[gst-devel] Daily IRC logs

[wim.taymans at chello.be]

[08:50] Nick change: taaz -> taazzzz
[09:24] omega_ (omega at omegacs.net) left irc: Read error to omega_[omegacs.net]: EOF from client
[09:57] thomas (thomas at urgent.rug.ac.be) joined #gstreamer.
[09:57] <thomas> happy easter
[10:03] Action: thomas is away - Automatically set away. - messages will be saved.
[10:20] <thomas>  
[11:07] Last message repeated 1 time(s).
[12:08] Last message repeated 1 time(s).
[12:08] <thomas> quit
[12:08] thomas (thomas at urgent.rug.ac.be) left irc: I'm outta here!
[12:25] Uraeus (Uraeus at c224s9h5.upc.chello.no) joined #gstreamer.
[12:26] <Uraeus> hi
[12:26] Nick change: wtay-sleeping -> wtay
[12:26] <wtay> yo
[12:26] <Uraeus> brb
[12:26] Uraeus (Uraeus at c224s9h5.upc.chello.no) left irc: 
[12:27] Uraeus (Uraeus at c224s9h5.upc.chello.no) joined #gstreamer.
[12:27] <Uraeus> weird
[12:28] <Uraeus> x-chat is misbehaving
[12:28] <wtay> hmm
[12:29] <Uraeus> wtay: I am mentioning in my GUADEC summary that you misrepresented yourself on IRC :)
[12:30] <wtay> uh?
[12:30] <Uraeus> your claims of being an antisocial caveman didn't turn out to be true :)
[12:30] <wtay> grr :)
[12:31] <Uraeus> wtay: do you know if it is possible to resize a ext2 partition?
[12:32] <wtay> Uraeus: I think so, ext2resize exists I think
[12:32] <Uraeus> ok thanks
[12:57] thomas (thomas at urgent.rug.ac.be) joined #gstreamer.
[12:57] <wtay> hello
[12:57] <thomas> hi
[12:57] <thomas> uhm...
[12:57] <thomas> I've been looking for that core dump error
[12:57] <thomas> generated when I use esdsink instead of osssink
[12:57] <thomas> the one where we disabled width = get_caps ...
[12:57] <thomas> remember that ?
[12:57] <wtay> in your plugin?
[12:57] <thomas> yes...
[12:57] <thomas> but also in stereo2mono
[12:58] <wtay> ok
[12:58] <wtay> does it work with osssink?
[12:58] <thomas> yes
[12:58] <thomas> so I think there's something wrong with
[12:58] <wtay> ok, esdsink is broken then
[12:58] <thomas> a) either esdsink
[12:58] <thomas> or b) the stereo2mono way of getting caps
[12:58] <thomas> so what I did...
[12:58] <thomas> I tried to get the mime type
[12:58] <thomas> of the pad
[12:58] <thomas> before getting the width
[12:58] <thomas> if I make a pipe with esdsink
[12:58] <thomas> it fails, because it can't assert the pad or something
[12:58] <thomas> oss works fine
[12:59] <thomas> so I inspected both sinks
[12:59] <thomas> seems that esd only has different fields from oss
[12:59] <thomas> it only has depth, not width
[12:59] <thomas> I tried addin that
[12:59] <thomas> but that doesn't change much
[12:59] <thomas> I don't know enough about this yet to fix it
[12:59] <wtay> hmm
[13:00] <thomas> or even know if that's wrong or not
[13:00] <thomas> who does most of the work on the audio ?
[13:00] <thomas> (sorry to be so talkative)
[13:00] <wtay> it looks like esdsink didn't want to use the provided audio format
[13:00] <thomas> (i have to get home quickly for easter dinner)
[13:00] <thomas> so...
[13:00] <wtay> esdsink was done by richard
[13:00] <thomas> that means I shouldn't really worry about it now ?
[13:00] <wtay> I'll take a look at it, it hasn't been updated for a while
[13:01] <wtay> thomas: capsnego also has some problems so...
[13:01] <thomas> So, you're pretty sure that the width getting in stereo2mono is ok ?
[13:01] <wtay> this can be very messy
[13:01] <thomas> yeah, I tried doing the same thin as in capsnego
[13:01] <wtay> thomas: yeah
[13:01] <thomas> first get the Caps in a ptr, then access it
[13:01] <thomas> because of the threads, gdb doesn't seem helpful
[13:01] <thomas> don't know, never did threads
[13:01] <thomas> not in c anyway
[13:01] <wtay> it's just that the capabilities of the pad is not set when the pads don't agree, hence the segfault
[13:02] <thomas> so...
[13:02] <wtay> gdb is useless
[13:02] <thomas> would it be hard to put a check on that when negotiating pads
[13:02] <thomas> in the gstreamer core ?
[13:02] <wtay> can you live with osssink for now?
[13:02] <thomas> yes of course...
[13:02] <thomas> I was just worried if it would work
[13:02] <thomas> plus I also put some time in making my xmms crossfade plugin
[13:02] <wtay> thomas: the core should fail to play the pipeline in that case
[13:02] <thomas> work with esd
[13:02] <thomas> so I don't have to turn off xmms all the time for testing ;(
[13:02] <thomas> but that doesn't matter
[13:03] sienap (synap at ipc379c0e6.dial.wxs.nl) joined #gstreamer.
[13:03] <wtay> sorry :(
[13:03] <thomas> no problem
[13:03] <thomas> not your fault
[13:03] <thomas> ok...
[13:03] <wtay> sorta
[13:03] <thomas> I'll clean up the passhtrough filter
[13:03] <wtay> ok
[13:03] <thomas> i've written a small shell script to copy a filter
[13:03] <sienap> thomasss!!!
[13:03] <thomas> and change all the function references
[13:03] <thomas> hi sienap !
[13:03] <sienap> thomas how is the stuff going ?
[13:03] <wtay> thomas: that is really cool
[13:03] <thomas> well...
[13:03] <thomas> bugs suck
[13:03] <thomas> so that I can quickly copy the filters
[13:04] <thomas> sienap: i have to go home and eat for easter
[13:04] <thomas> sienap: but i'll be back
[13:04] <thomas> sienap: and hopefully get something FINISHED today
[13:04] <sienap> he :)
[13:04] <sienap> cool :)
[13:04] <thomas> wtay: BTW, I couldn't find any way yesterday
[13:04] <sienap> we just had the eat for easter bla bla :)
[13:04] <thomas> wtay: to make C work the way I wanted to
[13:04] <sienap> wtay still busy with bonobo ?
[13:04] <thomas> wtay: define macros aren't useful either at the moment
[13:05] <thomas> wtay: for a more complex filter, the whole loop should be rewritten
[13:05] <thomas> wtay: i'll think about it some more
[13:05] <thomas> meanwhile i'll leave it like it is
[13:05] <wtay> ok
[13:05] <thomas> until I understand gstreamer more
[13:05] <thomas> there *has* to be a way
[13:05] <thomas> ok,
[13:05] <thomas> off to the home
[13:05] <thomas> till later
[13:05] <wtay> cya
[13:05] <thomas> bye wtay,sienap
[13:05] thomas (thomas at urgent.rug.ac.be) left irc: I'm outta here!
[13:06] <wtay> sienap: doing the mad plugin
[13:06] <sienap> aah cool
[13:06] <sienap> is it really fast ?
[13:06] <wtay> not really
[13:06] <sienap> he
[13:06] <wtay> same as mpg123
[13:06] <sienap> so it was some bla bla?
[13:06] <sienap> hmm
[13:07] <sienap> xmms seems to be GODDAMN fast in mp3 decoding
[13:07] <sienap> less then 1% on my 233
[13:07] <sienap> and mpg123 15%
[13:07] <sienap> no kidding
[13:07] <sienap> however xmms does some strange things with mp3's it also has an HUGE memory print
[13:07] <sienap> however FAST FAST FAST FAST FAST
[13:07] <sienap> would be nice if someone could rip out the routine :)
[13:07] <sienap> or look how they did it :)
[13:07] <wtay> hmm
[13:08] <sienap> 21517 root       0   0  8540 8540  5092 S     0.7  8.9   0:10 xmms
[13:08] <sienap> 0.7 % cpu
[13:08] <sienap> 21566 root      19   0  1900 1900  1080 R    17.7  1.9   0:00 plaympeg
[13:08] <sienap> plaympeg
[13:09] <sienap> 21568 root      16   0   804  804   588 S    14.3  0.8   0:00 mpg123
[13:09] <sienap> mpg123
[13:09] <wtay> gstmediaplay?
[13:09] <sienap> mom
[13:09] <sienap> the current one crashes
[13:09] <sienap> mom recompile the latest cvs
[13:10] <sienap> making :)
[13:10] <sienap> takes a big while
[13:10] <Uraeus> sienap: your getting your mom recompile the latest cvs?
[13:10] <sienap> he ?
[13:10] <sienap> i mean one moment
[13:10] <Uraeus> hehe
[13:11] <sienap> so ehm wtay
[13:11] <sienap> it is pretty clear
[13:11] <sienap> xmms's routine is somehow verry fast
[13:11] <sienap> but still uses mpg123
[13:11] <sienap> i don't really get it
[13:11] <wtay> mpg123 is pretty fast here too
[13:11] <sienap> he you have an athlon
[13:11] <sienap> everything is fast there :)
[13:11] <sienap> however as you see
[13:12] <sienap> xmms uses aroudn 9mb ram here
[13:12] <sienap> mpg123 only taks 1mb
[13:12] <sienap> and plaympeg 2mb
[13:12] <sienap> however xmms is die hard fast
[13:12] <sienap> i don't understand how they got it done this fast
[13:12] <sienap> since they use mpg123 
[13:12] <sienap> in some strange way
[13:12] <sienap> going to look at it ?
[13:13] <wtay> perhaps :)
[13:13] <wtay> I got lots of thing to do :)
[13:13] <wtay> the mad decoder somewhat works...
[13:14] <sienap> he :)
[13:14] <sienap> he
[13:14] Nick change: Uraeus -> Ura_shop
[13:14] <sienap> *GRRR*
[13:14] <sienap> we need more good develers
[13:14] <wtay> yeah
[13:14] <sienap> root at Oegaah mpg123]# cat *.c |wc -l
[13:14] <sienap>   10736
[13:14] <sienap> AUWWWWWWWW
[13:15] <sienap> 10k lines for an mp3 decoder
[13:15] <Ura_shop> sienap: we have many, I think that if you post your suggestions to the list there is a good chance more people than wtay can look at it
[13:15] <sienap> fucking hell
[13:15] <sienap> he :)
[13:15] <Ura_shop> ok, really going to the shop now or I will be starved
[13:16] <wtay> Ura_shop: l8r
[13:16] <sienap> starve you bast
[13:16] <sienap> wtay is it me
[13:17] <sienap> or is 10k lnies pretty much for an mpeg decoder
[13:17] <wtay> mpg123 also has id3/v2 and http code in it
[13:17] <wtay> and all audio output code
[13:18] <sienap> he
[13:18] <sienap> i still don't get it why xmms is faster then mpg123
[13:18] <sienap> they actually use mpg123
[13:18] <sienap> Note about the mpg123 licens!
[13:18] <sienap> We have permission to release mpg123 source inside xmms as GPL!
[13:18] <sienap> Using mpg123 source (at least in a commercial environment) may infringe 3rd party patents.
[13:18] <sienap> Also Michael Hipp (hippm at informatik.uni-tuebingen.de) would like to know if you modify
[13:18] <sienap> this sourcecode. send him an email if you do, he would like to have an eye on external developments.
[13:18] <sienap> Happy playing!
[13:19] <sienap> it is the mpg123 source that is included
[13:19] <sienap> but still it is faster then mpg123 alone
[13:19] <sienap> man why..
[13:20] <wtay> they modified it heavily AFAIK
[13:21] <sienap> hmmmm
[13:21] <sienap> in a pretty good way
[13:21] <sienap> you should look at it what they modified
[13:21] <wtay> man, it's huge!
[13:22] <sienap> he :)
[13:22] <sienap> maybe ask one of the developers
[13:22] <sienap> or the mpg123 author
[13:22] <sienap> i bet he knows what they changed
[13:22] <sienap> http://directfb.org/news/count/gtk-themed-alpha.png AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAHHHhhhhhhhhhhhhhhhhhhhhhhh *DREWL*
[13:23] <wtay> hehe, I saw that live at GUADEC :)
[13:24] <sienap> man
[13:24] <sienap> that stuff is cool :)
[13:24] <sienap> wtf is it ?
[13:24] <sienap> a replacement for X or what ?
[13:24] <sienap> xrenderer
[13:24] <sienap> i don't get it
[13:24] <wtay> for embeded devices
[13:25] <sienap> why embedded devices do get alpha transparancy and X doesn't :)
[13:25] <sienap> on a that cool level >:)
[13:26] <wtay> it can't do stipples or lines...
[13:26] <sienap> he ?
[13:26] <sienap> how do you mean ?
[13:27] <wtay> it's not as feature complete as X
[13:28] <sienap> he ok ok.. ..
[13:28] <sienap> you have any idea when xrender will be more mature
[13:28] <sienap> so we all can use the wonders of non dirty hacked good looking alpha transparency
[13:28] <wtay> no idea
[13:28] <wtay> I think it allready is
[13:28] <wtay> you can code for it now
[13:29] <wtay> xdpyinfo shows I have it
[13:29] <sienap> jr
[13:29] <sienap> he
[13:29] <sienap> however
[13:29] <sienap> go code again :)
[13:30] <wtay> mad now plays audio with gstreamer :)
[13:30] <sienap> great
[13:30] <sienap> ohjeah i was compiling gstreamer
[13:30] <sienap> btw the code bugged somewhere
[13:30] <sienap> i had to change some gfloat stuff to double
[13:30] <sienap> i remember
[13:31] <wtay> I fixed it
[13:31] <sienap> he okie 
[13:31] <sienap> me too :)
[13:31] <wtay> I fixed it in CVS :)
[13:31] <sienap> and me local :)
[13:31] <sienap> hmm
[13:31] <sienap> compiling will take another 30 mins :)
[13:31] <sienap> man i need a compile farm here :)
[13:32] <sienap> he however seen from the Moore law we wiil have 1 TIPS CPU's in 2010 :)
[13:32] <sienap> so i'll wait till then :)
[13:32] <sienap> Tera instructions per seconds *NJUM*  >:)
[13:33] <wtay> :)
[13:33] <sienap> he :)
[13:40] <sienap> hmm i am raelly wondering what the cpu usage is with gstmediaplay and testxmms
[13:44] Nick change: Ura_shop -> Uraeus
[13:50] Uraeus (Uraeus at c224s9h5.upc.chello.no) left irc: 
[13:51] Uraeus (Uraeus at c224s9h5.upc.chello.no) joined #gstreamer.
[13:56] <sienap> jeeh compiled
[13:56] <sienap> make install
[14:07] <sienap> hmm
[14:08] <sienap> jade:gst-plugin-writers-guide.sgml:370:14:E: element "EMPHASIS" undefined
[14:08] <sienap> jade:gst-plugin-writers-guide.sgml:372:15:E: element "PARA" undefined
[14:08] <sienap> jade:gst-plugin-writers-guide.sgml:373:14:E: element "EMPHASIS" undefined
[14:08] <sienap> jade:gst-plugin-writers-guide.sgml:374:15:E: element "PARA" undefined
[14:08] <sienap> jade:gst-plugin-writers-guide.sgml:375:14:E: element "EMPHASIS" undefined
[14:08] <sienap> jade:gst-plugin-writers-guide.sgml:379:19:E: element "ITEMIZEDLIST" undefined
[14:08] <sienap> jade:gst-plugin-writers-guide.sgml:380:17:E: element "LISTITEM" undefined
[14:08] <sienap> jade:gst-plugin-writers-guide.sgml:381:8:E: element "PARA" undefined
[14:08] <sienap> lot of this kind of crap while building gstreamer
[14:08] <sienap> howeve r:)
[14:08] <wtay> that's ok
[14:09] <sienap> rebuilding register
[14:09] <sienap> wtay he.. it delays the build pretty much..
[14:09] <sienap> ** CRITICAL **: file mpeg1parse.c: line 843 (plugin_init): assertion `plugin != NULL' failed.
[14:09] <sienap> INFO(11517:-1):gst_plugin_load_absolute:368: loading plugin "/usr/local/lib/gst/libmpeg_play.so"...
[14:09] <sienap> ** CRITICAL **: file gstmpeg_play.c: line 357 (plugin_init): assertion `plugin != NULL' failed.
[14:09] <sienap> INFO(11517:-1):gst_plugin_load_absolute:368: loading plugin "/usr/local/lib/gst/libmpeg1encoder.so"
[14:09] <wtay> that's ok too
[14:09] <sienap> A LOT of this with gstreamer-register
[14:10] <sienap> he
[14:10] <sienap> jesus :)
[14:10] <sienap> ** CRITICAL **: file system_encode.c: line 500 (plugin_init): assertion `plugin != NULL' failed.
[14:10] <sienap> ** CRITICAL **: file gstmpegaudio.c: line 184 (plugin_init): assertion `plugin != NULL' failed.
[14:10] <sienap> ** CRITICAL **: file mpeg1types.c: line 78 (plugin_init): assertion `plugin != NULL' failed.
[14:10] <sienap> ** CRITICAL **: file vcdsrc.c: line 374 (plugin_init): assertion `plugin != NULL' failed.
[14:10] <sienap> ** CRITICAL **: file rtjpeg.c: line 38 (plugin_init): assertion `plugin != NULL' failed.
[14:10] <sienap> Segmentation fault (core dumped)
[14:10] <sienap> [root at Oegaah gstreamer]# 
[14:10] <sienap> gstmediaplay
[14:10] <sienap> :)
[14:11] <wtay> ?
[14:11] <sienap> gstmediaplay segfaults....
[14:12] <wtay> hmm
[14:12] <sienap> where there more bugs in the yesterday build ?
[14:12] <rdj> yo sienap
[14:12] <sienap> that were fatal ?
[14:12] <sienap> hej rdj!!!
[14:12] <wtay> nope
[14:12] <sienap> hmmm..
[14:12] <sienap> then ehm yeah ;)
[14:13] <sienap> will do an ltrace
[14:13] <sienap> if xmms-test is compiled >:)
[14:13] <rdj> sienap: I HAD nautilus-gstreamer working ;)
[14:13] <sienap> wanna see the cpu usage
[14:13] <sienap> rdj he that already is great
[14:13] <sienap> how so the HAD ?
[14:13] <rdj> but it seems like oaf / bonobo / nautilus got a bit screwed on my machine now so it won't start most of the time
[14:13] <sienap> btw rdj i managed to compile file-sel however when using via glade i don't get anything
[14:13] <sienap> rdj send it to me
[14:13] <sienap> i'll check it..
[14:13] <rdj> sienap: then you compiled it wrong? :)
[14:14] <sienap> rdj he :)
[14:14] <sienap> how should i compile it ?
[14:14] <rdj> sienap: hold on
[14:15] <rdj> I'll make a tar.gz, but promise me you won't get angry if it screws up your setup too ;)
[14:15] <sienap> hmmx
[14:16] <sienap> xmms-test also segfaults..
[14:16] <sienap> rdj he it is ok :)
[14:16] <sienap> i expermentate with it everyday :)
[14:16] <sienap> wtay there is something wrong ...
[14:16] <rdj> ok one moment
[14:17] <rdj> run autogen.sh first
[14:17] <sienap> me not that stupid :)
[14:17] <sienap> can you also make an screenshot of the url..
[14:17] <wtay> hmm
[14:17] <rdj> screenshot of url?
[14:18] <rdj> excuse me? :)
[14:18] <sienap> ehm
[14:18] <sienap> he :)
[14:18] <sienap> ehm of the file sel
[14:18] <sienap> `/dcc
[14:18] <rdj> oh, the filesel
[14:19] <rdj> well, it looks just like the one I made, except it's a bonobo object
[14:19] <rdj> but ok
[14:19] <sienap> he
[14:19] <sienap> damn :)
[14:19] <sienap> really funny :)
[14:19] <sienap> that you and miguel started something like that at the same time
[14:20] <Uraeus> did you guys know that miguel is currently making a C# compiler?
[14:20] <sienap> write(2, "\n", 1
[14:20] <sienap> )                       = 1
[14:20] <sienap> open("/usr/local/lib/gst/librtjpeg.la", O_RDONLY|O_NONBLOCK|O_DIRECTORY) = -1 ENOTDIR (Not a directory)
[14:20] <sienap> open("/usr/local/lib/gst/librtjpeg.a", O_RDONLY|O_NONBLOCK|O_DIRECTORY) = -1 ENOTDIR (Not a directory)
[14:20] <sienap> open("/usr/local/lib/gst/libvideosink.so.0.0.0", O_RDONLY|O_NONBLOCK|O_DIRECTORY) = -1 ENOTDIR (Not a directory)
[14:20] <sienap> open("/usr/local/lib/gst/libvideosink.so.0", O_RDONLY|O_NONBLOCK|O_DIRECTORY) = -1 ENOTDIR (Not a directory)
[14:20] <sienap> open("/usr/local/lib/gst/libvideosink.so", O_RDONLY|O_NONBLOCK|O_DIRECTORY) = -1 ENOTDIR (Not a directory)
[14:20] <sienap> stat("/usr/local/lib/gst/libvideosink.so", {st_mode=S_IFREG|0755, st_size=568235, ...}) = 0
[14:20] <sienap> open("/usr/local/lib/gst/libvideosink.so", O_RDONLY) = 5
[14:20] <sienap> read(5, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0P4\0\000"..., 1024) = 1024
[14:20] <sienap> fstat(5, {st_mode=S_IFREG|0755, st_size=568235, ...}) = 0
[14:20] <sienap> old_mmap(NULL, 60160, PROT_READ|PROT_EXEC, MAP_PRIVATE, 5, 0) = 0x412b6000
[14:20] <sienap> mprotect(0x412c4000, 2816, PROT_NONE)   = 0
[14:21] <sienap> old_mmap(0x412c4000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 5, 0xd000) = 0x412c4000
[14:21] <sienap> close(5)                                = 0
[14:21] <sienap> mprotect(0x412b6000, 57344, PROT_READ|PROT_WRITE) = 0
[14:21] <sienap> mprotect(0x412b6000, 57344, PROT_READ|PROT_EXEC) = 0
[14:21] <sienap> --- SIGSEGV (Segmentation fault) ---
[14:21] <sienap> +++ killed by SIGSEGV +++
[14:21] <sienap> [root at Oegaah gstreamer]# 
[14:21] <sienap> it seems to crash on that wtay..
[14:21] <sienap> he :)
[14:21] <wtay> run with --gst-debug-mask=-1
[14:21] <rdj> sienap: it does resemble mine, doesn't it ;)
[14:22] <wtay> Uraeus: cool
[14:22] <sienap> wtay one mom
[14:22] <Uraeus> wtay: you like C#?
[14:22] <sienap> checking for GDK_PIXBUF - version >= 0.10.0... 
[14:23] <sienap> *** 'gdk-pixbuf-config --version' returned 0.9.0, but the minimum version
[14:23] <sienap> *** of GDK_PIXBUF required is 0.10.0. If gdk-pixbuf-config is correct, then 
[14:23] <sienap> s
[14:23] <sienap> *** best to upgrade to the required version.
[14:23] <sienap> *** If gdk-pixbuf-config was wrong, set the environment variable GDK_PIXBUF_
[14:23] <sienap> IG
[14:23] <sienap> *** to point to the correct copy of gdk-pixbuf-config, and remove the file
[14:23] <sienap> *** config.cache before re-running configure
[14:23] <sienap> no
[14:23] <sienap> configure: error: *** gdk_pixbuf 0.10.0 or better is required.
[14:23] <sienap> 0.10.1
[14:23] <sienap> mwha
[14:23] <sienap> rdj dat is lame :)
[14:23] <sienap> ik heb 0.10.1
[14:23] <sienap> :)
[14:23] <wtay> hmm
[14:23] <sienap> wtay ok looking at it now
[14:24] <rdj> sienap: try an mp3 or something in my viewer, it worked yesterday
[14:24] <sienap> still the same stuff
[14:24] <sienap> rdj mwha not really compiling huh
[14:24] <sienap> i've got gdkpixbuf 0.10.1
[14:24] <rdj> sienap: oh...
[14:24] <sienap> he ik snappu tal
[14:24] <sienap> mom
[14:24] <sienap> eens zien nu
[14:24] <rdj> you might have an old gdk-pixbuf-config
[14:25] <rdj> ok
[14:25] <sienap> correct :)
[14:25] <sienap> i had 2 installed :)
[14:25] <sienap> removed the wrong one
[14:25] <sienap> wtay any suggestion ?
[14:25] <rdj> oh fuck waiyt
[14:25] <sienap> for the gstreamer problem
[14:26] <rdj> it won't compile
[14:26] <sienap> EVERY gstreamer app crashes..
[14:26] Action: rdj made a mistake
[14:26] <sienap> he
[14:26] <rdj> wait
[14:26] <wtay> sienap: gstreamer-register too?
[14:27] <rdj> remove the x on line 104 in nautilus-gstreamer-content-view.c
[14:27] <rdj> then it will compile
[14:27] <rdj> I accidentally hit my keyboard I assume
[14:27] <sienap> he no
[14:27] <sienap> he will do
[14:28] <sienap> wtay yes.. still crashes
[14:28] <wtay> what does -register output?
[14:28] Action: rdj does not understand why my viewer brings down nautilus, without me changing anything
[14:28] <sienap> rdj still says i've got gdkpixbuf 0.9 instead of 0.10.1
[14:29] <rdj> code I tested yesterday does not work anymore
[14:29] <sienap> ** CRITICAL **: file gstmpegaudio.c: line 184 (plugin_init): assertion `plugin != NULL' failed.
[14:29] <sienap> INFO(1760:-1):gst_plugin_load_absolute:368: loading plugin "/usr/local/lib/gst/libmpeg1types.so"...
[14:29] <sienap> ** CRITICAL **: file mpeg1types.c: line 78 (plugin_init): assertion `plugin != NULL' failed.
[14:29] <sienap> INFO(1760:-1):gst_plugin_load_absolute:368: loading plugin "/usr/local/lib/gst/libvcdsrc.so"...
[14:29] <sienap> ** CRITICAL **: file vcdsrc.c: line 374 (plugin_init): assertion `plugin != NULL' failed.
[14:29] <sienap> INFO(1760:-1):gst_plugin_load_absolute:368: loading plugin "/usr/local/lib/gst/librtjpeg.so"...
[14:29] <sienap> ** CRITICAL **: file rtjpeg.c: line 38 (plugin_init): assertion `plugin != NULL' failed.
[14:29] <rdj> sienap: rm config.cache?
[14:29] <sienap> INFO(1760:-1):gst_plugin_load_absolute:368: loading plugin "/usr/local/lib/gst/libvideosink.so"...
[14:29] <sienap> Segmentation fault
[14:29] <sienap> for every plugin..
[14:29] <sienap> he ;)
[14:29] <sienap> hmm 
[14:29] <sienap> i was looking at the wrong console 
[14:29] <sienap> i guess
[14:29] <sienap> let's see
[14:29] <wtay> so libvideosink causes the crash
[14:29] <sienap> if yo usay so..
[14:30] <sienap> btw every plugin gives the critical
[14:30] <sienap> however guess you don't want me to paste that >:)
[14:30] <wtay> yeah, that because you did make install
[14:30] <wtay> it now tries to load the plugins from the build dir *and* /usr/local/lib/gst
[14:30] <sienap> he
[14:30] <sienap> ok
[14:30] <sienap> how to fix it ?
[14:30] <rdj> sienap: lemme know ok, I expect it to crash hard
[14:31] <sienap> checking for GDK_PIXBUF - version >= 0.10.0... ./configure: /usr/bin/gdk-pixbuf-config: No such file or directory
[14:31] <sienap> ./configure: /usr/bin/gdk-pixbuf-config: No such file or directory
[14:31] Last message repeated 3 time(s).
[14:31] <sienap> no
[14:31] <sienap> he ;)
[14:31] <sienap> mom
[14:31] <wtay> sienap: remove /usr/local/lib/gst/libvideosink.so
[14:31] <sienap> [root at Oegaah nautilus-gstreamer]# cp /usr/local/bin/gdk-pixbuf-config /usr/bin/
[14:31] <wtay> and do make clean in plugins/videosink
[14:31] <sienap> retrying
[14:31] <sienap> wtay will do
[14:32] <sienap> why a make clean ?
[14:32] <wtay> to remove the .so file
[14:32] <wtay> only in plugins/videosink
[14:33] <sienap> he
[14:33] <sienap> did
[14:33] <sienap> it seems to work now
[14:33] <rdj> hmmmm it works here again now
[14:33] <rdj> playing an mp3 with it now
[14:33] <sienap> NFO(12451:-1): Initializing GStreamer Core Library
[14:33] <sienap> INFO(12451:-1): CPU features: (00000000) MMX 
[14:33] <sienap> Xlib:  extension "XVideo" missing on display ":0.0".
[14:33] <sienap> ** ERROR **: GstElement: error in element 'disk_src': mmapping file
[14:33] <sienap> aborting...
[14:33] <sienap> Aborted (core dumped)
[14:34] <sienap> [root at Oegaah test]# 
[14:34] <rdj> I'm sure it will crash next time I try it
[14:34] <sienap> not great wtay ;)
[14:35] <wtay> hmm
[14:35] <wtay> xmmstest?
[14:36] <sienap> no
[14:36] <sienap> gstmediaplay
[14:36] <sienap> xmms test is even more funny :)
[14:36] <rdj> weird stuff, if I don't try it for a while, then start it, it works
[14:36] <rdj> after that it brings down nautilus
[14:36] <rdj> and my plugin doesn't even start
[14:36] <sienap> Song Info (text:length):ELF:-1
[14:36] <sienap> Segmentation fault
[14:36] <sienap> what is xmmstest
[14:36] <sienap> [root at Oegaah test]# ltrace ./xmmstest
[14:36] <sienap> "./xmmstest" is not an ELF binary object
[14:36] <sienap> and this is funny :)
[14:37] <wtay> heh
[14:37] <wtay> yeah ltrace doesn't work on libtool binaries
[14:37] <sienap> he :)
[14:37] <sienap> however..
[14:37] <wtay> use libtool strace ./xmmstest
[14:38] <sienap> wait a min
[14:38] <sienap> just did a make clean / make on the test dir
[14:39] <sienap> rdj: Compile error --------------------------------
[14:39] <sienap> gstplay.o: In function `gst_play_init':
[14:39] <sienap> gstplay.o: In function `gst_play_set_uri':
[14:39] <sienap>  /usr/src/nautilus-gstreamer/src/gstplay.c:269: undefined reference to `gst_pipeline_add_src'
[14:39] <sienap>  /usr/src/nautilus-gstreamer/src/gstplay.c:271: undefined reference to `gst_pipeline_autoplug'
[14:39] <wtay> ouch, that's old
[14:39] <sienap> he :))))
[14:39] <sienap> rdj cvsup the new gstreamer :)
[14:39] <sienap> and start over :)
[14:40] <sienap>  www.sourceforge.net/projects/gstreamer
[14:40] <sienap> rdj 0.1.1 is months old already
[14:40] <sienap> :)
[14:40] <sienap> you really don't want it :)
[14:41] <sienap> and i don't have it installed >:)
[14:41] <wtay> it 0.1.0 even
[14:41] <wtay> s/it/it's
[14:44] <sienap> he i think your file selector is more cute rdj :)
[14:44] <sienap> he :)
[14:44] <sienap> sick :)
[14:44] <sienap> also this one needs bonobo
[14:44] <sienap> it is overkill
[14:47] <sienap> ok
[14:47] <sienap> wtay
[14:47] <sienap> what i had to do on xmmstest ?
[14:47] <wtay> try to find out why it segfaults...
[14:48] <sienap> he
[14:48] <sienap> :)
[14:48] <sienap>  libtool strace ./xmmstest
[14:48] <sienap> this i mean
[14:49] <sienap> he :)
[14:50] <sienap> rdj:
[14:50] <sienap> [root at Oegaah nautilus-gstreamer]# cat AUTHORS 
[14:50] <sienap> Ian McKellar <ian at eazel.com>
[14:50] <sienap> sure about that ? :)
[14:51] <sienap> RDJJJJJJJJJ
[14:52] <sienap> wthe
[14:52] <sienap> wtay support me some how
[14:53] <sienap> he
[14:53] <sienap> where is rdj
[14:53] <wtay> dunno what goes on there
[14:53] <wtay> try gstreamer-launch fakesrc ! kafesink
[14:53] <sienap> wtay he :)
[14:53] <wtay> try gstreamer-launch fakesrc ! fakesink
[14:53] <sienap> mom
[14:54] <rdj> I'm back
[14:54] <sienap> fakesrc: ******* (fakesrc0:src)> 
[14:54] <sienap> fakesink: ******* (fakesink0:sink)< 
[14:54] <sienap> fakesrc: ******* (fakesrc0:src)> 
[14:54] <sienap> fakesink: ******* (fakesink0:sink)< 
[14:54] <sienap> fakesrc: ******* (fakesrc0:src)> 
[14:54] <sienap> works i gueszx
[14:54] <wtay> good
[14:54] <sienap> let me try a mpeg playing with it..
[14:54] <wtay> try gstreamer-launch disksrc location=some.mp3 ! mp3parse ! mpg123 ! osssink
[14:54] <sienap> how could i set the source for disksrc ?
[14:54] <rdj> I'll upgrade to cvs then, but it won't make a difference
[14:55] <wtay> location=...
[14:55] <sienap> aah location
[14:55] <sienap> :)
[14:55] <sienap> rdj he ?
[14:55] <Uraeus> hate hate hate
[14:55] <Uraeus> hate hate
[14:55] <Uraeus> I hate women
[14:55] <rdj> sienap: it won't make a difference, the crash is not gstreamer related\
[14:55] <wtay> Uraeus: oh oh
[14:55] <rdj> since it works sometimes
[14:56] <Uraeus> wtay: seems things will not work out between myself and the horse riding girl
[14:56] <sienap> rdj he i want to be able to compile it
[14:56] <sienap> i can't right now
[14:57] <rdj> sienap: I know, will upgrade to cvs this afternoon
[14:57] <sienap> ureaus why not dude ?
[14:57] <sienap> had a fight or something ?
[14:57] <wtay> Uraeus: pfff, those girls don't know what a fine man you are...
[14:57] <sienap> wtay LOL :)
[14:57] <sienap> ureaus is media god for sure :)
[14:57] <rdj> girls never appreciate coding skills :)
[14:57] <sienap> he >:)
[14:58] <Uraeus> well, maybe I am overreacting maybe am just lost a battle not a war or something
[14:58] <wtay> Uraeus: maybe you shouldn't give up that fast...
[14:58] Action: rdj will start the war again tomorrow
[14:58] <rdj> for a nice indonesian woman
[14:59] <sienap> ureaus tell us what happened
[14:59] Action: rdj takes a deep breath
[14:59] <wtay> Uraeus: what happened?
[14:59] <sienap> rdj lol :)
[14:59] <Uraeus> well, I called her asking her out
[14:59] <sienap> ureaus and she answered "no" ?
[14:59] <Uraeus> and got the I am busy all week and next weekend reply
[14:59] <Uraeus> but maybe some other time
[14:59] <wtay> Uraeus: and is she busy?
[15:00] <sienap> ureaus so ask her over 2 weeks again :)
[15:00] <Uraeus> yes, she is, but then again if she were interested I would guess she would make time
[15:00] <sienap> ureaus maybe she isn't intrested at that kind of level YET
[15:00] <sienap> :)
[15:00] <sienap> so be patient :)
[15:01] <sienap> wtay:
[15:01] <sienap> [root at Oegaah src]# gstreamer-launch disksrc location="/root/mp3/atr_deleteyself/01_20-_20Atari_20Teenage_20Riot_20-_20Start_20the_20Riot.mp3" ! mp3parse ! mpg123 ! osssink
[15:01] <sienap> INFO(9978:-1): Initializing GStreamer Core Library
[15:01] <sienap> INFO(9978:-1): CPU features: (00000000) MMX 
[15:01] <sienap> RUNNING pipeline
[15:01] <sienap> works
[15:01] <Uraeus> sienap: I hate being patient
[15:01] <sienap> ureaus he me too :)
[15:01] <wtay> sienap: does gstreamer-register work now?
[15:02] <sienap> happy thing for me was i thought it was hopeless for me and nathalie however she is sitting next to me right now >:)
[15:02] <sienap> wtay yes..
[15:02] <sienap> wtay however gstmediaplay / xmmstest say bye bye
[15:02] <wtay> sienap: so just gstmediaplay crashes
[15:02] <sienap> and xmmstest
[15:02] <Uraeus> hmm, guess I join some of my friends for a Diablo2 lan-party now instead, killing seem to ease my frustration level
[15:02] <sienap> gstmediaplay says:
[15:02] <sienap> Xlib:  extension "XVideo" missing on display ":0.0".
[15:02] <sienap> ** ERROR **: GstElement: error in element 'disk_src': mmapping file
[15:02] <sienap> aborting...
[15:02] <sienap> Aborted (core dumped)
[15:02] <sienap> [root at Oegaah src]# 
[15:03] <sienap> and xmmstest i can't get a clue on..
[15:03] Nick change: Uraeus -> Ura_Diablo
[15:04] <wtay> INFO(29472:-1): Initializing GStreamer Core Library
[15:04] <wtay> INFO(29472:-1): CPU features: (c1c7f9ff) MMX 3DNOW MMXEXT 
[15:04] <wtay> Song Info (text:length): üÿÑÏ  xüÑ¿    ýÿ¿:-1
[15:04] <wtay> Segmentation fault
[15:04] <sienap> hmm
[15:04] <sienap> >:)
[15:04] <sienap> wtay gstmediaplay works
[15:04] <wtay> ah
[15:04] <sienap> trynig to look up why it didn't
[15:05] <sienap> 01_People of the Sun.mp3: Could not find the beginning of MPEG data
[15:05] <sienap> think that explains a lot :)
[15:05] <sienap> it crashes on that..
[15:05] <sienap> ok gstmediaplay works
[15:05] <sienap> xmmstest doesn't
[15:05] <wtay> doesn't work here either
[15:05] <sienap> he
[15:05] <sienap> 17167 root       4   0  6468 6468  4248 S    20.4  6.7   0:03 gstmediaplay
[15:05] <sienap> he
[15:06] <sienap> xmms had an huge memory print / less cpu
[15:06] <sienap> mpg123 / plaympeg an small memory prnit / more cpu
[15:06] <sienap> gstmediaplay has both >:)
[15:06] <sienap> HMMMM
[15:06] Action: rdj uses 3dnow crap for decoding mp3, almost no cpu usage visible :)
[15:06] <rdj> it says 0.1%
[15:06] <sienap> >:)
[15:06] <sienap> he
[15:07] <sienap> rdj however if you port your view to the current cvs thing
[15:07] <sienap> i will be able to use it
[15:07] <rdj> sienap: yes?
[15:07] <rdj> I will
[15:07] <rdj> this afternoon
[15:07] <sienap> rdj he ok :)
[15:07] <sienap> wtay i am wondering how much cpu xmmstest will use on decoding mp3 ..
[15:07] <sienap> since it uses mpg123 thing from xmms
[15:07] <sienap> that uses 0.7 % cpu here
[15:08] <sienap> which is pretty nothing on a p233
[15:08] <wtay> yeah...
[15:08] <wtay> trying to fix xmmstest...
[15:08] Action: rdj is out for a while
[15:08] <sienap> rdj ok
[15:08] <sienap> cu later
[15:08] <rdj> have fun hacking ;)
[15:08] <rdj> later
[15:09] <wtay> sienap: hmm, xmmstest uses the mikmod plugin :)
[15:09] <sienap> he
[15:09] <sienap> ok
[15:09] <sienap> changing it
[15:09] <sienap>   src = gst_elementfactory_make("XMMS_INPUT_mikmod_player_1.2.4", "xmms_plugin");   
[15:09] <sienap> lol :)
[15:10] <sienap> you still have to make an detection system for that :)
[15:10] <wtay> no
[15:10] <sienap> he
[15:10] <sienap> bitch :)
[15:10] <sienap> what is the use of the xmms wrapper if you don't use it anyway ?
[15:10] <wtay> fame and glory
[15:10] <sienap> *smash you*
[15:11] <wtay> hmm, xmms is broken
[15:11] <sienap> ok
[15:12] <sienap> xmmstest mpg123 gives nothing else then ehm jeah *RUIS*
[15:12] <wtay> it doesn't set the capabilities (sample rate, depth etc...)
[15:12] <sienap> home comes so ?
[15:12] <wtay> historical reasons :)
[15:12] <sienap> 24848 root       0   0  6124 6124  4080 S     0.8  6.4   0:02 lt-xmmstest
[15:13] <sienap> 0.8 cpu
[15:13] <sienap> man
[15:13] <wtay> ah
[15:13] <sienap> and as much as ram as gstmediaplay :)
[15:13] <sienap> however gstmediaplay sucks out 20% cpu
[15:13] <wtay> yaeh, 4080 from gtk
[15:13] <sienap> on my p233
[15:13] <sienap> he detail :)
[15:13] <sienap> however
[15:13] <sienap> 0.8 cpu
[15:13] <sienap> and the native gstreamer plugin 20%
[15:13] <sienap> :)
[15:14] <sienap> however it was decoding *ruis*
[15:14] <sienap> mwha
[15:14] <sienap> fix it :)
[15:14] <wtay> shut up
[15:14] <sienap> he :)
[15:14] <wtay> :)
[15:15] <wtay> doing the mad plugin first
[15:15] <sienap> i want a 0.8% cpu usage for gstmediaplay too !!!!!!! :)
[15:15] <wtay> buy a faster CPU :)
[15:15] <sienap> he
[15:15] <sienap> he okie ;)
[15:15] <sienap> i just use xmms >:)
[15:15] <wtay> yeah
[15:15] <sienap> for the quality of gstreamer huh >:)
[15:16] Ura_Diablo (Uraeus at c224s9h5.upc.chello.no) left irc: Ping timeout for Ura_Diablo[c224s9h5.upc.chello.no]
[15:18] <wtay> cool, het hagelt hier volle bak
[15:18] <sienap> he :)
[15:18] <sienap> hier een klein beetje net
[15:18] <sienap> dus jij zit daar stiekem op zware materieele schade te hopen >:)
[15:20] <wtay> gewoon effe de vensters dichtdoen..
[15:20] <sienap> >:)
[15:26] <sienap> pwd
[15:26] <sienap> hmm
[15:26] <sienap> ;)
[15:35] <sienap> hmm
[15:40] <sienap> ls
[15:40] <sienap> aah bah
[15:40] <sienap> ik blijf hier maar in tikken :)
[15:42] <sienap> <Cactus> sienap: wtay's bonobo-media-video problem should be solved
[15:42] <sienap> cool :)
[15:43] <wtay> no it isn't
[15:44] <sienap> he
[15:45] <sienap> :)
[15:45] <sienap> he is cumming
[15:45] <sienap> >:)
[15:45] <wtay> oh cool
[15:45] Cactus (cactus at port059.bp1.telnet.hu) joined #gstreamer.
[15:45] <Cactus> hi
[15:45] <sienap> there he is
[15:45] <wtay> aha!
[15:45] <Cactus> wtay: so isn't it solved?
[15:45] <sienap> our bonobo-bast
[15:45] <Cactus> hi btw:)
[15:45] <wtay> nope, I had the exact code in there
[15:45] <sienap> hi :)
[15:45] <Cactus> wtay: hmmm
[15:46] <Cactus> wtay: it certainly worked for me
[15:46] <wtay> oh
[15:46] <Cactus> but I haven't tried it with the currently shipped media-player
[15:46] <wtay> with a button?
[15:46] <Cactus> Are you sure it even checks for the Video interface?
[15:46] <Cactus> wtay: yes
[15:46] <wtay> ah, maybe I should try the CVS version
[15:46] <wtay> Cactus: my gst_video_control method is called
[15:46] <Cactus> wtay: the media player is not checked into CVS yet
[15:47] <wtay> ok
[15:47] <Cactus> wtay: http://cactus.rulez.org/files/bonobo-video.png (this is done by modifiing the vorbis provider)
[15:48] <wtay> hmm, ok
[15:48] <Cactus> but I really should go now and have lunch
[15:48] <Cactus> catch you later
[15:48] <wtay> can you take a look at my current code?
[15:48] Nick change: Cactus -> CacFood
[15:48] <sienap> later
[15:51] <wtay> cya
[16:01] sienap (synap at ipc379c0e6.dial.wxs.nl) left irc: sienap has no reason
[16:11] CacFood (cactus at port059.bp1.telnet.hu) left irc: Tank, I need an exit
[16:35] thomas (thomas at urgent.rug.ac.be) joined #gstreamer.
[16:35] <thomas> wtay: are you there ?
[16:35] <wtay> yup
[16:36] <thomas> ok... I think I finished the passhtrough filter plus the script
[16:36] <thomas> to test it I made a swapchannel plugin
[16:36] <thomas> not very useful but it works
[16:36] <wtay> neat
[16:36] <thomas> but I wanted to test it against 8 bit audio
[16:36] <thomas> problem is, I can't get any good 8 bit audio
[16:36] <wtay> heh
[16:36] <thomas> I don't know how it works, since sox says there is no such thing as signed 8 bit
[16:37] <thomas> ah by the way, I solved the function problem as this :
[16:37] <thomas> include a "filter.func"
[16:37] <thomas> and this is the body of the function
[16:37] <thomas> and I made the function call with num_samples
[16:37] <thomas> instead of num_bytes
[16:37] <thomas> that's the best I could think of
[16:37] <wtay> ok
[16:37] <thomas> How did you test 8 bit audio ?
[16:38] <wtay> I didn't
[16:38] <thomas> so you're guessing it works ?
[16:38] <wtay> yup
[16:38] <thomas> ok...
[16:38] <thomas> then I don't have to worry for now either
[16:38] <wtay> neeh
[16:38] <thomas> or maybe...
[16:38] <thomas> I could do an 16 to 8 bit plugin
[16:38] <thomas> though that would be more difficult
[16:39] <wtay> we should have a general audio resampler/scaler plugin
[16:39] <thomas> that's for later
[16:39] <thomas> I don't understand enough yet
[16:39] <thomas> so, would it make sense to put the passthrough + script in CVS ?
[16:39] <wtay> sure
[16:39] <thomas> how do I do that then ?
[16:39] <wtay> are you planning to do a lot of work on gstreamer?
[16:39] <thomas> I'm not sure yet
[16:40] <thomas> I mean, yeah, I'd like to keep at it
[16:40] <thomas> but I have a schedule to follow
[16:40] <wtay> omega can give you CVS access
[16:40] <wtay> write access that is
[16:40] <thomas> Maybe it's best if I see if I can understand it enough
[16:40] <thomas> my goal is to get a mixer plugin working as you know
[16:40] <wtay> best way to understand is to use it :)
[16:40] <thomas> but since that's got to work with variable inputs
[16:40] <thomas> I have some learning to do
[16:40] <thomas> especially
[16:40] <thomas> since it's so hard to debug
[16:40] <wtay> true
[16:41] <wtay> printf is my friend
[16:41] <thomas> yeah... mine for years as well
[16:41] <thomas> but it does feel stupid !
[16:41] <thomas> You'd think after 30 years of programming languages
[16:41] <wtay> yeah, I know the feeling :)
[16:41] <thomas> people would've come up with something smarter
[16:41] <thomas> *ah well*
[16:41] <thomas> ok...
[16:41] <wtay> my previous RH install had a working gdb...
[16:41] <thomas> what do you mean, "working" ?
[16:42] <thomas> with threads as well ?
[16:42] <wtay> the one I have now is useless with threads
[16:42] <wtay> yeah
[16:42] <thomas> I didn't know it could
[16:42] <thomas> I can't get the hang of gdb anyway
[16:42] <wtay> it can
[16:42] <wtay> gdb is kinda awkward..
[16:42] <thomas> I started out on ms-dos for C, a borrowed version
[16:42] <thomas> of microsoft c 6.0
[16:42] <thomas> but it had a data error in the iDE
[16:42] <thomas> so I only did command-line compiling
[16:43] <thomas> and now on linux
[16:43] <thomas> I only use it for bt'ing
[16:43] <wtay> nothing changes :)
[16:43] <thomas> too much to learn
[16:43] <wtay> anyway, send me the patch and I'll commit it
[16:43] <thomas> how do I get the sample rate of the input data ?
[16:43] <thomas> just like the gst_caps ?
[16:43] <wtay> yup
[16:43] <thomas> ok...
[16:44] <thomas> ah would you check the copy script first ?
[16:45] <wtay> I'll check it
[16:45] <thomas> ok I'll put it online somewhere
[16:50] Action: thomas is away - Automatically set away. - messages will be saved.
[16:51] <thomas> ok...
[16:51] <thomas> it's at http://urgent.rug.ac.be/thomas/gst
[16:51] <thomas> just put it in plugins/filters and extract
[16:51] <thomas> there's a simple readme
[16:51] <thomas> don't know where the script should be put in a final version
[16:53] <wtay> ok, sec..
[16:55] <thomas> each plugin object has it's own memory space right ?
[16:55] <wtay> no
[16:55] <thomas> if I use statics, they don't mix across different actual plugins ?
[16:56] <wtay> yup
[16:56] <wtay> static means that the code is not exported from the .so file
[16:56] <wtay> but they all use the same address space
[16:56] <thomas> so if i have two volenv plugins active
[16:56] <thomas> they will share their static variable ?
[16:56] <wtay> never
[16:57] <thomas> ok...
[16:57] <thomas> just checkin
[16:57] <wtay> only the plugin_init should be exported 
[16:58] <thomas> so this way I can set a static timer on the volenv plugin
[16:58] <thomas> counting along to see how far it is in the stream ?
[16:58] <wtay> hmm, sec... I'm not sure abaout that
[16:59] <wtay> static is shared over all mmaps of the .so file AFAIK
[16:59] <wtay> so you can't do that
[16:59] <wtay> all element specific data should go into the instance
[16:59] <wtay> the plugin looks good...
[17:01] <wtay> I'm going to edit the .README file a bit.. users should also modify configure.in...
[17:01] <thomas> what do you need to change there ?
[17:01] <wtay> add a line so that the Makfile is created
[17:01] <wtay> only the Makefile.am goes into CVS
[17:02] <thomas> can I add the line automatically in the script ?
[17:02] <wtay> hmm, yes...
[17:02] <wtay> I'd rather not do that...
[17:02] <thomas> I just ran two volenv's consecutively...
[17:03] <thomas> time runs twice as fast then
[17:03] <thomas> so static's are shared ;(
[17:03] <thomas> but that's probably a good thing for other stuff
[17:03] <wtay> what I suspected
[17:03] <thomas> ok...
[17:03] <thomas> a pointer then
[17:03] <wtay> in the object instance
[17:03] <thomas> or is there some standard way you use to save data ?
[17:03] <thomas> in the struct ?
[17:03] <wtay> yes
[17:03] <wtay> where the sinkpad etc live
[17:04] <thomas> what was the MataAudioRaw for in the struct ?
[17:04] <wtay> are you familiar with Gtk+ object model?
[17:04] <thomas> it's commented
[17:04] <thomas> no not yet
[17:04] <thomas> so I should read up on it first
[17:04] <wtay> historical reasons :)
[17:04] <wtay> it can go
[17:05] <thomas> so then I should pass the runtime along with the chain function ?
[17:05] <thomas> or can the chain function get it...
[17:05] <thomas> let's see
[17:05] <wtay> gst_pad_get_parent
[17:06] <thomas> I also added a warning for the caps pointer...
[17:06] <thomas> but you probably also have a better warning system
[17:06] <thomas> than printf i suppose
[17:06] <wtay> uhm no... g_warning maybe
[17:06] <wtay> but it's an error anyway
[17:07] <thomas> do you use standard doubles or a custom type ?
[17:09] <wtay> for what?
[17:09] <thomas> for the runtime for example
[17:09] <wtay> standard doubles if any
[17:14] <wtay> Copying filter boilerplate PassThrough to new filter MyTest...
[17:14] <wtay> Bareword "passthrough" not allowed while "strict subs" in use at (eval 1) line 1.
[17:14] <wtay> Can't do inplace edit: CVS is not a regular file.
[17:14] Last message repeated 2 time(s).
[17:14] <thomas> yeah...
[17:14] <thomas> that's because of the dir
[17:14] <thomas> I could get that out with a little more work
[17:14] <thomas> perl can't edit a dir like a file
[17:14] <wtay> and the Bareword error?
[17:14] <thomas> huh ?
[17:14] <thomas> what one ?
[17:14] <wtay> passthrough isn't renamed
[17:15] <thomas> how did you call it ?
[17:15] <thomas> the cmd line
[17:15] <wtay> ./filterstamp.sh PassThrough MyTest
[17:15] <thomas> that looks right
[17:15] z- (dave at host213-122-58-239.btinternet.com) joined #gstreamer.
[17:15] <thomas> and should work
[17:15] <thomas> I don't get the bareword error
[17:15] <wtay> it's just the filename that is not renamed here
[17:16] <wtay> inside the file looks ok
[17:16] <thomas> can you copy the bareword error ?
[17:16] <wtay> copy?
[17:16] <thomas> cut and paste it
[17:16] <thomas> so I can see it
[17:16] <wtay>  <wtay> Bareword "passthrough" not allowed while "strict subs" in use at (eval 1) line 1.
[17:17] <thomas> seems like perl...
[17:17] <thomas> is doing some install-specific things
[17:17] <thomas> but I can't tell what
[17:17] <wtay> the rename command gives that error...
[17:17] <thomas> ah ok...
[17:17] <thomas> there are better ways than rename
[17:17] <thomas> I just like that command
[17:18] <thomas> but that's weird...
[17:18] <thomas> so it's your shell giving that
[17:18] <thomas> which one do you use ?
[17:18] <wtay> maybe quote it?
[17:18] <wtay> bash
[17:18] <thomas> and is your rename a binary or some sort of shell script ?
[17:19] <wtay> a perl script
[17:19] <thomas> weird... mine's a binary
[17:19] <thomas> ok... I'll think of something else to rename it
[17:19] <thomas> the rest of it is ok ?
[17:19] <wtay> perfect
[17:23] z- (dave at host213-122-58-239.btinternet.com) left irc: Client Exiting
[17:24] <thomas> would a run time make sense for all sort of audio plugins ?
[17:24] <thomas> then I'd better add it to the passthrough
[17:25] <wtay> I don't think so
[17:26] <wtay> this should be handled by the core clock system anyway
[17:26] <thomas> is that working ?
[17:26] <wtay> nope
[17:27] <thomas> then I'll stick to mine for now
[17:27] <wtay> you should be able to say: my clock runs 2x slower and everything is slowed down then..
[17:27] <wtay> or -1 for that matter
[17:27] <thomas> ;)
[17:27] <thomas> really ?
[17:27] <wtay> sure
[17:27] <thomas> why would you use that ?
[17:27] <thomas> sorry, I see
[17:27] <wtay> play an mpeg backwards
[17:28] <thomas> I can see all sorts of problems with that
[17:28] <thomas> good luck
[17:28] <wtay> me too :)
[17:28] <wtay> anyway, I think DS can do it
[17:28] <thomas> who's that ?
[17:29] <wtay> DirectShow?
[17:29] <thomas> you're going to integrate that ?
[17:29] <thomas> I'm glad I don't do video
[17:30] <wtay> backward playback you mean?
[17:30] <thomas> no : I'm glad I don't need to know things about directshow
[17:30] <thomas> for now
[17:31] <wtay> or DirectMedia whatever they call it today...
[17:31] <thomas> ok... if I'm going to keep track of the levels as well
[17:31] <wtay> anyway you should be able to playback an mp3 backward too...
[17:31] <thomas> then it would make more sense to pass the object pointer to the chain function...
[17:32] <thomas> yeah backwards audio could be useful as well
[17:32] <wtay> the chain function has the pad as an arg
[17:32] <wtay> the parent of the pad is the element
[17:33] <wtay> yaeh, the chain function definatly needs the element as an arg
[17:35] <thomas> hmmm... what part of the function does inline speed up specifically ?
[17:35] <thomas> only the loop code ?
[17:36] <wtay> only the call overhead is eliminated with inlining
[17:37] <thomas> meaning...
[17:37] <thomas> it pretty much does the same thing as a macro def ?
[17:37] <wtay> yes
[17:37] <wtay> it's a glorified macro
[17:37] <thomas> does that work in standard c then ?
[17:37] <wtay> not sure
[17:37] <wtay> I think it's gcc specific...
[17:39] <wtay> ok, glib redefines it in glib.h
[17:40] <wtay> to __inline or __inline__ :)
[17:41] <thomas> ok...
[17:41] <thomas> back to the plugin...
[17:41] <thomas> I was thinking
[17:41] <thomas> that it would be faster
[17:41] <thomas> if I only calculate the level change for each block of samples
[17:41] <thomas> which is ok for me now
[17:41] <thomas> but how high can you set the length
[17:41] <thomas> of the buffer ?
[17:42] <wtay> unlimited
[17:42] <thomas> oops
[17:42] <thomas> not a good idea then
[17:42] <wtay> it depends on the peer element
[17:42] <thomas> uhm...
[17:42] <wtay> typically 4096 bytes
[17:42] <wtay> you don't know
[17:42] <thomas> can a stereo buffer be set to have an odd number of bytes ?
[17:42] <wtay> no
[17:42] <thomas> good !
[17:42] <wtay> by convention that is
[17:43] <thomas> that's a good convention
[17:43] <wtay> yeah, let's be reasonable...
[17:44] <wtay> I gotta get some food...
[17:45] <wtay> back in about an hour...
[17:45] <thomas> ok...
[17:45] Nick change: wtay -> wtay-eating
[17:45] <thomas> meanwhile I'll see
[17:45] <wtay-eating> ok
[17:45] <thomas> if i can get a signal to work
[17:45] <thomas> can I pass data to it ?
[17:45] <thomas> as in "set volume to"
[17:45] <wtay-eating> yup
[17:45] <wtay-eating> passthrough is in CVS btw
[17:46] <thomas> ok thanks
[18:23] <rdj> hmmm gstreamer-register from cvs segfaults on my machine
[18:23] <rdj> ** CRITICAL **: file gstavidecoder.c: line 697 (plugin_init): assertion `plugin != NULL' failed.
[18:23] <rdj> INFO(19050:-1):gst_plugin_load_absolute:368: loading plugin "/usr/lib/gst/libaudioscale.so"...
[18:23] <rdj> I get this for all the plugins
[18:26] <rdj> it starts failing with this one:
[18:26] <rdj> INFO(19058:-1):gst_plugin_load_all:233: loading plugins from /usr/lib/gst
[18:26] <rdj> INFO(19058:-1):gst_plugin_load_absolute:368: loading plugin "/usr/lib/gst/libwincodec.so"...
[18:26] <rdj> ** CRITICAL **: file gstwincodec.c: line 65 (plugin_init): assertion `plugin != NULL' failed.
[18:31] <rdj> what the hell can be causing this :)
[18:33] Cactus (cactus at port059.bp1.telnet.hu) joined #gstreamer.
[18:33] <Cactus> hi
[18:39] <rdj> hey Cactus
[18:41] Action: rdj gets lots of errors with gstreamer-register :(
[18:58] <thomas>  
[19:04] thomas (thomas at urgent.rug.ac.be) left irc: I'm outta here!
[19:05] Nick change: wtay-eating -> wtay
[19:05] thomas (thomas at urgent.rug.ac.be) joined #gstreamer.
[19:05] <wtay> Cactus: ?
[19:05] <Cactus> wtay: hi
[19:06] <rdj> does anyone know a sollution to my gstreamer-register problem?
[19:06] <rdj> compiling and installing worked without problems
[19:07] <wtay> rdj: yeah, don't do make install or rename the src dir
[19:07] <Cactus> do you guys know MPlayer? (http://mplayer.sourceforge.net/)
[19:07] <wtay> rdj: the plugins are loaded from /usr/local/ib/gst *and* the build dir
[19:07] Uraeus (Uraeus at c224s9h5.upc.chello.no) joined #gstreamer.
[19:07] <Cactus> it has this nice feature that if you have a supported video card (I think only Matrox G[24]00 is currently supported) it can do the YUV->RGB conversion in hardware
[19:08] <Uraeus> hi again
[19:08] <thomas> everyone: is there a good example of a plugin with signals implemented ?
[19:08] <wtay> Cactus: gstreamer does that too
[19:08] <Cactus> wtay: nice
[19:08] <wtay> thomas: sec..
[19:09] <wtay> thomas: osssink fires a signal
[19:09] <Uraeus> hi Cactus nice to see that you wanted to integrate bonobo-media and Gstreamer
[19:09] <wtay> Cactus: can I send you my current code so you can take a look at it?
[19:09] <Cactus> wtay: sure
[19:09] <Cactus> wtay: is it working?
[19:09] sienap (synap at ipc379c0e6.dial.wxs.nl) joined #gstreamer.
[19:10] <wtay> Cactus: I can't get the button to show...
[19:10] <sienap> hi all
[19:10] <Cactus> wtay: OK I'll check it
[19:10] <sienap> hmm this fight for fun with my girlfriend actually hurt my back :)
[19:10] <sienap> she has sharp nails or something
[19:11] <thomas> wtay: it has a signal, but it doesn't do anything
[19:11] <thomas> wtay: is the "marshaller function" responsible for handling it ?
[19:11] vektor (vektor at HSE-Kitchener-ppp3504954.sympatico.ca) left #gstreamer.
[19:11] <wtay> thomas: it fires when a new buffer is handed off to oss
[19:11] <thomas> wtay: does that mean it signals a reopen of the audio output ?
[19:11] <wtay> Cactus: I'm copying a tarball to gstreamer.net
[19:12] <wtay> thomas: for every buffer that is written to it
[19:13] <wtay> Cactus: gstreamer.net/gstreamer-stream-provider.tgz
[19:13] <sienap> thomas!
[19:13] <thomas> wtay: I see the emit... but I don't see where it catches it
[19:13] <Cactus> wtay: d/l'ing
[19:13] <sienap> cactus what actually are your plans for bonobo-media ?
[19:13] <sienap> only use gstreamer for it and code further media features in gstreamer
[19:14] <sienap> or what ?
[19:14] <thomas> sienap: hi ! you should send your girlfriend home if you want to get something done ;)
[19:14] <sienap> he
[19:14] <wtay> thomas: that's the point, the app can listen to it...
[19:14] <sienap> i won't get something done without her either :0
[19:14] <sienap> so who cares :)
[19:14] <sienap> she once a 2 a 3 weeks her so ehm actually happy she is her :)
[19:14] <sienap> however her victory sucks :)
[19:14] <wtay> thomas: do you want an example of an app listening for the signal too?
[19:14] <thomas> wtay: ok, so that's not a good example.  how do I get data in ?
[19:15] <thomas> wtay: yes
[19:15] <Cactus> sienap: I think since GStreamer can potentially support any multimedia content, bonobo-media could be eventually integrated into it
[19:15] <thomas> wtay: I need a way to set the expected volume and time
[19:15] <thomas> sienap: that's ok then
[19:15] <thomas> sienap: I see mine every 2-3 hours if it's up to her
[19:15] <wtay> thomas: oh a property you mean... sec...
[19:15] <thomas> (getting food)
[19:16] <wtay> thomas: mp3encode/lame has plenty of properties to set...
[19:16] <Cactus> wtay: I'm updating GStreamer itself first
[19:16] <wtay> Cactus: ok, you'll have to do make install because it links to gstplay.la
[19:18] <sienap> cactus your plans are great :)
[19:18] <sienap> thomas nice :)
[19:18] <sienap> thomas be happy :)
[19:18] <sienap> 3 weeks without girlie is furstration++
[19:18] <rdj> wtay: renamed the build dir, I already did a make install... it doesn't give errors anymore but it still coredumps\
[19:19] <sienap> rdj did an register update ?
[19:19] <sienap> rdj does gstreamer-register coredumps ?
[19:19] <rdj> register update?
[19:20] <rdj> I'm talking about that
[19:20] <sienap> that is core dumping
[19:20] <rdj> gstreamer-register dumps core
[19:20] <sienap> same problem as i had i guess..
[19:20] <sienap> ehm wtay what file it was again
[19:20] <sienap> about videosink..
[19:20] <thomas> wtay: either I'm blind or I can't find any signals
[19:20] <thomas> wtay: it is in plugins/mp3encode/lame/gstlame.c right ?
[19:21] <sienap> rdj let me look
[19:21] <rdj> sienap: ps. the nautilus plugin should compile if you replace gstplay* in the src dir with the ones in gstreamer/gstplay/
[19:21] <wtay> thomas: you want an example of an argument or a signal?
[19:22] <thomas> well...
[19:22] <rdj> from the gstreamer dir in cvs
[19:22] <sienap>  /usr/local/lib/gst 
[19:22] <thomas> wtay: i want a way to tell my plugin :
[19:22] <thomas> wtay: set the volume to 1 in 5 seconds' time
[19:22] <sienap> there is an videosink something there
[19:22] <wtay> thomas: you need an argument then
[19:22] <sienap> remove that
[19:22] <thomas> wtay: you mean no signals ?
[19:22] <rdj> sienap: me?
[19:22] <sienap> and do a makeclean at ./plugins/videosink/
[19:22] <sienap> rdj yes
[19:22] <rdj> ok
[19:23] <thomas> wtay: i'll try that
[19:23] <wtay> thomas: signals are used to notify the app, args to set proerties of a plugin
[19:24] <rdj> sienap: that should do it?
[19:24] <rdj> sienap: it did
[19:24] <rdj> thanks :)
[19:25] <sienap> np :)
[19:26] <sienap> rdj btw did you found the bug with your plugin ?
[19:26] <wtay> Cactus: ...
[19:26] <wtay> ** WARNING **: Filename : '/opt/data/armageddon1.mpg'
[19:26] <wtay> Gtk-CRITICAL **: file gtkstyle.c: line 514 (gtk_style_attach): assertion `window != NULL' failed.
[19:26] <wtay> Bonobo-WARNING **: Reference counting error: Attempts to release CORBA_Object associated with 'BonoboControl' which still has a reference count of 1
[19:26] <sienap> your nauti view that is
[19:26] <sienap> you have armageddon in .mpg ?
[19:26] <sienap> lekker grooooooooot :)
[19:26] <rdj> sienap: nope
[19:26] <sienap> rdj he damn
[19:26] <sienap> however compiling the thingy right now..
[19:26] <rdj> sienap: you replaced the 3 files?
[19:26] <sienap> rdj it is all a mather of clicking media types
[19:26] <sienap> rdj 6
[19:27] <sienap> also the .h's :)
[19:27] <rdj> 6??
[19:27] <sienap> main.o: In function `main':
[19:27] <sienap> collect2: ld returned 1 exit status
[19:27] <sienap> make[2]: *** [nautilus-gstreamer-content-view] Error 1
[19:27] <rdj> lemme see
[19:27] <sienap>  /usr/src/nautilus-gstreamer/src/main.c:21: undefined reference to `gst_media_play_new'
[19:27] <sienap>  /usr/src/nautilus-gstreamer/src/main.c:24: undefined reference to `gst_media_play_start_uri'
[19:27] <sienap> what deos it need to be ?
[19:27] <rdj> you shouldn't replace those :)
[19:28] <sienap> hmm
[19:28] <sienap> they don'tw ork :)
[19:28] <rdj> only gstplay.c gstplay.h and gstplayprivate.h
[19:28] <sienap> ah..
[19:28] <sienap> nice..
[19:28] <sienap> mom
[19:28] <thomas> wtay: if i want the plugin to recalculate stuff
[19:28] <thomas> wtay: based on the argument setting
[19:29] <thomas> wtay: how can I do that ?
[19:29] <wtay> thomas: in the _arg_set function
[19:29] <sienap> okie
[19:29] <sienap> retrying
[19:30] <rdj> but it uses a videosink so it does not work now
[19:30] <rdj> I think :P
[19:30] <rdj> brb food
[19:30] <sienap> he
[19:30] <sienap> i think it does :)
[19:30] <sienap> wtay is videosink broken ?
[19:30] <sienap> let me check
[19:31] <wtay> sienap: very broken
[19:31] <wtay> sienap: it isn't compile naymore
[19:31] <sienap> mwha
[19:31] <sienap> nope :)
[19:31] <sienap> wtay FIX IT !
[19:31] <sienap> >:)))
[19:31] <sienap> wtay how far is xvideosink ?
[19:31] <wtay> that's why you had to remove the previously compiled version
[19:31] <wtay> xvideosink is done
[19:32] <sienap> hmm
[19:32] <sienap> why is it not implented ?
[19:32] <sienap> rdj hack xvideosink into it :)
[19:32] <wtay> it is implemented
[19:33] <wtay> rdj: you need to play a bit with gtk_plug (look at gstplay.c)
[19:33] <sienap> why gstmediaplay doesn't use it
[19:33] <wtay> it does
[19:33] thomas (thomas at urgent.rug.ac.be) left irc: Leaving
[19:33] <sienap> why does it not work here then ?
[19:33] <wtay> because it probably is not 100% perfect yet
[19:34] <wtay> and capsnego sucks too...
[19:35] <sienap> he :)_
[19:35] <sienap> *G*
[19:35] <sienap> rdj how can i use your view ?
[19:41] thomas (thomas at urgent.rug.ac.be) joined #gstreamer.
[19:41] <Cactus> oh my
[19:41] <Cactus> this gstreamer installation sure takes a while
[19:41] <sienap>  he
[19:41] <sienap> yeah
[19:41] <sienap> compiling takes more then 2 minutes :)
[19:42] <Cactus> OK so first let's try out gstplay before trying wtay's component
[19:42] <wtay> Cactus: it's not an easy build...
[19:42] <Cactus> WARNING **: gstplugin: registry needs rebuild: run gstreamer-register
[19:42] <Cactus> what does that mean?
[19:42] <wtay> Cactus: you need to run gstreamer-register
[19:42] <sienap> that you have to run gstreamer-register :)
[19:42] <Cactus> freaky :)
[19:42] <wtay> Cactus: to register the plugins :)
[19:43] <Cactus>  /usr/lib/xmms/Output//libcrossfade.so: undefined symbol: get_output_list
[19:43] <Cactus> that's not good, right?
[19:43] <sienap> it is
[19:43] <sienap> ignore it :)
[19:43] <wtay> not good..
[19:43] <sienap> just for the not used xmms wrapper :)
[19:43] <sienap> wtay you want my list of errors ?
[19:43] <sienap> with xmms ?
[19:43] <Cactus> sienap: well the next line is `Segmentation fault' so maybe it isn't that good after all
[19:43] <sienap> that is because video sink
[19:43] <sienap> do the same thing as i told rdj
[19:43] <wtay> sienap: sure
[19:44] <wtay> Cactus: do you have libxmms-dev or something?
[19:44] <Cactus> wtay: I think I do
[19:44] <wtay> Cactus: maybe it's an old version... did you do make install?
[19:44] <sienap>  /root/.xmms/Plugins/libspatial.so: undefined symbol: inside_widget
[19:44] <sienap>  /usr/lib/xmms/Visualization//libbscopemax.so: undefined symbol: xmms_dga_leave
[19:44] <sienap>  /usr/lib/xmms/Visualization//libxmms_speakers.so: undefined symbol: mainwin
[19:45] <sienap> that are mine
[19:45] <sienap> however it doesn't really mather
[19:45] <sienap> cactus i bet it is just videosink
[19:45] <Cactus> wtay: currently I'm reading the back-buffer about sienap's solution
[19:45] <sienap> mainwin is btw a reference of the xmms main win location..
[19:46] Action: thomas is away - Automatically set away. - messages will be saved.
[19:46] <sienap> i am almost sure it doesn't segfault on the xmms plugins
[19:46] <Cactus> should I remove all *videosink* from $prefix/lib/gst?
[19:46] <sienap> gstreamer-register takes the xmms plugins last
[19:46] <sienap> cactus yes
[19:46] <sienap> and also from ./plugins/videosink
[19:46] <sienap> make clean there
[19:46] <sienap> :)
[19:46] <Cactus> sienap: also the xvideosink?
[19:47] <sienap> nope
[19:47] <sienap> that is the new replacement
[19:47] <Cactus> 'k
[19:47] <Cactus> gstreamer-register segfaults as well...
[19:48] <wtay> Cactus: what's the last line it displays?
[19:48] <sienap> still does ?
[19:48] <Cactus> yup
[19:48] <Cactus> INFO(25847:-1):gst_plugin_load_absolute:368: loading plugin "/home/cactus/prog/gnome/gstreamer/./plugins/xmms/.libs/libgstxmms.so"...
[19:48] <Cactus> Segmentation fault
[19:48] <Cactus>  /usr/lib/xmms/Output//libcrossfade.so: undefined symbol: get_output_list
[19:49] <wtay> can you do make clean in plugins/xmms too then
[19:49] <sienap> he
[19:49] <Cactus> wtay: yup, I'm on it
[19:49] <Cactus> oh man
[19:49] <Cactus> why does it want to write to /etc?
[19:49] <Cactus> shouldn't that be $(prefix)/etc?
[19:49] <wtay> Cactus: yeah, it's a configure options :)
[19:51] <Cactus> there is no configure option (at least in configure --help) to change the configuration directory specifically
[19:51] <wtay> hmm
[19:51] <thomas> wtay: if I have a filter object (GstVolEnv *filter)
[19:51] <thomas> wtay: how do I  get back to the pad ...
[19:51] <thomas> wtay: ... so I can get the sample rate ?
[19:51] <wtay> filter->srcpad
[19:52] <wtay> Cactus: hmm, I thought there was an option...
[19:52] Action: Cactus tries to figure out a way to write to /etc
[19:52] <wtay> Cactus: --with-configdir
[19:53] <Cactus> oh
[19:53] <Cactus> sorry
[19:53] <Cactus> but that's something entirely different
[19:53] <thomas> wtay: works... so how can I set an argument
[19:53] <thomas> wtay: and can I do it from within the plugin itself to test ?
[19:53] <wtay> thomas: gtk_object_set (...)
[19:53] <Cactus> it is for storing cross-compilation settings etc
[19:54] <wtay> Cactus: hm?
[19:55] <wtay> Cactus: I don't think it is..
[19:55] <Cactus> wtay: try it out yourself...
[19:56] <wtay> Cactus: it works here: config.h line 21: #define GST_CONFIG_DIR "/tmp"
[19:56] <wtay> ./configure --with-configdir=/tmp
[19:57] <Cactus> hmm
[19:57] <Cactus> now it does
[19:57] <Cactus> oh I missed a `='
[19:57] <wtay> doh
[19:57] <sienap> so cactus & wtay you guys really think gnome-media is going to be intergrated =]
[19:57] <sienap> sounds fun for sureeee :)
[19:57] <wtay> sienap: ?
[19:58] <rdj> gstmediaplay doesn't work in cvs either sienap, it uses xvideosink but something else is broken it seems
[19:58] <wtay> rdj: do you have Xv support?
[19:58] <Cactus> OK now here's the next problem:
[19:58] <rdj> dunno
[19:58] <rdj> lemme see
[19:59] <sienap> rdj he is correct
[19:59] <Cactus> ** CRITICAL **: file vcdsrc.c: line 374 (plugin_init): assertion `plugin != NULL' failed.
[19:59] <Cactus> INFO(31293:-1):gst_plugin_load_absolute:368: loading plugin "/home/cactus/prog/gnome/sandbox//lib/gst/libv4lsrc.so".
[19:59] <sienap> i mean bonobo-media
[19:59] <sienap> damn :)
[19:59] <Cactus> lots of these
[19:59] <rdj> I forgot how to check
[19:59] <wtay> Cactus: that's ok
[20:00] <wtay> rdj: xdpyinfo
[20:00] <rdj> wtay: Xv should be enabled?
[20:01] <wtay> rdj: for maximum success yes, capsnego/colorspace converters don't work very well yet
[20:01] <rdj> hmmm
[20:01] <rdj> ok
[20:01] <rdj> lemme try
[20:02] rdj (rdj at a37030.upc-a.chello.nl) left irc: proud member of the anti movement...
[20:02] <sienap> he
[20:02] <sienap> wtay what kind of thing still need to be coded for QoS again ?
[20:02] <wtay> sienap: events
[20:02] hadess (hadess at pc213-gui2.cable.ntl.com) joined #gstreamer.
[20:02] hadess (hadess at pc213-gui2.cable.ntl.com) left irc: Read error to hadess[pc213-gui2.cable.ntl.com]: EOF from client
[20:02] hadess (hadess at pc213-gui2.cable.ntl.com) joined #gstreamer.
[20:02] <wtay> yo
[20:03] <sienap> owjeah :)
[20:03] <sienap> hej hadess
[20:03] <hadess> yo sienap, wtay
[20:04] <Cactus> hmm
[20:04] <sienap>  hej hadess
[20:04] <Cactus> when I run gstreamer-register
[20:04] <sienap> you code dockapps before huh ?
[20:04] <sienap> wmdock apps that is..
[20:04] <Cactus> it should create a bunch of files in $prefix/etc, right?
[20:05] <wtay> Cactus: only one file: reg.xml
[20:05] <hadess> sienap: dockapps are only for window maker afaik
[20:05] <sienap> i know
[20:05] <sienap> but did you create them before ?
[20:06] <hadess> yep
[20:06] <Cactus> wtay: so configdir should be $prefix/etc/gstreamer and not $prefix/etc, right?
[20:06] <Cactus> damn
[20:06] <sienap> nice help me with this then:
[20:06] <wtay> Cactus: gstplugin.c is going to write to $GST_CONFIGDIR/reg.xml
[20:07] <sienap> i hacked a wmdockapp xmms plugin however when i disable it it takes down xmms..
[20:07] <sienap>         RedrawWindow();  
[20:07] <Cactus> OK, let me reinstall it once again:)
[20:07] <sienap> he damn
[20:07] <sienap> fuck tab completion
[20:07] <sienap>         RedrawWindow();  
[20:07] <sienap> aaaaaaah fuck it..
[20:07] sienap (synap at ipc379c0e6.dial.wxs.nl) left irc: Read error to sienap[ipc379c0e6.dial.wxs.nl]: EOF from client
[20:07] <hadess> sienap: because it's shite
[20:08] <thomas> wtay: so the basic volume envelope plugin is ready
[20:08] <wtay> thomas: aha
[20:08] <thomas> wtay: you can set the level wished for and the rise wished for
[20:08] <thomas> wtay: but for me this is impractical
[20:08] <thomas> wtay: I know the full volume envelope before it starts playing
[20:08] <thomas> wtay: I have it in an xml on disk
[20:09] <thomas> wtay: should I write a "controlling" plugin to control the volume envelope ?
[20:09] <wtay> thomas: I would create a signal in the plugin and let the app adjust the envelope in the signal function
[20:09] <thomas> (I'm hating my plugin already... it rises to full volume then drops back to -10 dB
[20:09] sienap (synap at ipc379c0e6.dial.wxs.nl) joined #gstreamer.
[20:09] <sienap> ok i suck
[20:09] <sienap> wait a mom
[20:09] <thomas> (then continues again)
[20:09] <thomas> wtay: yes, but that is too low-level for me
[20:10] <sienap> hej thomas!
[20:10] <wtay> thomas: ?
[20:10] <thomas> wtay: I'd like to be able to somehow register a volume envelope
[20:10] <thomas> wtay: then set it off to play
[20:10] <thomas> wtay: and never have to control the plugin again
[20:10] <thomas> wtay: maybe I could do both at the same time ?
[20:10] <thomas> wtay: put it in the plugin and leave it there to use ?
[20:11] <wtay> thomas: maybe yes...
[20:12] <Cactus> OK now I really feel ready to rock
[20:12] <wtay> Cactus: hehe
[20:12] <wtay> Cactus: compiled/installed?
[20:15] <Cactus> OK
[20:15] <Cactus> doesn't work
[20:15] <Cactus> 20:22:27 [cactus at galaxy gstplay]$ ./gstmediaplay  /home/cactus/muzax3/Anime/FLCL/16\ -\ Beautiful\ Morning\ With\ You.mp3 
[20:15] <Cactus> INFO: Initializing GStreamer Core Library
[20:15] <Cactus> INFO: CPU features: (0c040843) MMX 
[20:15] <Cactus> ** CRITICAL **: file gstbin.c: line 172 (gst_bin_add): assertion `element != NULL' failed.
[20:15] <Cactus> ** CRITICAL **: file gstelement.c: line 577 (gst_element_connect): assertion `src != NULL' failed.
[20:15] <Cactus> ** CRITICAL **: file gstelement.c: line 326 (gst_element_get_pad): assertion `element != NULL' failed.
[20:15] <Cactus> ** CRITICAL **: file gstelement.c: line 273 (gst_element_add_ghost_pad): assertion `pad != NULL' failed.
[20:16] <wtay> Cactus: gstmediaplay?
[20:16] <Cactus> wtay: yup
[20:16] <wtay> Cactus: you need Hermes
[20:17] <hadess> Cactus: you have Herpes, you need a doctor
[20:17] <wtay> right :)
[20:17] <Cactus> hermes?
[20:17] <wtay> a color conversion library...
[20:17] <Cactus> could it be a problem in gstmediaplay, BTW?
[20:18] <wtay> gstmediaplay uses the color conversion plugin
[20:18] <wtay> but it's not build because Hermes was not found in configure.in
[20:18] <wtay> lemme find the URL...
[20:18] <Cactus> it's in Debian
[20:18] <wtay> yes
[20:19] <wtay> hermes1 and hermes1-dev
[20:19] <Cactus> and should the autoplug example work?
[20:19] rdj (rdj at a37030.upc-a.chello.nl) joined #gstreamer.
[20:19] <wtay> Cactus: try gstreamer-launch fakesrc ! fakesink
[20:20] <sienap> hej rdj!
[20:22] <Cactus> yayy
[20:22] Action: Cactus has mp3 audio working
[20:22] <wtay> aha!
[20:22] <wtay> video is the hardest part though...
[20:23] Action: Cactus has OGG Vorbis audio working!
[20:23] <Cactus> I'm getting the hang of it:)
[20:23] <sienap> he :)
[20:23] <sienap> gstreamer is cool :)
[20:23] <Cactus> OK let's try video
[20:23] <sienap> but the current cvs oooooooe man :) wild ride :)
[20:24] <thomas> wtay: I have an idea for the envelope data passing...
[20:24] <thomas> wtay: but I don't know if it's allowed
[20:24] <sienap> off
[20:24] <thomas> wtay: can I check with you ?
[20:24] <sienap> quake :)
[20:24] sienap (synap at ipc379c0e6.dial.wxs.nl) left irc: sienap has no reason
[20:24] <wtay> thomas: sure
[20:25] <thomas> wtay: the data structure is a list of time points and volume levels
[20:25] <thomas> wtay: so two floats
[20:25] <thomas> wtay: for each
[20:25] <wtay> thomas: a glist?
[20:25] <thomas> wtay: so I make a pointer to a data structure big enough to hold them
[20:25] <thomas> wtay: what's a glist ? sorry, not smart enough yet...
[20:26] <wtay> thomas: a glib linked list
[20:26] <thomas> wtay: I was thinking, "make the pointer", "pass it", "object frees it at end"
[20:26] <thomas> wtay: is a glist in the gtk tutorial ?
[20:26] <wtay> thomas: just make sure not to have a header file for the plugin
[20:26] <wtay> thomas: yup
[20:26] <thomas> why no header file ?
[20:27] <wtay> thomas: cause that would violate the gstreamer  plugin philosofy
[20:27] <wtay> thomas: plugins should be self contained and completely introspactable
[20:27] <thomas> wtay: I don't understand.  Do you mean the normal .h header ?
[20:27] <Cactus> could someone show me an example pipeline to get mpeg1 video working?
[20:27] <wtay> thomas: yup
[20:27] hadess (hadess at pc213-gui2.cable.ntl.com) left irc: Read error to hadess[pc213-gui2.cable.ntl.com]: EOF from client
[20:28] <thomas> wtay: there's one for stereo2mono
[20:28] <wtay> Cactus: try gstmediaplay?
[20:28] <wtay> thomas: it's not installed, apps are not supposed to link to it
[20:28] <Cactus> wtay: I'm currently using gstreamer-launch
[20:28] <thomas> wtay: okay.  but why would I use a header file then ?
[20:29] <wtay> Cactus: can't do mpeg1 with -launch yet :(
[20:29] <wtay> thomas: I dunno, to pass a struct to the plugin?
[20:29] <thomas> wtay: I think I should check what a glist is first...
[20:29] <thomas> wtay: it should cover my needs, no ?
[20:29] <wtay> thomas: yeah
[20:30] <wtay> thomas: but an arg would do the same albeit a bit more awkward
[20:30] <Cactus> wtay:  would something like gstreamer-launch disksrc \"location=cowbe-ed.mpg\" ! mpeg1parse | gstmpegaudio ! osssink work to get sound from the video?
[20:30] hadess (hadess at pc213-gui2.cable.ntl.com) joined #gstreamer.
[20:30] <wtay> Cactus: nope
[20:30] <wtay> Cactus: can't do mpeg1 with -launch at all
[20:30] <Cactus> wtay: why not?
[20:30] <wtay> Cactus: the mpeg1 demuxer has dynamic pads...
[20:31] <wtay> -launch doesn't know about dynamic pads at all..
[20:31] <Cactus> wtay: OK so I will wait for gstreamer to recompile so that I have hermes support
[20:31] <wtay> dynamic pad: pads created at runtime
[20:31] <wtay> Cactus: yup, your best bet..
[20:31] Action: rdj still can't get gstmediaplay working again
[20:31] <wtay> rdj: what happens?
[20:32] <thomas> wtay: OK, checked the glist.  so what's the best thing...
[20:32] <rdj> sorry for the paste:
[20:32] <rdj> ** CRITICAL **: file gstbin.c: line 172 (gst_bin_add): assertion `element != NULL' failed.
[20:32] <rdj> ** CRITICAL **: file gstelement.c: line 577 (gst_element_connect): assertion `src != NULL' failed.
[20:32] <rdj> ** CRITICAL **: file gstelement.c: line 326 (gst_element_get_pad): assertion `element != NULL' failed.
[20:32] <rdj> ** CRITICAL **: file gstelement.c: line 273 (gst_element_add_ghost_pad): assertion `pad != NULL' failed.
[20:32] <thomas> wtay: have the app create the list and pass it on
[20:32] <wtay> rdj: same problem: install Hermes
[20:32] <thomas> wtay: or have the app pass arguments and have the arguments create the list on the fly ?
[20:32] <rdj> Hermes?
[20:32] <rdj> lemme read
[20:32] <wtay> thomas: I prefer the second option
[20:32] <rdj> ok will try
[20:32] <wtay> rdj: a color conversion lib
[20:33] <thomas> wtay: I know I can set two arguments at once with gtk_object_set...
[20:33] <thomas> wtay: but the set arg functions split it, right ?
[20:33] hadess (hadess at pc213-gui2.cable.ntl.com) left irc: Read error to hadess[pc213-gui2.cable.ntl.com]: EOF from client
[20:33] <wtay> thomas: yup
[20:33] <thomas> wtay: so how can I make sure BOTH arguments are passed ?
[20:33] <wtay> thomas: I still think you need something like "1-5:.01"
[20:34] <thomas> wtay: I would let the app do that
[20:34] hadess (hadess at pc213-gui2.cable.ntl.com) joined #gstreamer.
[20:34] <thomas> wtay: the idea right now is :
[20:34] <thomas> wtay: register control points on the envelope curve
[20:34] <thomas> wtay: let the plugin calculate volumes between control points
[20:34] <thomas> wtay: the way you draw it in cooledit
[20:34] <wtay> thomas: cool
[20:34] <rdj> building hermes
[20:34] <wtay> thomas: so what do you need to pass to the plugin then?
[20:34] <thomas> wtay: the control points
[20:35] <wtay> thomas: time/volume pairs?
[20:35] <thomas> wtay: yes
[20:35] <thomas> wtay: so your way of passing it would work
[20:35] <wtay> thomas: ok, the best option IMO
[20:36] <wtay> thomas: just pass them one after the other and create a glist inside the plugin
[20:36] <thomas> wtay: ok, I'll try that now
[20:36] <thomas> (sigh)
[20:36] <wtay> thomas: glib2.0 interfaces will help with this in  the future...
[20:36] <thomas> wtay: oh, then i'll sleep some two weeks and do it then ;)
[20:37] <wtay> thomas: I thought you had a scedule to keep :)
[20:37] <thomas> is a glist GTK_TYPE_GLIST ?
[20:37] <wtay> thomas: noe
[20:37] <wtay> nope
[20:37] <wtay> it's untyped
[20:37] <wtay> GTK_TYPE_POINTER will do
[20:37] <wtay> err
[20:37] <thomas> wtay: oh right... doesn't matter anyway, right ? I should have normal arguments
[20:38] <wtay> yeah, just strings
[20:38] <thomas> wtay: but if I want to pass on 1.0:1.0, I should do it as a string
[20:38] <wtay> yes
[20:38] <wtay> and use g_split to parse it or sscanf
[20:38] <thomas> so a string is GTK_TYPE_STRING ?
[20:38] <wtay> I think so..
[20:39] <wtay> yup
[20:39] Action: rdj cries for having to rebuild all of gstreamer :P
[20:39] <wtay> rdj: hehehehehehe HEHEHE
[20:39] Action: wtay need to redo capsnego again...
[20:39] <rdj> oh well, if it works then, which seems likely cause I redid my oaf file and it behaves better now, I'm a happy man
[20:40] <wtay> rdj: this happens with unpolished unstable software :)
[20:40] <rdj> wtay: I know, I'm not complaining ;)
[20:40] <rdj> it may sound that way, but I'm not :)
[20:40] Action: rdj is really thankfull for the extensive help I got on this channel
[20:41] <wtay> rdj: I was not suggesting anything of the kind...
[20:41] <rdj> ok :)
[20:41] <thomas> wtay: one other problem I had in the past...
[20:41] <thomas> wtay: since a level of 0 means -infinite dB
[20:41] <thomas> wtay: i need a practical level of 0
[20:42] <thomas> wtay: would it be best to have this as a settable variable for the plugin ?
[20:42] <wtay> thomas: yyup
[20:44] Action: rdj goes to the coffeeshop now :P
[20:44] <rdj> seeyou ppl :P
[20:46] <wtay> rdj: damn you ;)
[20:47] <thomas> wtay: there isn't a disksink element is there ?
[20:47] <wtay> thomas: hmm, no, only fdsink
[20:47] <wtay> thomas: need to pass it a file descriptor...
[20:56] <thomas> wtay: if you have a glist called envelope ...
[20:56] <thomas> wtay: ... can you say envelope->data to get your data back ?
[20:56] <wtay> yup
[20:56] <thomas> can you also do that with g_list_nth ?
[20:57] <wtay> this will return a single g_list element I think
[20:57] Uraeus (Uraeus at c224s9h5.upc.chello.no) left irc: 
[20:57] <thomas> the prototype says GList * g_list_nth, so...
[20:58] <wtay> yu
[20:58] <wtay> yup
[21:00] <thomas> wtay: oh, there's a g_list_nth_data
[21:00] <thomas> returning a pointer
[21:00] <wtay> ah
[21:00] <thomas> but I can't cast it as double...
[21:01] <wtay> nope, you need to add the pointer to a struct in the glist
[21:01] <wtay> and malloc the struct...
[21:12] <thomas> wtay: any reason why a glist starts out as length 1 ?
[21:12] <wtay> uhm, no...
[21:12] <wtay> you did set it to NULL did you?
[21:12] <thomas> no, I allocated it in the init section
[21:13] <wtay> no need to allocate a list...
[21:13] <thomas> you're right...
[21:13] <wtay> GList *mylist = g_list_append (NULL, somepointer);
[21:13] <thomas> so what does the g_list_alloc do then ?
[21:14] <wtay> internal kitchen 
[21:14] <wtay> space for one element, the docs say
[21:15] <thomas> what docs are you looking at ?
[21:15] <wtay> http://developer.gnome.org/doc/API/glib/glib-doubly-linked-lists.html#G-LIST-ALLOC
[21:17] <thomas> oh... those are much better
[21:17] <thomas> I probably had old ones
[21:22] <thomas> wtay: ok... so I use g_malloc to allocate space for doubles
[21:22] <thomas> then when I register
[21:23] <thomas> pass these allocated pointers to the g_list_append ?
[21:23] <thomas> it seems to be working
[21:23] <wtay> yup
[21:24] <thomas> cool ! I just put a volume envelope registration in the plugin init
[21:24] <thomas> it prints out
[21:24] <thomas> now have it actually do work
[21:29] <thomas> wtay: can a plugin have these values registered before it actually starts but after init ?
[21:30] <thomas> I need to set the starting volume...
[21:30] <wtay> why not in init?
[21:30] <thomas> because it's the app that registers the starting point...
[21:30] <thomas> or am I wrong ?
[21:31] <wtay> you create the plugin, set the properties and then you run...
[21:31] <wtay> you can add a state change function if you want..
[21:31] <thomas> ok...
[21:31] <thomas> is there a good example of that ?
[21:32] <wtay> osssink has one
[21:32] <Cactus> OK
[21:32] <Cactus> one jackie chan movie later,
[21:32] <Cactus> I finally have gstmediaplay working :)))
[21:32] <wtay> Cactus: cool, a succes story!
[21:33] <Cactus> now let's try video
[21:33] sienap (synap at ipc379c0e6.dial.wxs.nl) joined #gstreamer.
[21:33] <sienap> Hi al!
[21:33] <sienap> l
[21:33] <Cactus> uh-oh
[21:33] <Cactus> gotta go
[21:33] <sienap> he :)
[21:33] <sienap> why ?
[21:33] <wtay> Cactus: ok
[21:33] <sienap> hej wtay!
[21:34] <wtay> hi again
[21:34] <sienap> still busy with mad ?
[21:34] <wtay> sienap: mad is done
[21:34] <wtay> working on capsnego now
[21:35] <wtay> try gstreamer-launch disksrc location=some.mp3 ! mad ! osssink :)
[21:38] <sienap> he >:)
[21:38] <sienap> mad doesn't have a parser ?
[21:38] <sienap> is mad cool in someway btw ?
[21:38] <wtay> pretty
[21:39] <thomas> wtay: you said "set arguments then init plugin", was that right ?
[21:40] <wtay> thomas: er, no I said, init, set args, run
[21:40] <thomas> ok...
[21:40] <wtay> brb
[21:40] <thomas> wtay: I'm using a next_time and next_level var in my plugin
[21:40] <thomas> wtay: don't know how to initialize them in the beginning
[21:43] <wtay> in the init
[21:43] <hadess> wtay: w00t ? mad is done ?
[21:43] <thomas> wtay: what I mean is :
[21:43] <wtay> hadess: that's one beer for me :)
[21:43] <thomas> wtay: first I register point 1 (e.g. level .001 at 0 sec)
[21:43] <hadess> wtay: wheee !
[21:43] <thomas> then point two (level 1 at 10 sec)
[21:43] <thomas> but registering is after the init
[21:44] <wtay> thomas: yup
[21:44] <thomas> it's only at the second point that I know the first next_level
[21:44] <thomas> but I need to set it before it plays
[21:44] <wtay> thomas: why don't you calculate it in the arg set function then?
[21:44] <thomas> so how do I know that it does this only at the SECOND call of arg_set ?
[21:45] <thomas> use a cheat variable ?
[21:45] <wtay> thomas: check the length of the list?
[21:45] <hadess> is the Changelog ever changed ?
[21:45] <wtay> hadess: try it if you can
[21:45] <wtay> hadess: nope
[21:45] <thomas> ok... that's a good clean way to cheat. will do.
[21:45] <hadess> wtay: yeah, i'm gonna recompile it
[21:47] <hadess> wtay: i'm telling you how it works at the end of the film
[21:47] Nick change: hadess -> hds-tv
[21:49] sienap (synap at ipc379c0e6.dial.wxs.nl) left irc: sienap has no reason
[22:14] z- (dave at host213-122-121-172.btinternet.com) joined #gstreamer.
[22:14] <wtay> yo
[22:15] <z-> hey
[22:34] <thomas> quit
[22:49] <Cactus> hi
[22:50] <Cactus> I think I have only one more problem
[22:50] <wtay> Cactus: ah
[22:50] <Cactus> apparently, it can't find the DLL's I've installed for libaviplay
[22:51] <Cactus> Video in DivX ;-) format
[22:51] <Cactus> gstwindec:: ICDecompressBegin failed
[22:51] <Cactus> gst_windec: could not init decoder
[22:51] <wtay> Cactus: oh, they should go into /usr/lib/win32
[22:51] <wtay> Cactus: maybe it's something else...
[22:51] <Cactus> wtay: isn't it good enough if libaviplay can find them? (I don't want to pollute /usr)
[22:51] <Cactus> I get sound output from the video
[22:52] <wtay> Cactus: We don't use libaviplay yet so the paths might be different...
[22:52] <wtay> you can set the paths with a configure option though..
[22:52] <Cactus> wtay: looking at the strace output, it doesn't try to open any dll's
[22:53] <wtay> Cactus: hmm
[22:53] <Cactus> 15297 open("/mnt/stuff2/anime/FLCL/flcl_ep02.avi", O_RDONLY) = 12
[22:53] <Cactus> fn=0xf00, child_stack=0x80c7710, flags=CLONE_FILES|CLONE_PID|CLONE_PTRACE|CLONE_VFORK|CLONE_PARENT|0xbfff0074, args=0xbffff3d015297 --- SIGRT_0 (Real-time signal 0) ---
[22:53] <Cactus> this is the last open() call
[22:54] <wtay> you have the latest windows dlls?
[22:54] <Cactus> it doesn't even look for them
[22:54] <wtay> hmm yeah..
[22:56] <wtay> are you sure there's a /usr/lib/win32 dir with the dlls?
[22:57] <wtay> maybe it does a stat first
[22:57] <Cactus> it's in /usr/local/lib/win32
[22:57] <Cactus> 23:04:26 [cactus at galaxy cactus]$ grep win32 /tmp/gst.strace 
[22:57] <Cactus> 23:04:29 [cactus at galaxy cactus]$ 
[22:57] <wtay> you can run configure  --with-win32-libdir=/usr/local/lib/win32
[22:58] <wtay> strace -f?
[22:58] <wtay> it's all threaded so you need the -f flag
[22:58] <Cactus> yes
[22:58] <Cactus> I am smarter than I look :)
[22:59] <wtay> grr :)
[22:59] <Cactus> strace -F -f   -o /tmp/gst.strace gstmediaplay /mnt/stuff2/anime/FLCL/flcl_ep02.avi
[22:59] <Cactus> I don't really know about -F, but it can't hurt:)
[22:59] <wtay> yeah, smart :)
[23:00] <wtay> I'd suggest to try with the configure option...
[23:00] <Cactus> is it enough to recompile/reinstall only the avi plugin?
[23:00] <wtay> actually libs/winloader is enough
[23:02] <Cactus> nope, still the same
[23:02] <wtay> darn
[23:05] <wtay> hmm, no idea, maybe it's the .registry in $HOME
[23:06] <Cactus> but like I said, it doesn't even try to see if there are any DLLs lying around
[23:06] <wtay> strange
[23:07] <thomas> DAMN DAMN DAMN
[23:07] <thomas> my volume envelope overshoots 1.0 level by a narrow margin
[23:07] <thomas> but it shouldn't
[23:07] <thomas> I hate floating point calculations
[23:07] <thomas> *sigh*
[23:07] <wtay> hehe
[23:08] <wtay> Cactus: only two function calls before actually opening the dll
[23:08] <wtay> the first one is an alloc..
[23:12] <thomas> wtay: one question...
[23:12] <thomas> wtay: right now I check at the start of each block...
[23:13] <thomas> wtay: if the point has been reached
[23:13] <wtay> ok
[23:13] <thomas> wtay: if it is, I reset the increase value (which is added to the level at each sample)
[23:13] <thomas> well of course...
[23:13] <thomas> it only knows at point 5.002 (for example)
[23:13] <thomas> that it's past 5 (if 5 is a control point)
[23:13] <thomas> so the level is 1.002 (for example)
[23:13] <thomas> if I put the check INSIDE the loop
[23:14] <thomas> that's bound to slow things down a lot
[23:14] <thomas> so...
[23:14] <thomas> what would be a good solution here ?
[23:14] <wtay> switch out in two cases
[23:14] <wtay> n samples with processing M samples without
[23:15] <thomas> so you mean : first calculate at what sample the transition will be
[23:15] <thomas> and then do two loops if necessary ?
[23:15] <wtay> yup
[23:15] <thomas> ok... should do it
[23:15] <thomas> I want to get this finished
[23:16] <thomas> (I hope the normal plugin still works by the time volumes are complete !)
[23:17] <thomas> wtay: one problem I've been having since some time today...
[23:17] <thomas> I used a caps ptr to get the caps of the src pad
[23:17] <thomas> but since some time (don't know since when)
[23:18] <thomas> it fails to get the pointer
[23:18] <wtay> hmm
[23:18] <thomas> I don't know how this can happen
[23:18] <thomas> so... how does it get the caps of a pad ?
[23:18] <thomas> the src pad is there cause when i print the name, it says "src"
[23:18] <thomas> is there some other check I can do ?
[23:18] <wtay> it happens during caps negotiation (when two plugins connect)
[23:19] <thomas> the weird thing is : I do it in two places in my plugin
[23:19] <thomas> the one always works
[23:19] <thomas> the other one doesn't
[23:19] <wtay> what does your pipeline look like
[23:19] <wtay> oh
[23:19] <thomas> tools/gstreamer-launch disksrc location=/tmp/arid.mp3 ! mp3parse ! mpg123 ! volenv ! osssink
[23:19] <thomas> I used fakesrc up until the point where it started failing
[23:19] <thomas> i mean fakesink
[23:20] <wtay> but the sound still plays fine?
[23:20] <thomas> with osssink, yes.
[23:20] <thomas> but only because i set the sample rate manually
[23:21] <wtay> aha
[23:21] <thomas> and the rate is what i need the caps for
[23:22] <wtay> lemme check something...
[23:23] <z-> what do those '!'s do ?
[23:24] <wtay> z-: in gstremaer-launch? they separate plugins :)
[23:24] <z-> ah
[23:24] <z-> i thought they ewre handled by the shell
[23:25] <z-> maybe thats just csh..
[23:25] <thomas> are there other validity checks i can do on a pad ?
[23:25] <wtay> thomas: ther must be something wrong with the negotiate functions of your plugin
[23:26] <thomas> ok...
[23:26] <thomas> but what if I never did anything with the negotiate functions ?
[23:26] <thomas> hmm....
[23:26] <thomas> how can I check ?
[23:26] <wtay> you should have a negotiate function on the src pad and the sink pad that just proxy the result to the peer
[23:27] <thomas> ok that's chinese to me
[23:27] <wtay> like passthrough has
[23:27] <thomas> well it should have the same thing
[23:27] <thomas> since i used passthrough
[23:27] <wtay> ok, I'm compilein passthrough to check it out
[23:27] <wtay> s/compilein/compiling
[23:28] <hds-tv> wtay: ...
[23:28] Nick change: hds-tv -> hadess
[23:28] <wtay> yup
[23:28] <hadess>  /plugins/filters/colorspace is broken, tries to compile yuv2rgb_mmx16.s
[23:29] <wtay> hadess: oh yeah...
[23:29] omega_ (perw at omegacs.net) joined #gstreamer.
[23:29] <wtay> hi omega_
[23:29] <omega_> yo
[23:29] Action: omega_ is attempting to get transparent proxying *and* a local httpd to work at the same time
[23:29] <thomas> hello
[23:30] <thomas> wtay: ok... the overshoot calculation works as well
[23:30] <thomas> wtay: did your "yup" five lines ago signify anything ?
[23:30] <wtay> it never does :)
[23:32] <omega_> whee, got it working, mostly
[23:34] <wtay> thomas: ok, passthrough works here, with capsnego
[23:35] <thomas> wtay: ok, so I did something else wrong then ?
[23:35] <omega_> what is the 'passthrough' plugin, anyway?
[23:36] <wtay> omega_: a dummy plugin with a shell script to create an audio effect plugin
[23:36] <omega_> shell script?
[23:36] <wtay> only implement one single function and off you go
[23:36] <thomas> just an easy way of replacing every occurence
[23:36] <thomas> of a function name with a new function name
[23:36] <wtay> it copies/modifies the code
[23:36] <omega_> shell script as in /bin/bash ??
[23:36] <omega_> ah, ok, a poor man's filterfactory
[23:37] <wtay> yes :)
[23:37] <thomas> sorry, I hope to get rich later
[23:37] <thomas> ;)
[23:37] <wtay> heh
[23:37] <omega_> yeah, well, that's something we have to worry about as far as portability, the FilterFactory will *have* to make use of the varargs macros....
[23:38] <omega_> the new padtemplate stuff removes that problem there, we can solve DEBUG and INFO with varargs functions...
[23:38] <thomas> wtay: the anti-overshoot works
[23:38] <thomas> wtay: i should also do it for every control point, not just level 1.0
[23:38] <omega_> overshoot?
[23:38] <thomas> omega: I'm making a volume envelope plugin
[23:38] z- (dave at host213-122-121-172.btinternet.com) left irc: Confucius not: know what to say!
[23:39] <omega_> define overshoot in that contex?
[23:39] <thomas> omega: and it's tricky to get the floating point stuff right
[23:39] <wtay> omega_: do you have problems with the padfactory gone now?
[23:39] <omega_> heh, yeha <g>
[23:39] <thomas> omega: well, overshoot being in this case "volume level getting past 1.0"
[23:39] <omega_> wtay: nope, though I had to fix the videosink...
[23:39] <omega_> thomas: ok
[23:39] <thomas> omega: which could result in distortion
[23:39] <wtay> omega_: it's not compiled anymore...
[23:39] <omega_> thomas: only if the resulting level is over 1.0
[23:39] <omega_> wtay: yeah, well, I tried to last night, er, this morning
[23:40] <thomas> yes, that's right.
[23:40] <thomas> but I'm not going to check EVERY sample value
[23:40] <thomas> anyway, the point is that it should work for general envelope control points
[23:40] <omega_> thomas: actually, the gain filter is the very best place to check level
[23:40] <omega_> or at least peak
[23:40] <thomas> how does that work ?
[23:41] <omega_> just take the max of all resultant samples
[23:41] <omega_> fire a signal every N samples with this peak value
[23:41] <wtay> omega_: thomas uploads the envelope to the plugin hat do you think about that?
[23:41] <omega_> depends on how he does it <g>
[23:41] <thomas> uhm... I register it through control point arguments ?
[23:41] <thomas> adding them to a glist ?
[23:42] Cactus (cactus at port059.bp1.telnet.hu) left irc: Tank, I need an exit
[23:42] <thomas> omega_ : is the gain filter already in place ?
[23:42] <omega_> um, /me will be back, I have to switch machines
[23:42] <omega_> brb in 5min
[23:42] omega_ (perw at omegacs.net) left irc: [x]chat
[23:42] <thomas> wtay: what can i do about the caps problem ?
[23:42] <wtay> dunno, I have to look at your code...
[23:43] <thomas> shall I put it somewhere then ?
[23:43] <wtay> I you can, yes
[23:43] <thomas> so you can see if it works as well
[23:43] <thomas> don't pay attention to the mess
[23:43] <wtay> only if you don't pay attention to my mess :)
[23:43] <thomas> ;)
[23:44] <thomas> it should be at http://urgent.rug.ac.be/thomas/gst
[23:44] <wtay> ok thx
[23:45] <thomas> and it still prints out too much debug info
[23:45] <wtay> err, is that done with the current CVS version?
[23:45] <thomas> current as in yesterday evening...
[23:45] <thomas> why ?
[23:46] <wtay> hmm GstPadFactory is gone
[23:46] <thomas> what ? in volenv.c ?
[23:46] <wtay> as of yesterday 11:56pm :)
[23:47] <wtay> anyway..
[23:47] <thomas> huh ? you mean I don't need to specify it anymore ?
[23:47] <wtay> yes you do, just a different API
[23:47] <thomas> ok...
[23:47] <thomas> so would that be the problem then ?
[23:47] <wtay> I'll do that for you
[23:47] <wtay> no it shouldn't be a problem...
[23:48] <wtay> although I'm not sure I changed the negotiation too..
[23:48] <wtay> ok, I'll try with current CVS version and adjust to make it compile...
[23:50] omega_ (omega at omegacs.net) joined #gstreamer.
[23:50] dave (dave at host213-122-121-172.btinternet.com) joined #gstreamer.
[23:50] <omega_> yo
[23:51] <dave> hi
[23:51] Nick change: dave -> z-
[23:51] <wtay> hi again
[23:51] <omega_> http://gstreamer.net/guadec-pics2/
[23:52] <wtay> cool
[23:53] <omega_> now /me has another thread interlock issue to work through
[23:53] <wtay> and /me has to redo capsnego again *sigh*
[23:53] <omega_> oh?
[23:54] <wtay> proxy stuff again and passing a gpointer to the pad negotiate function...
[23:54] <omega_> hmm
[23:54] <thomas> wtay: should I get the new CVS and see if I still have the problem ?
[23:55] <wtay> thomas: wait just a bit, I'll send you one that compiles with the new CVS version first
[23:56] <omega_> wtay: did you deal with the pending mail to the list?
[23:56] <wtay> omega_: yeah
[23:56] <omega_> ok
[23:56] <wtay> spam
[23:56] <omega_> cool
[23:56] <wtay> unless you wanted a loan or something :-)
[23:56] <omega_> um, no
[23:58] <hadess> and /me wants colorspace to be fix0red
[23:58] <wtay> hadess: just a sec.. :)
[23:58] Action: omega_ isn't overly happen with the Hermes dependency in the colorspace converter
[23:58] <wtay> ok, got volenv compiling...
[23:58] Action: omega_ would like to find some way to remove that dep
[23:58] <omega_> er, s/happen/happy/
[23:59] <omega_> I think colorspace is a sufficently core filter that it shouldn't require outside libs
[23:59] <wtay> omega_: hmm, copy the code?
[23:59] <wtay> true
[23:59] <hadess> the core gstreamer shouldn't depend on anything but glib/gtk+
[23:59] <omega_> make gstreamer/deplibs/Hermes/ ?
[23:59] <omega_> hadess: right, same with the core plugins, of which colorspace should be one
[00:00] --- Mon Apr 16 2001
[00:00] <wtay> thomas: perfect here, and very cool too !
[00:01] <thomas> wtay: great to know it works ;)
[00:01] <thomas> wtay: don't CVS it yet...
[00:01] <wtay> making a new tarball
[00:01] <thomas> wtay: so does it complain about the caps ?
[00:01] <wtay> not at all
[00:01] <thomas> ok great....
[00:01] <thomas> that means I should get new CVS code now right ?
[00:01] <wtay> hmm, yes it does... :(
[00:02] <thomas> well that's not a problem....
[00:02] Action: hadess plays with the gal shortcut bar
[00:02] <thomas> but does CVS do an update or a full download ?
[00:02] <wtay> oh, you check the caps before the pipeline plays!
[00:02] <thomas> yes
[00:02] <thomas> I'm starting to like the whole irc/cvs way of developing
[00:02] <wtay> you can't do that
[00:02] <thomas> instant gratification
[00:03] <wtay> yeah
[00:03] <thomas> hmmm... why can't I do that ?
[00:03] <thomas> the rate is not set yet ?
[00:03] <thomas> is that it ?
[00:03] <wtay> if my company would only understand that...
[00:03] <wtay> thomas: the caps is only set at runtime (when mpg123 knows the media properties)
[00:03] <thomas> ok that's understandable
[00:03] <wtay> yeah
[00:03] <omega_> wtay: does the libmad plugin work?
[00:03] <thomas> but how can I get at the sample rate then
[00:04] <wtay> not when you're not running the pipeline
[00:04] <thomas> but if the pipeline's set up, there must be some way of telling the volenv plugin what type of input sound there is
[00:04] <thomas> no ?
[00:04] <wtay> yes
[00:05] <thomas> how ?
[00:05] <wtay> there is a function call (newcaps) that is fired when the caps are known and set..
[00:05] <thomas> as in a signal ?
[00:05] <wtay> the _set_args functions should not care yet
[00:06] <wtay> as in: function callback
[00:06] <thomas> ok... so in set_args I can't use sample rate
[00:06] <wtay> nope
[00:06] <thomas> but I can convert them later on ?
[00:06] <wtay> sure
[00:06] <thomas> ok... I'll check how I can re-do that.
[00:06] <thomas> did you have to change much in the plugin ?
[00:06] <wtay> the newcaps callback is perfect for that
[00:07] <wtay> not much
[00:07] <wtay> like:
[00:07] <wtay> static GstPadTemplate*
[00:07] <wtay> volenv_src_template_factory (void)
[00:07] <wtay> {
[00:07] <wtay>   static GstPadTemplate *template = NULL;
[00:07] <wtay>   if (!template) {
[00:07] <wtay>     template = gst_padtemplate_new (
[00:07] <wtay>       "src",
[00:07] <wtay>       GST_PAD_SRC,
[00:07] <wtay>       GST_PAD_ALWAYS,
[00:07] <wtay>       gst_caps_new (
[00:07] <wtay>        "test_src",
[00:07] <wtay>        "audio/raw",
[00:07] <wtay>        gst_props_new (
[00:07] <wtay>          "channels", GST_PROPS_INT (2),
[00:07] <wtay>          NULL)),
[00:07] <wtay>       NULL);
[00:07] <wtay>   }
[00:07] <wtay>   return template;
[00:07] <wtay> }
[00:07] <wtay> sorry bout that :)
[00:07] <thomas> no problem...
[00:08] <thomas> so you might want to do that in passthrough as well...
[00:08] <thomas> can I download the new version ?
[00:08] <wtay> did that
[00:08] <wtay> sec...
[00:08] <thomas> so that I can modify it further
[00:08] <thomas> I'm doing the overshoot right now
[00:08] <wtay> I'll check it in
[00:09] <thomas> so when i do a check-out, does it overwrite stuff ?
[00:09] <wtay> It'll merge, but just move it to somewhere else to be sure...
[00:10] <thomas> I backed it up.
[00:10] <omega_> make sure to do a cvs update, not cvs checkout
[00:10] <thomas> I'd rather not do a full check out
[00:10] <wtay> cvs update -d in the plugin/filters dir should do
[00:11] <thomas> are you sure ? I don't have the CVS version with the dropped gstpadfactory
[00:11] <wtay> aahh, damn
[00:11] <thomas> at least I don't think so
[00:11] <thomas> can I check
[00:11] <thomas> ?
[00:11] <wtay> there's no other way than a complete update
[00:12] <omega_> -z9
[00:12] <wtay> just cvs update -d in the root
[00:12] <thomas> I have ADSL
[00:12] <omega_> -dPA
[00:12] <wtay> thomas: you're not like on a modem line..
[00:12] <omega_> cvs -z9 update -dPA
[00:12] <thomas> omega_: i'll try your options
[00:13] <wtay> ok, checked in
[00:13] <wtay> I didn't do the configure.in yet...
[00:13] <wtay> thomas: what CPU?
[00:15] <thomas> ok I don't know much about CVS
[00:15] <thomas> here's what I type :
[00:15] <thomas> cvs -z9 update -d:pserver:anonymous at cvs.gstreamer.sourceforge.net:/cvsroot/gstreamer gstreamer
[00:15] <thomas> so I got the order wrong I guess
[00:16] <omega_> no, just cd into the gstreamer directory
[00:16] <omega_> and run cvs -z9 update -dPA
[00:16] <omega_> the working copy has the repository path built into it
[00:16] <thomas> ok that works
[00:16] <thomas> does that mean I have to do a full recompile now ?
[00:16] <wtay> you'll see lots of files being patched..
[00:16] <wtay> yes
[00:16] <omega_> run ./autogen.sh first, of course
[00:17] <thomas> god... that'll take ages again.
[00:17] Action: omega_ wonders if autogen.sh should rm -f config.cache ?
[00:17] <wtay> did you change configure.in yourself?
[00:17] <thomas> not yet
[00:17] <thomas> what should I change ?
[00:17] <wtay> omega_: yes please
[00:17] <wtay> thomas: just a sec.. I'll commit the configure.in too
[00:17] <omega_> stw showed me that KDE has some hacks to speed up the config.status bit with perl, I'll look into those at some point
[00:17] <wtay> I'm doing a full compile (3mins)
[00:18] Action: omega_ needs more RAM
[00:18] Action: wtay need more CPUs
[00:18] Action: hadess needs a g4 733
[00:18] Action: omega_ wants a titanium powerbook
[00:18] <thomas> another dumb irc question :
[00:19] <thomas> how do you do those actions ?
[00:19] <omega_>  /me
[00:19] <wtay>    /me
[00:19] <z-> hm
[00:19] <omega_>    /me
[00:19] <omega_>  /hehe
[00:19] Action: thomas wants his work pc
[00:19] <z-> i get SIGSEGV in gst_caps_append
[00:19] <omega_> in videosink.so, right ?
[00:19] <z-> i dont think so
[00:19] <omega_> make clean in plugins/videosink
[00:19] <z-> ok..
[00:19] <omega_> also icecastsend, if you haven't updated recently
[00:19] <wtay> I should remove the old videosink
[00:19] <z-> actually, i just checked it all out
[00:20] <omega_> wtay: we should modify the old padfactory stuff to have a #warning or #error in it ?
[00:20] <omega_> hmmmm
[00:20] <wtay> omega_: everything is updated now
[00:20] <omega_> ok
[00:20] <wtay> except for some make check programs (I have them fixed locally)
[00:20] <wtay> ok
[00:20] <wtay> configure.in comming up
[00:20] <omega_> oooh, thread interlock problem was a typo
[00:21] <wtay> phew
[00:21] <wtay> NFO(30714:-1):gst_plugin_load_all:238: loaded 64 plugins with 83 elements and 10 types
[00:21] <wtay> whoohoo!
[00:21] <omega_> if (spinning) FLAG_SET
[00:21] <omega_> else FLAG_SET
[00:21] <omega_> oops
[00:21] <wtay> doh
[00:21] <thomas> 64 plugins ? where are they ?
[00:21] <thomas> i only have 52
[00:21] <thomas> boohoo
[00:22] <omega_> you probably don't have all the dependencies installed
[00:22] <omega_> so it doesn't build some of the plugins
[00:22] <thomas> that's right
[00:22] <wtay> new revision: 1.132; previous revision: 1.131
[00:22] <omega_> maybe we should have a *** notice: plugin xyz won't build on your machine *** from configure.in ?
[00:22] <thomas> say... would it be very hard to make a gstreamer lite with only audio ?
[00:22] <omega_> nope
[00:22] <wtay> or a gstreamer ultra light with no plugins..
[00:22] <omega_> that's what rr is shipping
[00:23] <omega_> so far..
[00:23] <thomas> wtay: i started compiling anyway
[00:23] <wtay> eventually all the plugins go out of the core..
[00:23] <omega_> not all
[00:23] <thomas> what did you change in it ?
[00:23] <wtay> configure.in?
[00:23] <thomas> yes ?
[00:23] <omega_> did you add volenv to plugins/filters/Makefile.am ?
[00:24] <wtay> I added the volenv/passthrough Makefiles to the output targets
[00:24] <thomas> ok, so I can do that manually for now
[00:24] Action: wtay thinks the amount of plugins is getting ridiculous
[00:24] <omega_> quite
[00:24] <thomas> why ? does it slow down things ?
[00:25] <thomas> I can think of at least ten more that are useful
[00:25] <omega_> a lot of them should be merged into a single plugin, with groups of related filters
[00:25] <thomas> and that's only for audio
[00:25] <wtay> ah I checked in a new mp3 decoder: mad, it should work on ARM
[00:25] <omega_> thomas: more than that <g>
[00:25] Action: thomas is getting something to drink while compiling
[00:25] Action: thomas forgot to return a video
[00:25] <omega_> oops
[00:26] <wtay> keep it, and never go back to that shop :)
[00:26] <omega_> heh
[00:26] <wtay> I implemented the bufferpool passing...
[00:26] <omega_> cool
[00:27] <wtay> will try to get mpeg2dec use it...
[00:27] <wtay> I'm not sure if I can allocaed multiple xvimages though...
[00:27] <omega_> no, Xv sucks
[00:27] <wtay> only one?
[00:27] Action: rdj ends up using almost every line of code in gstmediaplay.c
[00:27] <omega_> it's irrelevant, Xv sitll copies
[00:27] <omega_> doh
[00:28] <wtay> only less copy never hurts..
[00:28] <omega_> wtay: huh?
[00:28] <omega_> oh, right
[00:28] <wtay> yeah well, it's the idea that counts..
[00:28] Action: thomas is off to slashdot while waiting
[00:30] <wtay> hadess: you drink beer on every picture :)
[00:30] <omega_> wtay: that surprises you?
[00:30] <hadess> wtay: and i'm drinking right now
[00:30] <wtay> hehe
[00:30] Action: omega_ isn't surprised <g>
[00:30] <wtay> oh no, the matches... <g>
[00:31] <omega_> hehehehee
[00:31] <wtay> http://gstreamer.net/guadec-pics2/dscn0021.jpg
[00:31] <omega_> thomas, rdj, z-: never give zaheer a box of matches and some spare time.  bad idea.
[00:31] <hadess> omega_: and absolutely _no_ pictures you're drinking
[00:31] <omega_> hadess: yup
[00:31] <omega_> wheeee!
[00:32] <thomas> compile bombed on esdsink
[00:32] Action: wtay sees zaheer making an osssink :)
[00:32] <omega_> eh?
[00:32] <thomas> esdsink.c:51: parse error before `<'
[00:32] <thomas> esdsink.c: In function `gst_esdsink_get_type':
[00:32] <thomas> esdsink.c:175: `gst_esdsink_class_init' undeclared (first use in this function)
[00:32] <wtay> hmm
[00:33] <thomas> <<<<<<< esdsink.c
[00:33] <wtay> show me that line..
[00:33] <thomas> that's what that line says
[00:33] <omega_> rm -f esdsink.c
[00:33] <omega_> cvs update esdsink.c
[00:33] <thomas> maybe I shouldn't have interrupted autogen.sh at first ?
[00:33] <thomas> ;)
[00:33] <wtay> a cvs conflict
[00:33] <thomas> oh right...
[00:33] <thomas> i remember trying to add a width cap to esd
[00:33] <wtay> you probably changed it
[00:34] <z-> hm, does osssink in gstmediaplay segfault for anyone else?
[00:34] <wtay> z-: yes
[00:34] <z-> ok :)
[00:34] <wtay> caps nego problems
[00:34] <wtay> I think
[00:34] <omega_> works for me
[00:34] <wtay> eh?
[00:35] <omega_> I have some M's in cvs update, though
[00:35] <wtay> in esdsink?
[00:35] <rdj> woohoo
[00:35] <thomas> wtay: what are you asking whom ?
[00:35] <rdj> it works excellent now
[00:35] <wtay> rdj: cool
[00:36] <rdj> gstreamer inside nautilus
[00:36] Action: omega_ bows before rdj
[00:36] Action: rdj made a very dumb error
[00:36] <wtay> grr I almost got video in bonobo-media too...
[00:36] <rdj> omega_: hah! I just ported gstmediaplay :)
[00:36] Action: omega_ tells eazel to hand rdj a $5000 check
[00:36] <rdj> no gstreamer code by me
[00:36] <rdj> just porting... copy&pasting that is
[00:36] <omega_> even better <g>
[00:36] <wtay> rdj: why not link against gstplay.so?
[00:37] <rdj> wtay: ugh, it exists
[00:37] <rdj> :)
[00:37] <rdj> well that's easy now
[00:37] <omega_> gstmediaplay.c uses gstplay.so, right?
[00:37] <rdj> I did not change a single line in it
[00:37] <wtay> yeah
[00:37] Action: omega_ is rebuilding gstreamer from scratch on the netwinder at rr
[00:37] <rdj> oh? I thought it used the gstplay.c in it's own dir
[00:38] <omega_> hmm, maybe
[00:38] <rdj> oh never mind
[00:38] <rdj> it makes a lib out of it
[00:38] <wtay> rdj: nope current CVS installs gstplay.h and libgstplay.so
[00:38] <rdj> I see :)
[00:38] <omega_> oooooh!!!!!
[00:38] <wtay> or gstmediaplay.so, not sure..
[00:38] Action: wtay ducks
[00:38] <omega_> examples/autoplug/autoplug asdf.mp3 works in incsched!!
[00:38] <rdj> ok I'll change that soon, low priority
[00:38] <wtay> omega_: whooho
[00:39] <omega_> now /me has a massive merge to do
[00:39] <wtay> rdj: first get your money :)
[00:39] <wtay> omega_: you bet
[00:39] <rdj> hehehe :))
[00:39] <thomas> so what is nautilus ?
[00:39] <thomas> gnome's explorer ?
[00:39] <omega_> basically
[00:39] <omega_> see www.eazel.com
[00:39] <wtay> eyecandy
[00:39] <rdj> it's very nice
[00:39] <omega_> RAMhog
[00:40] <wtay> works pretty good here
[00:40] <rdj> wtay: nah, nice tech
[00:40] <rdj> bonobo = sweet
[00:40] <wtay> yup
[00:40] <wtay> actually gob is nice too
[00:40] <rdj> never messed with it
[00:40] <omega_> gob as in gtk object builder?
[00:40] <wtay> yup
[00:40] <thomas> wtay did you put the volenv tarball somewhere ?
[00:40] <wtay> has to use it for bonobo-media
[00:40] <wtay> thomas: it's in CVS
[00:41] <thomas> sure ? I can't get it
[00:41] <thomas> shouldn't be in cvs anyway
[00:41] <wtay> thomas: you should subscribe to gstreamer-cvs
[00:41] <thomas> it's not done yet
[00:41] <wtay> thomas: no prob, just send a patch
[00:41] <wtay> easier to manage in vs
[00:41] <wtay> s/vs/cvs
[00:42] Action: thomas feels bad about admitting not to have used patches much
[00:42] <thomas> wtay: so how can I get it then ?
[00:42] <wtay> thomas: you did cvs update -d?
[00:42] <omega_> since sourcexchange is down, someone will have to email eazel directly to see if that competition/prize is still valid
[00:43] <thomas> cvs update -d volenv.c
[00:43] <thomas> cvs server: nothing known about volenv.c
[00:43] <hadess> thomas: cvs update -Pd
[00:43] <wtay> thomas: nono, just go to the gstreamer topdir and do cvs update -Pd
[00:44] <omega_> doh
[00:44] <thomas> ok... that worked
[00:44] <thomas> I did keep my old filter.func and header
[00:44] <omega_> I didn't talk with Jim Getty's about the iPAQ stuff, specifically Xfree 4 on it
[00:45] <wtay> oops
[00:45] <omega_> I'll email him...
[00:45] <thomas> ok... so gstreamer-register coredumps...
[00:45] <thomas> hmmm... looks like it did on my swapchannels, let's update that
[00:46] <thomas> I can't, it's got the same errors
[00:46] <thomas> I'm recreating it then
[00:47] <wtay> thomas: same prob: do make clean in plugins/videosink...
[00:48] <wtay> hum, esdsink is broken.. fixing...
[00:49] <omega_> #%$@!#$!^
[00:49] <omega_> I just did an update -A in my incsched copy
[00:49] <thomas> how can you generate Makefile from Makefile.am ?
[00:50] <wtay> omega_: uh oh...
[00:50] <wtay> omega_: is it broken?
[00:50] <omega_> um, you could say that
[00:50] <wtay> do you have a backup?
[00:50] <omega_> um
[00:51] <omega_> not really
[00:51] <omega_> it's all there, but it won't be fun to fix
[00:51] <wtay> oh man
[00:51] Action: hadess wonders if he should put the radio from his computer or from the hifi 1 meter away
[00:52] <omega_> hadess: so when do we see a screenshot of soundbox?
[00:53] <hadess> omega_: http://www.hadess.net/guadec/soundbox-cddb-enabled.jpeg <- from 10 days ago
[00:53] <thomas> omega_: what's soundbox ?
[00:53] <omega_> hadess's itunes clone
[00:54] <wtay> hadess: heh cool, is that a gtk theme?
[00:54] <hadess> wtay: no, that's the app i'm coding
[00:54] <wtay> hadess: I mean the tab panels?
[00:55] <hadess> wtay: xenostep iirc
[00:55] <wtay> ok
[00:55] <wtay> cool
[00:55] <hadess> the notebook tabs will disappear in the final version
[00:56] <omega_> hadess: are you going to run soundbox past Anna and crew for a UI sanity check?
[00:56] <omega_> since we all know that Apple recently tossed their own guidelines out the window....
[00:57] <thomas> what is itunes ?
[00:57] <omega_> hadess: you have the url for your itunes shot handy?
[00:57] <hadess> omega_: i don't think it will be needed, i'm copying apple's ui to the label
[00:57] <wtay> that volenv filter works very well..
[00:57] <omega_> hadess: um, that's what I meant <g>
[00:57] <thomas> wtay: are you playing with it?
[00:57] <hadess> http://www.hadess.net/files/shots/20-3-2001.1.jpg
[00:57] <wtay> thomas: yeah :)
[00:57] <hadess> http://www.hadess.net/files/shots/20-3-2001.2.jpg
[00:57] <omega_> hadess: compare quicktime player to their own guidelines....
[00:58] <thomas> ok... I got printf DEBUG messages all over...
[00:58] <thomas> what's the standard gstreamer way of doing this ?
[00:58] <wtay> esdsink should be fixed now
[00:58] <hadess> omega_: even if it doesn't apply to their guidelines, they did user testing on it
[00:58] <omega_> hadess: so?
[00:59] <omega_> hadess: they did user testing on the quicktime player too....
[00:59] <wtay> thomas: GST_DEBUG (0, "testing %d", number);
[00:59] <hadess> omega_: i'll try to do some haker testing as well =
[00:59] <hadess> )
[00:59] <thomas> what's the 0 for ?
[00:59] <wtay> thomas: prints thread/cothread info as well as method etc..
[00:59] <hadess> fuck it i can't spell
[00:59] <wtay> thomas: it's a category
[01:00] <thomas> is that the bit mask you can set on the command line ?
[01:00] <thomas> i mean according to instead of is
[01:00] <wtay> run gstreamer-launch with --gst-debug-mask=-1 to see the debugging
[01:00] <omega_> hadess: http://www.iarchitect.com/qtimeno.htm
[01:00] <wtay> thomas: yup
[01:00] <thomas> yeah, I tested that
[01:00] <wtay> thomas: --help shows all of them, 0 is the lazy mans value :)
[01:01] <wtay> uhm and it has a bug... fixing...
[01:01] <omega_> wtay: we should eventually have a list of things like that (the 0's) that are invalid and easy to check for, and have a script that spews all that stuff at us
[01:01] <hadess> omega_: itunes has a much saner interface
[01:01] <omega_> hadess: one hopes
[01:02] <wtay> omega_: yup, I added GST_CAT_PLUGIN_INFO...
[01:03] <wtay> and NEGITIATION too
[01:04] <wtay> NEGOTIATION even
[01:04] <omega_> great, this unmerge isn't going to be easy
[01:04] <wtay> that sucks
[01:04] Action: omega_ slaps himself with a large slimy trout
[01:04] Action: wtay slaps omega_ with a larger trout
[01:05] <omega_> don't hit me too hard, or I'll end up all the way back in your timezone again
[01:05] <wtay> hehe
[01:05] <omega_> I'm still halfway there....
[01:06] Nick change: taazzzz -> taaz
[01:06] <omega_> yo
[01:06] <wtay> hey
[01:06] <taaz> hi
[01:07] <omega_> ok, there are backups of the mangled files, it seems
[01:07] Action: wtay is wondering how to get in the "top 10 most active projects" sourceforge list..
[01:08] <omega_> hack like mad, send lots of mail to the list
[01:08] <wtay> hmm mail to list list must be our problem :)
[01:08] <z-> commit one line at a time i guess
[01:08] <omega_> taaz: yeah, a wiki would work, but I'd rather be more formal about maintaining the live projects list
[01:08] <omega_> z-: yup <g>
[01:08] <omega_> like me, 5min ago
[01:09] <wtay> bah
[01:09] <wtay> isn't there a wike page allready for app development?
[01:10] <omega_> this is why we need a docs manager.... <g>
[01:11] Action: thomas starts wondering if it was a good idea to make the plugin stereo only...
[01:11] Action: wtay is adding bonobo-media to the wiki
[01:12] <omega_> once I get incsched merged (today, I hope, but I need a good list of regressions to try out), I'll get the chainfunc replacement stuff working, and ppl can start dealing with changing caps
[01:13] <wtay> changing caps?
[01:13] <omega_> well, repluggable chainfuncs, for whatever reason
[01:13] Action: hadess just tried tigert's rotating scripts for nautilus
[01:13] <hadess> on omega_'s photos
[01:14] <omega_> cool
[01:14] <wtay> omega_: I think you'll be quite feature complete if gstmediaplay still works on mpeg1 system/video
[01:14] <omega_> yup
[01:15] <taaz> need to increase email volume?  we should close this channel...
[01:15] <omega_> before I try to fully undo my -A fiasco, I'm going to make a full dupe of the working copy <g>
[01:15] <omega_> taaz: hrm, yeah
[01:16] <omega_> or just send out hourly IRC logs instead of daily <g>
[01:16] <taaz> yup
[01:16] <taaz> please don't ;)
[01:16] <wtay> oh yeah!
[01:17] <wtay> people can participate in IRC with the mailing list too.. I'll have the bot copy the mails into IRC etc.. <g>
[01:17] <omega_> hm
[01:17] <hadess> omega_: on the group photos ! i'm not drinking =)
[01:17] <omega_> heheheh
[01:17] <omega_> in 17 days of logs, I count 950KB
[01:18] Action: omega_ thinks we talk too much
[01:18] <wtay> hmm, that's a pretty heavy book already...
[01:18] <omega_> um, make that 15 days
[01:18] <taaz> oh my... this is going to get published?
[01:19] <taaz> hi mom!
[01:19] <z-> what's the syntax for disksrc in gstreamer-launch?
[01:20] <omega_> disksrc location=xyz
[01:20] <z-> thx
[01:20] <z-> i always forget that one :\
[01:20] Action: hadess is waiting for colorspace to be fixored
[01:20] Action: omega_ has the merge from hell waiting for him
[01:22] <taaz> i bet you wish you had an automatic test suite now
[01:22] <omega_> quite
[01:22] <omega_> taaz: what are you waiting for?
[01:22] <thomas> ok...
[01:22] <taaz> do you really want to know? ;)
[01:22] <z-> aha, i got it working, at last
[01:22] <thomas> volume envelope seems to work
[01:23] <thomas> and touch control points exactly
[01:23] Action: thomas starts cleaning code
[01:23] <omega_> taaz: no, but if you're going to keep pointing out the lack of a test suite, please, please help us fix it ;-)
[01:24] <taaz> omega_: i've been writing one for my school work ;)  its kinda neat... just hit go and everything tests itself.  easy to test if refactoring broke anything..
[01:24] <omega_> cUnit style?
[01:24] <taaz> some in pyunit, some in junit.
[01:24] <omega_> hmm
[01:25] <omega_> did you look at the stuff I messed with in docs/random/omega/testing/ ?
[01:25] Action: taaz hides
[01:25] <thomas> I wish fakesink would not print data
[01:25] <omega_> thomas: I agree, maybe it should be an arg that defaults to FALSE
[01:25] <wtay> hadess: f1x0r1ng colorspace..
[01:26] <hadess> wtay: th4nk j00
[01:26] <taaz> why do so many hackers have broken keyboards?
[01:26] <wtay> hadess: uhm, not easy
[01:26] Action: omega_ turns on the rot13 anti-31337 filter
[01:27] <thomas> so you only have plugin_loading and plugin_errors as debug output ?
[01:27] <thomas> hmm... that's not much.
[01:27] <omega_> ok, I think I've recovered from my bout of -A stupidity
[01:28] <wtay> hadess: care if it only compiles?
[01:28] <hadess> wtay: well, for tonight, i could just hack the Makefile.am
[01:28] <hadess> wtay: leave the real fix for another dya
[01:28] <wtay> okie
[01:30] <taaz> omega_: why do you need xml test descriptions?
[01:30] <omega_> eh?
[01:30] <taaz> that 'framework' file
[01:31] <omega_> seems like a sane way to describe thousands of tests, given a tool to help generate them (GUI, probably)
[01:32] <omega_> though if you can think of a better way to accomplish the same concepts, please let me know <g>
[01:33] <taaz> hmm... you're thinking large
[01:33] <omega_> yes
[01:33] <taaz> you'll need a test framework for the test framework too ;)
[01:33] <omega_> unit testing of gstreamer will run into the many hundreds, possibly thousand or more tests
[01:33] <omega_> um, yes
[01:34] <hadess> wtay: you have a test for the mad plugin in the tests ?
[01:34] <wtay> hadess: nope, use gstreamer-launch instead
[01:34] <hadess> wtay: replacing mpg123 by mad in the mp3 test should work, right ?
[01:35] <wtay> hadess: you can't use the mp3parser, for some reason mad fails to decode something then
[01:35] <thomas> (yikes)
[01:35] <hadess> wtay: ok. i need to strip this part out as well then
[01:35] <omega_> wtay: huh?
[01:35] <thomas> why doesn't this work :
[01:35] <thomas>       GST_DEBUG (MSG_MASK, "volenv : control point at sample %d\n",
[01:35] <thomas>                  stop_sample);
[01:35] <thomas> if I defined MSG_MASK as 0x00004000
[01:35] <thomas> ?
[01:36] <wtay> gstreamer-launch disksrc location=some.mp3 ! mad ! osssink works here
[01:36] <omega_> it's not a mask
[01:36] <thomas> how does it relate to the mask ?
[01:36] <wtay> it's an index
[01:36] <omega_> thomas: the debug value is a shifter
[01:36] <wtay> 1<<index wise
[01:36] <omega_> if DEBUG_MASK && 1<<index, print
[01:37] <omega_> wtay: what version of libmad are you using?
[01:37] <thomas> ok, so the index is the bit position fo the mask ?
[01:37] <omega_> header->sfreq doesn't exist in mine
[01:37] <omega_> yes
[01:38] <wtay> oh
[01:38] <wtay> sec..
[01:38] <omega_> thomas: it must be between 0 and 31
[01:38] <wtay> 0.12.5b or something I think
[01:38] <rdj> dumb question, but I need a test mpg / whatever as long as it contains video and can be played using gstreamer
[01:38] <omega_> I have 0.13.0b
[01:38] <thomas> omega_: I'm counting the 0's from the right times four and add a few for the actual number
[01:38] <omega_> rdj: with audio?
[01:39] <rdj> anyone have an url for one that is guaranteed to work?
[01:39] <rdj> omega_: yeah
[01:39] <omega_> gstreamer.net/media/AlienSong.mpg in 1min
[01:39] <rdj> or, that should work
[01:39] <rdj> ok
[01:39] <rdj> thanks
[01:39] <omega_> uploading now
[01:39] <wtay> and it's funny too :)
[01:39] <omega_> but only if it remains in sync....
[01:40] <rdj> I can only play things without video at the moment
[01:40] <rdj> I want to find out why, it might be just my brothers porn collection
[01:40] <rdj> :)
[01:41] <rdj> mp3's and such work fine now in nautilus, atleast I got that working today :P
[01:43] <omega_> hrm, examples/autoplug works, gstmediaplay doesn't
[01:43] <omega_> but I haven't merged HEAD in yet, so...
[01:43] <omega_> however, the error I get is not nice
[01:43] <omega_> cothread: trying to switch to same thread, legal but not necessary
[01:43] <wtay> hmm
[01:52] <omega_> heh, someone suggested that if I ever get a cat [or dog], I name it 'fsck' <G>
[01:54] <wtay> heh
[01:55] Action: wtay has apt-getted flac 0.9...
[01:55] <hadess> why the heck is gstreamer-register loading the plugins from the src dir instead of the installed ones ?
[01:56] <omega_> because you used autogen.sh, which adds --enable-plugin-srcdir
[01:56] <hadess> ok....
[01:56] <omega_> which btw is one of many things that arts doesn't have...
[01:56] <rdj> omega_: where did you place it? :P
[01:56] <omega_> rdj: which?
[01:57] Action: rdj still gets a 404
[01:57] <hadess> omega_: am i supposed to like it ?
[01:57] <omega_> should be in gstreamer.net/media/AlienSong.mpg
[01:57] <rdj>  gstreamer.net/media/AlienSong.mpg
[01:57] <omega_> oh
[01:57] <omega_> fixed
[01:57] <rdj> ah cool, thanks
[01:57] <omega_> media/ didn't exist, it gave that name to the file ;-(
[01:58] Action: thomas slaps himself on the forehead
[01:58] Action: thomas made things more complicated when it was easy
[01:58] <omega_> wtay: any idea why autoplug example would work but getmediaplay doesn't ?
[02:00] <wtay> omega_: autoplug doesn't do video?
[02:00] <omega_> no, it fails on an mp3 file in incsched1
[02:01] <wtay> hmm
[02:01] <wtay> any idea where or why?
[02:01] <omega_> still trying to figure that out
[02:02] <omega_> a scheduling issue, but I'm unsure why one would cause it and the other not
[02:02] <wtay> gstmediaplay uses two target elements audio/video to do autoplugging
[02:02] <omega_> and doesn't autoplug as well?
[02:02] <wtay> only one audio output AFAIK, checking
[02:03] <omega_> it creates a window for me
[02:03] <hadess> $ sudo gstreamer-register
[02:03] <wtay> oh yes...
[02:03] <hadess> <snip>
[02:03] <hadess> INFO(17596:-1):gst_plugin_load_absolute:368: loading plugin "/home/hadess/src/gstreamer/./plugins/au/.libs/libgstparseau.so"...
[02:03] <hadess> Segmentation fault
[02:03] <omega_> have you updated and rebuilt recently?
[02:03] <omega_> right now, that's a sign of old padfactory stuff still existing in some .so
[02:04] <hadess> it's sources from 4h ago
[02:04] <omega_> hadess: try a make clean;make in that dir, just in case
[02:04] <wtay> omega_: autoplug runs the top pipeline in the gtk idle loop while gstplay uses a thread...
[02:04] <omega_> hmmmm
[02:05] <hadess> still the same...
[02:05] <omega_> hadess: not sure, I don't have that problem here
[02:05] <omega_> wtay: don't you work tomorrow?
[02:05] <wtay> hadess: isn't ther a stale libgst.so in /usr/local/lib or something?
[02:05] <hadess> omega_: estaer monday
[02:05] <wtay> omega_: nope easters holiday
[02:05] <omega_> ah
[02:06] Action: omega_ struggles to stay awake, goes to get some more sugar
[02:07] <hadess> bah
[02:07] <hadess> time for bed
[02:07] <hadess> cya gang
[02:08] hadess (hadess at pc213-gui2.cable.ntl.com) left irc: sleep
[02:11] <rdj> well ok, no video in nautilus today, I can't find the error :P
[02:14] <wtay> rdj: too bad 
[02:15] <rdj> yeah well, tomorrow's another day :)
[02:18] <thomas> is there a good way to profile plugins ?
[02:19] <thomas> sorry for asking difficult questions at difficult times
[02:19] <wtay> hmm
[02:19] <thomas> wtay: is there an example of that callback function you mentioned back then ?
[02:19] <thomas> all I have to do now is get the sample rate
[02:19] <thomas> then I'm done for the moment
[02:19] <wtay> sec..
[02:20] <wtay> ok, xvideosink has it
[02:20] <wtay> the gst_pad_set_newcaps_function call and the gst_xvideosink_newcaps function
[02:21] <thomas> and it's called as soon as the capabilities are known, but before playing, right ?
[02:21] <wtay> right before the first buffer is passed to the element
[02:22] <wtay> while playing
[02:34] <wtay> omega_: any news on the arts plugin?
[02:39] <wtay> thomas: btw the esdsink should be fixed now... It stil had old caps...
[02:39] <thomas> wtay: ok
[02:39] <thomas> wtay: i think the envelope plugin is ready
[02:39] <thomas> wtay: it seems to work fine
[02:39] <wtay> cool
[02:40] <thomas> wtay: whatever I throw at it
[02:40] <thomas> wtay: tomorrow I'll check if the signalling still works
[02:40] <wtay> do you have a way of restting the plugin too?
[02:40] <thomas> wtay: resetting ?
[02:40] <wtay> like throw the current glist
[02:40] <thomas> wtay: no not yet
[02:40] <wtay> ok
[02:40] <thomas> wtay: would that be useful ?
[02:41] <thomas> wtay: you can turn the volume envelope off
[02:41] <thomas> wtay: no that's entirely correct
[02:41] <thomas> wtay: there's a envelope_active var
[02:41] <thomas> wtay: if it's unset, no changes will be made to the level
[02:41] <thomas> wtay: yeah that sounds right
[02:41] <thomas> it's late
[02:41] <wtay> yeah
[02:41] <wtay> I'm falling asleep...
[02:42] Action: thomas sighs and wonders how long it will take to get a mixing plugin working with variable src's
[02:43] <wtay> someone is doing that, so says the wiki..
[02:43] <thomas> wtay: do you want me to upload it like it is ?
[02:43] <wtay> sure
[02:45] <thomas> ok, it's back up at http://urgent.rug.ac.be/thomas/gst
[02:45] <wtay> ok
[02:49] <wtay> thomas: getting the hang of it :)
[02:50] <thomas> it's about time !
[02:50] <thomas> i like gstreamer
[02:50] <thomas> it's a nice way of doing things
[02:50] <wtay> good :)
[02:50] <wtay> damn my mp3 sounds weird
[02:50] <thomas> i just put some quick in and out fades
[02:51] <thomas> I was thinking of doing a stereo envelope as well
[02:51] <thomas> but first the mixing
[02:51] <thomas> how hard would it be to add real filters ?
[02:51] <thomas> they should have a "memory" of samples passed through
[02:51] <thomas> that doesn't seem easy
[02:51] <wtay> no idea
[02:52] <wtay> I'm an idiot with audio
[02:52] <wtay> commited
[02:52] <thomas> well, it works that way for video too, no ?
[02:52] <thomas> I mean, how do you do a fourier transform in video ?
[02:53] <wtay> I'm an idiot with video :)
[02:53] <thomas> 3 AM ?
[02:53] <thomas> time to get some sleep
[02:54] <thomas> jeez
[02:54] <wtay> yup, me too
[02:54] <thomas> well... maybe till tomorrow if I can get up
[02:54] <thomas> then clean up all the shit I put in today
[02:54] <wtay> ok, 'night
[02:54] Action: omega_ got hijacked
[02:55] <wtay> omega_: ?
[02:55] <omega_> intended to come right back, ended up having to deal with lots of stuff
[02:55] <wtay> ah
[02:57] <wtay> going to sleep now.. cya
[02:57] Nick change: wtay -> wtay-sleeping
[02:57] <omega_> l8r
[03:01] <thomas> sleep tight everyone
[03:01] thomas (thomas at urgent.rug.ac.be) left irc: I'm outta here!
[03:30] iGN_ (ign at login1.simplemente.net) joined #gstreamer.
[04:52] z- (dave at host213-122-121-172.btinternet.com) left irc: .
[06:00] ajmitch (me at p52-max1.bal.ihug.co.nz) joined #gstreamer.