IRC Logs

IRC <[email protected]>
Newsgroups gmane.comp.video.gstreamer.daily
Message-ID <[email protected]>
*******************************************************************
[03:06] sub_pop ([email protected]) left irc: "Client exiting"
[03:07] Lovechild ([email protected]) left irc: "Leaving"
[03:18] thomasvs ([email protected]) left irc: Read error: 110 (Connection timed out)
[03:20] thomasvs ([email protected]) joined #gstreamer.
[03:21] sub_pop ([email protected]) joined #gstreamer.
[03:45] Marsupilami23 ([email protected]) left irc: "I don't wanna grow up, I'm a Toys R Us Kid. There's a million toys at Toys R Us that I can play with!"
[03:56] herzi ([email protected]) left irc: Read error: 113 (No route to host)
[04:13] hyriand_ ([email protected]) joined #gstreamer.
[04:24] hyriand ([email protected]) left irc: Read error: 113 (No route to host)
[04:39] jaiserca ([email protected]) left irc: "Terminando cliente"
[04:40] jcsston ([email protected]) joined #gstreamer.
[04:58] mxpxpod ([email protected]) joined #gstreamer.
[05:13] BBB ([email protected]) joined #gstreamer.
[05:13] <BBB> whee!!!!!!!!!!!!!!!!!
[05:13] Action: BBB cheers
[05:13] <BBB> MY LAPTOP WORKS!
[05:15] <NFusi0n> woo woo
[05:15] Action: BBB will take his laptop to work tomorrow and do all updates (CVS, distro, ...) tomorrow
[06:40] thaytoo ([email protected]) left irc: "leaving"
[06:50] hyriand_ ([email protected]) got netsplit.
[06:50] nemoglobine ([email protected]) got netsplit.
[06:50] md` ([email protected]) got netsplit.
[06:50] hyriand_ ([email protected]) returned to #gstreamer.
[06:50] nemoglobine ([email protected]) returned to #gstreamer.
[06:50] md` ([email protected]) returned to #gstreamer.
[06:58] BBB ([email protected]) left irc: Read error: 110 (Connection timed out)
[07:01] hyriand_ ([email protected]) got netsplit.
[07:01] nemoglobine ([email protected]) got netsplit.
[07:01] md` ([email protected]) got netsplit.
[07:01] hyriand_ ([email protected]) returned to #gstreamer.
[07:01] nemoglobine ([email protected]) returned to #gstreamer.
[07:01] md` ([email protected]) returned to #gstreamer.
[07:05] ds-work ([email protected]) left irc: Read error: 60 (Operation timed out)
[07:43] sub_pop ([email protected]) left irc: "Client exiting"
[08:12] Nick change: mathrick|sleep -> mathrick|Uni
[08:42] jcsston` ([email protected]) joined #gstreamer.
[08:55] md` ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[08:55] md` ([email protected]) joined #gstreamer.
[09:05] jcsston ([email protected]) left irc: Read error: 110 (Connection timed out)
[09:10] alley_cat ([email protected]) joined #gstreamer.
[09:37] markey ([email protected]) joined #gstreamer.
[09:40] kmaraas ([email protected]) left irc: "Leaving"
[09:40] markey ([email protected]) left irc: Client Quit
[10:01] Nick change: jcsston` -> jcsston
[10:26] teuf ([email protected]) joined #gstreamer.
[10:27] <teuf> hi
[10:28] jcsston ([email protected]) left irc: "VDubMOD watchout TCME is coming!"
[10:34] thomasvs ([email protected]) left irc: Read error: 110 (Connection timed out)
[10:36] jdahlin ([email protected]) joined #gstreamer.
[10:41] wheels ([email protected]) left irc: "work"
[10:47] dolphy ([email protected]) joined #gstreamer.
[10:49] <dolphy> morning
[10:49] <teuf> hi dolphy
[10:51] sysan ([email protected]) joined #gstreamer.
[10:55] iain ([email protected]) joined #gstreamer.
[10:59] thomasvs ([email protected]) joined #gstreamer.
[11:02] iain ([email protected]) left irc: "leaving"
[11:03] pb_ ([email protected]) joined #gstreamer.
[11:24] markey ([email protected]) joined #gstreamer.
[11:26] Shoragan ([email protected]) joined #gstreamer.
[11:27] sysan ([email protected]) left #gstreamer.
[11:33] <jdahlin> Why is gst-launch-ext using vorbisfile instead of oggdemux ! vorbisdec?
[11:33] <thomasvs> because nobody bothered to update it yet
[11:34] <thomasvs> don't forget to add float2intnew :)
[11:35] <jdahlin> after or before?
[11:44] <thomasvs> after
[11:53] thaytan ([email protected]) joined #gstreamer.
[12:05] sxpert|work ([email protected]) joined #gstreamer.
[12:06] natyaht ([email protected]) joined #gstreamer.
[12:06] sxpert|work ([email protected]) left irc: Remote closed the connection
[12:07] sxpert|work ([email protected]) joined #gstreamer.
[12:07] thaytan ([email protected]) left irc: "leaving"
[12:07] Nick change: natyaht -> thaytan
[12:10] sxpert|work ([email protected]) left irc: Remote closed the connection
[12:11] sxpert_work ([email protected]) joined #gstreamer.
[12:12] sxpert_work ([email protected]) left irc: Remote closed the connection
[12:16] Action: bilboed is back (gone 10:01:06)
[12:23] n3m0gl081n3 ([email protected]) joined #gstreamer.
[12:51] sysan ([email protected]) joined #gstreamer.
[12:54] <sysan> Any idea which plugin I am missing I get
[12:54] <sysan> WARNING **: Unrecoverable syntax error while parsing pipeline
[12:54] <sysan> ERROR: pipeline could not be constructed: no element "filesrc".
[12:54] <thomasvs> sysan: did you run gst-register ?
[12:55] <sysan> yep 
[12:56] Rotty ([email protected]) joined #gstreamer.
[12:56] <sysan> it works with gst-launch osssrc ! osssink
[13:26] earthw0rm ([email protected]) left #gstreamer (""I allege that SCO is full of it." - Linus Torvalds").
[13:43] markey ([email protected]) left irc: "leaving"
[13:46] thaytan ([email protected]) left irc: Read error: 113 (No route to host)
[13:48] thaytan ([email protected]) joined #gstreamer.
[13:51] sysan ([email protected]) left #gstreamer.
[14:13] pb_ ([email protected]) left irc: "bbl"
[14:19] <teuf> thomasvs: your mail to d-d-l doesn't have an attachment...
[14:21] <bilboed> a project dev question : is there a free online groupware (schedules, tasks, publication, ... (all but code)) site better than sourceforge ?
[14:22] mathrick|Uni ([email protected]) left irc: Read error: 60 (Operation timed out)
[14:26] Action: bilboed is away: I'm busy
[14:26] sysan ([email protected]) joined #gstreamer.
[14:43] n3m0gl081n3 ([email protected]) left irc: "IceChat IRC Client - Download at www.IceChat.net"
[14:58] markey ([email protected]) joined #gstreamer.
[15:13] Action: bilboed is back (gone 00:46:36)
[15:18] ChriHJW_log ([email protected]) joined #gstreamer.
[15:19] Action: bilboed is away: gone to school
[15:21] ChrisHJW_log ([email protected]) left irc: Read error: 60 (Operation timed out)
[15:35] Nick change: hyriand_ -> hyriand
[15:35] nemoglobine ([email protected]) left irc: Remote closed the connection
[15:36] BBB ([email protected]) joined #gstreamer.
[15:45] sysan ([email protected]) left #gstreamer.
[15:55] nemoglobine ([email protected]) joined #gstreamer.
[15:57] bilboed_school ([email protected]) joined #gstreamer.
[16:02] Shoragan ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[16:03] bilboed ([email protected]) left irc: Read error: 110 (Connection timed out)
[16:03] Shoragan ([email protected]) joined #gstreamer.
[16:03] bilboed ([email protected]) joined #gstreamer.
[16:04] BBB ([email protected]) left irc: Remote closed the connection
[16:18] tromey ([email protected]) joined #gstreamer.
[16:36] jaiserca ([email protected]) joined #gstreamer.
[16:51] Nick change: markey -> AFK
[16:55] jcsston ([email protected]) joined #gstreamer.
[17:02] yippi ([email protected]) joined #gstreamer.
[17:07] jcsston ([email protected]) left irc: "VDubMOD watchout TCME is coming!"
[17:11] <jdahlin> When using mad ! spider ! osssink, I get errors like this GStreamer-WARNING **: could not load valid typefind function for feature 'image/*'
[17:11] <jdahlin> err, not mad, just a filesrc pointing to a mp3
[17:21] kwm ([email protected]) joined #gstreamer.
[17:51] dolphy ([email protected]) left irc: "Network down, IP Packets delivered via UPS"
[17:58] KoRnouille ([email protected]) joined #gstreamer.
[17:58] ibsteveog ([email protected]) joined #gstreamer.
[18:04] Nick change: AFK -> markey
[18:11] teuf ([email protected]) left irc: "Client exiting"
[18:16] iain ([email protected]) joined #gstreamer.
[18:17] sublett ([email protected]) joined #gstreamer.
[18:18] pmfp ([email protected]) joined #gstreamer.
[18:23] md` ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[18:28] lupusBE ([email protected]) joined #gstreamer.
[18:34] jcsston ([email protected]) joined #gstreamer.
[18:38] trow ([email protected]) left irc: "Leaving"
[18:42] bilboed_school ([email protected]) left irc: "Leaving"
[18:50] BBB ([email protected]) joined #gstreamer.
[18:50] <BBB> thomasvs: you would really be my hero-of-the-day if you could fix the "offline docs build"
[18:51] <thomasvs> BBB: if you have fedora core 1 and install my openjade rpm it just works
[18:51] <BBB> I have RH9, sorry
[18:51] <thomasvs> BBB: other than that, it boils down to broken catalogs
[18:51] <jdahlin> maybe you can rebuild the srpm
[18:51] <BBB> good idea...
[18:51] <thomasvs> if you give me access to your machine, however, I can check
[18:51] <BBB> I'm behind a gateway (local network), so I'm affraid that's not possible... :/
[18:51] <BBB> where can I find your srpms?
[18:52] <BBB> I can send you my catalog file and install your srpm, then we have the same setup
[18:52] <thomasvs> which srpm ?
[18:52] <BBB> openjade
[18:52] <thomasvs> there are two issues
[18:52] <thomasvs> the openjade rpm fixes some messup in fedora core 1's resolver through proxies
[18:52] <thomasvs> the other issue is "the catalog doesn't have the files that the docbook xml file needs"
[18:52] <thomasvs> which is due to a catalog problem
[18:53] <thomasvs> but I don't know how catalogs work :)
[18:53] <BBB> neither do I ;)
[18:53] <BBB> I tried setting up some stuff to make it work locally, but didn't get far :(
[18:53] <thomasvs> well
[18:53] <thomasvs> I think you can make it work locally by not specifying an online URL for the system id, but a local one
[18:54] <BBB> where's the systemID
[18:54] <BBB> >
[18:54] <BBB> ?
[18:55] <thomasvs> second line inside the main xml file
[18:56] Action: BBB tries that
[18:56] Rotty ([email protected]) left irc: "Bye"
[18:56] Rotty ([email protected]) joined #gstreamer.
[18:56] Action: BBB goes offline to try
[18:56] <BBB> brb
[18:56] <BBB> :)
[19:00] <BBB> seems to work :)
[19:00] <BBB> so do you know how to 'emulate' this in the catalog file so I don't need to have a changed pwg.xml (w.r.t. the CVS one)?
[19:01] <thomasvs> no, that's the problem
[19:01] <BBB> hm...
[19:01] <thomasvs> I hope to spend some time on it when i upgrade to fc2 test 1
[19:01] <BBB> ok, well I'll leave it at this then
[19:01] <thomasvs> because I'm assuming I'llh ave the same problem
[19:01] Action: BBB cannot compile totem
[19:01] <BBB> and gst-player still doesn't display video
[19:02] sublett ([email protected]) left irc: Read error: 113 (No route to host)
[19:02] <BBB> am I the only one having this issue?
[19:02] <taaz> thomasvs: are there some inconsistancies in the doc changes?  like "this is 0.7.5" then "changes since 0.7.5"  "bugs fixed since 0.7.5"?  either should be since 0.7.4 or in 0.7.5
[19:02] <thomasvs> did I screw up again ? where ?
[19:04] <taaz> plugins/NEWS
[19:04] <thomasvs> great
[19:04] <thomasvs> too late :)
[19:04] <taaz> gstreamer/NEWS too
[19:04] md` ([email protected]) joined #gstreamer.
[19:05] <thomasvs> taaz: thanks for ruining my day :)
[19:05] <taaz> heh ;)
[19:05] <taaz> i'm sure people can figure it out
[19:08] <taaz> i suppose we still don't have an official gstreamer pgp key?
[19:08] <thomasvs> no, we don't
[19:08] <thomasvs> maybe we should do that soon
[19:08] Action: bilboed is back (gone 03:05:33)
[19:09] Action: bilboed is away: bbl
[19:09] #gstreamer: mode change '+o thomasvs' by ChanServ!ChanServ@services.
[19:09] Topic changed on #gstreamer by [email protected]: GStreamer "In A Bar" 0.7.5 and GStreamer Plugins "Under The Sea" 0.7.5 are out - see http://gstreamer.freedesktop.org/
[19:10] <thomasvs> guess I need to do a gst-ffmpeg release too now
[19:11] ds-work ([email protected]) joined #gstreamer.
[19:11] <taaz> .asc files would fit in the src dir nicely next to the .md5s
[19:11] <thomasvs> yeah, they would
[19:11] <jdahlin> congrats thomasvs
[19:11] <thomasvs> man
[19:11] <jdahlin> :-)
[19:11] <thomasvs> each time I think "releasing will be easy this time"
[19:11] Action: taaz thinks .sig is better naming than .asc...
[19:12] <thomasvs> each time it just takes shitloads of time
[19:26] sxpert_work ([email protected]) joined #gstreamer.
[19:27] jcsston ([email protected]) left irc: "VDubMOD watchout TCME is coming!"
[19:28] BBB ([email protected]) left irc: "Client exiting"
[19:31] <KoRnouille> hi
[19:31] teuf ([email protected]) joined #gstreamer.
[19:31] <KoRnouille> in python, how to I set a window that's the size of the screen_width + screen_height ?
[19:32] <jdahlin> KoRnouille: pygtk?
[19:32] <KoRnouille> I have something like this
[19:32] <KoRnouille> self.window.set_usize(gtk.screen_width(), gtk.screen_height())
[19:32] <KoRnouille> yeah, pygtk
[19:33] <jdahlin> use set_size_request
[19:33] <jdahlin> there is a #pygtk channel FYI
[19:33] <KoRnouille> aaah ok, thanks
[19:34] <mxpxpod> thomasvs: I get an error compiling gstreamer-0.7.5
[19:34] <KoRnouille> jdahlin: the channel is pretty empty FYI
[19:34] <KoRnouille> :)
[19:34] <thomasvs> mxpxpod: tell met it's not true
[19:34] <thomasvs> mxpxpod: what error ?
[19:34] <mxpxpod> ../../gst/.libs/libgstreamer-0.7.so: undefined reference to `g_value_take_boxed'collect2: ld returned 1 exit status
[19:35] smoke ([email protected]) joined #gstreamer.
[19:35] <thomasvs> mxpxpod: you need glib 2.3.0
[19:36] <mxpxpod> then you need to update the configure.in
[19:36] <mxpxpod> GLIB2_REQ="2.2"
[19:36] <mxpxpod> AC_SUBST(GLIB2_REQ)
[19:36] <thomasvs> yep, just did
[19:36] <mxpxpod> ok
[19:37] <mxpxpod> so it's fixed in the tarball?
[19:37] <thomasvs> nope
[19:37] <thomasvs> the tarball is already out
[19:37] <mxpxpod> suck
[19:38] <thomasvs> we can do a new release soon enough if need be
[19:38] <mxpxpod> thomasvs: because I'm sure you're going to get that question a lot
[19:40] <thomasvs> mxpxpod: yeah
[19:40] Topic changed on #gstreamer by [email protected]: GStreamer "In A Bar" 0.7.5 and GStreamer Plugins "Under The Sea" 0.7.5 are out - see http://gstreamer.freedesktop.org/ - you need glib 2.3 for GStreamer 0.7.5
[19:40] <mxpxpod> thomasvs: not trying to be an ass
[19:41] <bitshifter> thomasvs: just out of curiosity: what are the chances that you bump up the requirements to glib-2.4 once it's out?
[19:41] <thomasvs> bitshifter: not sure - but given we need 2.3 anyway now, pretty good
[19:41] <thomasvs> esp. since we want to use the gatomic stuff
[19:42] <bitshifter> great
[19:43] <ds-work> we use glib-2.3?
[19:43] <thomasvs> somebody added g_value_takes_boxed
[19:44] <ds-work> that's a stupid reason to require glib-2.3
[19:45] <thomasvs> I know
[19:46] <thomasvs> I just found out myself because mxpxpod reported it
[19:46] <thomasvs> if we want to revert, we can do so for the next release
[19:46] <thomasvs> decide that one for me, I'm going home :)
[19:46] <mxpxpod> thomasvs: later
[19:52] casaxno ([email protected]) joined #gstreamer.
[19:55] <taaz> KoRnouille: #pygtk is on irc.gnome.org if you hadn't figured that out yet ;)
[19:56] <jdahlin> taaz: it's here too
[19:56] <taaz> it is?
[19:56] <taaz> hmm.
[19:56] <KoRnouille> taaz: thanks, I did. jdahlin told me
[19:56] <KoRnouille> hey, I have more a python question, more than a gstreamer question
[19:56] <KoRnouille> about writing phylosophy
[19:57] <KoRnouille> I'm seeing many examples around (and in python it self) using class bla: def blabla(), etc...
[19:57] <KoRnouille> in my video_wall case
[19:58] <KoRnouille> should I create a class window, to create my window, and then a class pipeline, to build my pipeline ?
[19:58] <KoRnouille> or just 1 class containing def window and def pipeline ?
[20:03] iamthelukas ([email protected]) joined #gstreamer.
[20:04] <iamthelukas> i get this errror: DEBUG: gstgconf: error parsing pipeline osssink
[20:04] <iamthelukas> Unrecoverable syntax error while parsing pipeline
[20:04] <iamthelukas> GConf audio sink not found, using osssink
[20:04] <iamthelukas> while trying to play rhythmbox
[20:09] thomasvs ([email protected]) left irc: No route to host
[20:13] jdahlin ([email protected]) left irc: "Klienten avslutas"
[20:33] Company ([email protected]) joined #gstreamer.
[20:38] BBB ([email protected]) joined #gstreamer.
[20:38] Action: BBB looks for dolphy
[20:46] iamthelukas ([email protected]) left irc: Read error: 110 (Connection timed out)
[20:46] sub_pop ([email protected]) joined #gstreamer.
[20:46] sbw_ibiblio ([email protected]) left irc: Read error: 110 (Connection timed out)
[20:48] apoc ([email protected]) joined #gstreamer.
[20:48] <apoc> yo
[20:49] <BBB> hi
[20:52] <BBB> nice, SVQ3/SVQ1/divx/everything works
[20:52] <BBB> now what's next?
[20:54] <apoc> BBB : got a new laptop ?
[20:54] <BBB> new screen
[20:55] gatto ([email protected]) joined #gstreamer.
[20:56] <BBB> ... mplayerhq.hu CVS is down...
[20:56] <BBB> *sigh*
[20:58] Action: BBB continues autotoolification of ffmpeg
[20:59] <ds-work> can we take ramon's CVS access away?
[20:59] <BBB> sure, but why?
[21:00] casaxno ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[21:00] <ds-work> so he stops doing things that are Just Plain Wrong?
[21:01] <BBB> my current opinion is pretty much that he will make a lot of mistakes and that I'm just hoping that in the end it will be fine... if it's not according to our standards/needs, we will not use his RTP stuff in our CVS... and apart from that, as long as he doesn't touch RTP stuff... well... whatever
[21:01] <BBB> er, !RTP stuff
[21:02] <ds-work> could you try running videotestsrc ! xvimagesink in valgrind?
[21:02] <ds-work> I've noticed that valgrind tends to catch problems that cause gdb to go crazy
[21:02] <BBB> yes
[21:03] <BBB> RUNNING pipeline ...
[21:03] <BBB> Segmentation fault
[21:03] <BBB> hm
[21:03] <BBB> valgrind ~/projects/gstreamer/gstreamer/tools/.libs/lt-gst-launch videotestsrc ! xvimagesink
[21:05] <Company> disInstr: unhandled instruction bytes: 0xF 0xD 0x8A 0xCC
[21:05] <Company> w00t?
[21:05] <BBB> what's that?
[21:05] <ds-work> that's an mmx/sse instruction, iirc
[21:05] <ds-work> what CPU?
[21:05] <Company> athlon
[21:05] <BBB> mine?
[21:06] <BBB> P3
[21:06] <Company> -march=athlon-xp was never a problem though :/
[21:07] <ds-work> what is disInstr from?
[21:08] <Company> that's a valgrind error
[21:08] <Company> (sorry if i didn't make it clear)
[21:17] <ds-work> -march=athlon-xp doesn't change the set of instructions used by the compiler
[21:17] ChrisHJW ([email protected]) joined #gstreamer.
[21:18] gatto ([email protected]) left irc: "Lost terminal"
[21:18] n3m0gl081n3 ([email protected]) joined #gstreamer.
[21:21] jimmy_dean_ ([email protected]) left irc: "Carrier Signal Lost"
[21:26] mathrick ([email protected]) joined #gstreamer.
[21:31] <Company> -march does change the set of instructions used
[21:32] <ds-work> Company: only -march=i686
[21:33] sublett ([email protected]) joined #gstreamer.
[21:33] <ds-work> all other arch settings are equivalent to i386 or i686
[21:33] jcsston ([email protected]) joined #gstreamer.
[21:33] <Company> oh
[21:34] jcsston ([email protected]) left irc: Client Quit
[21:35] <Company> it works with valgrind cvs
[21:36] <Company> someone should use helgrind with gst :/
[21:37] <teuf> Company: last time I tried, rhythmbox ate something like 600MB for the initial import of my library using gstreamer for metadata ;)
[21:37] Action: BBB has a make-test sort of thing
[21:37] <BBB> in ffmpeg :)
[21:38] <Company> teuf: that's because schedulers aren't freed on pipeline destruction
[21:38] <BBB> ds-work: if I give you a patch, could you fix some ASM for me in ffmpeg?
[21:38] <apoc> Someone know where i can find some docs about float audio data, samples, ...
[21:39] jcsston ([email protected]) joined #gstreamer.
[21:41] <ds-work> BBB: I suppose
[21:41] <ds-work> BBB: is it the part that's currently disabled in gst-ffmpeg?
[21:41] <Company> valgrind has no problems with videotestsrc ! xvimagesink here
[21:41] <Company> apart from memleaks
[21:43] mathrick ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[21:44] mathrick ([email protected]) joined #gstreamer.
[21:46] <Company> do we have a properly set up (autogen, configure and friends) project for plugins in gst-sandbox?
[21:47] <ds-work> for gst-autoplug
[21:48] <Company> yeah, i didn'T want to use a working project
[21:48] <Company> just thought we might have a working dumptser
[21:48] <ds-work> no
[21:48] <ds-work> gst-bitrotten
[21:48] <ds-work> but I didn't care to set that up
[21:48] <Company> i wanted to develop in there ;)
[21:50] <ds-work> so far, we've been creating a new project for each [set of] plugin[s]
[21:50] Action: ds-work comments that gstreamer compiles just fine with glib-2.2.3
[21:50] <BBB> ds-work: no
[21:51] <BBB> ds-work: it's just latest CVS of ffmpeg, I've been developing against it (using quilt, very nice stuff!)
[21:51] <BBB> so I've got some patches here and there
[21:51] <BBB> I want to use those in gst-ffmpeg
[21:51] <ds-work> BBB: can we use quilt in conjunction with CVS?
[21:51] <BBB> and I want to do more work on gst-ffmpeg for new plugins, supporting existing codecs better, etc.
[21:51] <BBB> yes
[21:51] <BBB> quilt is a patchset manager
[21:51] <BBB> so it can be used for CVS, but you can also provide ptches to people who use tarballs (no CVS)
[21:52] <BBB> and you can maintain multiple patchsets per module
[21:52] <BBB> which is really cool
[21:52] <BBB> I currently have two
[21:52] <BBB> (one for matroska support in ffmpeg, one for autotoolification of ffmpeg)
[21:54] <BBB> anyway, the MMX issue (code doesn't compile using PIC) makes make distcheck fail
[21:54] <BBB> and I want to fix make distcheck
[21:54] <BBB> so I have something to send to the list and (assuming they reject it) to put in gst-ffmpeg
[21:54] <BBB> :)
[21:55] <ds-work> ok, check it in, and I'll look at it
[21:55] <ds-work> probably later tonight
[21:55] <BBB> no, I'll send it to you, you should first fix "vanilla" ffmpeg CVS compilation
[21:55] <BBB> if that works, I'll start messing around in gst-ffmpeg
[21:55] <BBB> (I hate gst-ffmpeg, but ohwell...)
[21:56] <ds-work> ok
[21:56] <BBB> so if you have a patch, just send it to me and I'll continue my work on it :)
[21:56] <BBB> if you want
[21:57] ramon ([email protected]) joined #gstreamer.
[21:57] <ramon> Hi.
[21:58] <ramon> I have just read the email of thomasvs saying not to use private branches for the development of UDPsrc.
[21:58] <ramon> Of course, I have no problem respecting an agreed on policy.
[21:58] <ramon> But I would like to listen other opinions.
[21:59] <BBB> ds-work: DCC doesn't work?
[21:59] <BBB> email then?
[21:59] <BBB> ramon: hm... isn't bugzilla or regularly posting patches to gst-devel a solution then?
[21:59] <BBB> that's how most of us do it
[21:59] <ds-work> BBB: yeah, email
[21:59] <BBB> [email protected]?
[22:00] <ds-work> BBB: yes
[22:00] <ds-work> ramon: please add a project to gst-sandbox if it's experimental code
[22:02] <ramon> What is exactly gst-sandbox? A CVS directory?
[22:02] <ds-work> yes
[22:02] sxpert_work ([email protected]) left irc: "Leaving"
[22:02] <ramon> Thanks.
[22:03] <BBB> it's a module for these sort of experimental stuff :)
[22:04] <BBB> I've got some applications in there, and developed the gnome volume control application in there, too, before it got into gnome-media
[22:04] <BBB> ds-work: sent
[22:06] sxpert_work ([email protected]) joined #gstreamer.
[22:06] <ramon> Thansk. I will see what I can do tonight.
[22:06] ramon ([email protected]) left irc: "Terminando cliente"
[22:07] <ds-work> BBB: how do I use quilt?  (In 3 lines or less :))
[22:07] <BBB> apt-get install quilt, quilt import /path/to/patch, quilt push -a -f
[22:07] <BBB> I think
[22:07] <BBB> or well, first do checkout ffmpeg
[22:08] <BBB> thomas has a PDF explaining it
[22:08] <BBB> and RPMs
[22:08] <BBB> and you can also just patch -p1 < /path/to/patch if you want, that works just as well
[22:08] <BBB> quilt is nothing more than a script collection around patch/diff/bla :)
[22:09] <BBB> (that was more than three lines! :o)
[22:13] <ds-work> thanks
[22:15] Action: bilboed is back (gone 03:06:35)
[22:16] ibsteveog ([email protected]) left irc: "quitting"
[22:19] <ds-work> i386/cputest.c:5:21: dsputil.h: No existe el fichero o el directorio
[22:23] <ds-work> hrm, nm
[22:24] <ds-work> BBB: I need some m4 files please
[22:25] <BBB> oh... hm... another issue
[22:25] <BBB> which ones?
[22:25] <BBB> I think I use SDL and one other, right?
[22:25] ChrisHJW ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[22:26] <ds-work> AC_FF_ALLOW_DISABLE
[22:26] <BBB> ft2, imlib
[22:26] <ds-work> AC_FF_ALLOW_OPT 
[22:26] <BBB> that's in configure.ac
[22:26] <ds-work> ok
[22:26] <BBB> grep AC_DEFUN configure.ac :)
[22:26] <ds-work> I'm getting strange errors then
[22:27] <BBB> like?
[22:27] ChrisHJW ([email protected]) joined #gstreamer.
[22:27] <BBB> oh, btw, I really don't want to use glib-2.3/gtk-2.3 :(
[22:27] Nick change: mathrick -> mathrick|sleep
[22:27] Action: BBB is very unhappy about installing unstable dependencies
[22:27] Zeenix ([email protected]) joined #gstreamer.
[22:28] <ds-work> http://www.schleef.org/~ds/ffmpeg.log
[22:28] <Zeenix> hello
[22:31] <BBB> hm
[22:31] <BBB> which autotool versions? :)
[22:31] <ds-work> 2.59
[22:31] <ds-work> 1.7.9
[22:32] markey ([email protected]) left irc: "leaving"
[22:32] Action: BBB has 2.57 and 1.7.8...
[22:32] <BBB> do you have sdl.m4?
[22:32] <BBB> maybe the other errors are caused by sdl missing
[22:33] <BBB> you can copy it from gst-plugins/m4/gst-sdl.m4 if you need it
[22:34] spyder482 ([email protected]) joined #gstreamer.
[22:37] <ds-work> heh
[22:37] <apoc> Company: around ?
[22:37] <ds-work> it was all because sdl.m4 was missing
[22:38] <ds-work> config.status: error: cannot find input file: libavcodec/mlib/Makefile.in
[22:38] <BBB> hm...?
[22:39] Action: BBB must have forgotten something there
[22:40] <BBB> see your mail
[22:40] <BBB> :)
[22:40] <BBB> I've added it to the patch locally and sent you the Makefile.am itself
[22:41] <ds-work> and tests/Makefile.am?
[22:42] <BBB> grr
[22:42] <BBB> coming up
[22:43] <BBB> File tests/Makefile.am is already in patch autotools
[22:43] <BBB> ?
[22:43] <ds-work> odd
[22:43] <BBB> well, I've emailed it anyway
[22:44] <BBB> Index: ffmpeg/output_example.c
[22:44] <BBB> Index: ffmpeg/tests/dsptest.c
[22:44] <BBB> Index: ffmpeg/tests/Makefile.am
[22:44] <BBB> aha
[22:45] arana ([email protected]) joined #gstreamer.
[22:45] Uraeus ([email protected]) joined #gstreamer.
[22:45] jcsston ([email protected]) left irc: Read error: 110 (Connection timed out)
[22:46] <ds-work> we have a patch in gst-ffmpeg to disable the code that causes register allocation problems
[22:46] <Uraeus> hello chiefs
[22:46] <BBB> yes, but I don't want it disabled
[22:46] <BBB> I want it to work ;)
[22:47] <BBB> if we disable random stuff, they'll definately never accept the patch
[22:47] <ds-work> I'll fix it, but that's the quickest way to continue on
[22:48] thaytoo ([email protected]) joined #gstreamer.
[22:49] <BBB> hm
[22:49] <BBB> k
[22:50] <Company> apoc: pong
[22:51] <BBB> mpegvideo_mmx_template.c do I need that patch too?
[22:52] spyder482 ([email protected]) left irc: Read error: 60 (Operation timed out)
[22:53] <BBB> hm, somehow I cannot seem to get that tests/Makefile.am imported in my patch
[22:53] <BBB> why not?
[22:53] <BBB> mpegvideo_mmx_template.c: In function `dct_quantize_SSE2':
[22:53] <BBB> mpegvideo_mmx_template.c:92: can't find a register in class `BREG' while reloading `asm'
[22:53] <BBB> :)
[22:53] apoc_ ([email protected]) joined #gstreamer.
[22:55] <BBB> I'm using the exact same patch as the one in gst-ffmpeg CVS
[22:56] pmfp ([email protected]) left irc: "Leaving"
[22:56] <Company> booo
[22:57] <Company> typedef (*my_func) (my_func *function);
[22:57] <Company> that's not allowed :(
[22:58] <BBB> sure is
[22:58] <BBB> we use that everywhere in GStreamer
[22:58] <BBB> but you need to specify a return value
[22:58] <BBB> a
[22:58] <BBB> return type
[22:58] <Company> even then it's not allowed
[22:59] <Company> you can't give your function a pointer to a function of same type
[22:59] <BBB> typedef GstCaps*                (*GstPadGetCapsFunction)        (GstPad *pad);
[22:59] <Company> yeah
[22:59] <Company> but not this:
[22:59] <Company> typedef GstCaps*                (*GstPadGetCapsFunction)        (GstPadGetCapsFunction *func);
[22:59] <BBB> aha! I get it
[23:00] <BBB> you need a dummy declaration first
[23:00] <BBB> as we do for structs
[23:00] <BBB> but I don't know how to do that ;)
[23:03] Zeenix ([email protected]) left #gstreamer ("The whole problem with the world is that fools and fanatics are always so certain of themselves, but wiser people so full of ).
[23:04] apoc ([email protected]) left irc: Read error: 110 (Connection timed out)
[23:04] Nick change: apoc_ -> apoc
[23:07] ChrisHJW ([email protected]) left irc: Read error: 110 (Connection timed out)
[23:14] <BBB> =============================================
[23:14] <BBB> ffmpeg-0.4.8.tar.gz is ready for distribution
[23:14] <BBB> =============================================
[23:14] <BBB> *proud*
[23:15] <BBB> so can I put this in gst-ffmpeg and play around with random features and break random stuff?
[23:16] <Uraeus> argh I built cvs gstreamer yesterday and was going to build plugins today...wrong wrong wrong :)
[23:16] <BBB> never do that ;)
[23:16] Action: BBB still uses yesterday's CVS since that works with glib-2.2.x
[23:17] <Uraeus> well glib isn't a problem for me, but plugins now depend on 0.7.5 gstreamer :)
[23:18] <BBB> aha ;)
[23:19] wheels ([email protected]) joined #gstreamer.
[23:20] Action: BBB feels like ffmpeg needs more love but this is enough for now
[23:21] <Uraeus> more ffmpeg love is only ok if it directly contributes to get more video formats playing perfectly in Totem :)
[23:22] <BBB> pff.
[23:22] <sxpert_work> Uraeus: /me has a nice "build gstreamer overnight" script
[23:23] <Uraeus> sxpert_work: that would crash with my nice 'turn of computer at night routine' :)
[23:23] Action: wheels has a nice "build GStreamer in about 30 minutes" computer.  :-)
[23:23] <sxpert_work> Uraeus: ah, I don't have such routine
[23:23] <thaytoo> wheels: 30 minutes? slowpoke ;)
[23:23] Action: sxpert_work has a "gstreamer needs several hours to build" carpc
[23:23] Action: teuf has a nice "use xine instead of gstreamer" habit
[23:24] Action: BBB slaps teuf
[23:24] Action: Uraeus aquires a nice beat up teuf for being a Quisling habit
[23:24] <wheels> thaytoo: Well, that was a random guess, but it's actually not a particularly new machine.  But well, GCC 2.95 helps.  :-)
[23:24] Action: wheels notes that RB is currently doing some evil macro thingers that don't work on 2.95.
[23:24] <teuf> wheels: which kind of evil things ?
[23:24] Action: sxpert_work has a nice "I'm against nuclear power" ecologist glued on a power-generating bike
[23:25] Action: Uraeus notes that gcc 2.95 is evil in itself
[23:25] <wheels> Uraeus: Yes.  But just less evil than everything that came later.  ;-)
[23:25] Action: sxpert_work shows uraeus his nice and shiny gcc 3.4
[23:26] Action: BBB needs gtk-2.3/glib-2.3 for gnome-media, gstreamer *and* totem now
[23:26] <BBB> :'(
[23:26] <wheels> Uraeus: I really don't know for C, but for C++ GCC 3 is really slow and every version seems to get a whole new batch of compilation errors.
[23:26] <Uraeus> BBB: apt-get upgrade as I did to Fedore Test
[23:27] jimmy_dean ([email protected]) joined #gstreamer.
[23:27] thomasvs ([email protected]) joined #gstreamer.
[23:28] <Uraeus> wheels: yes, but it actually tries to follow the C++ spec which you can't say for 2.95 :) and slow is the name of the game for C++ isn't it? ;)
[23:28] <Uraeus> hi thomasvs
[23:28] <sxpert_work> Uraeus: it's C++ that's evil
[23:29] <wheels> Uraeus: Well, 2.95 has known limitations but almost everything that it claims to support works.  3.x is just buggy.  As for speed, well, 3.x is roughly twice as slow as 2.95.
[23:29] Action: wheels makes sxpert_work wash his mouth out with soap.
[23:29] <sxpert_work> wheels: C++ is evil
[23:30] <tromey> there's always PCH, supposedly that helps
[23:30] <wheels> tromey: Well, but GCC's implementation is kind of crappy since it only allows one PCH for each compilation unit.
[23:30] <sxpert_work> well, there's ccache... that helps a lot :D
[23:31] <thomasvs> so, no big messups yet with the release ?
[23:31] <Uraeus> hmm, there is a VRML editor using Motif on freshmeat, I guess if you are doing a editor for a dead file format using a dead gui toolkit is the way to go
[23:32] <Uraeus> thomasvs: no, you did well, my blessings go with you
[23:32] <sxpert_work> Uraeus: lol
[23:32] <wheels> sxpert_work: Like I can go along with an argument that C++ is evil is you're assuming that C is evil too, but assuming that C is good and C++ evil is kind of silly.
[23:32] Zeenix ([email protected]) joined #gstreamer.
[23:32] <sxpert_work> wheels: I have my reasons /D
[23:33] <Uraeus> personally I am conviced that evil things are evil
[23:33] <wheels> Uraeus: the killers are to blame
[23:34] <Uraeus> yeah, they tend to be
[23:34] <BBB> thomasvs: it seems like sound profiles in gnome-media requires gtk-2.3 but the configure.in reqs say 2.2
[23:34] <BBB> thomasvs: can you update that?
[23:34] <thomasvs> BBB: ah, yeah, but not now
[23:34] <wheels> Uraeus: (Don't know if you're familiar with the old Dan Quale quote)
[23:34] <thomasvs> BBB: mail me that
[23:34] <BBB> it's not urgent ;)
[23:34] <BBB> bugzilla good too?
[23:34] <thomasvs> sure
[23:34] <thomasvs> assign to me
[23:34] <Uraeus> wheels: no, I try to keep sane and never quote Dan Quale :)
[23:34] <BBB> will do
[23:35] <wheels> Uraeus: "When I have been asked during these last weeks who caused the riots and the killing in L.A., my answer has been direct and simple: Who is to blame for the riots? The rioters are to blame. Who is to blame for the killings? The killers are to blame."
[23:38] <Uraeus> :)
[23:38] <Zeenix> ?
[23:38] <Uraeus> Zeenix: its about having insight into the obvious :)
[23:39] <wheels> Ah, the wisdom of the great Quayle.
[23:40] <Zeenix> are you people discussing that 'exit code' or someone (equating guns with spoons)?
[23:40] <Zeenix> s/or/of
[23:41] Action: bitshifter was going to throw in a Sartre quote for good measure, but now finds it kind of hard to top the Quayle quote
[23:41] <wheels> http://ourworld.compuserve.com/homepages/nagorski/quayle.htm
[23:44] <Zeenix> Uraeus: orkit aint good, only boys add me to their friend's list :(
[23:45] <Uraeus> Zeenix: well you can add some girls to your list then :) it doesn't need to be a one way train you know :)
[23:47] Action: Zeenix starts adding girls from Uraeus' list...
[23:47] Action: iain never got invited to orkit...
[23:48] <iain> but, I guess I never really cared that much about it
[23:48] <Zeenix> who is incharge of www these days?
[23:48] <Uraeus> to all the girls I used to know, who now are someone else's wife
[23:48] <iain> and I don't really see the point
[23:48] <Uraeus> iain: um, you added me to your friends list on orkut so I think you are there actually
[23:48] <iain> Uraeus: I did?
[23:48] <iain> Uraeus: I'm not on it
[23:48] <Zeenix> Uraeus: i like the idea of stealing away other's share :)
[23:49] <iain> Uraeus: unless someone has forged me
[23:49] <Uraeus> iain: ah, wrong Ian then :) ok, well that is maybe why nobody invited you, we all thought it was you who added us to the list :)
[23:49] <sxpert_work> wow, I already have one orkut fan :D
[23:50] <iain> Uraeus: I don't really want to be invited anyway, I don't see the point in it
[23:50] <Uraeus> iain: it lets you know when people have a birthday :)
[23:50] <iain> I know when the birthdays that are important to me are
[23:51] <iain> mine is on monday, roger's is in april, alex's in may and my sister's in july
[23:51] <iain> all the others my mother will remind me :)
[23:51] <Uraeus> iain: your mother know when I have my birthday?
[23:52] <iain> yup
[23:52] <iain> she knows everything
[23:52] <ds-work> thomasvs: where exactly do we use g_value_takes_boxed()?
[23:56] <Zeenix> the links to video-wall project page and the video-wall howto are both broken on our website
[23:59] <Zeenix> www seems to be above my head, how do i fix it (if possible)
[00:00] --- Fri Feb 27 2004
[00:00] <Misirlou> mmmm
[00:00] <Misirlou> Has anyone here played with libextractor?
[00:00] teuf ([email protected]) left irc: "Client exiting"
[00:03] Company ([email protected]) left irc: Remote closed the connection
[00:04] kwm ([email protected]) left irc: "Leaving"
[00:04] <Uraeus> night
[00:04] Uraeus ([email protected]) left irc: "Client exiting"
[00:15] Zeenix ([email protected]) left irc: "The whole problem with the world is that fools and fanatics are always so certain of themselves, but wiser people so full of 
[00:24] harshy ([email protected]) joined #gstreamer.
[00:25] n3m0gl081n3 ([email protected]) left irc: Read error: 60 (Operation timed out)
[00:30] BBB ([email protected]) left irc: "Client exiting"
[00:35] <thomasvs> ds-work: not sure, mxpxpod reported it
[00:36] <ds-work> thomasvs: I grepped the source and found nothing
[00:36] <mxpxpod> ds-work: lemme get the line and file
[00:36] <ds-work> and did a full compile, and found no errors with 2.2.3
[00:37] <thomasvs> mxpxpod: you can'tt give the line and file
[00:37] <mxpxpod> thomasvs: why not?
[00:38] <thomasvs> gcc -O2 -g -pipe -march=i386 -mcpu=i686 -o .libs/spidertest spidertest-spidertest.o -Wl,--export-dynamic  -pthread ../../gst/.libs/libgstreamer-0.7.so /usr/lib/libxml2.so -lpthread -lz -lm -lgobject-2.0 -lgthread-2.0 -lgmodule-2.0 -ldl -lglib-2.0 /usr/lib/libpopt.so../../gst/.libs/libgstreamer-0.7.so: undefined reference to `g_value_take_boxed'collect2: ld returned 1 exit statusmake[4]: *** [spidertest] Error 1make[4]: Leaving dir
[00:38] <thomasvs> ectory `/usr/src/rpm/BUILD/gstreamer-0.7.5/gst/autoplug'
[00:38] <thomasvs> mxpxpod: because it's a link-time error
[00:38] herzi ([email protected]) joined #gstreamer.
[00:38] <mxpxpod> ohhh
[00:38] <thomasvs> ds-work: that's from a friend who makes packages
[00:39] <ds-work> thomasvs: that's someone else's problem, not ours
[00:39] <ds-work> it's a problem in a different package, likely
[00:39] <mxpxpod> hold on, lemme check something
[00:40] <thomasvs> ds-work: according to the changelog for glib, it got added going to 2.3.0 though
[00:41] <ds-work> but we don't use g_value_take_boxed, and don't require any library version that does
[00:43] <mxpxpod> ds-work: this is strange... it's still dying... no clue why, though
[00:44] sxpert_work ([email protected]) left irc: "Leaving"
[00:44] <ds-work> mxpxpod: you have a broken library
[00:44] <mxpxpod> I guess so
[00:44] <ds-work> like, say, libxml2
[00:44] <mxpxpod> you sure it's libxml2?
[00:44] <ds-work> or popt
[00:44] <mxpxpod> hrmm..
[00:44] <mxpxpod> I was thinking more like glib
[00:45] <ds-work> those are the only choices in the above line
[00:45] <ds-work> it would odd for glib to require a newer version of itself
[00:46] sxpert_work ([email protected]) joined #gstreamer.
[00:49] <mxpxpod> ds-work: I'm looking into it
[00:51] <thomasvs> mxpxpod: what dist do you run ?
[00:51] <mxpxpod> thomasvs: gentoo :)
[00:52] <Misirlou> SURPRISE
[00:53] <mxpxpod> Misirlou: I've never had this kind of problem before
[00:53] <thomasvs> mxpxpod: well, the friend runs fedora core 2
[00:54] <thomasvs> so it's safe to say it's not an isolated problem
[00:54] <mxpxpod> thomasvs: huh?
[00:56] <mxpxpod> thomasvs: ask your friend if he's using gcc3.3
[00:58] <mxpxpod> thomasvs: ok, it's something to do with the marshal stuff
[01:00] <mxpxpod> thomasvs: aha!!!
[01:00] <mxpxpod> thomasvs: gstmarshal.c and gstmarshal.h are distributed with the tarball
[01:01] <mxpxpod> thomasvs: someone with glib2.3 must have generated those
[01:11] <mxpxpod> thomasvs: after regenerating those two files, everything is ok
[01:13] Action: bilboed is away: I'm busy
[01:14] walters ([email protected]) joined #gstreamer.
[01:18] jcsston ([email protected]) joined #gstreamer.
[01:22] Action: wheels slaps walters around with an evil GCC > 3 macro.
[01:22] <wheels> Oh, and hi.  :-)
[01:24] <walters> hey wheels
[01:24] <walters> wheels: which is that?
[01:24] <wheels> rb-debug.h:34: badly punctuated parameter list in `#define'
[01:25] <walters> hm
[01:26] <wheels> walters: I don't think 2.95 supports the (...) stuff in macros.
[01:26] <walters> ah.
[01:26] <walters> hm.
[01:26] <ds-work> hrm, interesting bug
[01:26] <ds-work> that should probably be fixed in glib
[01:27] <ds-work> gcc-2.95 supports both ISO and GNU style variadic macros
[01:28] Action: wheels certainly isn't a macro wizard.  Too much C++ conditioning.  :-)
[01:28] <ds-work> however, it has a parsing bug^Wfeature where you have to surround commas with spaces before ##
[01:29] iain ([email protected]) left irc: "! don't go !"
[01:30] Action: walters is trying to figure out if __GNUC__ is actually >= 3 with gcc 3.x
[01:30] <ds-work> should be
[01:30] <walters> iirc cpp -dM /dev/null gave you a good dump of standard defines, but GNUC isn't among them
[01:31] <walters> ds-work: how would it be fixed in glib?
[01:32] <ds-work> walters: fix glib-genmarshal to not output glib-2.4-dependent code
[01:32] <ds-work> or even glib-2.2-dependent code, imo
[01:32] <walters> oh, you were talking about a different bug
[01:34] <bitshifter> walters: http://gcc.gnu.org/onlinedocs/gcc-3.3.3/cpp/Common-Predefined-Macros.html#Common%20Predefined%20Macros
[01:34] wheels ([email protected]) left irc: Read error: 113 (No route to host)
[01:35] <walters> bitshifter: thanks
[01:35] <walters> there should be a way to get this out of gcc directly
[01:37] BBB ([email protected]) joined #gstreamer.
[01:38] <walters> wheels: * committed [email protected]/rhythmbox--mainline--0.7--patch-219
[01:38] <walters> ah
[01:38] <walters> he left.
[01:42] sublett ([email protected]) left irc: "I like food, food is good!"
[01:44] yippi ([email protected]) left irc: "Client exiting"
[01:45] KoRnouille ([email protected]) left irc: Client Quit
[01:48] smoke ([email protected]) left irc: Read error: 110 (Connection timed out)
[01:49] KoRnouille ([email protected]) joined #gstreamer.
[01:50] wheels ([email protected]) joined #gstreamer.
[01:57] jcsston ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[02:04] jcsston ([email protected]) joined #gstreamer.
[02:26] crazney ([email protected]) joined #gstreamer.
[02:26] <crazney> hi there, is there a basic memory source?.. in the docu and examples all i can see is a disk source..
[02:27] <crazney> also, anyone care to write this? http://www.freedesktop.org/~gstreamer/data/doc/gstreamer/head/pwg/html/chapter-other-source.html ;p - or at least point me in the right direction?
[02:31] <BBB> "memory source"?
[02:31] <BBB> like fakesrc?
[02:31] <BBB> I mean, you need to provide data from someplace
[02:31] <BBB> what exactly do you want?
[02:31] <BBB> and I wrote part of that, but it's not online yet
[02:31] <crazney> well i have a big lump of memory.. that contains an audio file i want to play
[02:31] <BBB> because it's unfinished
[02:31] <taaz> i think it's a good idea too
[02:31] <BBB> fakesrc can do that, right?
[02:32] <taaz> makes it easier to integrate gst with other code
[02:32] <BBB> fakesrc ;)
[02:32] <crazney> is there any docu of the builtin sources?.. like i've never heard of fakesrc before :)
[02:32] <BBB> fakesrc, fakesink and identity are your friends
[02:32] <BBB> each of the three has a handoff signal to which you connect
[02:32] <BBB> using g_signal_connect
[02:33] <BBB> then, simply modify the memory in that function that handles the signal
[02:33] <crazney> uhm..
[02:33] <crazney> you've confused me :)
[02:33] <taaz> plus you really need support to set caps of that data
[02:35] <BBB> sure, it could be improved
[02:35] <BBB> I'd say: suggest it! :)
[02:35] <taaz> and fakesrc/sink is a poor name for this functionality ;)
[02:35] <BBB> crazney: I'll write you something in the PWG... for now, write an email to the list, I'll reply tomorrow, ok?
[02:35] <BBB> taaz: true again... same thing: suggest something better ;)
[02:36] <taaz> rawsrc/sink?
[02:36] <BBB> na
[02:36] <BBB> it's not 'raw'
[02:36] <taaz> nakedsrc?
[02:36] <BBB> 'fake' at least specifies that it's not normalyl useful ;)
[02:36] <BBB> naked? :p what's that supposed to mean? ;)
[02:36] <taaz> i dunno ;)
[02:37] <taaz> extsrc? ie, external
[02:37] <BBB> I was just gonna suggest that ;)
[02:37] <taaz> i dunno.  if i were cool i'd whip up an example python plugin to do it ;)
[02:37] <BBB> I think you'll have to wait until 0.10 since we're way past our freezes for 0.8.0... but I like the idea
[02:37] <BBB> lazy ass
[02:38] <BBB> write the PWG part on python please ;)
[02:38] <ds-work> we can always add elements
[02:38] <BBB> *nod*
[02:38] <taaz> it's not just me being lazy.  the code to do that is broken right now
[02:39] <taaz> jdahlin was working on that.  the boxed gstdata patch is part of that
[02:39] <BBB> hmk
[02:40] lupusBE ([email protected]) left irc: Remote closed the connection
[02:40] jaiserca ([email protected]) left irc: "Terminando cliente"
[02:40] <taaz> silly stuff like the not-OO-but-treated-like-OO-using-casts-and-other-evil-Cness data/buffer/event code
[02:41] <taaz> i wonder how hard it would be to actually try coverting those to GObjects to test performance
[02:42] <ds-work> it wouldn't be that hard
[02:42] <ds-work> the problem is that GObject doesn't have atomic reference counting
[02:42] <taaz> doh...
[02:43] <BBB> please... gobjects are really rather slow
[02:43] <BBB> boxed types are good
[02:43] <taaz> BBB: do you have a test i can run to prove it?  i'm curious of the actual performance difference
[02:44] <taaz> i just think the d/b/e code is a poor design.  and i'm not sure how to explain why ;)
[02:45] <ds-work> we could make them like GObjects, but with a separate implementation that is optimized for GstBuffers
[02:45] <ds-work> and thread-safe
[02:47] <BBB> taaz: no... I'm just reproducing what omega said when he designed that
[02:47] <BBB> I'm sure he *did* test it
[02:48] <taaz> yeah, but that was a -long- time ago and gst has accumulated features that probably have slowed it down
[02:48] <ds-work> keep in mind that wtay and omega never had a really working implementation to decide which features were important to optimize
[02:48] <ds-work> fakesrc ! fakesink is not an interesting test case for optimization
[02:49] <taaz> and fakesrc/sink have code paths that are not optimized themselves
[02:49] <taaz> too many features
[02:49] <taaz> which is fine, just not for testing other parts of the code
[02:50] <BBB> crap... anyone have a public SMTP for me
[02:50] <BBB> ?
[02:50] Action: BBB cannto send emails
[02:50] <ds-work> mail.nyc.rr.com?
[02:51] lupusBE ([email protected]) joined #gstreamer.
[02:52] Shoragan ([email protected]) left irc: "Leaving"
[02:52] <BBB> I get a relaying denied
[02:54] harshy ([email protected]) left irc: "I quit for now"
[02:56] sub_pop ([email protected]) left irc: "Client exiting"


-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.