From: IRC <wt...@us...> - 2003-10-29 06:42:10
|
******************************************************************* [03:00] sublett (~rv...@21...) left irc: "I like food, food is good!" [03:09] Spundun (~Sp...@cy...) left irc: "Leaving" [04:02] abraar (~goo@AToulouse-105-1-16-147.w80-11.abo.wanadoo.fr) left irc: Read error: 110 (Connection timed out) [04:02] abraar (~goo@AToulouse-105-1-13-104.w80-11.abo.wanadoo.fr) joined #gstreamer. [04:22] walters (wa...@ve...) joined #gstreamer. [05:01] Nick change: harshyWORK -> harshy [05:06] walters_ (wa...@ve...) joined #gstreamer. [05:07] walters (wa...@ve...) left irc: Nick collision from services. [05:07] Marsupilami23 (~Mar...@15...) joined #gstreamer. [05:07] Nick change: walters_ -> walters [06:26] walters (wa...@ve...) left irc: "out" [06:29] ds (~ds...@ad...) joined #gstreamer. [07:15] ds (~ds...@ad...) left irc: Read error: 110 (Connection timed out) [07:20] Marsupilami23 (~Mar...@15...) left irc: Remote closed the connection [07:20] <taaz> anyone want to help debug a doc bug? [07:20] Marsupilami23 (~Mar...@15...) joined #gstreamer. [07:37] thaytoo (ja...@ad...) left irc: "leaving" [07:52] Marsupilami23 (~Mar...@15...) left irc: Remote closed the connection [07:53] Marsupilami23 (~Mar...@15...) joined #gstreamer. [08:05] Marsupilami23 (~Mar...@15...) left irc: Remote closed the connection [08:09] ChristianHJW (~chr...@p5...) joined #gstreamer. [08:16] Nick change: harshy -> harshyBETT [08:20] ChristianHJW (~chr...@p5...) left irc: [08:29] Surphaze (~bl...@ad...) joined #gstreamer. [08:30] <Surphaze> question, is it worthwhile to make an app work for .7.x to be ready for .8 or is .8 going to be much different and out soon? [08:41] <thomasvs> not going to be out soon [08:41] <thomasvs> and 0.8 will be the logical evolution of 0.7 [08:43] <Surphaze> thomasvs: alright thanks [09:07] wheels (~sc...@ds...) left irc: Read error: 113 (No route to host) [09:08] Nick change: harshyBETT -> harshy [09:13] kmaraas (~km...@14...) left irc: "Leaving" [09:16] swentel (sw...@d5...) joined #gstreamer. [09:23] athom (~the...@so...) joined #gstreamer. [09:27] chocobo525 (~cho...@pc...) joined #gstreamer. [09:31] <chocobo525> Hi there, does anyone know where to find the fingerprint for the Gstreamer packages available via apt for RPM? [09:31] <chocobo525> I'm trying to upgrade via synaptic but the packages are rejected... [09:31] <Surphaze> thomasvs: where would a good reference for changes be located, our app doesn't seem to like .7 so much [09:32] <chocobo525> Or perhaps they are unsigned, do you know a way to force update ? [09:32] <Surphaze> thomasvs: basically cvs logs the only reference? [09:44] Surphaze (~bl...@ad...) left irc: "Client exiting" [09:48] md` (ill...@md...) left irc: [10:05] dolphy (~do...@po...) joined #gstreamer. [10:07] md` (ill...@md...) joined #gstreamer. [10:11] gernot (~gernot@81.10.172.83) joined #gstreamer. [10:11] gernot77 (~gernot@81.10.172.83) joined #gstreamer. [10:11] gernot (~gernot@81.10.172.83) left irc: Client Quit [10:12] gernot77 (~gernot@81.10.172.83) left irc: Client Quit [10:42] chocobo525 (~cho...@pc...) left irc: "Leaving" [11:00] Surphaze (~bl...@ad...) joined #gstreamer. [11:05] herzi (~he...@ki...) joined #gstreamer. [11:15] dolphy (~do...@po...) got netsplit. [11:15] Company (~Company@pD958B020.dip.t-dialin.net) got netsplit. [11:15] taaz (~dlehn@66.37.66.32) got netsplit. [11:15] sxpert (~sx...@sx...) got netsplit. [11:15] makki2 (~yv...@fw...) got netsplit. [11:15] thomasvs (~th...@58...) got netsplit. [11:15] dolphy (~do...@po...) returned to #gstreamer. [11:15] sxpert (~sx...@sx...) returned to #gstreamer. [11:15] makki2 (~yv...@fw...) returned to #gstreamer. [11:15] taaz (~dlehn@66.37.66.32) returned to #gstreamer. [11:15] thomasvs (~th...@58...) returned to #gstreamer. [11:15] Company (~Company@pD958B020.dip.t-dialin.net) returned to #gstreamer. [11:20] smoke (sm...@ch...) joined #gstreamer. [11:20] jeroen_ (~je...@hm...) joined #gstreamer. [11:25] Rotty (~an...@ch...) joined #gstreamer. [11:50] foser (d0...@22...) joined #gstreamer. [12:08] gernot (~gernot@81.10.172.83) joined #gstreamer. [12:08] <gernot> guadn morgn [12:09] <md`> moin gernot [12:10] <gernot> md`: Und es war Stille auf den Wassern von #gstreamer ... ;) [12:10] <md`> lol [12:15] <Company> morgens is hier nie was los [12:15] Action: Company is auch nur da, weil er heute wegen arzt frei hat [12:17] <gernot> Company: ah, du arbeitest ? Ich dachte, du wärst Student - meine Herren, dann tust du dir aber ordentlich was an mit GStreamer [12:20] <Company> ich studiere, ja [12:22] <gernot> aber wieso hast du dir dann vom Arzt freigeben lassen müssen ? Fachhochschule ? [12:22] <Company> aber studieren tut man morgens normalerweise ;) [12:23] <gernot> Company: najaa, alles Auslegungssache ;) - ich hatte auch Vorlesungen zwischen 8 und 10 in der Früh, die sind aber fast immer unter den Tisch gefallen [12:25] <gernot> gah: ffmpeg/libavformat/ogg.c:12:21: ogg/ogg.h: No such file or directory [12:25] <gernot> ffmpeg/libavformat/ogg.c:13:30: vorbis/vorbisenc.h: No such file or directory [12:25] <gernot> doesn't it autodetect that ? [12:30] jeroen__ (~je...@hm...) joined #gstreamer. [12:33] Action: gernot installs ogg-devel and vorbis-devel [12:33] <Company> ich hab mir freigenommen, damit ich hingehen kann [12:34] <gernot> Company: Ah, dann versteh ich, dann hab ich Ursache und Wirkung vertauscht ;-) [12:34] <gernot> What does this mean: devel(libm) is needed by libvorbis0-devel-1.0-12mdk [12:35] <gernot> (from rpm -i /inbound/libvorbis0-devel-1.0-12mdk.i586.rpm ) [12:38] <Company> gernot: could you submit a bug for the ffmpeg/ogg stuff? [12:38] <gernot> Company: jepp; how ? Homepage ? [12:39] <gernot> Company: got it [12:39] <Company> bugzilla.gnome.org :) [12:45] <gernot> Company: done - my first bug report with bugzilla ! [12:46] <gernot> didn't know any keyword, though [12:46] <gernot> ok, now back to business: I want to move my GStreamer 0.6 glsink to HEAD CVS - is there any more elegant way than copying ? [12:47] <dolphy> oh no [12:47] <dolphy> gst head is broken again on ppc [12:48] <dolphy> ffmpeg [12:49] jeroen_ (~je...@hm...) left irc: Read error: 110 (Connection timed out) [12:57] <Company> dolphy: what's broken in there? [12:58] <Company> (considering we didn't even change ffmpeg stuff, or did we?) [12:58] <Company> gernot: keywords are mostly used by developers when they categorize their bugs, not when submitting [12:59] Action: Company just spend 20 minutes until he figureed out his bug: not having run gst-register [12:59] <Company> d'oh [13:00] Action: gernot is glad that this doesn't only happen to himself ;) [13:02] <dolphy> Company: i think i ve seen the TAG changed [13:02] <dolphy> Company: and now ppc building of ffmpeg fails [13:06] <Company> hm [13:06] <Company> i'll debug it when TYPEFIND branch is merged [13:06] <dolphy> can't we just revert to a previous TAG ? [13:07] <Company> no idea [13:07] <Company> probably BBB wants to keep up with ffmpeg [13:07] <Company> and considering he did the colorspace thing, he probably pulled a recent version [13:08] <dolphy> revision 1.6 [13:08] <dolphy> date: 2003/10/26 11:51:16; author: rbultje; state: Exp; lines: +1 -1 [13:08] <dolphy> Update FFMPEG snapshot to this morning's. Asked for by apoc. [13:10] <Company> yeah [13:10] <Company> we should probably talk to BBB about this and/or fix it upstream [13:12] <gernot> is it known that MPEG2 doesn't play anymore in head using this command gst-launch-0.7 --verbose filesrc location=/mnt/video2/stream.dump ! mpeg2dec ! xvideosink [13:14] <Company> i think it's known [13:14] <gernot> ok. Then I will move over glsink now, and hope that I won't this problem there ;) [13:15] <Company> good luck [13:15] <Company> you're free to fix all bugs you stumble upon in HEAD :) [13:16] Action: gernot suspects that this was the insidious plan all along :-P [13:16] <Company> sure :p [13:17] sublett (~rv...@21...) joined #gstreamer. [13:18] <Company> dang, i don't have a single mpeg2 stream on my disk [13:18] <Company> everything is mpeg1 [13:19] <gernot> Company: I can generate one for you, if you like [13:20] <gernot> Company: ah, I think this is a problem of _interlacing_ [13:21] Action: Company has close to no clue about video [13:22] <gernot> (process:13720): GStreamer-WARNING **: clock is way behind: 2038252000s (max allowed is 2000000000s [13:22] <gernot> That could be the reason [13:23] <gernot> Wait, I upload some megs of a dump that doesn't work [13:23] <gernot> http://www.lysator.liu.se/~gz/streamp.dump [13:23] <gernot> http://www.lysator.liu.se/~gz/stream.dump [13:24] <gernot> ok 2 meg is enough [13:24] <gernot> this here to test: gst-launch-0.7 --verbose filesrc location=stream.dump ! mpeg2dec ! xvideosink [13:25] <Company> could it be that mpeg2dec requires properly packetized buffers? [13:26] <gernot> huh ? [13:26] <gernot> has that changed to 0.6.x ? [13:27] wheels (~sc...@ds...) joined #gstreamer. [13:27] <gernot> s/to/compared to [13:27] <Company> hm, if it works in 0.6 probably not [13:27] Action: Company shuts up again [13:28] <gernot> the problem is that I currently have no alternative sink where I could see the result - or is there any ? [13:30] <gernot> where, who, did I ... create vtexsink ? [13:30] <Company> gernot: sdlvideosink? [13:30] <Company> (though i dunno if that works either ;)) [13:30] phubuh (~us...@c-...) joined #gstreamer. [13:30] <gernot> Company: no result; will check with working file [13:31] <gernot> vtexsink - only in 0.6, my local source tree, ok [13:31] wingo (~wi...@ma...) joined #gstreamer. [13:31] <phubuh> I've hooked a file source to a MAD decoder and the MAD decoder to an OSS sink, yet iterating the pipeline doesn't actually produce any sound. Using gst-launch to run the same sequence works fine. [13:33] <Company> phubuh: you're following the examples in C code? [13:34] <phubuh> Yup. [13:34] <gernot> Company: no, sdlvideosink is totally silent; will debug in glsink, since I know my way 'round that code [13:38] <Company> gernot: according to the cvs log someone updated mpeg2dec to compile with current CVS of mpeg2dec - maybe they broke something [13:40] <phubuh> I don't really know where to start debugging... running with --gst-mask=-1 gives an overwhelming amount of information. [13:42] <gernot> phubuh: maybe you can hook in a sinesrc instead ? [13:43] <gernot> Company: aha ! Who was that ? [13:43] <Company> gernot: BBB [13:43] <Company> though even going some versions back doesn'T seem to fix it [13:43] <phubuh> That doesn't make any noise either. [13:43] <phubuh> Err, maybe because I'm trying to decode it as MP3. :-) [13:44] <gernot> phubuh: ;) [13:44] <Company> there seem to be some errors still with colorspace conversions [13:44] <Company> but i'm not going to debug that as i have no clue about it [13:44] <phubuh> Nope, hooking the sinesrc directly to the osssink doesn't make any sound either. [13:45] <gernot> Company: that's ok, I will base my debugging on glsink, it can take both RGB and YV12 [13:45] <phubuh> I'll try sending it to a disksink instead of oss. [13:45] <Company> phubuh: you set the element to PLAYING, you connected the elements, you added them to the pipeline, all of that produced no errors and it still doesn't work? [13:46] <gernot> phubuh: which version do you use ? [13:46] <gernot> phubuh: can you upload the code somewhere ? [13:47] <phubuh> I use 0.6, and it's at http://www.phubuh.org/~phubuh/playmp3.c [13:48] <phubuh> Company: I think so. [13:50] <gernot> phubuh: do you do an add_bin ? [13:50] <Company> no, he didn'T :) [13:50] <phubuh> ... *smack* [13:50] <phubuh> Thanks guys. :-) [13:51] <gernot> phubuh: you're welcome (one more proof that reading the source directly sometimes helps faster ;) ) [13:52] <phubuh> Hooray! [13:53] <phubuh> I'm starting to like this. :-) [13:56] <gernot> phubuh: yes, GStreamer is very elegant once you got the hang of it [13:57] <Company> at least as long as you code applications ;) [13:57] Action: Company is at the doctor now [14:00] <wingo> the doctor is in. [14:01] <gernot> Company: true ;) .... in other news: cc1: warnings being treated as errors [14:01] <gernot> huh ? Quite brutal policy [14:01] <gernot> ah well ;) [14:02] <phubuh> I like to keep warnings around during development... they're like FIXMEs, but they whine at you at every compile. :-) [14:04] <Surphaze> gst-play really needs some intermediate thing between creation and failing. Its unusable to anyone without gnome [14:04] <Surphaze> would be great if it would at least create, as it is currently if it doesn't have gnomevfssrc it returns NULL on new [14:05] <Surphaze> it has a method to change source element but can't use it if you can't create it ;) [14:13] Action: gernot keeps his mouth shut, has no clue on these things :-) [14:20] <phubuh> Phew! I read the section about static sinks and MIME types and thought I was in for pain, and then I discovered the spider. [14:20] <phubuh> Switching the MPEG decoder out for a spider was amazingly easy. I'm starting to love this API. :-) [14:29] thaytan (~ja...@ad...) joined #gstreamer. [14:32] <wingo> howdy thaytan [14:32] <thaytan> hi, wingo :) [14:36] thomasvs (~th...@58...) left irc: Read error: 104 (Connection reset by peer) [14:41] Lefungus (~l@ARennes-303-1-10-104.w80-14.abo.wanadoo.fr) joined #gstreamer. [14:41] <Lefungus> Hello all [14:42] <Lefungus> i wonder, if a musepack input plugin has ever been considered by any dev here ? [14:46] <gernot> Lefungus: musepack ? [14:46] <phubuh> Also known as MPC. It's a lossless audio compression format, IIRC. [14:48] <Lefungus> no [14:48] <phubuh> Actually, I'm thinking about FLAC. MPC is lossy. [14:48] <Lefungus> it's a lossy audio format [14:49] <Lefungus> known for its performance at high bitrate, above mp3/aac/ogg [14:50] <Lefungus> source code is open [14:50] <Lefungus> but its support is very limited under linux platforms, unfortunately [14:51] thomasvs (~th...@59...) joined #gstreamer. [14:52] <Lefungus> http://www.personal.uni-jena.de/~pfk/mpp/index2.html [14:54] <Lefungus> so a gstreamer plugin for it would open many possibilities [14:54] thaytan (~ja...@ad...) left irc: Remote closed the connection [14:57] kmaraas (~km...@19...) joined #gstreamer. [14:59] wingo (~wi...@ma...) left irc: "I like core dumps" [15:10] Zahir (Tw...@54...) joined #gstreamer. [15:10] <Zahir> hi all [15:17] <gernot> Lefungus: that's a good idea, but I think the devs are maxed for the moment - they will give you help if you want to implement such a plugin, though :-) [15:17] <Lefungus> i'm reading documentation [15:18] <Lefungus> but my skills are fairly limited [15:22] <gernot> Lefungus: in regard to GStreamer, or in general ? [15:24] <gernot> Gaaah, it seems like the whole PAD_TEMPLATE-System has changed, right ? [15:24] <Lefungus> in regard to gstreamer [15:25] <gernot> get_caps is still called, but gst-launch claims that it couldn't link mpeg2dec and glsink [15:25] <gernot> Lefungus: ah, well then, that's good :-) ... have you developed test apps with musepack before ? [15:27] <Lefungus> i haven't coded anything related to video/audio actually [15:27] <gernot> OH HECK - NavigationInfo in xvideosink.c, etc., bah ! [15:27] <Lefungus> maybe i'll begin here [15:28] <gernot> Lefungus: In that case I recommend to make a little test app first that uses musepack to read from a file and output to a wav or raw audio (that you can play with sox) or sends something to OSS soundout [15:28] <Lefungus> this app already exist [15:28] <gernot> Lefungus: _or_ you have a look around GStreamer, make a little MP3 player or something to get acquainted with the architecture :-) [15:28] <Lefungus> mppdec [15:29] <gernot> Lefungus: ah, good ! So you could transform mppdec into a GStreamer plugin, you know mppdec well ? [15:29] <Lefungus> i know how to use it, now i plan to learn how it works [15:30] <gernot> Lefungus: the point I want to come down to is to keep "one foot fixed" whilst exploring things, to avoid frustration [15:30] <Lefungus> i know that :), i'll just go slowly [15:31] <gernot> Lefungus: well then - there are a lot of devs here in the evening, you can ask them then .... [15:31] <Lefungus> ok thanks [15:33] <phubuh> Gah, I've listened to the same song 10 times in a row now. Debugging music players is a pain. :-) [15:33] <gernot> GST_TYPE_INTERFACE and GST_TYPE_NAVIGATION and GST_TYPE_X_OVERLAY, do I need that for operation as a video outputplugin [15:33] <gernot> ? [15:33] <gernot> phubuh: change to video ;) [15:34] <gernot> phubuh: or switch to some NewAge stuff, hard to grow tired of ;) [15:35] <phubuh> Thanks for the tips. :-) By the way, can I set the default arguments of gdb on startup? I don't find anything about it in the man page, and -- doesn't work. :/ [15:36] The_Company (~Co...@p5...) joined #gstreamer. [15:37] <gernot> phubuh: sorry, not good at that [15:37] <gernot> Company: and, are you still allowed to carry on ? ;) [15:38] Rotty (~an...@ch...) left irc: Read error: 104 (Connection reset by peer) [15:39] Company (~Company@pD958B020.dip.t-dialin.net) left irc: Read error: 110 (Connection timed out) [16:05] Oskuro (~jo...@11...) joined #gstreamer. [16:05] <Oskuro> oi! [16:05] gernot (~gernot@81.10.172.83) left irc: "Client exiting" [16:06] <Oskuro> sublett: here? [16:07] <Oskuro> sublett: you helped me some weeks ago to confirm Debian bug #210318 (Bugzilla #124580). It'd be great if someone here using JACK could tell me if 0.6.4 fixed this in some way. [16:07] <Oskuro> basically, it'd allow much stuff to enter Sarge :) [16:10] kmaraas (~km...@19...) left irc: "Leaving" [16:11] <sublett> Oskuro: I can take a look later today [16:12] <Oskuro> sublett: thanks. If you can reply to bugzilla, I'll get the Cc: [16:18] Nick change: The_Company -> Company [16:24] Rotty (~an...@ch...) joined #gstreamer. [16:35] Rotty (~an...@ch...) got netsplit. [16:35] thomasvs (~th...@59...) got netsplit. [16:35] jeroen__ (~je...@hm...) got netsplit. [16:35] dolphy (~do...@po...) got netsplit. [16:35] taaz (~dlehn@66.37.66.32) got netsplit. [16:35] sxpert (~sx...@sx...) got netsplit. [16:35] makki2 (~yv...@fw...) got netsplit. [16:35] Lefungus (~l@ARennes-303-1-10-104.w80-14.abo.wanadoo.fr) got netsplit. [16:35] Surphaze (~bl...@ad...) got netsplit. [16:35] abraar (~goo@AToulouse-105-1-13-104.w80-11.abo.wanadoo.fr) got netsplit. [16:35] matroska_mirrorb (~ch...@p5...) got netsplit. [16:35] alley_cat (AlleyCat@pD9E61E20.dip.t-dialin.net) got netsplit. [16:35] phubuh (~us...@c-...) got netsplit. [16:35] ds-work (ds...@co...) got netsplit. [16:35] sack (~sa...@po...) got netsplit. [16:35] Galt (ga...@ca...) got netsplit. [16:36] Rotty (~an...@ch...) returned to #gstreamer. [16:36] thomasvs (~th...@59...) returned to #gstreamer. [16:36] Lefungus (~l@ARennes-303-1-10-104.w80-14.abo.wanadoo.fr) returned to #gstreamer. [16:36] phubuh (~us...@c-...) returned to #gstreamer. [16:36] jeroen__ (~je...@hm...) returned to #gstreamer. [16:36] ds-work (ds...@co...) returned to #gstreamer. [16:36] sack (~sa...@po...) returned to #gstreamer. [16:36] Galt (ga...@ca...) returned to #gstreamer. [16:36] alley_cat (AlleyCat@pD9E61E20.dip.t-dialin.net) returned to #gstreamer. [16:36] matroska_mirrorb (~ch...@p5...) returned to #gstreamer. [16:36] abraar (~goo@AToulouse-105-1-13-104.w80-11.abo.wanadoo.fr) returned to #gstreamer. [16:36] Surphaze (~bl...@ad...) returned to #gstreamer. [16:36] dolphy (~do...@po...) returned to #gstreamer. [16:36] sxpert (~sx...@sx...) returned to #gstreamer. [16:36] makki2 (~yv...@fw...) returned to #gstreamer. [16:36] taaz (~dlehn@66.37.66.32) returned to #gstreamer. [16:40] yippi (~br...@nw...) joined #gstreamer. [16:53] ain (~ia...@us...) joined #gstreamer. [16:57] <Company> yay, another stupid bug for me [16:57] <Company> creating mmaped buffers in GstFileSrc doesn't ref the filesrc [17:09] <thomasvs> yeah, I noticed that too [17:09] jeroen__ (~je...@hm...) left irc: "Client exiting" [17:09] <thomasvs> but it was slightly more complex than that [17:09] <Company> i made it ref the filesrc now [17:09] <Company> it works now [17:10] phubuh (~us...@c-...) left irc: Remote closed the connection [17:11] <Company> thomasvs: did you test the new typefind stuff? [17:12] swentel (sw...@d5...) left irc: [17:12] athom (~the...@so...) left irc: [17:16] <ain> Company: have you merged it into HEAD yet? [17:16] <Oskuro> sublett: I'm back, just in case you want to contact me here :) [17:18] <Company> ain: not yet, but i don't have any reason to not merge it i think [17:18] <ain> argh, my university wants me to sit a "computer proficiency test" [17:18] <Company> ain: apart from having no testsuite :/ [17:18] <ain> to see if I need to attend a "learn computers" class [17:19] <ain> Company: dunno how you make a test suite for it, other than have small files of each type? [17:19] <Company> you shut probably refresh your windows knowledge ;) [17:19] <Company> should even [17:19] Action: Company thinks about different things while writing [17:19] <ain> yeah, but I never need to use windows [17:19] <ain> I write my essays in latex :) [17:24] <Company> the test probably wont ask that though ;) [17:27] <Company> how do i compute 1-(1-1/2^(65*2)^50000 accurately? [17:27] <Company> doubles aren't good enough for that [17:27] <Company> 1-(1-1/2^(65*2))^50000 even [17:28] <ain> "slowly" [17:28] <ain> muhahahaha [17:31] gernot (~gernot@81.10.172.83) joined #gstreamer. [17:33] kmaraas (~km...@14...) joined #gstreamer. [17:38] <wheels> Company : Well, if you're doing that a lot I'm pretty sure that the MM instructions on the newer processors have instructions for integer exponentiation... [17:40] Action: wheels grabs his P4 reference manual... [17:41] <Company> wheels: i only want that number, once, without rounding errors [17:42] <Company> wheels: it must be << 0,001 :) [17:42] <Company> wheels: it's the possibility that random data is identified as video/mpeg [17:43] <wheels> Ah. :-) [17:44] <wheels> Company : Do you know lisp? [17:44] <wheels> (Lisp works on "real" numbers and tends to give as much precision as you have RAM for. :-) ) [17:44] <Company> that's an idea [17:45] Action: Company doesn't remind how to start it though [17:46] bilboed (~bilboed@80.15.146.121) left irc: Read error: 110 (Connection timed out) [17:46] BBB (~rb...@ph...) joined #gstreamer. [17:48] bilboed (~bilboed@AMontsouris-108-1-17-49.w80-15.abo.wanadoo.fr) joined #gstreamer. [17:48] Marsupilami23 (~Mar...@15...) joined #gstreamer. [17:48] <BBB> dolphy: is gst-player already working with the new X interface? [17:50] <gernot> BBB: gotcha ;) [17:50] <dolphy> BBB: sorry i have a company to run during the day :) [17:52] <gernot> argl, what is this: /* Loading the library containing GstVideoSink, our parent object */ [17:52] <gernot> so xvideosink now derives from an object ? [17:53] <BBB> gernot: it does, but we're considering removing that again ;) [17:53] <gernot> phew ... [17:53] <BBB> dolphy: I have an internship to do during the day ;) [17:54] <BBB> gernot: however, all v4l/v4l2 elements derive from other elements ;) [17:54] <BBB> so do the oss elements [17:54] <gernot> a question from earlier: GST_TYPE_INTERFACE/NAVIGATION/X_OVERLAY, is that obligatory now ? [17:54] <BBB> no [17:55] <gernot> what does derivation mean in glib-object context ? [17:55] <BBB> but they're itnerfaces that you can derive from [17:55] <BBB> derivation is parent/child classing in c++/java [17:55] Action: gernot guesses he has to read more on glib-objects [17:55] <BBB> interfaces are extensions outside the derivation tree [17:55] <gernot> BBB: like in java, ok [17:55] <BBB> glib is sweet... took me a while to fully get it, but it's honestly perfect [17:56] <gernot> BBB: I believe that first when glsink runs again in 0.7.x ;) [17:56] Action: wheels notices that emacs's lisp interpreter aparently only uses longs and as such overflows quickly... [17:56] <gernot> get_caps/set_caps have disappeared ? [17:56] <BBB> no? [17:56] <wheels> BBB : I'd go with you that glib is quite nice, but I wouldn't call gobject "perfect" :-) [17:56] <BBB> for my goals, it's close to ;) [17:57] <BBB> but yeah, I know you like c++ too ;) [17:57] <BBB> (poor you :D) [17:57] <gernot> BBB: okej, so xvideo only restructured, takes a while for me to rev.engineer [17:57] <wheels> BBB : Well, I've yet to see a base object implementation that I would call "perfect" :-) [17:58] <BBB> true, true... [17:58] <BBB> gernot: exactly [17:58] <BBB> and we might reverse it [17:58] <BBB> talk to dolphy about that [17:58] <BBB> I'll be online tonight [17:58] Action: BBB goes home [17:58] <BBB> later! [17:59] BBB (~rb...@ph...) left irc: "Client exiting" [18:04] <wheels> Company : But for kicks I'll evaluate it when I get home -- there I have both a ti-89 and a lisp interpreter and reference -- one of them should be able to handle it. :-) [18:15] <taaz> thomasvs: ping? [18:17] <gernot> Hmmm, why plugin_init provide a dummy sink in 0.6.x and why doesn't that work in 0.7.x ? [18:17] <gernot> s/why/why could [18:22] Lefungus (~l@ARennes-303-1-10-104.w80-14.abo.wanadoo.fr) left irc: Remote closed the connection [18:24] <Company> ha, it works [18:25] Action: Company just started a remote mathematica [18:26] Action: wheels used to do that in Uni too. :-) [18:26] <Company> gah [18:26] <Company> no license [18:27] Action: wheels hasn't used Mathematica or Lisp much since Uni though... [18:30] <Company> hm, why do these sites i browse to to get a mthematica serial want to send me .exe files? [18:33] <thomasvs> taaz: pong [18:34] <taaz> thomasvs: found doc bug problem [18:35] ds (~ds...@ad...) joined #gstreamer. [18:35] <thomasvs> taaz: shoot [18:36] <taaz> #125539 is in 0.7.1 and 0.6.4 [18:38] <taaz> problem is in htmlinstall.mak. if XMLTEX isn't found then XMLTO is redefined to be false and then HTML stuff is not built and then HTML_DAT is empty so the cp command in htmlinstall.mak fails since you don't have any source files listed [18:40] <taaz> i guess fix is to eaither test if HTML_DAT is empty before doing the cp or maybe use one of the conditionals to take out that whole install-data-local rule if it shouldn't be used [18:44] <thomasvs> taaz: ok, I added to bugzilla, will look at it asap [18:44] <thomasvs> ack [18:44] <taaz> oops.. i ust did too ;) [18:44] <thomasvs> I have one file, and another file. I want every line from A that is not a line in B [18:44] <thomasvs> I don't want to use grep -v -f A B because that takes five hours [18:44] <thomasvs> any suggestions ? [18:45] <taaz> get a faster computer ;) [18:45] <Company> sort A and B and then diff them [18:46] <taaz> or even sort and uniq depending on the data [18:46] <Oskuro> sublett: please use bugzilla, or jo...@de... if you don't have an account. Thanks :) [18:46] Oskuro (~jo...@11...) left #gstreamer. [18:47] <ain> Company: they're probably a pornspam dialer program [18:47] <Company> ain: that sounds reasonable [18:47] <Company> any objections to me merging TYPEFIND with HEAD? [18:47] kmaraas (~km...@14...) left irc: "Leaving" [18:48] <ain> Company: nope, none :) [18:48] <ain> Company: then, once you've done that and I come back from bowling ( :/ ) I'll add the image/* stuff [18:49] <Company> ain: i've done video/mpeg streams already [18:49] <ain> Company: differently to the way I did it? [18:49] <ain> Company: or did I not commit that? [18:50] <Company> ain: stream detection, not file detection [18:50] <ain> oh right [19:00] <wheels> Company : Oh, and got it converted to scheme -- running now: [19:00] <wheels> (- 1 (expt [19:00] <wheels> ) [19:01] <wheels> oops [19:01] <wheels> (- 1 (expt (- 1 (/ 1 (expt 2 (* 65 2)))) 50000)) [19:02] Action: wheels also had to grab MIT-Scheme when he noticed that it also overflowed the implementation of the only interpreter that SuSE had packaged... [19:03] <thomasvs> oops [19:03] <thomasvs> my roommate asked me when I was going to cook salmon and endives again [19:03] <thomasvs> I said "as soon as you run around the house naked" [19:03] <thomasvs> fifteen minutes later, I have to cook [19:03] <dolphy> lool [19:04] <thomasvs> wish I had my camera at the ready though :) [19:04] Nick change: thomasvs -> thomasvs-cook [19:04] <dolphy> should do the same trick for my meat :) [19:05] <ain> thomasvs-cook: did you mean the inside or the outside of the house? [19:08] ds (~ds...@ad...) left irc: Read error: 110 (Connection timed out) [19:08] chillywilly (danielb@CPE-24-167-193-166.wi.rr.com) left irc: Read error: 54 (Connection reset by peer) [19:08] chillywilly (danielb@CPE-24-167-193-166.wi.rr.com) joined #gstreamer. [19:16] Zeenix (~zak@203.135.12.158) joined #gstreamer. [19:16] <Zeenix> hmm... connected to freenode.net at last.. [19:16] <Zeenix> hi guys [19:18] <makki2> anyone working on a bluefish444 sink ? [19:20] chillywilly (danielb@CPE-24-167-193-166.wi.rr.com) left irc: "leaving" [19:20] chillywi1ly (danielb@CPE-24-167-193-166.wi.rr.com) joined #gstreamer. [19:23] <gernot> hej zeenix ! [19:25] ChristianHJW (~chr...@p5...) joined #gstreamer. [19:39] <Company> how does merging work in CVS? [19:43] <wheels> Company : Ok -- 40 minutes of CPU time later still no answer. :-) I'm about to head home, we'll know the exciting result tomorrow. :-) [19:44] Zeenix (~zak@203.135.12.158) left irc: Remote closed the connection [19:45] <Company> wheels: hehe [19:46] <Company> wheels: mathematica told me it's < 10E-30 though so i'm happy :) [19:46] <Company> but i still have no exact answer [20:06] <dolphy> gernot: your glsink handles hardware scaling ? [20:10] <gernot> dolphy: sure ! [20:10] apoc (~ap...@dy...) joined #gstreamer. [20:10] <gernot> dolphy: through texture uploads [20:11] <dolphy> gernot: interesting [20:11] <dolphy> gernot: and you handle the capacity of the video card [20:12] <dolphy> gernot: i mean what happens if the card does not have the correct features to do acceptable scaling [20:12] smoke (sm...@ch...) left irc: "Mind leaves body" [20:12] smoke (sm...@ch...) joined #gstreamer. [20:19] <ds-work> Company: use an approximation [20:20] <taaz> damnit. docs are still failing to build for me [20:20] <ds-work> (1+epsilon)^N = 1 + N * epsilon + (1/2) N^2 epsilon^2 + ... [20:22] <gernot> dolphy: I have a standard RGB renderer, any card with OpenGL 1.2 can do that. I do expect the presence of OpenGL/Mesa for the moment, maybe that should be checked. [20:22] <ds-work> er, actually [20:22] <gernot> dolphy: with standard RGB I mean that it uses RGB texture uploads instead of YUV ones. [20:22] <ds-work> (1+epsilon)^N = 1 + N * epsilon + (1/2) N (N-1) epsilon^2 + ... [20:22] <dolphy> gernot: right so any GL enabled card should have hw scaling [20:23] <gernot> dolphy: yes, since I use textures, which implicitly scale on drawing [20:23] <Company> ds-work: i did already - it's close enough to 0 [20:23] <gernot> dolphy: actually you can check it out now if you feel for it, I just checked it into HEAD *proud* [20:23] <Company> gernot: is it possible to have multiple glsinks running at once in different windows? [20:24] <dolphy> gernot: the reason why i m askin is that directfb can provide a GL context from a surface [20:24] <gernot> Company: _that_ is a challenge, I check [20:24] <Company> gernot: i'm asking because it's not possible with Xvideo [20:24] <dolphy> gernot: could be fun to have a gstreamer video output as a GL surface in a directfb window [20:25] Action: Company would have to fix direct rendering for it, but at least he'D have a reason [20:25] <taaz> anyone understand docs ps building errors? [20:25] <dolphy> time to go back home [20:25] <dolphy> see you later aligator [20:25] dolphy (~do...@po...) left irc: "Network down, IP Packets delivered via UPS" [20:26] <ds-work> Company: it's approximately 50000 * (1/2)^130 [20:27] <Company> ds-work: good enough :) [20:27] <gernot> http://www.lysator.liu.se/~gz/glsink_shot.jpg [20:28] <gernot> ok, there are better shots, but it proves the principle ;) [20:28] <taaz> anyone? please? ;) http://crib.lehn.org:8080/~dlehn/tmp/glog2.gz [20:29] <gernot> dol: yes, possible :-) [20:29] <gernot> Company: I leave soon, please repost this to dolphy when he returns :-) [20:30] <Company> gernot: i hope i'll remember it [20:31] <gernot> Company: no prob, he'll ask, I guess [20:32] Misirlou (~ale...@c-...) joined #gstreamer. [20:32] <gernot> Company: would you have time to test-compile glsink from HEAD ? [20:33] smoke (sm...@ch...) left irc: Read error: 104 (Connection reset by peer) [20:33] <Company> gernot: i'll test it when i'm done merging [20:34] <gernot> Company: ah sure, thanks ! :-) [20:35] m_wheels (~sc...@ds...) joined #gstreamer. [20:35] wheels (~sc...@ds...) left irc: "[BX] Wouldn't it be great if the dog could walk itself?" [20:35] <taaz> hmm.. i have some dtd fixes to make... maybe that will help [20:35] Nick change: m_wheels -> wheels [20:41] Zeenix (~zak@203.135.12.131) joined #gstreamer. [20:41] <Zeenix> hi [20:41] <Zeenix> hello Zahir! good to see you back [20:44] pb_ (~pb...@cp...) joined #gstreamer. [20:48] jeroen_ (~je...@je...) joined #gstreamer. [20:48] phubuh (~us...@c-...) joined #gstreamer. [20:49] mg_ (~mg...@19...) joined #gstreamer. [20:49] <phubuh> Hmm. How do I make Gentoo's ebuild build v4l support? [20:49] <mg_> is there a HOWTO anywhere to add mp3 to rh distros (fedora test 3 to be exact)? [20:50] <mg_> already downloaded and install the gstreamer-plugins-mad [20:50] <mg_> but some filetype association is still missing, piping through spider doesnt work and neither do rythmbox nor nautilus music view [20:51] <phubuh> Oh, and does "qcamsrc" mean "QuickCam source"? [20:53] <Zeenix> phubuh: yeah [20:53] <phubuh> Would it work with one of those newfangled USB QuickCam Webs? [20:54] <phubuh> (If not, I've got a V4L module) [20:55] <mg_> anyone} [20:55] <mg_> ? [20:55] <gernot> mg_: Sorry, I have no clue about such things :/ [20:55] <taaz> ds-work: sorry to pick on you but... have you built the ps docs on a debian box lately? i really don't know what to do about a latex "Emergency Stop!" message. [20:56] <ds-work> taaz: no [20:59] BBB (~rbultje@213.160.215.2) joined #gstreamer. [20:59] <Zeenix> phubuh: it didnt work with my qcam either, i used v4l [21:01] <phubuh> Okay. Well, I suppose that's the proper way to do it, anyway. [21:01] kmaraas (~km...@14...) joined #gstreamer. [21:03] mg_ (~mg...@19...) left irc: "Leaving" [21:08] Zahir (Tw...@54...) left irc: "mañana será otro día" [21:14] dolphy (~do...@21...) joined #gstreamer. [21:14] <Company> BBB: did you know that you broke the device property for oss when you redid that stuff? [21:16] <BBB> I did? [21:16] <BBB> no I didn't [21:16] <BBB> did I? [21:17] <BBB> case ARG_DEVICE: [21:17] <BBB> /* disallow changing the device while it is opened [21:17] <BBB> get_property("device") should return the right one */ [21:17] <BBB> if (gst_element_get_state (GST_ELEMENT (oss)) != GST_STATE_NULL) { [21:17] <BBB> g_free (oss->device); [21:17] <BBB> oss->device = g_strdup (g_value_get_string (value)); [21:17] <BBB> } [21:17] <BBB> oops [21:17] <BBB> that should be '==', not '!=' [21:18] <ds-work> could dparams be reimplemented as an interface? [21:18] Action: BBB quickly commits fix and does as if nothing was wrong [21:18] <BBB> I don't know how... but they might [21:18] <phubuh> libtool: install: error: cannot install `libgstv4lsrc.la' to a directory not ending in /usr/local/lib/gstreamer-0.6 [21:19] <Company> yeah, dparams is a good example of an interface [21:19] <phubuh> Why this restriction? I want to install in /usr, and I set --prefix as such, but make install fails at that library. [21:19] <phubuh> (This is gst-plugins-0.6.3) [21:21] <phubuh> I suppose it's okay to install in /usr/local/, but how do I make GStreamer look there for modules? gst-inspect-0.6 isn't showing any of the new ones. [21:21] <phubuh> Oh, gst-register-0.6, duh. Thanks! [21:23] <ds-work> you're thanking us for what exactly? Did we somehow telepathically impart knowledge over IRC? [21:23] <ds-work> :) [21:23] <phubuh> I suppose! [21:23] <Company> we patiently listened ;) [21:25] <BBB> Company: OSS works again [21:25] <BBB> oops etc. [21:25] <Company> BBB: thx [21:26] <BBB> does anyone already have a brilliant idea on how to make a HAL-like system for GStreamer? [21:26] <BBB> or do we ask the application to use HAL? [21:26] <Company> HAL-like? [21:26] <BBB> see freedesktop.org [21:26] <phubuh> HAL? As in 2001? [21:26] <BBB> "how to know what devices are on the system without the user entering the /dev/ entry" [21:26] <phubuh> Oh. [21:26] <BBB> i.e., a uniform way of accessing devices [21:26] <Company> dolphy: I was supposed to show you http://www.lysator.liu.se/~gz/glsink_shot.jpg [21:26] <ds-work> BBB: probing [21:27] <BBB> without the app having specific knowledge [21:27] <BBB> ds-work: in the plugin? [21:27] <ds-work> BBB: sure [21:27] <BBB> or by the app? [21:27] wheels (~sc...@ds...) left irc: Excess Flood [21:27] <Company> that's a question i'll think about post-0.8 [21:27] <ds-work> BBB: I'm actually designing an interface now [21:27] wheels (~sc...@ds...) joined #gstreamer. [21:28] <BBB> Company: well, I'd like to solve it now since I need it for gst-mixer/gst-rec [21:29] <Company> no idea though [21:29] Action: BBB neither... [21:29] <BBB> I want a device-num property [21:29] <BBB> but then, each plugin needs knowledge of where to find devices [21:30] <BBB> this isn't so bad [21:30] <BBB> but the question is: do we want that? [21:30] <BBB> we're basically doing HAL-inside-elements then [21:30] <ds-work> it's not like devices don't follow a standard [21:30] <BBB> yeah, but the app is not supposed to know [21:30] <ds-work> the app doesn't need to know [21:30] <BBB> that's what device-num as property means ;) [21:31] <BBB> I say device-num=0 [21:31] <BBB> and the plugin translates it to /dev/dsp1 [21:31] <BBB> or /dev/sound/dsp1 if I use devfs [21:31] <BBB> or ... whatever my freaking fucked up system uses [21:31] <ds-work> use enums [21:31] <Company> yeah, the plugin can do magic anyway [21:31] <Company> ;p [21:32] <BBB> ds-work: enums? [21:32] <BBB> as in...? [21:32] <ds-work> use enums for the object property [21:32] <BBB> why? [21:32] <ds-work> that way, it's not a "number", you're chosing an item from a list [21:33] mg_ (~mg...@19...) joined #gstreamer. [21:33] <BBB> how do I generate the list? [21:33] <ds-work> GstPropertyProbe interface [21:33] <BBB> probing [21:33] <mg_> any idea why libgstmad.so fails to load when i run gst-register? [21:33] <BBB> d'oh [21:33] <BBB> ;) [21:33] <BBB> mg_ check --gst-debug-level=5 [21:33] <Company> uh [21:34] <Company> --gst-debug=GST_PLUGIN*=5 is better i think [21:34] <BBB> whatever ;) [21:34] Action: Company loves the new debugging system btw [21:34] <Company> since I did the typefinding, I love it [21:34] <BBB> ds-work: well, explain yourself when it's finished, but I suppose I like it ;) [21:34] <BBB> brb [21:34] <Company> I liked it before [21:35] <mg_> gst-register --gst-debug=GST_PUGIN_MAD=5? [21:36] <Company> gst-register --gst-debug=GST_PUGIN*=5 [21:36] <mg_> ah, using 0.6.x, not 0.7.x :) [21:37] <mg_> aparently no gst-debug option supported [21:37] <Company> oops :/ [21:37] <Company> no, it's called --gst-mask=<nr> there [21:37] <Company> --gst-mask=-1 prints everything [21:37] <mg_> theres a gst-debug-mask however [21:37] <Company> but finding the right line there is quite difficult [21:38] <phubuh> Is v4lelement what I use to use v4l as a source? [21:38] Marsupilami23 (~Mar...@15...) left irc: Remote closed the connection [21:39] <BBB> no [21:39] <BBB> v4lsrc is [21:39] <BBB> v4lelement is the generic parent element of v4lsrc and other v4l elements [21:39] <phubuh> Oh. I don't have that. :( [21:39] <phubuh> v4lelement is the only thing that shows up when I grep gst-inspect for v4l. [21:40] <BBB> :? [21:40] <Company> v4lelement shouldn't show up though, should it? [21:40] <BBB> yeah, it should [21:40] <BBB> it's like ossmixer [21:40] <BBB> a generic element that is useful for overlay-only [21:41] <BBB> includign channel tuning etc. [21:41] <Company> gst_element_make ("v4lelement") has a purpose? [21:41] <phubuh> Do I need to pass any special options to ./configure of gst-plugins? I didn't pass anything. [21:42] <phubuh> But it did say it would build v4l. >_< [21:42] <BBB> locate libgstv4lsrc.so [21:43] <phubuh> As soon as updatedb finishes. :-) By the way, does v4lsrc give me data suitable for dumping directly into a video sink? [21:43] <BBB> yes [21:43] <BBB> v4lsrc ! xvideosink is one of my primary testcases in v4l plugin development ;) [21:44] <phubuh> Hee hee. I thought I'd play around with EffecTV and my QuickCam :-) [21:45] <phubuh> I do have that file, in /usr/local/lib/gstreamer-0.6/ [21:46] <phubuh> I'll try kludging around some symlinks... [21:47] <phubuh> Oh, that did make v4lsrc appear. [21:48] <phubuh> Yay! [21:49] phubuh (~us...@c-...) left irc: Remote closed the connection [21:54] #gstreamer: mode change '+o Company' by ChanServ!ChanServ@services. [21:55] <dolphy> BBB: i m wondering about this interface and video sinks [21:55] <dolphy> BBB: if we are planning to make some more changes [21:55] <dolphy> BBB: to the way video sinks are working [21:55] Topic changed on #gstreamer by Company!~Co...@p5...: GStreamer 0.6.4 "Moo" is OUT ! -> http://gstreamer.net/releases/0.6.4/src ! GStreamer 0.7.1 "2.6 or Bust" is out -> http://gstreamer.net/releases/0.7.1/src (this is not an upgrade path from 0.6.4) | new typefinding hit HEAD, anyone wanna release 0.7.2? [21:55] #gstreamer: mode change '-o Company' by Company!~Co...@p5... [21:55] <dolphy> BBB: why don't we do all that now ... [21:56] <dolphy> BBB: i m a bit afraid of having lot of changes in API [21:56] <ds-work> yeah, we'll release 0.7.2 in time for the next gnome release [21:57] <taaz> can someone fix ps builds on debian boxes before you release another version ;) [21:57] <BBB> dolphy: I'm fine with it... however, don't count on me today [21:57] <BBB> I'm @ work [21:57] <taaz> i'm too confused to figure this out right now... [21:57] <mg_> ok, now i can play mp3 files from the command line, anyone know how to get the mime type thing working? [21:57] <ds-work> taaz: complain to thomasvs [21:57] <dolphy> BBB: i m still confused with the API [21:57] <BBB> and matroskademux is broken once again... this is starting to get really annoying [21:57] <dolphy> BBB: i mean your interface is ok but very trivial [21:57] <BBB> yes! exactly! you understand! /me bows [21:58] <BBB> finally... someone fully understands! [21:58] <dolphy> lol [21:58] <dolphy> i understood that since the beginning but still we have the signals in GstVideoSink [21:58] <BBB> ok, so remove them [21:59] <dolphy> BBB: how to get the video size then ? [21:59] <dolphy> BBB: is xvideosink listening to resize events ? [21:59] <mg_> anyone? how do i associate mp3 files in gstreamer for the type-find system to work? [22:01] phubuh (~ph...@c-...) joined #gstreamer. [22:01] <BBB> dolphy: no, not yet... but it should [22:01] <BBB> mg_: happens atuomatically with the right plugin installed [22:02] <BBB> dolphy: try resizing the window of gst-overlay (see sandbox) with v4l and with the videotestsrc ! xvideosink backend [22:02] <BBB> the one workd fully [22:02] <BBB> the other doesn't resize properly [22:02] <BBB> but it should! [22:02] <BBB> that's when it becomes extremely easy [22:02] <BBB> the plugin doing all the work [22:02] <BBB> and the app being a stupid moron [22:02] <BBB> perfection! [22:03] <phubuh> How would I hook up a v4lsrc to an EffecTV effect? v4lsrc ! colorspace ! edgeTV ! colorspace ! xvideosink doesn't want to play. [22:03] kcongerWORK (~kc...@so...) joined #gstreamer. [22:04] <BBB> phubuh: why doesn't it want to play? [22:04] <kcongerWORK> I created a bug for my darwin build problems, 125727 [22:04] <phubuh> It just says it doesn't want to. :( Perhaps I can trick gst-launch into giving me more information. [22:05] <BBB> --gst-mask=-1 2>&1 | grep -i v4l [22:05] <dolphy> BBB: ok so our goal is set_xid and then gdk_window_move_resize [22:06] <dolphy> BBB: the plugin listens to configure events [22:06] <dolphy> BBB: if it can do hardware scaling it does [22:06] <dolphy> BBB: if it can't it emit upstream resize request event [22:06] <BBB> yes [22:06] <dolphy> BBB: so that video scaler does software scaling [22:06] <BBB> this can probably be done automatically using videoscale [22:07] <BBB> if it needs software scaling, it (sink) sets a new caps on the peer element's src pad with a list of width/heigth [22:07] <dolphy> BBB: yes but the video output plugin has to communicate with video scaler [22:07] <BBB> first the new width/height [22:07] <BBB> and then, as second caps, the old width/height [22:07] <BBB> if it's a scale, it will chooce the first and we'll have software scaling [22:07] smoke (sm...@ch...) joined #gstreamer. [22:08] <BBB> if it's something else, it'll continue using the second and everything will be nice (just unscaled ;) ) [22:08] <dolphy> but that force video scaler to be connected directly with the sink [22:08] <BBB> you can do resize GstEvents too [22:08] <BBB> it's just an idea... all is fine with me [22:08] <BBB> as long as its automatic [22:09] <dolphy> BBB: was there a plan to cleanup and simplify xvideosink ? [22:09] <BBB> originally, yes [22:09] <BBB> but consider that post-0.8 stuff [22:09] <dolphy> BBB: i heard about a split for xv and non xv [22:09] <BBB> yeah, I'd love such a split [22:09] <ds-work> it's definitely post-January-1 stuff [22:09] <BBB> but that'd first require some autodetection [22:10] <BBB> and we still don't have that [22:10] <BBB> I've got some ideas on that [22:10] <BBB> using element ranks and all that [22:10] <BBB> but not now [22:10] <BBB> I've got more important stuff to work on first [22:10] <dolphy> autodetection of XV support ? [22:10] <BBB> if you want to, please go ahead ;) [22:10] <BBB> no [22:10] <BBB> autodetection of the right sink to use [22:10] <BBB> autodetection of xv is easy [22:10] <BBB> you want a xvideosink and a ximagesink [22:11] <BBB> mplayer calls them x11 and xv [22:11] <dolphy> yup [22:11] <BBB> xvideosink has a higher priority than ximagesink [22:11] <BBB> priority being the element rank [22:11] kcongerWORK (~kc...@so...) left #gstreamer. [22:11] <BBB> sdlvideosink is somewhere in between [22:12] <dolphy> we would need something similar for colorspaces [22:12] <BBB> why? [22:13] <BBB> just use ffcolorspace [22:13] <BBB> it's probably the best of the few [22:13] <dolphy> that will make distribution issue [22:13] <BBB> dolphy: can you work on that? [22:13] <dolphy> BBB: i first want to fix video sinks :) [22:13] <BBB> can you try to make a piece of code (in libgstplay) that selects all Video/Sink category elements based on rank, and tries each of them? [22:14] <BBB> you might also want to make an output selection dialog in gst-player ;) [22:14] <BBB> with 'autodetect' being the default' [22:14] <dolphy> yeah [22:14] <taaz> thomasvs-cook: when you get a chance, could you look at that log link i posted earlier here and help me fix that ps building problem? [22:14] <phubuh> Hmm. Is it okay if I paste about 5 lines of debugging information? [22:15] <phubuh> Maybe in #flood? :-= [22:15] <Company> sure [22:15] <Company> just put it here [22:15] <phubuh> DEBUG(28329: 0)gstelement.c(2156): gst_element_set_state: : [v4lsrc0] element cl [22:15] <phubuh> aimed state-change success, but state didn't change (now READY, PAUSED <-> PAUSE [22:15] <phubuh> D [22:15] <phubuh> DEBUG(28329: 0)gstbin.c(661): gst_bin_change_state: : child 'v4lsrc0' failed to [22:15] <phubuh> go to state 4(PAUSED) [22:15] <phubuh> DEBUG(28329: 0)gstelement.c(2105): gst_element_set_state: : [v4lsrc0] element is [22:15] <phubuh> already in requested state READY [22:15] <phubuh> DEBUG(28329: 0)gstelement.c(2156): gst_element_set_state: : [pipeline0] element [22:15] <phubuh> claimed state-change success, but state didn't change (now READY, READY <-> PAUS [22:15] <phubuh> ED [22:15] <phubuh> pipeline doesn't want to play [22:16] Action: ds-work realizes that glib doesn't handle dynamic enums [22:16] mg_ (~mg...@19...) left irc: "Leaving" [22:16] <Company> lol [22:17] <Company> C doesn't either ;) [22:17] <BBB> ds-work: exactly ;) [22:17] <ds-work> well, GEnums [22:17] <BBB> phubuh: I need the lines before that, the ones where it sayd why buffer allocation failed [22:17] <phubuh> This might be what you want: [22:17] <phubuh> INFO (28329: 0)gstpad.c(1408): gst_pad_try_set_caps_func: : got reply REFUSED (- [22:17] <phubuh> 1) from link function on pad colorspace0:sink [22:17] <phubuh> INFO (28329: 0)gstpad.c(1418): gst_pad_try_set_caps_func: : pad colorspace0:sink [22:17] <phubuh> doesn't accept caps [22:17] <phubuh> INFO (28329: 0)gstpad.c(1485): gst_pad_try_set_caps: : tried to set caps on peer [22:17] <phubuh> pad colorspace0:sink but couldn't, return value -1 [22:17] <phubuh> DEBUG(28329: 0)gstcaps.c(276): gst_caps_unref: : unref 0x808a070 (2->1) 1 [22:18] <phubuh> DEBUG(28329: 0)v4lsrc_calls.c(543): : : [v4lsrc0] V4LSRC: gonna try out palette [22:18] <phubuh> format 15 (YUV-4:2:0 (planar)) [22:18] <phubuh> INFO (28329: 0)gstpad.c(1408): gst_pad_try_set_caps_func: : got reply REFUSED (- [22:18] <phubuh> 1) from link function on pad v4lsrc0:src [22:18] <phubuh> INFO (28329: 0)gstpad.c(1418): gst_pad_try_set_caps_func: : pad v4lsrc0:src does [22:18] <phubuh> n't accept caps [22:18] <phubuh> Ugh. [22:18] <BBB> that means your card didn't like it [22:19] mxpxpod (~bryan@65.243.233.2) joined #gstreamer. [22:19] <phubuh> Card? [22:19] <BBB> your v4l card [22:19] <BBB> or cam [22:19] <BBB> whatever [22:19] <phubuh> Oh, I'm just grabbing from a webcam. [22:19] <phubuh> Am I screwed? [22:19] <BBB> well, it means v4lsrc didn't find a format that your webcam likes [22:19] <BBB> which formats does your webcam support? [22:20] <phubuh> Well, when I remove the EffecTV it works, and it says YUV-4:2:2 [22:20] <BBB> colorspace doesn't have a yuv422-to-rgb conversion [22:21] <BBB> one of the reasons why I truely hate colorspace [22:21] <BBB> it's so unfinished [22:21] <BBB> put two instances of colorspace ;) [22:21] <phubuh> v4lsrc ! colorspace ! xvideosink works... [22:21] <phubuh> colorspace ! colorspace? :) [22:21] <BBB> i.e., v4lsrc ! colorspace ! colorspace ! effectv ! colorspace ! xvideosink [22:21] <BBB> yes [22:21] <phubuh> I'll try [22:22] <phubuh> Still doesn't want to play, but I get four of these warnings: [22:22] <phubuh> (process:28339): GStreamer-WARNING **: trying to set non fixed caps on pad colorspace0:src, not allowed [22:23] <dolphy> BBB: so is the xv x11 split done or not finally ? [22:24] <BBB> it should be done [22:24] <BBB> but I didn't do it [22:24] <BBB> (yet) [22:24] <BBB> phubuh: ok, nevermind then ;) [22:24] <BBB> phubuh: colorspace is broken... use HEAD and use ffcolorspace... or so ;) [22:24] <phubuh> Aha. [22:25] <dolphy> BBB: any code on the way on your comp ? [22:25] <BBB> no [22:26] <dolphy> BBB: ok is it ok with you if i do that ? [22:26] <BBB> yeah, please do [22:26] <dolphy> BBB: ximagesink and xvideosink [22:26] <phubuh> Is there something I can read to understand what colorspace and the other weird stuff is? :-) [22:26] <BBB> phubuh: hm... www.mir.com? [22:27] <BBB> http://www.mir.com/DMG/ [22:27] <BBB> everything oncolorspaces [22:27] <phubuh> Cool. Thanks. [22:31] walters (wa...@ve...) joined #gstreamer. [22:33] <taaz> hey walters [22:34] <walters> hi taaz [22:39] Action: dolphy starts GstXImageSink [22:39] <BBB> walters: sorry I bumped away last week, accident with the router here [22:39] <BBB> you were giving comments on GstInterface ;) [22:39] <BBB> dolphy: are you going to revert gstvideosink? or do you want to keep that in there for now? [22:40] <BBB> GstXImageSink sounds good :) [22:41] <BBB> after that (yeah, I'm giving away tasks for free ;) ), you want to give ranks to the video sinks (sdlvideosink gets secondary, xvideosink gets primary, ximagesink gets a lower rank, maybe aasink gets one lower than ximagesink? :D) and then make an autodetection method for this [22:41] <BBB> :) [22:41] <BBB> and put that in gst-player and make sure they all support the xoverlay interface [22:41] <BBB> if you get that done, you totally rock ;) [22:42] <dolphy> BBB: i m starting from scratch [22:42] <dolphy> BBB: to get a clean ximagesink [22:42] <BBB> good idea... [22:42] <dolphy> BBB: then i ll copy that and make it xv enable [22:42] <dolphy> BBB: that will be the new xvideosink [22:42] <BBB> :) [22:43] <BBB> sounds good [22:43] <dolphy> BBB: they will both implement navigation and xoverlay interface [22:43] <BBB> make sure they both support xoverlay, though [22:43] <BBB> good, thanks :) [22:43] <dolphy> BBB: and listen to configure events [22:43] <BBB> maybe you can share the navigation support for both, if it's much code [22:43] <BBB> it it's little code, don't bother [22:43] <dolphy> yup [22:43] Action: BBB loves dolphy [22:44] <dolphy> ;-) [22:44] <BBB> tell noelle she needs to treat you very weel if you've finished this ;) [22:44] <dolphy> BBB: she is reading :) [22:45] <BBB> what is she reading? [22:46] <BBB> I still owe here a dutch pancake recepee [22:47] ow3n (~ow...@ti...) joined #gstreamer. [22:47] <BBB> ow3n!!!!!!!!!!!!!!!!!!!!!!!!!!!! [22:47] <ow3n> Hi. [22:47] <ow3n> Long time no see! [22:48] <BBB> true, true... is your girlfriend keeping you in custody or so? ;) [22:48] <BBB> how are you? :) [22:48] <BBB> oh, btw, I fixed asfdemux myself, you took so long ;) [22:48] <BBB> and I wrote an asfmux, too [22:48] <ow3n> :) I'm good thanks. [22:48] <BBB> :p [22:48] <ow3n> Yeah, I noticed. [22:48] <ow3n> Sorry about that... [22:48] <ow3n> Actually, I have been doing some gst stuff in the meantime. [22:48] <BBB> heh, well, your basis was good... I now see why it never worked for you :) [22:49] <BBB> you missed an alignment thingy somewhere [22:49] <BBB> I only noticed it after a few guesses [22:49] <ow3n> Doh! [22:49] <BBB> it's a very stupid weird mistake... ohwell... [22:49] <BBB> stupid mistake of gcc, that is [22:49] <BBB> it should align by default [22:49] <BBB> ohwell... [22:50] <BBB> now, since chris is preparing to leave oslo, are you going to stay there? [22:50] <BBB> that's no place for a brit, is it? [22:50] <ow3n> Yeah. I've no plans to leave in the near future. [22:50] <phubuh> Does the install make target in CVS install a pkg-config thing? [22:50] <ow3n> It's a pretty good place for a brit though. [22:50] <ow3n> The younger generation worship English football and the older generation are still grateful from the war. [22:51] <Zeenix> where would the next guadec be held? [22:51] <BBB> phubuh: yes [22:52] <BBB> Zeenix: I don't think it's decided yet [22:52] <BBB> ow3n: heh, that sounds acceptable... well, I guess it's pretty much ok if you like skiing ;) [22:52] <ow3n> I've been working on an idea I had after the media summit. [22:52] <Zeenix> BBB: whats your gues? [22:52] <ow3n> BBB: Yup. And the ski season should start any minute :) [22:52] <BBB> I heard italy wanted it [22:52] <BBB> I hope they will [22:53] <BBB> I love Rome [22:53] <BBB> and else, well, probably germany [22:53] <BBB> which isn't that bad either [22:53] <BBB> but I won't be there anyway... [22:53] <BBB> I'll have to skip it for one year [22:53] <BBB> I'll be back in next year's fall summit, though [22:53] <ow3n> The idea's similar to what's described in the plugin guide as a FilterFactory. [22:54] jeroen_ (~je...@je...) left irc: "sleep" [22:54] <Zeenix> BBB: i am trying to get admission in UK and chances are quite high that i'll be landing in UK before Feb. [22:54] <BBB> what are you doing exactly? would you mind explaining a bit? we're nosy here ;) [22:54] <BBB> cool! [22:54] <BBB> will you stay permanently? [22:55] <Zeenix> BBB: no student visa [22:55] <BBB> ah, cute [22:55] <ow3n> BBB: was "what are you doing..." to me or Zeenix? [22:55] <BBB> ow3n: you! [22:55] <ow3n> Ah, right. [22:55] <Zeenix> BBB: but i'll like to prolong my stay to my death if possible [22:55] <BBB> *grin* [22:55] <ow3n> Well, firstly, I thought it would be a good idea to design an XML language to describe binary file formats. [22:55] <BBB> we could explain that in a wrong way given the country you come from [22:55] <Zeenix> BBB: IF they like me :) [22:55] <ow3n> E.g. the ASF format. [22:56] <BBB> (terrorists also prolong their stay 'till death...) [22:56] <BBB> ow3n: wouldn't that be slow? [22:56] <ow3n> Next I decided it would be a good idea to write a meta framework for turning descriptions of plugins which use the binary format descriptions to generate plugin code. [22:57] <ow3n> Yes, very slow but it's pre-compile time. [22:57] <BBB> oh, ok [22:57] <BBB> sounds interesting... will it work? ;) [22:57] <ow3n> So basically you start off with asf.xml which describes the whole asf.xml format. [22:58] <ow3n> Then you write gstasfdemux.xml which says little more than "I'm a demuxer for the format described in asf.xml" and kaboom, you get gstasfdemux.[ch] [22:58] <BBB> hm... sounds interesting... [22:58] <BBB> I'm interested in how you want to do that ;) sounds like an awful hard job to me :o [22:58] <ow3n> I've pretty much got the format language sorted out and I've written a view for it too. [22:59] <ow3n> And I'm about half way with the second part. [22:59] <ow3n> I've got as far as being able to describe the ... [truncated message content] |