings, but unless you're using the dubbing features of the cam, you won't lose anything (you will gain better audio quiality though)
[09:16] <lilo_booter> do you have smilutils installed by any chance?
[09:20] markey ([email protected]) joined #gstreamer.
[09:21] LeRoutier ([email protected]) left irc: Read error: 60 (Operation timed out)
[09:23] jcsston ([email protected]) left irc: ""Matroska Not Included.""
[09:24] steve_b__ ([email protected]) joined #gstreamer.
[09:26] steve_b ([email protected]) joined #gstreamer.
[09:29] LeRoutier ([email protected]) joined #gstreamer.
[09:31] tlg ([email protected]) joined #gstreamer.
[09:35] nick_m ([email protected]) joined #gstreamer.
[09:38] steve_b_ ([email protected]) left irc: Read error: 110 (Connection timed out)
[09:44] <LeRoutier>  
[09:44] <LeRoutier> ds: can gstreamer renegotiates caps in the middle of a video ?
[09:44] <ds> LeRoutier: it can, but it's probably not recommended
[09:45] <lilo_booter> LeRoutier: did you confirm if the dv audio info was fluctuating?
[09:45] <ds> LeRoutier: it's how videoscale ! xvimagesink works during resizing
[09:45] steve_b__ ([email protected]) left irc: Read error: 110 (Connection timed out)
[09:46] <lilo_booter> issue here is related to audio
[09:46] <LeRoutier> lilo_booter, i confirm that audio format changes inside this .dv
[09:46] <lilo_booter> but i think it's not a specific issue with your audio playback... seems like libdv is interpreting some dv frames in his capture as 0khz...
[09:47] <lilo_booter> ok
[09:47] <lilo_booter> well, 0khz is gonna screw up most apps
[09:48] <LeRoutier> not kino but it's quite normal as it was made with dv in mind
[09:49] steve_b ([email protected]) left irc: Read error: 110 (Connection timed out)
[09:49] <lilo_booter> kino (and mlt) force the issue - if there's an error in the frame, generate silent audio samples of the requested size
[09:49] bilboed ([email protected]) left irc: Read error: 110 (Connection timed out)
[09:50] <lilo_booter> (the number of samples also fluctuates per frame :-))
[09:50] <lilo_booter> (but that's not an error - very much designed to do that...)
[09:51] bilboed ([email protected]) joined #gstreamer.
[09:51] <lilo_booter> my gut reaction would be to recommend that gstreamer dvdec does the same i guess... 
[09:51] <LeRoutier> i'll try to avoid raw DV for now, and OpenDML avis too
[09:52] <lilo_booter> well, dvgrab and kino is open dml only
[09:52] <lilo_booter> and raw dv should not be avoided :-)
[09:52] <LeRoutier>   --opendml         use the OpenDML extensions to write large (>1GB)
[09:52] <LeRoutier>                       'Type 2' DV AVI files (default off).
[09:53] <lilo_booter> oh - sorry - yeah - same option will be in kino then
[09:53] <LeRoutier> ok, so now trying to generate an DV type 2 avi without OpenDML
[09:53] <lilo_booter> i don't use avi for this (nasty, nasty thing :-))
[09:53] <LeRoutier> well, i have no choice if i want to play them
[09:54] <lilo_booter> should be simple to fix in the code :-)
[09:54] Action: LeRoutier needs a new/big harddisk because DV takes so much space
[09:54] <LeRoutier> lilo_booter, my C is crap, better not see my code
[09:54] <lilo_booter> heh - ok :-)
[09:55] <lilo_booter> with ffmpeg, you might be able to do on the fly encoding to divx/mjpeg/mpeg or mpeg2
[09:55] <lilo_booter> depending on your hardware and of course, whether it likes a dirty audio stream...
[09:55] <LeRoutier> lilo_booter, or i'll bug BBB some more time to have dv1394src fixed in gst-recorder 
[09:56] <lilo_booter> the other way is to repair the dv using smilutils...
[09:56] <lilo_booter> you can extract a wav and then dub it back - lengthy process and doubles your disk space :-)
[09:59] sub_pop ([email protected]) left irc: "Client exiting"
[10:08] ChriHJW_log ([email protected]) joined #gstreamer.
[10:13] ChrisHJW_log ([email protected]) left irc: Read error: 110 (Connection timed out)
[10:16] nick_m ([email protected]) left irc: "Client exiting"
[10:19] dolphy ([email protected]) joined #gstreamer.
[10:21] tlg ([email protected]) left #gstreamer ("Client exiting").
[10:21] tlg ([email protected]) joined #gstreamer.
[10:25] bitshifter ([email protected]) joined #gstreamer.
[10:38] thomasvs ([email protected]) left irc: Read error: 110 (Connection timed out)
[10:38] ploum ([email protected]) joined #gstreamer.
[10:40] aflinta ([email protected]) joined #gstreamer.
[10:41] <aflinta> hi :)
[10:41] <aflinta> if it possible to use with gst-ffmpeg system ffmpeg instead of builtin?
[10:44] <LeRoutier> nope
[10:45] <LeRoutier> aflinta, wonder why so many apps use their own ffmpeg version and links statically to it instead of linking to ffmpeg lib dynamically
[10:45] <aflinta> LeRoutier: is ffmepg included in gst-ffmpeg specially hacked/patched or it is regular?
[10:46] <LeRoutier> patched, but mostly to make it autotools aware
[10:46] <LeRoutier> i don't know what other patches it applies to it
[10:46] <aflinta> chmmm, I'll try to write small patch to use system ffmpeg, it is horrible to link statically ffmepg for every app
[10:46] <LeRoutier> better ask BBB but he's not there at this tim
[10:47] <aflinta> LeRoutier: ok, thx :)
[10:48] Nick change: markey -> markey|afk
[10:49] thomasvs ([email protected]) joined #gstreamer.
[10:53] jdahlin ([email protected]) joined #gstreamer.
[11:13] <jdahlin> woho http://bugzilla.gnome.org/show_bug.cgi?id=140058 was fixed
[11:23] steve_b ([email protected]) joined #gstreamer.
[11:35] Rotty ([email protected]) joined #gstreamer.
[11:37] bilboed_school ([email protected]) joined #gstreamer.
[11:43] markey|afk ([email protected]) left irc: "leaving"
[11:44] ensonic ([email protected]) joined #gstreamer.
[11:50] sublett ([email protected]) joined #gstreamer.
[12:00] <LeRoutier> thomasvs, you can access to my files ?
[12:00] iain ([email protected]) joined #gstreamer.
[12:00] <LeRoutier> hi iain
[12:00] <iain> hi
[12:03] steve_b_ ([email protected]) joined #gstreamer.
[12:09] bitshifter ([email protected]) left irc: "Client exiting"
[12:10] bitshifter ([email protected]) joined #gstreamer.
[12:10] steve_b__ ([email protected]) joined #gstreamer.
[12:21] steve_b ([email protected]) left irc: Read error: 110 (Connection timed out)
[12:28] smoke ([email protected]) joined #gstreamer.
[12:29] steve_b_ ([email protected]) left irc: Read error: 110 (Connection timed out)
[12:36] Company ([email protected]) joined #gstreamer.
[12:36] steve_b__ ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[12:41] nick_m ([email protected]) joined #gstreamer.
[12:42] Action: sxpert_work is away: alasoup
[12:49] lilo_booter ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[12:53] kmaraas ([email protected]) joined #gstreamer.
[13:06] lilo_booter ([email protected]) joined #gstreamer.
[13:06] <LeRoutier> bbl
[13:06] LeRoutier ([email protected]) left irc: "Leaving"
[13:07] nick_m ([email protected]) left irc: "Client exiting"
[13:30] Company ([email protected]) left irc: Remote closed the connection
[13:33] alley_cat ([email protected]) joined #gstreamer.
[13:45] <jdahlin> ffmpeg can't decode qdmc/qdmc2 ?
[13:46] Action: sxpert_work is back (gone 01:03:03)
[13:51] jonathang ([email protected]) joined #gstreamer.
[13:53] <jdahlin> dolphy: is http://bugzilla.gnome.org/show_bug.cgi?id=140046 a totem bug?
[13:53] <jdahlin> I've seen it here too
[13:54] markey ([email protected]) joined #gstreamer.
[13:59] <dolphy> jdahlin: bugzilla is away
[14:00] <jdahlin> dolphy: it's back again
[14:00] <jdahlin> very slow though
[14:02] ensonic ([email protected]) left irc: Read error: 110 (Connection timed out)
[14:06] ensonic ([email protected]) joined #gstreamer.
[14:09] Company ([email protected]) joined #gstreamer.
[14:18] <jdahlin> cd
[14:18] <jdahlin> d'oh
[14:22] <dolphy> bash: cd: d'oh: No file or directory of that kind
[14:22] <Company> wrong
[14:22] <Company> bash says "> "
[14:22] <thomasvs> I think we need to stick some food into jdahlin
[14:23] <Company> at least on this sun box :o
[14:24] <jdahlin> thomasvs: sure, paella? :)
[14:24] Action: jdahlin tries to clean up bugzilla
[14:51] markey ([email protected]) left irc: "leaving"
[14:58] BBB ([email protected]) joined #gstreamer.
[15:00] <BBB> thomasvs: ?
[15:11] <thomasvs> BBB: yeah ?
[15:12] <jdahlin> hello BBB
[15:12] <jdahlin> I got some work for you :)
[15:13] sublett ([email protected]) left irc: "I quit, therefore I was"
[15:15] <BBB> thomasvs: about those patches, do you want a RLE sample file?
[15:16] <BBB> jdahlin: ho! I'm only here to quickly poke thomasvs, I need to go to work after that
[15:16] <jdahlin> BBB, I've started to go through the mplayer samples directory
[15:16] <jdahlin> 5 new media files that I couldn't play
[15:16] <BBB> I noticed, good job :)
[15:16] <jdahlin> we need to write a policy for submitting "this file doesn't work"
[15:16] <jdahlin> bugs
[15:17] <thomasvs> BBB: I just got one from LeRoutier
[15:17] <BBB> oh, ok
[15:17] <BBB> it works for me, but gst-player fails horribly in keeping steady playback
[15:17] <BBB> it has a constant 100% CPU use, even at a 1fps movie
[15:17] Company ([email protected]) left irc: Remote closed the connection
[15:18] <jdahlin> maybe gstplay needs to be rewritten/rethinken
[15:18] <jdahlin> "rethinken", nice word :)
[15:18] <BBB> rethought
[15:19] <thomasvs> BBB: have you read the guide to updating ffmpeg snapshots yet ?
[15:19] <thomasvs> I was waiting for you to say "ew, so much work" :)
[15:20] #gstreamer: mode change '+o thomasvs' by ChanServ!ChanServ@services.
[15:20] Topic changed on #gstreamer by [email protected]: 0.8.1 of core and plugins is OUT --> http://gstreamer.freedesktop.org/ | if you're trying to debug gst-launch and fail, try gdb gst-launch-0.8 | Media playback tracker bug: #138435
[15:20] #gstreamer: mode change '-o thomasvs' by [email protected]
[15:21] <BBB> thomasvs: tonight, I need to go to work ;)
[15:21] <BBB> ohwell
[15:21] <BBB> I'll commit my fixes tonight, I'm too late :p
[15:22] <BBB> jdahlin: 139819 can be closed as invalid yes
[15:23] <BBB> jdahlin: and re:140152: we *have* an avi muxer ;)
[15:23] <BBB> avimux :p
[15:24] <jdahlin> BBB, do we? :)
[15:24] <jdahlin> doh :)
[15:25] Action: BBB kicks jdahlin :p
[15:25] <BBB> and we have an asf muxer as well
[15:25] <BBB> asfmux!
[15:25] <jdahlin> closed
[15:25] <BBB> and even an mpeg muxer
[15:25] <BBB> mplex!
[15:26] Action: BBB goes to work
[15:26] <BBB> later!
[15:26] BBB ([email protected]) left irc: "Client exiting"
[15:30] <ensonic> respective to the gst-manual, what do you think about bug #133057, would it be okay to commit that (better that an empty page)
[15:32] foser ([email protected]) joined #gstreamer.
[15:44] rener ([email protected]) joined #gstreamer.
[15:48] _246tNt_pc_ ([email protected]) joined #gstreamer.
[15:50] <rener> hi. I'm trying to extract frames from a video file, for display in a gtk+ app. How can I get the decoded image data? Is there a "obj.get_image()" like function in one of the sinks?
[15:53] <thomasvs> rener: connect to the handoff signal on fakesink ?
[15:53] <rener> thomasvs, I'll see if I can find that. I'm new to gstreamer, so I'm confused
[15:54] <thomasvs> rener: well, that's the basics.  you create a pipeline, send your video pad to fakesink, and connect to the handoff signal, which will give you data buffers of video
[15:54] <thomasvs> rener: but, it depends what you want to do
[15:54] <thomasvs> rener: if you want to show video from inside a gtk app, you should really use the play library
[15:55] <rener> thomasvs, pretty much something like a the "appple quicktime" program
[15:55] <thomasvs> rener: you mean the player ?
[15:55] <rener> thomasvs, yes
[15:56] <thomasvs> rener: why would you want to do that if there are already players ?
[15:56] <rener> thomasvs, I want to be able to play, to play reverse and to pause
[15:56] <thomasvs> anyway, in that case you shoulnd't do what I just said.  You should use the play library, or osmething equivalent, and let gstreamer handle the output
[15:57] <rener> thomasvs, I need this to work on both windows and linux
[15:57] <thomasvs> rener: yes, and ?
[15:57] <rener> thomasvs, I have a server sending the images to the clients using a socket. 
[15:57] <rener> does gstreamer work on windows?
[15:58] markey ([email protected]) joined #gstreamer.
[15:58] <thomasvs> rener: the initial porting has started, and it's already capable of doing stuff like mp3 encoding and so on
[15:58] <thomasvs> rener: for video playback, someone will have to write a windows-specific display sink though
[15:58] <rener> hmmmm. that would be nice
[15:59] aflinta ([email protected]) left irc: "domek"
[16:00] kwm ([email protected]) joined #gstreamer.
[16:00] iain ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[16:00] <rener> thomasvs, I guess I'll stick to the socket solution until that happens
[16:02] kmaraas ([email protected]) left irc: Client Quit
[16:06] <jdahlin> what is the ffmpeg mpeg4 decoder plugin called?
[16:06] <rener> where is fakesink documented? I can't find it anywhere.
[16:06] <thomasvs> ffdec_mpeg4 ?
[16:07] <thomasvs> rener: gst-inspect fakesink
[16:07] <jdahlin> thanks
[16:07] <rener> thanks
[16:15] iain ([email protected]) joined #gstreamer.
[16:17] markey ([email protected]) left irc: "bbl"
[16:17] <iain> whats supposed to be used instead of try_set_caps?
[16:35] Company ([email protected]) joined #gstreamer.
[16:39] jonathang ([email protected]) left irc: "Leaving"
[16:46] The_Company ([email protected]) joined #gstreamer.
[16:47] The_Company ([email protected]) left irc: Client Quit
[16:48] The_Company ([email protected]) joined #gstreamer.
[16:48] bilboed_school ([email protected]) left irc: "Leaving"
[16:50] bitshifter ([email protected]) left irc: "Client exiting"
[16:54] Company ([email protected]) left irc: Read error: 60 (Operation timed out)
[17:07] trow ([email protected]) joined #gstreamer.
[17:09] thomasvs ([email protected]) left irc: Remote closed the connection
[17:12] thomasvs ([email protected]) joined #gstreamer.
[17:14] alley_cat ([email protected]) left irc: Read error: 110 (Connection timed out)
[17:16] alley_cat ([email protected]) joined #gstreamer.
[17:20] <thomasvs> jdahlin: could you fix buildbot ? :/
[17:21] <jdahlin> thomasvs: what needs to be done? restart all slaves?
[17:21] <thomasvs> yeah
[17:25] <jdahlin> thomasvs: dolphy has to restart his by himself
[17:27] <thomasvs> jdahlin: you're closer to him, poke him
[17:27] <jdahlin> thomasvs: Actually I think you are
[17:27] <jdahlin> :-)
[17:30] <thomasvs> argh :)
[17:30] Action: jdahlin curses sleeping belgians
[17:30] spider-man ([email protected]) joined #gstreamer.
[17:35] sub_pop ([email protected]) joined #gstreamer.
[17:37] Action: tlg happy not to be asleep
[17:52] spyder482 ([email protected]) joined #gstreamer.
[17:57] sub_pop ([email protected]) left irc: "Client exiting"
[18:07] walters ([email protected]) joined #gstreamer.
[18:11] Nick change: The_Company -> Company
[18:16] Nick change: mxpxpod -> mx|gone
[18:18] kmaraas ([email protected]) joined #gstreamer.
[18:28] LeRoutier ([email protected]) joined #gstreamer.
[18:28] <LeRoutier> hello
[18:34] <taaz> walters: cdbs problem.  where should i put the rules that always need to run to create debian/*.install files and so on?  i tried common-build:: but that doesn't seem to run for whatever reason
[18:34] sub_pop ([email protected]) joined #gstreamer.
[18:39] lilo_booter ([email protected]) left irc: Remote closed the connection
[18:39] <walters> taaz: you probably want common-build-arch::
[18:39] <walters> taaz: or -indep::
[18:39] ChrisHJW ([email protected]) joined #gstreamer.
[18:43] BBB ([email protected]) joined #gstreamer.
[18:43] <BBB> thomasvs: I'm sending some introduction message with some gnome-media-2.8 ideas to gnome-multimedia@, want to read it before I send it?
[18:45] <thomasvs> BBB: no, I trust you :)
[18:46] ensonic ([email protected]) left irc: Read error: 110 (Connection timed out)
[18:46] ensonic ([email protected]) joined #gstreamer.
[18:47] <taaz> walters: i tried something like "common-build-indep common-build-arch::" but that didn't work.  of course i really have no idea if that is valid syntax or not ;)
[18:48] <ensonic> 2nd try: respective to the gst-manual, what do you think about bug #133057, would it be okay to commit that (better that an empty page)
[18:50] pb_ (~pb@2002:5160:4550:1:240:95ff:fe30:aa84) joined #gstreamer.
[18:51] <BBB> url?
[18:54] <ensonic> BBB: http://gstreamer.freedesktop.org/data/doc/gstreamer/head/manual/html/chapter-clocks.html
[18:56] <BBB> no, the bug url
[18:56] <BBB> I asked martin soto to fill that page
[18:56] <BBB> but apparently, he didn't yet
[18:56] <BBB> :/
[18:56] <ensonic> http://bugzilla.gnome.org/show_bug.cgi?id=133057
[18:56] <BBB> ah right
[18:57] <BBB> the problem is that the clocks introduction is too short... it doesn't explain what a clock is and how to keep in sync
[18:57] <ensonic> BBB: it is just a start and maybe that motivates them to do it
[18:57] <BBB> martin wrote a full intro some time ago
[18:57] <BBB> he needs to convert it to XML
[18:57] <BBB> that's all
[18:57] Action: BBB tries to find it
[18:58] <ensonic> BBB: I know, as I said it is just a start and should be extended, deprecated stuff needs to be marked, etc. - I just want to reduce the number of open issues I have in my list - maybe you can pass that URL along to martin
[19:00] alley_cat ([email protected]) left irc: "May the Source be with you!"
[19:01] dolphy ([email protected]) left irc: "Network down, IP Packets delivered via UPS"
[19:02] <BBB> hm, cannot find it
[19:02] alley_cat ([email protected]) joined #gstreamer.
[19:02] <BBB> anyway, the text over there is not good... I have a much better text, I'll try to work something out tonight
[19:02] <BBB> but it's a bit busy
[19:06] <LeRoutier> installing xorg, bbl
[19:06] LeRoutier ([email protected]) left irc: "Leaving"
[19:06] <BBB> also bbl
[19:06] BBB ([email protected]) left irc: "Download Gaim: http://gaim.sourceforge.net/"
[19:15] <ensonic> ciao
[19:15] ensonic ([email protected]) left #gstreamer ("Leaving").
[19:18] <walters> taaz: hmm.  you might have to do what cdbs does to group them together
[19:24] Uraeus ([email protected]) joined #gstreamer.
[19:24] <jdahlin> hello Uraeus
[19:24] Spundun ([email protected]) joined #gstreamer.
[19:25] <Spundun> whoa... a lot of subsystem reworked! why was that? adding functionality?performance? or code cleanup?
[19:26] <Uraeus> hey jdahlin
[19:26] <Uraeus> Spundun: memleak fixes mostly I think
[19:27] <thomasvs> Spundun: they are features of the 0.8.x series, not necessarily of this particular release
[19:27] <Spundun> wow! :)) wonder how it worked at all ..before
[19:27] <taaz> walters: what does that mean? ;)
[19:27] <Spundun> thomasvs: ohh okkey
[19:27] <Uraeus> Spundun: it worked with high memory usage :)
[19:28] <Spundun> yeah I thought this is toomuch in the release notes for a x.x.1 release
[19:28] <taaz> walters: i really don't grok this makefu yet
[19:29] <Spundun> okkey is this  a good time to ask a question? :D I know I know but last time I asked.. someone said that noone was there
[19:29] <Spundun> asking anyway
[19:29] <Spundun> I use gstreamer on mac with rhythmbox
[19:29] <taaz> walters: if you do "foo::" does that override previous definitions of that rule?  seems that would make lots of things break if you use a rule that's already being used for something else
[19:29] <Spundun> I mean OSX Panther
[19:29] <Uraeus> thomasvs: so has Kristien made up her mind on what to do with the job offer?
[19:29] <walters> taaz: no
[19:29] <Spundun> and I get glitches everytime I do any ui event...
[19:29] <thomasvs> Uraeus: I'm not quite sure :=)
[19:29] <Uraeus> Spundun: great to hear, hopefully we get Mac Native plugins soon :)
[19:30] <walters> taaz: info make 'Double-Colon'
[19:30] <Spundun> sometimes even on mouse movement or keyboard activity
[19:30] <thomasvs> gstreamer and rhythmbox compile and work out of the box on osx ?
[19:30] <thomasvs> scary :)
[19:30] <Spundun> Uraeus: you have any guesses whats wrong and what level I will need to go to ..to fix this?
[19:31] <Spundun> thomasvs: nope.... people worked their asses off on it :D
[19:31] <Uraeus> Spundun: how does you play sound on MacOSX? 
[19:31] <Spundun> actually gstreamer doesnt have any patches (I think ) meaning it worked out of the box
[19:31] <Spundun> but gst-plugins did
[19:31] <Uraeus> Spundun: are those patches submited upstream? (to us?)
[19:31] <Spundun> Uraeus: esd sink.... does that answer your question
[19:31] <Spundun> >
[19:31] <Spundun> ?
[19:32] <Spundun> Uraeus: ummmm
[19:32] <taaz> walters: ah, thanks.  it still isn't running them for me though.  hmm.
[19:32] Action: Spundun checks what the patches are again
[19:32] <Uraeus> Spundun: yes, well could be that esdsink is part of the problem, having a macOs sound system plugin would be much better
[19:32] <Spundun> Uraeus: is that easy? writing a sink plugin?
[19:32] <taaz> or rather, it looks like it should while debugging but then doesn't when i do a full pbuilder build.
[19:34] <Uraeus> Spundun: easy is a relative term :), but I would say that a basic output plugin is probably not a big technical challenge, getting it to support 'everything' gstreamer offers might be a medium amount of work I would think
[19:36] <Spundun> Uraeus: thomasvs: ohhh actually gstreamer... gst-plugins AND rhythmbox... all work without any patches!!
[19:36] <Uraeus> Spundun: if you are a reasonable experienced programmer I would think you could have something basic running in a day or two
[19:36] <Spundun> the only thing is this env varialbe is set GCONF_DISABLE_MAKEFILE_SCHEMA_INSTALL=1 when building gst-plugins
[19:37] <Spundun> Uraeus: nice
[19:37] Action: Spundun pokes at the web site for a plugin howto or faq
[19:37] <Uraeus> Spundun: I think the plugin writers guide details an output plugin
[19:38] <Spundun> Uraeus: cool.... thanx
[19:39] <Spundun> Uraeus: so you think having a native sink will remove those glitches?
[19:42] <Spundun> Uraeus: question.... I thought gstreamer was to be used to stream media over the network... so if I write a native sink then even if I start it from else where I wont hear the sound there right? probably I am not understanding a lot of things
[19:42] <Uraeus> Spundun: probably yes (it will remove the glitches) at least that is my best guess
[19:43] <Uraeus> Spundun: yes, a local sink wouldn't give you the esd tranparency, so if that is what you want you probably want to try and debug esdsink on macosX
[19:44] <Spundun> Uraeus: I guess I dont want that....  just like to be purist when talking... action is harder ;)
[19:45] <Uraeus> Spundun: I think as long as we don't have a MacOSX using developer things working on MacOSX will be a little based on 'luck', cause there are always some platform specfic small issues that needs resolving
[19:46] <Spundun> Uraeus: I understand
[19:50] <Spundun> Uraeus: btw I use fink packages to install stuff... so gstreamer is there in fink(right now only the 0.6 series in the unstable) but soon 0.8 should be there
[19:50] <Spundun> just to let you know... in case some mac guy comes asking about gstreamer on mac 
[19:53] Action: iain looks for a clever person
[19:53] Action: Uraeus is really clever
[19:55] <iain> Uraeus: so...opt isn't working for me :)
[19:56] <jdahlin> cool, Uraeus will fix opt!
[19:57] Action: Uraeus is clever enough to leave opt fixing to jdahlin
[19:58] <Spundun> heh
[19:59] <Spundun> or may be gst is fine and the problem is with esd implementation for mac?
[19:59] <foser> thomasvs: nautilus media view crashes on some bad files which gst-launch handles with errors... is there some error logic missing in n-m ?
[20:00] <iain> I e it when marlin/gst messes up the sample rate, and my samples play at like 1000 hz
[20:00] <iain> you get such cool noises, that I really want to sample and play guitar over
[20:00] <Spundun> bbl
[20:01] <iain> (yes, I know trent reznor made a living out of that)
[20:01] <foser> it's 'done' now iain.. you won't make a dime..
[20:01] <foser> or are you really disturbed and sad as well? bad childhood and all that..
[20:01] <iain> no, but I want to be creative
[20:07] <foser> a strong emotional state helps with that
[20:07] <foser> like you know.. thinking of suicide and such
[20:14] Makki ([email protected]) joined #gstreamer.
[20:15] jdahlin ([email protected]) left irc: "Leaving"
[20:15] thomasvs ([email protected]) left irc: No route to host
[20:17] <iain> ds-work: ping?
[20:19] Spundun ([email protected]) left irc: Read error: 110 (Connection timed out)
[20:34] sublett ([email protected]) joined #gstreamer.
[20:34] markey ([email protected]) joined #gstreamer.
[20:38] apoc ([email protected]) joined #gstreamer.
[20:41] <apoc> hi
[20:41] thomasvs ([email protected]) joined #gstreamer.
[20:41] <apoc> hey thomasvs
[20:41] <thomasvs> evening apoc
[20:43] Spundun ([email protected]) joined #gstreamer.
[20:54] Uraeus ([email protected]) left irc: Remote closed the connection
[21:05] jdahlin ([email protected]) joined #gstreamer.
[21:33] <taaz> walters: what would be the difference between building with pbuilder and running make -f debian/rules binary with respect to common-build-* being executed?
[21:33] ChriHJW_log ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[21:33] <walters> taaz: pbuilder runs dpkg-buildpackage, which runs debian/rules build-arch, then debian/rules binary, IIRC
[21:34] <taaz> is there some other cdbs rule i should be using?  this is just some code to copy gstreamer-foo.install to gstreamer0.8-foo.install
[21:35] <walters> taaz: pre-build maybe?
[21:35] Nick change: mx|gone -> mxpxpod
[21:35] <taaz> i'm really confused why this doesn't run... i don't know how to easilly debug it either...
[21:36] <jdahlin> Tack, no further luck with gstreamer?
[21:41] spider-man ([email protected]) left irc: Read error: 60 (Operation timed out)
[21:43] spider-man ([email protected]) joined #gstreamer.
[21:44] markey ([email protected]) left irc: "leaving"
[21:46] <ds-work> iain: pong
[21:53] julo ([email protected]) joined #gstreamer.
[21:53] <julo> hi
[21:53] <julo> I get this error trying to yum update:
[21:53] <julo> Error getting file http://gstreamer.net/pkg/fedora/1/i386/yum/gst/RPMS/gstreamer-plugins-audio-0.8.1-0.fdr.1.1.i386.rpm
[21:53] <julo> [Errno 4] IOError: HTTP Error 403: Forbidden
[21:53] <julo> Any idea ?
[21:55] walters ([email protected]) left irc: "out"
[21:58] steve_b ([email protected]) joined #gstreamer.
[21:58] <thomasvs> julo: hm, probably my mistake again
[21:58] <thomasvs> lemme check
[21:58] <julo> no problem
[21:59] <taaz> are we expected to have lots of translations?
[22:00] <taaz> i ask wondering if it's worth splitting off into seperate debian package like the rest of debian does
[22:00] <thomasvs> taaz: expecting, yes
[22:00] <thomasvs> julo: try again
[22:02] jdahlin ([email protected]) left irc: Read error: 110 (Connection timed out)
[22:04] <julo> I'm trying...
[22:06] m_wheels ([email protected]) joined #gstreamer.
[22:07] <foser> thomasvs: shouldnt nautilus-media be using the gst-plugins media-info ?
[22:18] Nick change: m_wheels -> wheels
[22:20] <ds-work> taaz: no
[22:21] <taaz> no?
[22:22] <ds-work> don't split translations.  it's stupid
[22:24] <taaz> it's not that stupid is it?  reduce duplicate arch indep files from the archive?
[22:26] <ds-work> 2 binary packages almost always requires more resources than 1
[22:27] aflinta ([email protected]) joined #gstreamer.
[22:27] <aflinta> hi
[22:27] <taaz> eh ok, easier to not change it ;)
[22:27] <aflinta> I have problems during compillation gstreamer 0.8.1 - erro during docs build, in version 0.8.0 was fine
[22:28] <julo> thomasvs: it works, thanks !
[22:28] jdahlin ([email protected]) joined #gstreamer.
[22:28] julo ([email protected]) left irc: "Leaving"
[22:28] <ds-work> aflinta: could you file a bug with a build log?
[22:28] <taaz> would only take a few mins if someone starts to complain i guess.  most of the debian-devel thread from a while ago about that was for large large translations though.  i guess ours don't compare to openoffice or whatever
[22:29] <aflinta> ds-work: ok
[22:29] markey ([email protected]) joined #gstreamer.
[22:30] <ds-work> keep in mind that the Packages.gz file is downloaded by _everyone_, and often every day, whereas a new package, although larger, happens once per month and by fewer people
[22:32] <taaz> yeah, archive size vs bw tradeoffs
[22:33] smoke ([email protected]) got netsplit.
[22:36] <aflinta> ds-work: http://bugzilla.gnome.org/show_bug.cgi?id=140205
[22:38] Zeenix ([email protected]) joined #gstreamer.
[22:39] smoke ([email protected]) got lost in the net-split.
[22:43] <ds-work> oh right.  I saw that last night
[22:44] <Tack> jdahlin: I noticed that the ffmpeg bug got fixed, so I'll give that a try.   I have problems when using alsasink (sound pulses on and off about once a second), although osssink works.
[22:44] <Misirlou> The whole translation splitting thing seems pretty prejudicial to me.
[22:44] <jdahlin> Tack, oh. haven't seen that
[22:46] <Misirlou> Here's package foobar, but you won't be able to use it functionally until you download foobar-xx.
[22:46] <Zeenix> hi guys
[22:46] <Tack> jdahlin: Not surprising. :)  
[22:47] <Tack> jdahlin: The fix for that avi was in gst-plugins, and not gst-ffmpeg, right?
[22:48] <jdahlin> Tack: yes
[22:51] <Tack> Which cvs module builds alsasink?
[22:51] <Tack> Oh, gstalsa in gst-plugins?
[22:53] <ds-work> yes
[22:57] kwm ([email protected]) left irc: "Leaving"
[22:59] <jdahlin> what about correction errors related to gstreamer
[22:59] <Tack> Okay, that bug is fixed.  Now I have more :)
[23:00] <Tack> I believe it's related to the fact that this video resolution isn't multiple of 16 or something.
[23:01] <Tack> Oh, or not...
[23:01] <Tack> It works with ximagesink but not xvimagesink -- with the latter the video is all skewed like there's some offset issues.
[23:01] <jdahlin> divisible resolution frequencies are possibly the cause of the elementary disruption
[23:02] <ds-work> yeah, many elements don't do video width % 8 != 0 correctly
[23:03] <Tack> http://sault.org/~tack/clips/animatrix-clip.avi
[23:03] <Tack> Plays okay with ximagesink (slow and choppy of course), but with xvimagesink, crazy kalleidescope fun :)
[23:03] <jdahlin> marvelous :)
[23:04] <Tack> And, in addition to the screwed up video, it doesn't play smoothly.  Every second or so there seems to be like a "pulse" where the framerate slows down.
[23:04] <Tack> Probably a sync issue since it doesn't happen when I use fakesink for audio.
[23:04] <Company> ds-work: got some time?
[23:05] <ds-work> Company: a bit
[23:05] <Tack> jdahlin: Should I file a bug?
[23:05] <Company> ds-work: what do you think would be the easiest way to ensure a DISCONT is sent before any buffer?
[23:05] <jdahlin> pixel pixies are probably responsible for the disturbance
[23:06] <Tack> Pixel plxies? :)
[23:06] <Company> ds-work: that should include replugging playing elements of course
[23:07] harshy ([email protected]) left irc: "I quit for now"
[23:08] <jdahlin> woo woo gigastreamer is the cause for the interplantary frequency disruption
[23:08] <ds-work> Company: one idea: put a flag in GstPadLink that says whether the link has been consummated
[23:09] <Company> ds-work: that would require help from schedulers
[23:09] <ds-work> why?
[23:09] <ds-work> put a check in gst_pad_push()
[23:09] <Company> ds-work: get functions
[23:10] <jdahlin> sitting in barcelona considering how where and whats going on with you other friendly folk
[23:10] <Company> ds-work: you either need to hook into get functions and gst_pad_push or chain functions and gst_pad_push
[23:11] <jdahlin> that was my good english friend hijaking the keyboard
[23:11] <jdahlin> if you had any doubts :-)
[23:11] <ds-work> Company: what about a check in both push and pull?
[23:11] <ds-work> er, nm
[23:11] <Company> ds-work: still doesn't work on get=>chain
[23:12] <ds-work> yeah, you'll need to have scheduler junk somewhere
[23:12] <ds-work> sorry
[23:12] ChrisHJW_log ([email protected]) joined #gstreamer.
[23:12] <Company> ds-work: and it leaves the question where to leave the buffer while you invent an event
[23:13] <ds-work> also, remember to check if the first buffer is itself a discont event
[23:13] <Company> yeah
[23:13] <ds-work> well, in gst_pad_push(), you keep it in a local variable
[23:14] <Company> which doesn't work when you destroy the cothread ;)
[23:14] <Tack> jdahlin: That reminds me of http://bash.org/?5775 :)
[23:14] <ds-work> darn.  so it leaks
[23:14] <ds-work> you could keep it in the GstPadLink
[23:15] ChrisHJW_log ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[23:15] <Company> ds-work: that would work
[23:16] <Company> ds-work: we need to keep this in min when doing the iteration functions - the system might require a discont event at any time
[23:18] <Company> ds-work: i'll put the stuff in the GstPadLink and implement gst_pad_call_(chain|get)_function and make the schedulers use those
[23:18] ChrisHJW_log ([email protected]) joined #gstreamer.
[23:19] <jdahlin> hey party people does anyone feel like dancing??
[23:19] <ds-work> jdahlin-whomever: how do you know we're not dancing?
[23:20] <jdahlin> well that was what i was hoping as i have just commenced a scottish reel in my undies whilst singing 
[23:20] tlg ([email protected]) left irc: "Client exiting"
[23:20] Action: ds-work dances to Vivaldi
[23:20] <jdahlin> can anyone teach me how to tango??
[23:22] <jdahlin> i dance with my cat but tend to lose step as the catwalk is hard to keep up with
[23:24] <jdahlin> guys you're just simply too boring
[23:24] <jdahlin> damn programmers!
[23:24] <jdahlin> :-)
[23:25] <Company> nah, we'Re really interesting
[23:25] <Company> we just don't tango with cats
[23:26] <jdahlin> Tack, this wasn't as bad as that one
[23:26] <Tack> jdahlin: Quite :)
[23:26] <jdahlin> Tack, did I miss anything? :P
[23:27] <ds-work> we waltz with bears, though
[23:27] <ds-work> in the woods
[23:27] <ds-work> at night
[23:27] <ds-work> when we can't sleep
[23:28] <jdahlin> haha
[23:28] <jdahlin> Tack, it took a couple of lines for you too realize that something was wrong though
[23:28] <jdahlin> Tack, maybe my grammar was too good
[23:29] markey ([email protected]) left irc: "bbl"
[23:29] <Tack> jdahlin: I dunno.  It was hard to tell.  You made no sense, which is usual. :)
[23:29] Action: jdahlin kicks Tack
[23:29] Action: Tack blocks. :P
[23:31] spyder482 ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[23:31] <jdahlin> smack tack because he smokes crack :p
[23:31] <jdahlin> alice is once again in wonder land
[23:32] <ds-work> 10-foot-tall alice, or little alice?
[23:32] <jdahlin> little alice who has eaten many many mushrooms
[23:33] <jdahlin> the pixel monkey has come out to play
[23:34] <jdahlin> where are all the forest folk hiding among the numbers??
[23:35] Nick change: jdahlin -> triptic
[23:35] <triptic> any news? ;)
[23:36] <triptic> has anyone seen kill bill pt 2?
[23:38] <Tack> Not yet.
[23:38] <Tack> Just watched volume 1 the other night.
[23:39] earthw0rm ([email protected]) joined #gstreamer.
[23:39] <earthw0rm> would it be hard to make my gnome program play a wav file via gstreamer?
[23:43] <triptic> blame it on the gik
[23:43] <earthw0rm> ?
[23:43] <ds-work> earthw0rm: no
[23:44] <earthw0rm> will my sounds then be able to play at the same time as Rhythmbox this way?
[23:44] <ds-work> earthw0rm: create a GstPlay object and play it
[23:44] <ds-work> earthw0rm: if you are running a sound server, yes
[23:44] <earthw0rm> i'm not
[23:45] <triptic> earthw0rm: if you beat it with the rhythm stick
[23:45] <ds-work> ignore the drunk
[23:45] <earthw0rm> if i run esd, the rhythmbox won't play
[23:45] <earthw0rm> ds-work: i was doing  :)
[23:45] <triptic> esd inhibitors
[23:45] <triptic> am very sober
[23:45] <ds-work> earthw0rm: you need to configure rb for esd output
[23:46] <triptic> configuration settings??
[23:46] <earthw0rm> didn't see that in the options ...
[23:46] <ds-work> earthw0rm: gnome-properties
[23:46] <thomasvs> triptic: I can teach you how to tango
[23:46] <triptic> really?
[23:46] <thomasvs> triptic: sure.
[23:47] <earthw0rm> is it a gconf thing?
[23:47] <triptic> do you give classes?
[23:47] <ds-work> earthw0rm: yes
[23:47] <triptic> or seive glasses
[23:47] <thomasvs> triptic: heh, no :)
[23:47] <triptic> :)
[23:47] <earthw0rm> great
[23:48] <triptic> am having trouble with python - incompatible pointer
[23:48] earthw0rm ([email protected]) left #gstreamer (""I allege that SCO is full of it." - Linus Torvalds").
[23:49] <triptic> what are you doing this weekend?
[23:55] Nick change: triptic -> JDHALIN
[23:55] Nick change: JDHALIN -> jdahlin
[23:55] <jdahlin> damn limeys
[23:57] <jdahlin> thomasvs: so you want to give the limey some tango lessons? :)
[23:59] <Tack> Features of the 0.8.1 release says "Parallel installability with 0.6.x series" which reads an _awful_ lot like "parallel instability with 0.6.x series" :)
[00:00] --- Fri Apr 16 2004
[00:01] <jdahlin> heh
[00:01] <jdahlin> Tack: now you're being negative!
[00:02] <ds-work> anyone want to write a media bug buddy?
[00:02] <ds-work> i.e., media fails to play so it automatically asks if you want to file a bug
[00:07] <Company> that sounds way useful
[00:07] <Company> but please only do that after you wrote the automatic closer for "my mp3 doesn't play back on FC2"
[00:08] <ds-work> yeah
[00:08] <ds-work> I'm not sure how to handle unknown media errors
[00:09] <ds-work> I'll probably write a gstreamer bug buddy as a spinoff from a swfdec bug buddy
[00:10] walters ([email protected]) joined #gstreamer.
[00:11] <ds-work> basically, I'm thinking only of errors where _gstreamer_ realized there was a problem, not the user
[00:12] <Company> hm, there aren't many cases where gstreamer figures that out
[00:12] <walters> ds-work: how reproducible is your rb bug?
[00:12] <ds-work> 'caps negotiation problem, file a bug';
[00:12] <ds-work> walters: 100%
[00:12] <walters> ds-work: ah, ok, great
[00:13] <walters> ds-work: this is on your smp machine?
[00:13] <ds-work> yes
[00:13] <walters> ds-work: can i have a shell?
[00:21] md` ([email protected]) left irc: 
[00:21] ploum ([email protected]) left irc: Read error: 110 (Connection timed out)
[00:24] spyder482 ([email protected]) joined #gstreamer.
[00:26] <thomasvs> ds-work: it already does that
[00:26] <thomasvs> walters: I'm pretty sure that if you install an smp kernel you'll get it too
[00:26] <ds-work> thomasvs: ?
[00:27] <thomasvs> ds-work: gstreamer tells you about caps nego problems with extra bug info, no ?
[00:27] <ds-work> thomasvs: it doesn't file a bug for you, though
[00:28] <thomasvs> ds-work: heh, true; not that far away though.
[00:28] <thomasvs> main problem is figuring out what browser to use for it I guess :)
[00:28] <ds-work> we need to figure out how to start up bug-buddy and feed it the media file
[00:28] <thomasvs> do we really want 700mb attachments to bug reports ?
[00:29] <ds-work> we could chop it at 1 MB or so
[00:29] <walters> thomasvs: why would a SMP kernel make a difference?
[00:35] <thomasvs> walters: different code paths
[00:35] <walters> thomasvs: in what?  the kernel?
[00:35] <thomasvs> walters: yeah
[00:44] jdahlin ([email protected]) left irc: "Leaving"
[00:49] <taaz> did someone fix 137833? i thought i saw a related commit
[00:51] <Zeenix> thomasvs: i just translated the vw howto to xml, should i send it to you or the sgml would be ok?
[00:51] <ds-work> ah, yes
[00:51] Action: ds-work closes
[00:53] Nick change: spyder482 -> spyder_reading
[00:58] <Company> we still don'T have a bug related to not releasing the audio device, do we?
[00:59] smoke ([email protected]) joined #gstreamer.
[01:14] <sri> hi guys
[01:25] trow ([email protected]) left irc: "Leaving"
[01:29] pb_ (~pb@2002:5160:4550:1:240:95ff:fe30:aa84) left irc: "Client exiting"
[01:40] md` ([email protected]) joined #gstreamer.
[01:40] <iain> ds-work: can you tell me if I need to write a fixate function for my src?
[01:42] spider-man2010 ([email protected]) joined #gstreamer.
[01:42] <ds-work> what src?
[01:44] spider-man ([email protected]) left irc: Read error: 110 (Connection timed out)
[01:45] foser ([email protected]) left irc: "[ I want to believe ]"
[01:45] <iain> ds-work: marlinsamplesrc
[01:46] <iain> ds-work: it needs to set the sample rate and buffer-frames of the pad caps when the sample changes
[01:46] <ds-work> ok, then no
[01:46] <iain> okay, so how/when do I set them?
[01:46] <ds-work> it's like a decoder source pad
[01:47] <ds-work> use gst_pad_use_explicit_caps() and set_explicit_caps()
[01:47] <iain> okay
[01:48] <iain> in both the link function and when the sample is changed?
[01:49] <ds-work> you won't have a link function then
[01:49] <ds-work> or getcaps
[01:49] <iain> oh, ok, even better :)
[01:49] <ds-work> explicit_caps code takes care of that
[01:49] <ds-work> basically, it's "I'm in charge of the format, deal with it"
[01:55] <iain> thanks
[01:55] Action: iain copies what gstmad.c does, recompiles and hopes it works
[01:56] <iain> perfect...
[01:56] <iain> woohoo, marlin plays again
[02:02] sublett ([email protected]) left irc: "I quit, therefore I was"
[02:04] BBB ([email protected]) joined #gstreamer.
[02:06] Rotty ([email protected]) left irc: Remote closed the connection
[02:18] smoke ([email protected]) got netsplit.
[02:24] smoke ([email protected]) got lost in the net-split.
[02:42] <Company> what's happening with the fork now btw?
[02:42] <BBB> I don't think there's a fork
[02:42] <Company> i want to have a nice build on windows out of the box :)
[02:43] <BBB> with provided .exe installer?
[02:43] <ds-work> heh, that requires things like popt to build out of the box
[02:43] <ds-work> I wish someone would get a good distro going for Windows
[02:44] <ds-work> I'm tired of downloading and compiling source and searching for patches
[02:44] <Company> the provided exe installer would come second
[02:44] Action: BBB smiles
[02:44] <Company> after someone removed bonobo from rb and ported it
[02:44] <BBB> do we tell in the installer that our job is to take over the world?
[02:45] <BBB> or do you think that'd trigger some process in win32 to tell the FBI of our plans?
[02:45] <ds-work> what's wrong with bonobo?
[02:45] <iain> "its too hard"
[02:45] <Company> it doesn't work on windows afaik
[02:45] Action: BBB seconds iain
[02:45] <Company> bonobo is ugly
[02:46] <Company> it shouldn't be
[02:46] <iain> BBB: I'm just quoting...I never really found bonobo too hard at all
[02:46] <BBB> it should wrap gobject
[02:46] <walters> Company: i'm going to port to the new Gtk bits soon
[02:46] <BBB> but it defines it's own object system, it's own property system
[02:46] <BBB> I don't understand why
[02:46] <BBB> even its own type system
[02:46] <Company> bonobo should generate all needed stuff from header files
[02:46] <walters> BBB: GObject wasn't designed to be remoted...
[02:46] <Company> bonobo-generate-idl header.h
[02:47] iain ([email protected]) left irc: "leaving"
[02:47] <BBB> walters: that doesn't mean we just copy the whole system and redo it
[02:47] <BBB> it's horrible
[02:47] <BBB> bleh, and checkins are slow today
[02:48] <Company> the huge problem with bonobo is that it's not easy to code against and it requires an abstraction layer
[02:49] <Company> so people don't use it
[02:49] <Company> and bonobo would only have worked if everyone used it
[02:49] <BBB> like COM? :)
[02:49] Action: BBB smiles again
[02:49] <Company> yeah, like com :/
[02:49] <BBB> bonobo is ok for some things, I guess...
[02:50] <Company> +yeah, 7k live events
[02:50] <Company> i think there are still some leaks with discont events left
[02:50] <walters> so does anyone here grok alsa?
[02:51] <Company> depends on your definition of "grok"
[02:51] <walters> i am wondering whether i should default the gstreamer sink in FC2 to "alsasink device=hw:0"
[02:51] <Company> i've written an overly complicated sink that uses it
[02:51] <walters> right now it's just alsasink
[02:51] <walters> and people are getting errors
[02:52] <walters> is their alsa config just messed up?
[02:52] <Company> what alsa version is in FC2?
[02:52] <Company> alsalib?
[02:52] <walters> i would think that detecting the device should be an alsalib level issue
[02:52] <walters> [root@nexus walters]# rpm -qa | grep alsa
[02:52] <walters> alsa-lib-1.0.3a-1
[02:52] <walters> alsa-utils-1.0.3-1
[02:52] <walters> alsa-lib-devel-1.0.3a-1
[02:53] <Company> http://bugzilla.gnome.org/show_bug.cgi?id=134007
[02:53] <walters> i am definitely not a fan of alsalib's apparent habit of calling assert() for configuration failures
[02:53] <Company> there's a link to an alsalib bug in there
[02:53] <walters> Company: ah ok
[02:53] <Company> which was closed because everyone i bugged about it didn't follow up on it
[02:54] <Company> so if you get someone who can reproduce that and work with the alsa guys on a solution, i would be very happy
[02:54] <walters> hm
[02:54] <walters> apparently you need an account
[02:54] Action: BBB uses alsa with oss emulation in any app
[02:54] <walters> that's a little lame
[02:54] <Company> you can just use guest
[02:54] <BBB> alsalib is so broken
[02:55] <ds-work> they've only had _8_YEARS_
[02:55] <BBB> only?
[02:55] <ds-work> (didn't alsa start in about 96?)
[02:56] <BBB> after 8 years, I'd know how to sed away assert() to a decent fault handler
[02:56] <Company> depends on if you want it to be an assertion or not
[02:56] <BBB> not to mention that I'd learn that hiding posix is a stupid idea anyway
[02:56] <Company> it definitely is not
[02:56] <BBB> for users it's not, for developers it is
[02:57] <BBB> I'm *used* to posix
[02:57] <Company> yeah
[02:57] <BBB> app developers don't use the kernel stuff, they use gstreamer or so
[02:57] <Zeenix> thanks for telling me that bonobo IS hard guys, or else I would have died in the guilt i feel for not being able to understand/code it
[02:57] <Company> which makes v4l and oss so super portable
[02:57] <BBB> alsalib is an overly complicated and bloated stupid idea
[02:57] <Company> no, alsalib is a great _idea_
[02:57] Action: BBB disagrees
[02:57] <Company> an audio abstraction lib
[02:57] <Company> that's great
[02:58] <BBB> if it were inside gstreamer, sure
[02:58] <BBB> but god
[02:58] <ds-work> not at the level it's done
[02:58] <BBB> we'd have v4l2lib
[02:58] <Company> we would just need one audiosink if people had thought about that earlier
[02:58] <BBB> scsilib
[02:58] <BBB> whee
[02:58] <BBB> one lib for every kenrel layer
[02:58] <BBB> that's useful!
[02:58] <ds-work> there should be a v4llib
[02:58] <ds-work> BBB: there should be
[02:58] <Company> yeah
[02:58] <BBB> and then we'll have a wrapperlib for alsalib because alsalib sucks too
[02:58] <BBB> yay!
[02:58] Action: BBB doesn't get that
[02:58] Action: walters has a sinking feeling he's going to get a lot of dupes of this bug when FC2 ships
[02:58] <BBB> it simply means these people did their posix interface wrong
[02:59] <BBB> v4l2 has a reasonable posix interface, which is why I fully understand it
[02:59] <BBB> I understand OSS as well
[02:59] <BBB> I'll never understand ALSA
[02:59] <Company> i prefer calling ftell to ioctl (fd, magic_number, pointer)
[02:59] <Company> alsa is way to complicated


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