IRC Logs
IRC <[email protected]>
| Newsgroups | gmane.comp.video.gstreamer.daily |
|---|---|
| Message-ID | <[email protected]> |
******************************************************************* [03:01] pb_ ([email protected]) joined #gstreamer. [03:15] pb_ ([email protected]) left irc: "Client exiting" [03:22] earthworm ([email protected]) left irc: ""I allege that SCO is full of it." - Linus Torvalds" [03:26] Action: taf2 gstreamer makes my head hurt [03:31] harshy ([email protected]) joined #gstreamer. [03:44] <taf2> is gstreamer/play/play.h depreciated? [03:55] <taf2> how does gst_gconf_get_default_audio_sink determine the correct sound sink? [04:04] <taf2> is GstTagList new to gstreamer-0.7.x ? or does anyone know what header its in? [04:05] aeyakovenko ([email protected]) joined #gstreamer. [04:07] <bitshifter> taf2: I think it's been put in shortly after 0.7.2 [04:24] dilinger ([email protected]) left irc: Read error: 110 (Connection timed out) [04:29] <taaz> any docbook peoples here? [04:30] <taaz> wondering if it's ok to change gst-editor help doc s/artheader/articleinfo/ [04:30] <taaz> to close http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=218594 [04:30] Action: taaz changes it anyway [04:34] <aeyakovenko> can someone explain to me how to link a tee? i am trying to do this link(src, tee, null); link(tee, sink, null); [04:47] <aeyakovenko> should this work? [04:47] <aeyakovenko> gst_element_link (v4lsrc, tee ); [04:47] <aeyakovenko> gst_bin_add_many (GST_BIN (bin1), v4lsrc, tee, NULL); [04:47] <aeyakovenko> gst_element_link (bin1, xvideosink); [04:47] <aeyakovenko> gst_bin_add_many (GST_BIN (thread), bin1, xvideosink, NULL); [04:50] <bitshifter> aeyakovenko: according to http://www.gstreamer.net/docs/cvs/manual/html/sec-pads-type.html it looks like you have to use gst_element_get_request_pad(tee, "src%d") or so to make the tee element create a new source pad [04:51] <bitshifter> (however, feel free to ignore me, as I'm completely new to gstreamer myself) [04:52] <aeyakovenko> bitshifter: thanks a bunch for your help [05:00] Miko5881 ([email protected]) joined #gstreamer. [05:04] dilinger ([email protected]) joined #gstreamer. [05:08] <aeyakovenko> so do i need to create a bin to be able to link with that pad? [05:17] dilinger ([email protected]) left irc: "Terminated with extreme prejudice - dircproxy 1.0.5" [05:17] dilinger ([email protected]) joined #gstreamer. [05:18] <bitshifter> srcpad = gst_element_get_request_pad(tee, "src%d"); [05:18] <bitshifter> sinkpad = gst_element_get_pad(typefind, "sink"); [05:18] <bitshifter> gst_pad_link(srcpad, sinkpad); [05:18] ChrisHJW ([email protected]) joined #gstreamer. [05:18] <bitshifter> aeyakovenko: that's how I did it (worked, but don't know if that's the right/best way) [05:20] <aeyakovenko> like this? [05:20] <aeyakovenko> gst_element_add_ghost_pad(bin2, teesrc, gst_pad_get_name(teesrc)); [05:20] <aeyakovenko> gst_element_link_pads(bin2, gst_pad_get_name(teesrc), xvideosink, "sink"); [05:20] <aeyakovenko> [05:20] Action: bitshifter has no idea, he just knows that what he pasted worked for him [05:31] <aeyakovenko> how do i add the pads to the pipeline? [05:41] <bitshifter> as far as I understand it: you don't. Pads belong to elements and are used to connect elements. So you just add the elements to the pipeline and either link the elements (in which case the pads will be linked automatically) or (where that doesn't work) link the pads 'manually' [05:43] <aeyakovenko> thats what i thought as well, do you need to unlink the pads and unref them? [05:49] fridge ([email protected]) joined #gstreamer. [05:51] <aeyakovenko> for some reason it doesnt look like gst_pad_link is working [05:55] <aeyakovenko> i am trying to make a simple pipeline v4lsrc->tee->xvideosink, v4lsrc->xvideosink works great, but with the tee i am unable to get any output [05:58] Company ([email protected]) joined #gstreamer. [06:00] <fridge> in rhythmbox, when I change songs it sits in futex() for up to a minute, could someone tell me why, or help me further diagnose the problem? I'm not sure if its in a gstreamer part or not [06:01] <smoke_> why in devils name is rhythmbox not able to have a working time-slider :/ [06:01] <smoke_> everytime i fast forward in a song it CRASHES [06:02] <Company> that's easy - you can debug seeking with gst-launch [06:02] <Company> can'T [06:03] <aeyakovenko> is there any way to check if two pads linked correctly? [06:03] <Company> what do you mean "correctly"? [06:03] <Company> either they link or they don't [06:04] <aeyakovenko> well i am trying to build a simple pipe v4lsrc->tee->xvideosink, it works without the tee, but with the tee i cant get it output [06:05] <Company> 0.6 or HEAD? [06:05] <aeyakovenko> 0.6 [06:05] <aeyakovenko> is there a problem with tee's in 0.6? [06:05] <aeyakovenko> i noticed i cant get that pipe to work in the editor either [06:05] <Company> none that i know of [06:05] <Company> but i won't help debugging 0.6 ;) [06:06] <aeyakovenko> hehe, well i guess i have to install 0.7 [06:09] Action: taf2 todd finally makes break through in understanding gstreamer library :) [06:09] <Company> grats :) [06:13] <aeyakovenko> when i link two pads do i need somehow add them to a pipeline? or is it enough that their elements are in a pipeline [06:14] <Company> only the elements need to be added [06:15] <aeyakovenko> do i need to unlink and unref the pads when i am done with them? if i want to still keep the pipeline [06:15] <Company> uh, keeping pipelines is a buggy business [06:16] <Company> it's bad now, it was worse in 0.6 ... [06:16] <aeyakovenko> also, if i want to use gst_thread, the thread bin needs to have all the other elements in it? [06:16] <Company> but no, you don'T need to unref or unlink [06:16] <Company> the thread bin needs to have all elements in it that should run in that thread [06:17] <aeyakovenko> is adding and removing sinks from a pipeline buggy? [06:18] Nick change: danb_ -> danb [06:18] <Company> that should work [06:19] <aeyakovenko> from my understanding the correct way to do it would be to use a tee, and link and unlink the sinks to it as needed. is that correct? [06:19] <Company> yes [06:20] <aeyakovenko> cool [06:34] ChrisHJW ([email protected]) left irc: Read error: 110 (Connection timed out) [06:49] Miko5881 ([email protected]) left irc: "ChatZilla 0.8.31 [Mozilla rv:1.4/20030928]" [06:56] desrt ([email protected]) left irc: "leaving" [06:58] thaytoo ([email protected]) left irc: "leaving" [07:01] taf2 ([email protected]) left irc: "Leaving" [07:15] <aeyakovenko> hmm, i cant seem to compile gst-plugins-0.7.3, i get a bunch of errors in gstxvidenc.c [07:44] <aeyakovenko> i think i got it now [07:45] apoc__ ([email protected]) joined #gstreamer. [07:52] Action: sxpert is back (gone 05:33:46) [07:54] <aeyakovenko> doh, well i cant seem to get oss or v4l plugins compiled [08:00] mathrick|sleep ([email protected]) left irc: Remote closed the connection [08:01] somex1 ([email protected]) joined #gstreamer. [08:03] apoc_ ([email protected]) left irc: Read error: 110 (Connection timed out) [08:16] apoc__ ([email protected]) left irc: "Leaving" [08:32] aeyakovenko ([email protected]) left #gstreamer. [08:42] fridge ([email protected]) left irc: Remote closed the connection [09:03] Shoragan ([email protected]) joined #gstreamer. [09:28] rk ([email protected]) joined #gstreamer. [09:30] KoRnouille ([email protected]) joined #gstreamer. [09:32] Kurosu_ ([email protected]) joined #gstreamer. [09:34] swentel ([email protected]) joined #gstreamer. [09:39] somex1 ([email protected]) left #gstreamer. [09:50] Kurosu ([email protected]) left irc: Read error: 110 (Connection timed out) [09:50] Nick change: Kurosu_ -> Kurosu [09:52] thaytan ([email protected]) joined #gstreamer. [09:57] BBB ([email protected]) joined #gstreamer. [10:00] mathrick ([email protected]) joined #gstreamer. [10:00] <mathrick> mornin [10:01] <BBB> hi [10:01] <thaytan> hi! [10:13] jdahlin ([email protected]) left irc: "Leaving" [10:15] thomasvs ([email protected]) left irc: Read error: 110 (Connection timed out) [10:32] <BBB> thaytan: these colour tables that you talked about, have you found a solution yet? [10:33] dolphy ([email protected]) joined #gstreamer. [10:33] <mathrick> yo dolphy [10:34] thomasvs ([email protected]) joined #gstreamer. [10:34] <BBB> hi fluendo's [10:34] <thaytan> BBB: not really - they need to go from the dvdnavsrc, through mpegdemux to the mpeg subtitler [10:34] <thaytan> so, I see 2 solutions [10:34] <mathrick> thomasvs: http://www.w3.org/TR/html4/charset.html#h-5.2.2 :\ [10:35] <mathrick> thomasvs: apparently, mozilla's behavir is not a bug [10:35] <thaytan> 1) Change dvdnavsrc to produce a new mime type that is actually nav packets interleaved with mpeg data [10:35] <mathrick> thomasvs: and hi btw :) [10:35] <dolphy> hi [10:35] <BBB> I'd rather see MPEG data and nav data separated [10:35] <BBB> mangling those two is not good [10:35] <thaytan> 2) make dvdnavsrc also an mpegdemuxer that has only audio, video and subtitle pads [10:35] Action: dolphy decides he will try to go to bed early tonight :-/ [10:36] <BBB> can't you add two pads to dvdnavsrc? [10:36] <thaytan> BBB: I don't know how to issue the data from dvdnavsrc so the streams can be synchronised [10:36] <BBB> dvdnavsrc -> mpegdemux -> mpeg2dec -> .. && dvdnavsrc0.colourtables_00 -> mpeg2subt -> .. [10:36] <BBB> can you timestamp them? [10:37] <thaytan> perhaps, if mpegdemux preserves the timestamps [10:37] <BBB> it doesn't [10:37] <BBB> it takes MPEG PTS values [10:37] <BBB> or, if these aren't available, SCR values [10:38] <thaytan> so I don't see how to do it [10:38] <BBB> can you explain exactly how it works at all? [10:38] <BBB> I mean, how is menu data transported? [10:38] <BBB> is it MPEG? [10:38] <BBB> is it nav data? [10:39] <BBB> both? [10:39] <thaytan> both [10:39] <BBB> what's in the nav packets? [10:39] <thaytan> there is a subtitle issued from mpegdemux, which has the overlay for all buttons in a menu [10:39] <thaytan> then dvdvnavsrc gets told 'now, set this colour table', 'now crop to this region' [10:40] <thaytan> and gets those as you move over each button position [10:40] <thaytan> and different ones when you actually click, to show activation [10:40] <thaytan> in my local tree, I'm sending an event downstream for each of those [10:40] <thaytan> and mpeg2subt handles them [10:41] <thaytan> but I need a different way [10:41] water ([email protected]) joined #gstreamer. [10:42] <thaytan> substantially, the event I send is equivalent to interleaving a buffer which mpegdemux knew to handle specially [10:43] <thomasvs> mathrick: I know - because the server cannot use the meta tags to set the encoding, since the header is already sent by then [10:43] <thomasvs> mathrick: anyway, I'll ask the freedesktop guys to switch their apache, and in the meantime I use htaccess files for it [10:44] <mathrick> thomasvs: will switchinga apache help? [10:44] <mathrick> thomasvs: and anyway, I think it's dumb for html spec to set priorities that way. It should be content which sets encoding, not server [10:45] <thomasvs> mathrick: well, the server has no way of knowing without parsing content [10:46] <thomasvs> BBB: btw, there are three symbols you cannot use literally in docbook/xml [10:46] <thomasvs> < > & [10:46] <thomasvs> you have to write them ALWAYS as [10:46] water ([email protected]) left #gstreamer ("The water has parted."). [10:47] <thomasvs> < > & [10:47] <thomasvs> BBB: don't make me smack your ass again :) [10:47] <thomasvs> BBB: your new pwg is online now right now bw [10:47] <thomasvs> http://www.freedesktop.org/~gstreamer/data/doc/gstreamer/head/pwg/html/index.html [10:48] teuf ([email protected]) joined #gstreamer. [10:48] <mathrick> thomasvs: I know, that's why I think it's dumb. Besides, setting encoding with server requires higher privileges than setting the content, very often you don't have access to server config [10:48] <mathrick> yo teuf [10:48] <teuf> morning all [10:48] <thomasvs> mathrick: agreed [10:48] <thomasvs> mathrick: but .htaccess will do nice in the meantime [10:49] <mathrick> thomasvs: what exactly do you mean by switching apache? 2.0? [10:49] <thomasvs> mathrick: no, have them change the server config to serve utf8 by default [10:49] <mathrick> thomasvs: ahh, but it's still workaround for stupid spec IMHO :\ [10:50] <bitshifter> are there any known problems with the mp3parse/mpegaudioparse element in cvs? [10:51] <thomasvs> bitshifter: well, apart from them being unmaintained and having a high chance of not working, no [10:51] <thomasvs> bitshifter: what do you need them for ? [10:51] <thomasvs> man, my fonts are completely different today [10:51] <thomasvs> I must have killed a kitten [10:51] <bitshifter> thomasvs: filtering out random rubbish and tags from an mpeg audio stream [10:52] <BBB> thomasvs: but it compiles for me? [10:52] <thomasvs> BBB: then that's a bug [10:52] <BBB> I'm trying to make all & an &, same for </>, but I might miss some here and there [10:52] <BBB> I'm not getting compile errors, so that's probably my issue [10:52] <BBB> sorry... [10:52] <thomasvs> well, fix your tools :) [10:53] <thomasvs> BBB: so how do you like your stuff on the web ? [10:53] <bitshifter> gst-launch-0.7 filesrc location=in.mp3 ! mp3parse ! filesink location=out.mp3 produces mp3s with noticable skips for me, both with 0.7 and 0.6 [10:53] <thomasvs> bitshifter: that's an interesting application for mp3parse [10:54] <thomasvs> bitshifter: don't think anyone tried that before [10:54] <thomasvs> bitshifter: but I agree that it should work [10:54] <thomasvs> bitshifter: so please file a bug for it [10:54] <bitshifter> thomasvs: what's the intended application? [10:54] <thomasvs> bitshifter: making sure one buffer is one frame, was the basic intent [10:54] <thomasvs> but it's not really needed anymore with the mad plugin [10:55] <BBB> thomasvs: well, I'm still updating it, but I think it looks a lot better than our old PWG :) [10:55] <BBB> but it's really cool! [10:55] <teuf> BBB: is your updated pwg online ? [10:56] <BBB> well, the one from yesterday, I guess... but yes [10:56] Action: BBB is getting very happy with the PWG now [10:57] <teuf> BBB: the one from yesterday is the one at http://www.freedesktop.org/~gstreamer/data/doc/gstreamer/head/pwg/html/index.html ? [10:59] <BBB> yes [11:01] <BBB> currently, autoplugging (3.3.2), clocking (3.5), midi (3.7), all specific interfaces (3.8.*) and everything after that (4.*, 5.*) is not written yet [11:01] <BBB> the rest is pretty much done [11:02] <thomasvs> BBB: on the matter of code fragments though [11:02] <thomasvs> I'd like it if we used the same system as for the manual [11:02] <thomasvs> ie, put in compilable fragments [11:02] <thomasvs> so that we can at least check if the fragments stay valid [11:02] <BBB> most stuff I provide is example code, so not a full plugin, but rather small fragments from a full plugin [11:03] <thomasvs> BBB: right, so they should all be fragments from one complete source file [11:03] <thomasvs> BBB: so that that source can be extracted and compiled [11:03] <thomasvs> to make sure the code is compilable [11:03] <thomasvs> I have a very good hunch it won't be right now for example :) [11:04] <BBB> I'm sure there will be things wrong [11:04] <BBB> but that's hy it's 0.7.4, not 0.8.0 ;) [11:04] <BBB> I'll try to work on that, ok? [11:04] <BBB> for now, I first want to provide some documentation on interfaces... once that's done, I'll work on making C files for each code fragment [11:05] <BBB> it'd be very helpful if someone that's not too familiar with gst internals could give me a clue on whether the provided code samples are actually helpful [11:05] <BBB> i.e., is there anything missing, am I not explaining stuff correctly, that sort of things [11:05] Company ([email protected]) left irc: Remote closed the connection [11:06] <thomasvs> BBB: just like in the manual, it should be done from the start right so that's maintainable [11:06] <thomasvs> BBB: you're not going to check the code on each release in the manual/pwg [11:06] <thomasvs> so you need to make sure its' done automatically [11:06] <thomasvs> it's not that hard, it just needs to be done [11:06] <BBB> I know [11:06] <BBB> as I said, I will do it [11:07] <BBB> I'll first finish the interfaces documentation, and then I'll work on making C files [11:07] <BBB> I'll look in the appdevmanual on how to do that [11:07] <thaytan> does alsamixer have to open /dev/snd/pcm.... ? [11:08] <thaytan> gst-mixer doesn't work here because there's too many things that already have that open [11:08] <BBB> didn't I commit a fix for that? [11:08] <BBB> oh [11:08] <BBB> wait [11:08] <BBB> that was ossmixer [11:14] <thaytan> I suspect there is also no good reason for it to g_assert when one sets state 'PAUSED' or 'RUNNING' [11:14] Action: BBB really doesn't know [11:14] <BBB> I don't use ALSA [11:14] <BBB> Leif said he'd work on ALSA... I hope he will [11:14] <BBB> Leif really knows ALSA a lot better than me [11:29] Nick change: mathrick -> mathrick|Uni [11:42] <thomasvs> whee [11:42] <thomasvs> so now hte website autorebuilds itself when someone commits to cvs [11:42] <thomasvs> on to the other bits [11:46] Rotty ([email protected]) joined #gstreamer. [11:51] wheels ([email protected]) left irc: Remote closed the connection [11:53] <BBB> thomasvs: also when I commit PWG stuff? [11:53] <BBB> so when I commit my current, it'll all work? [11:57] alley_cat ([email protected]) left irc: Read error: 110 (Connection timed out) [11:59] alley_cat ([email protected]) joined #gstreamer. [12:00] <thomasvs> BBB: nope, that's completely separate [12:00] <thomasvs> BBB: but i'll explain that [12:00] <thomasvs> BBB: I could maybe set up a nightly build though [12:05] <BBB> nah, if you make a script that's good enough [12:05] <BBB> a script that I can call, like we had on SF [12:08] <thomasvs> it'll be make upload in the pwg dir [12:08] <thomasvs> well, it already is [12:08] <thomasvs> but I need to fix it up a little [12:12] walken ([email protected]) joined #gstreamer. [12:17] <BBB> oh, ok [12:17] <BBB> well, /me plays with interfaces [12:17] <BBB> do you know that the PDF doc from PWG is nearly a 100 pages? [12:19] <thomasvs> yep [12:19] <thomasvs> lots of white :) [12:19] <swentel> white is pretty :) [12:19] <thaytan> thomasvs: I can fix that [12:19] Action: thaytan has plenty of coloured paper [12:22] <BBB> thomasvs: not anymore [12:23] <BBB> thomasvs: most stuff is there now [12:23] <BBB> I just said what parts need to be done... interfaces, clocking, midi, autoplugging and everything in section 4 and 5 [12:23] <BBB> I'm doing interfaces right now [12:24] <thomasvs> it's great that you're doing this [12:24] Action: thomasvs fixes the upload targets for docs [12:24] jdahlin ([email protected]) joined #gstreamer. [12:26] <BBB> well, someone needs to do it ; [12:26] <BBB> ;) [12:26] <BBB> and I baically wanted to do this in september already [12:26] <BBB> I just didn't have time [12:28] <thomasvs> heh, me too :) [12:28] <thomasvs> BBB: so, we need some good docs on a) time and clocking b) seeking [12:28] <thomasvs> and how plugins do it [12:29] <BBB> I'll probably make an extra page on events (source pad event handling) in the advanced section [12:30] <BBB> I'll explain all events there [12:30] <BBB> like I'm doing with interfaces now [12:35] Company ([email protected]) joined #gstreamer. [12:37] <thaytan> hi Company [12:42] <Company> hi [12:42] <Company> our uni is on strike [12:42] <Company> but apparently > 80% don't care [12:43] <Company> so the president warned that the others will be punished for blocking entrance to courses [12:43] <Company> and as it looks, still noone cares [12:43] <Company> yay for being unpolitic [12:45] <thaytan> sounds like every strike we ever had [12:45] <thaytan> the one where we broke into the old bank building and then called a glass repairman was fun [12:46] <thaytan> the security guards were baffled as to how we got in, since it was fixed before they arrived ;) [12:47] <Company> well, the problem this time is they organized it really shitty [12:47] markey ([email protected]) joined #gstreamer. [12:47] <Company> there awareness raising didn't work [12:47] <Company> s/there/their/ [12:49] <bitshifter> Company: as opposed to all the other times? Say 1998 (or was it 1997)? ;-) [12:50] <Company> from my point of view 98 was done far better [12:50] <Company> they at least got courses to completely stop [12:50] <Company> and not just in 2 buildings [12:50] <Company> (out of 10+) [12:52] <thaytan> getting Toohey's to sponsor a keg-up always worked for us for stopping classes [12:55] <Company> if they wouldn't suck big this time, the president threatening them would at least have caused an uproar [12:55] <Company> but it doesn't look as if anything happened [12:57] sublett ([email protected]) joined #gstreamer. [13:01] <BBB> Company: wanna write clocking docs? [13:01] Action: thomasvs cleans out all of the built docs from fdo and runs "make upload" in gst/docs to test [13:03] <Company> BBB: not now [13:05] Action: BBB has PWG mixer interface docs [13:07] <Company> BBB: have you made anyone read the docs and comment if they understand it? Preferrably people with no clue about gst? [13:07] <BBB> I want to ask some people to do that, yes [13:07] <BBB> but no, we haven't done that yet [13:08] <BBB> I think it's too early, large parts still need writing anyway [13:08] <thomasvs> I don't mind reading it [13:08] <thomasvs> I can pretend I don't have a clue [13:08] <BBB> you know too much [13:08] <BBB> :p [13:08] Action: bitshifter is happy to volunteer for reading [13:08] <BBB> please! [13:08] Action: swentel will try it too, I really need to get into gst someday :) [13:08] <BBB> if you checkout gstreamer from CVS, simply 'make' and then point your browser to $(cvscheckoutdir)/gstreamer/docs/pwg/html/index.hmtl [13:09] <BBB> s/mtl/tml/ [13:09] <thomasvs> BBB: dude [13:09] <thomasvs> BBB: at least point them to the online docs :) [13:09] <BBB> oh wait [13:09] <BBB> I'll upload it [13:09] <BBB> :X [13:09] <thomasvs> BBB: you think I'm doing this for nothing ??? [13:09] Action: BBB stupid [13:09] <thomasvs> BBB: hang on [13:09] <thomasvs> BBB: wait five secs [13:09] <BBB> sorry, I forgot [13:09] <thomasvs> BBB: I'll tell you how to after I commit [13:09] <BBB> I have a really bad short-term memory [13:09] <thomasvs> BBB: it's pretty much done, and you can test it in a minute [13:09] <BBB> ok [13:09] <BBB> I'll commit my latest changes and try it out then [13:10] <thomasvs> yep [13:10] <thomasvs> almost done, proofchecking [13:11] <Company> huh, our docs autobuild online now after i check them in? [13:11] <BBB> no [13:12] <BBB> but it's one command to put them online [13:12] <BBB> 'make upload' or so [13:12] <Company> ah [13:12] <Company> that's not as good as i still need to fix my docbuild then ;) [13:13] markey ([email protected]) left irc: "bbl" [13:13] <BBB> you can ask one of us to upload... [13:14] <thomasvs> Company: well, you need to fix them anyway [13:14] <thomasvs> Company: it would be completely wrong to autobuild docs on commit if you can't even get them to build yourself :) [13:14] <Company> BBB: yeah, but that's like checking in uncompiled code... [13:15] <Company> thomasvs: i know, i'll bite the bullet some day (and i can still use my debian laptop i suppose) [13:16] <thomasvs> Company: anyway, I've done quite some work to make sure all of our website content is now very easy to update, so it should help in stimulating people to work on it [13:16] <thomasvs> BBB: ok, so [13:16] <thomasvs> update from cvs in docs/ [13:16] <Company> the bad part was that you changed the doc build system after i bit the bullet last time and spend one or two days getting the doc build running... [13:16] <thomasvs> read the first paragraph from HACKING [13:16] <BBB> will do [13:16] <thomasvs> Company: yeah, I'm sorry for that - but this time I know we're doing the futureproof thing [13:16] <thomasvs> BBB: it should tell you everything you need [13:16] <thomasvs> BBB: then upload your changes using that system, and check if your latest changes are online [13:17] <BBB> and let others proofread it ;) [13:17] <thaytan> all of a sudden my toolbar in gst-editor doesn't work, and I have no idea why [13:17] <thomasvs> Company: I spent a lot of time comparing with other projects and asking around what the future of docbuilding looks like [13:17] <thaytan> nor, looking at the source, do I have any idea how it ever worked :) [13:18] <thomasvs> Company: so - why was the clocking changed to read absolute machine time ? [13:18] <BBB> what's wrong? [13:18] <thomasvs> Company: and where should we start investigating it ? [13:18] <thaytan> BBB: me? [13:18] <thaytan> it doesn't show any buttons [13:19] <thaytan> OK, I now see how it _should_ work - it's all in the glade file [13:19] <thaytan> just not showing, for some reason [13:19] <BBB> crap, I updated in top_srcdir and now configure wants to rebuild [13:19] <BBB> bah [13:19] sublett ([email protected]) left irc: "I like food, food is good!" [13:19] <thomasvs> BBB: heh :) [13:20] <thomasvs> BBB: abort it [13:20] <thomasvs> BBB: cd docs [13:20] <thomasvs> make upload [13:20] <thomasvs> maybe that helps :) [13:20] <BBB> nah, I'll let it run [13:20] <BBB> need to re-run it anyway [13:20] <BBB> it'll just take a second longer [13:20] Action: thomasvs is happy with the new docs upload [13:20] <thomasvs> on to fixing the rest of the site [13:20] Action: BBB is happy too [13:21] Action: thomasvs adds scripts to sync data/ dir to and from online site [13:23] <BBB> do I make upload in docs/ or docs/pwg? [13:23] <thomasvs> BBB: you can choose either [13:23] <thomasvs> of course docs/pwg only uploads the pwg [13:24] <BBB> that's good enough for me... :) [13:24] <BBB> Uploading docs to freedesktop.org:/home/projects/gstreamer/www/data/doc/gstreamer/head/pwg [13:24] <Company> thomasvs: because changing the time of a clock is evil [13:24] <BBB> chown html : Operation not permitted [13:24] <BBB> rsync error: some files could not be transferred (code 23) at main.c(620) [13:25] <BBB> same for pwg.ps and pwg.pdf [13:25] <Company> thomasvs: a clock might be used in more than one pipeline and resetting the clock while one pipeline is already playing would've caused havoc [13:25] <BBB> http://www.freedesktop.org/~gstreamer/data/doc/gstreamer/head/pwg/html/index.html - for whoever wants to proofread it [13:26] <thomasvs> Company: so why not just make sure a clock is only used by one pipeline ? [13:26] <thomasvs> BBB: ok, let me check [13:26] <thomasvs> BBB: probably chmod settings online [13:26] <Company> thomasvs: that would still not solve the problem when you run 2 independant streams inside one pipeline [13:27] <Company> thomasvs: and apart from that the systemclock only exists once [13:33] <BBB> http://www.freedesktop.org/~gstreamer/features/ - the link to the PWG is wrong [13:33] <BBB> http://www.freedesktop.org/~gstreamer/features/docs/current/pwg/ [13:34] <thomasvs> BBB: sure, lots of things are still wrong, don't worry [13:34] <BBB> ok [13:34] <BBB> :) [13:35] <thomasvs> Company: well, we need to figure out something soon so we can unfuck seeking [13:36] <Company> that's just bugfixing [13:36] <thomasvs> just bugfixing ? [13:36] <thomasvs> sure [13:36] <bitshifter> BBB: do you want the comments by mail or here? [13:36] <BBB> well, it's an important feature, company [13:36] <BBB> bitshifter: mail please... if it's small, put it here [13:36] <thomasvs> but no-one's doing it, and all our apps display 24903842834328 as the current position in the song [13:36] <Company> i know, i'll look at it and fix it [13:37] <BBB> bitshifter: or better yet, bugzilla please [13:37] <bitshifter> k [13:37] <Company> or you'll look at it and fix it [13:37] <thomasvs> so we need to fix it before release or revert clocking [13:37] <thomasvs> and since we want to release this week ... [13:37] <BBB> do we? [13:37] <Company> reverting clocking will freak out a/v sznc completely [13:37] <Company> but at least you can seek somehow ;) [13:37] <thomasvs> BBB: yep [13:38] <BBB> oh [13:38] <Company> i'd be delighted if anyone else takes care of clocking [13:38] <thomasvs> Company: i know it's not ideal, but having broken seeking for over two weeks and all our apps people complaining about it a month and a half before 0.8 is not good either [13:38] <Company> it's not only hard but hard to debug, too [13:38] <BBB> does that mean the docs will have to be largely finished by the end of this week? [13:38] <thomasvs> Company: I want to give it a shot, but I'll need a lot more verbose help [13:38] <thomasvs> BBB: not necessarily [13:38] <BBB> Company: I don't even know how clocking works right now [13:38] <thomasvs> Company: so if you're up for teaching me, feel free [13:38] <BBB> Company: I barely understoof the old system [13:38] <thomasvs> Company: I don't mind digging through the code either [13:39] <Company> the old system did 2 things at once and failed at both [13:39] <thomasvs> BBB: ok, reupdate from cvs and retry make upload [13:39] <Company> the new thing just does one thing and the other is unsolved but works sort of [13:39] mathrick|Uni ([email protected]) left irc: Remote closed the connection [13:39] <Company> that's how i'd start looking at it :) [13:40] <thomasvs> Company: well, I can read the code, but to know your opinions on both the old wrong system and your ideas for the new it'd be nice if you send a mail explaining that [13:40] <Company> btw: seeking wasn't broken for 1.5 months (apart from the fact that it takes time if you seek through loads of queues) [13:40] <thomasvs> that'll help me when looking at the code and figuring out what is wrong [13:41] <thomasvs> Company: no, I said two weeks broken, 1.5 months before 0.8 [13:41] <Company> ah, understood [13:41] <Company> we need a simple app to test seeking with [13:41] <thomasvs> I don't really know how long it's been broken for, two weeks is a rough estimate [13:41] <thomasvs> agreed [13:41] <thomasvs> I was thinking of the following [13:41] <Company> anyone wanna write gst-seeker? :) [13:41] <thomasvs> * write a sirensrc [13:41] <BBB> chmod: changing permissions of `/home/projects/gstreamer/www/data/doc/gstreamer/head/pwg': Operation not permitted [13:41] <BBB> chmod: changing permissions of `/home/projects/gstreamer/www/data/doc/gstreamer/head/pwg/html': Operation not permitted [13:41] <BBB> apart from that, it works [13:41] <thomasvs> * it would implement a siren on which you can seek [13:42] <thomasvs> * the pitch of the siren would tell you if it works ok :) [13:42] <thomasvs> maybe we can have a length property too so it eos's after a given length [13:42] <thomasvs> so it acts like a "generated file" [13:42] <thomasvs> how does that sound ? [13:42] <thaytan> Company: I was thinking tonight of adding it to gst-editor [13:42] <BBB> thomasvs: that's what SEEK_SEGMENT did, and I removed that [13:42] <thaytan> then I got distracted trying to figure out why the toolbar isn't displaying [13:42] <thomasvs> BBB: SEEK_SEGMENT ? that's not an element is it ? [13:42] <BBB> thomasvs: it's not right... there's better ways to do that then in elements themselves [13:42] <BBB> it's an event [13:42] <thomasvs> BBB: to do what ? [13:42] <thomasvs> I'm not following [13:42] <BBB> it used to be an event [13:43] <Company> thomasvs: dunno - we need a simple app to test sync with, too [13:43] <BBB> to seek to a point and specify an EOS point where the stream should stop [13:43] <thomasvs> all I'm saying, if we have an element that acts as a filesrc serving audio, it's easy to test seeking on it [13:43] <thomasvs> anyways, the whole seeking/eos/discont/flush/length mess is very much undocumented, and everyone seems to have differing opinions on it [13:43] <Company> testing seeking is easy [13:43] <thomasvs> I just want a nice simple design document that I canverify the core against [13:43] <Company> just print out the current time [13:44] <Company> after seeking that should have changed [13:45] <Company> dead easy to do actually - sinesrc ! audiosink in one thread, videotestsrc ! ximagesink in another thread, print timestamps [13:46] <Company> maybe add a button to synchronize both [13:46] <Company> that sounds like i should do it [13:46] Company ([email protected]) left irc: Remote closed the connection [13:52] <BBB> thomasvs: if you find some time, please browse quickly through the docs, too... :) [13:53] Action: thomasvs starts by going through the 0.6.4 source and figuring out how clocking and seeking worked there [13:53] <thomasvs> BBB: can you update from cvs in docs/ and run make upload again ? I put in some other fixes, want to know if you get errors [13:54] <dolphy> any clues if Company fixed the element's time? [13:59] <BBB> he said he'd work onit [14:00] <BBB> chmod: changing permissions of `/home/projects/gstreamer/www/data/doc/gstreamer/head/pwg': Operation not permitted [14:01] <BBB> chmod: changing permissions of `/home/projects/gstreamer/www/data/doc/gstreamer/head/pwg/html': Operation not permitted [14:01] <BBB> still those two [14:01] <BBB> the rest works fine [14:22] walken ([email protected]) left irc: "zzzZZZzZzzzz" [14:44] hyriand ([email protected]) joined #gstreamer. [15:02] <BBB> I suppose tagging should become its own chapter rather than being a section in the interfaces chapter, and the same for filterfactory/element-construction [15:05] <BBB> dolphy: are you busy? [15:05] <BBB> thaytan: and you? [15:05] Action: BBB is getting really evil idea [15:07] <taaz> dolphy: ximagesink fix sort of works [15:07] <taaz> dolphy: xvimagesink over remote X segfaults: [15:07] <taaz> gst_xvimagesink_check_xshm_calls (xcontext=0x81dcc20) at xvimagesink.c:120 [15:07] <taaz> 120 xvimage->xvimage->data = xvimage->SHMInfo.shmaddr; [15:08] <taaz> and some other issue poped up with videotestsrc ! *sink with resizing causing segfaults [15:09] <taaz> eek... aasink just died with: [15:09] <taaz> Invalid buffer sizes! [15:09] <taaz> exited so don't know where that is [15:10] <taaz> ximagesink quick resizes sometimes causes color bars to be slanted then dies something like this: [15:10] <taaz> memset_str2 (dest=0x40b81000 <Address 0x40b81000 out of bounds>, val=0 '\0', [15:10] <taaz> n=75) at videotestsrc.c:73 [15:10] <taaz> 73 *dest = val; [15:10] <taaz> with some various val [15:11] <BBB> aasink might need some fixing [15:12] <taaz> and cacasink doesn't resize the image [15:14] Action: taaz runs off [15:15] <BBB> taaz! hold on! [15:15] <taaz> eh? [15:15] <BBB> I need you to do something for me [15:16] <BBB> there's an empty chapter 'appendix_python.xml' in the PWG - should I keep it there, waht should be in there and will you write that for me? [15:17] <taaz> uh... i'm not going to write it in the near future [15:17] <taaz> i didn't know that was there. i had planned on adding a tutorial in the gst-python package itself [15:18] <BBB> well, it's the *plugin* writer's guide [15:18] <BBB> not appdev manual [15:18] <BBB> so should I keep it or remove it? [15:18] <taaz> hey, i can write plugins too [15:18] <BBB> in python? [15:18] <taaz> they suck right now, but they do work ;) [15:18] <BBB> oh my ... [15:18] <BBB> ok, then I'll keep it [15:18] <dolphy> taaz: i fixed that some minutes ago [15:18] <taaz> heh. [15:18] <BBB> (why? WHY?!) [15:18] <BBB> ;) [15:19] <dolphy> taaz: for the xvimagesink on remote display [15:19] <taaz> to see if it would work. what other reason could there be? ;) [15:19] <dolphy> taaz: anyway xvimagesink on remote display won't work [15:19] <BBB> taaz: is it useful? [15:19] <BBB> I mean, will you write elements that my C apps can use? [15:19] <dolphy> taaz: because xv over network is impossible :) [15:19] <dolphy> taaz: regarding resizing [15:20] <dolphy> taaz: ximagesink triggers caps renegotiation when resized [15:20] <dolphy> taaz: so generating too much caps renego makes some segfaults [15:20] <BBB> dolphy: are you using videobalance in your gst-player pipeline? [15:20] <dolphy> yup [15:20] <BBB> is it fast enough after my changes? [15:20] <taaz> BBB: it's sort of useful. i use it in vdv to take in the text stream from vbisink and print it ;) [15:21] <BBB> vbisink? [15:21] <taaz> BBB: i haven't tried to do C->py usage [15:21] <taaz> closed captions decoder [15:21] <BBB> hmk.... [15:22] <BBB> dolphy: oh, and [15:22] Action: taaz really runs off now [15:22] <BBB> dolphy: will you write documentation on xoverlay interface? it's in advanced_interface.xml [15:22] sublett ([email protected]) joined #gstreamer. [15:28] <dolphy> BBB: if you want me to yes [15:28] <BBB> please [15:28] <BBB> you know it a lot better than me, and it gives me time to work on other parts of the docs, too [15:28] <BBB> it only takes an hour or so [15:29] <BBB> and I'm trying to get thaytan to do the navigation interface documentation [15:29] <BBB> I'll do tagging then [15:29] <dolphy> ok [15:30] <BBB> (note that all this is from a plugin's point of view - so don't write how apps will use it. instead, write what a plugin should do, where, why and how) [15:30] <BBB> :) [15:31] <dolphy> understood :) [15:31] Action: BBB pings thaytan [15:31] <dolphy> what tool are you using to write docs ? [15:32] <BBB> vim :p [15:32] <BBB> and sometimes anjuta [15:32] Action: BBB is trying to get used to anjuta [15:32] Action: thomasvs cleaned the devhelp docs a bit [15:33] Action: thomasvs runs make puload [15:33] markey ([email protected]) joined #gstreamer. [15:33] Action: BBB hopes thomasvs won't overwrite his new PWG with an old one [15:36] <thomasvs> BBB: doesn't really matter, it's in cvs no ? :) [15:37] <BBB> on the website ;) [15:38] <rk> cd [15:38] <rk> wrong terminal :) sorry [15:39] <thomasvs> BBB: yeah, but since your changes are in cvs, then when I do upload it's the same result [15:39] <BBB> ok [15:40] Action: thomasvs is really happy with the new docs and website push [15:40] <swentel> mmm is gstreamer.net going to move? [15:40] <thomasvs> swentel: temporarily [15:40] <thomasvs> we don't really own .net [15:41] <thomasvs> and we can't get it back either :) [15:41] <thomasvs> don't ask [15:41] <swentel> ok [15:41] <swentel> so new domain too ? [15:41] <thomasvs> well [15:41] <thomasvs> gstreamer.org is owned by the creator of gst [15:42] <thomasvs> still being discussed really [15:42] <BBB> why don't we have a gstreamer foundation to take care of all this? ;) [15:42] <swentel> btw, bidding on a arcade cabinet again ;) [15:42] <swentel> or is it betting ? or whatever :p [15:46] ct_ ([email protected]) joined #gstreamer. [15:46] iain ([email protected]) joined #gstreamer. [15:46] <thomasvs> swentel: where are hte two you bought ? [15:47] <swentel> didn't buy them, they were almost falling apart, and one monitor wasn't working [15:56] <thomasvs> man, most of our links on the site were broken anyway :) [15:57] Action: thomasvs is fixing up manual id's so we have consistent naming and linking [15:58] <sxpert_work> yesss /D [15:58] <BBB> ? [16:04] markey ([email protected]) left irc: "leaving" [16:08] <thomasvs> BBB: I just fixed up consistent id naming in the manual [16:08] taf2 ([email protected]) joined #gstreamer. [16:08] <thomasvs> BBB: I want to do the same right now for the pwg [16:09] <thomasvs> BBB: can you commit your local pwg changes so I cna do that now ? [16:09] <BBB> yup [16:09] <BBB> you mean the <sect1/chapter id=".."/> ids? [16:11] <BBB> committed, go ahead [16:12] <thomasvs> ok, thanks [16:12] Action: BBB takes a break [16:13] Action: taf2 ponders how to initialize a gstreamer app with gtk+ e.g. his example program has tried different combinations of gst_init and gtk_init with little success.... [16:14] <taf2> gst-player seems to initialize using gnome but i'd like to just use gtk for now [16:14] <teuf> taf2: just call both init functions and that should work I think [16:14] <BBB> no [16:15] <BBB> both quit if arguments of the other are given [16:15] <BBB> ;) [16:15] <taf2> teuf: so if i get errors then its my program [16:15] <BBB> teuf: gst_init_get_popt_tables (); [16:15] <taf2> ok so like the way gst-player does it [16:15] <taf2> options[3].arg = (void *) gst_init_get_popt_table (); [16:15] <BBB> something similar for gtk [16:15] Action: teuf hides ;) [16:15] <BBB> and use popt [16:15] <taf2> gst_scheduler_factory_set_default_name ("opt"); [16:15] <thomasvs> it's in the manual actually [16:16] <taf2> thomasvs: didn't find it on my first pass i'll check again. [16:16] <BBB> shall I fix the appdevman too when I'm done with the PWG? [16:16] <taf2> thomasvs: i was looking in the api :) [16:16] <thomasvs> BBB: well, discuss with me, I have a lot of pending changes for the manual [16:16] <thomasvs> BBB: I restructured some bits too [16:17] Misirlou ([email protected]) joined #gstreamer. [16:18] <taf2> http://gstreamer.net/docs/current/manual/html/ch28.html [16:18] <BBB> I'll leave it for now then [16:18] <BBB> I'll work on it when I find time, not earlier [16:22] <thomasvs> ok, let me know [16:23] <BBB> I do see that the ADM needs updating, too [16:23] Nick change: rk -> rk|away [16:23] rk|away ([email protected]) left irc: "leaving" [16:27] <dolphy> BBB: you are not talking about interface signals in your doc right ? [16:27] <dolphy> BBB: for the mixer [16:27] <BBB> no... I might add some pieces about that [16:27] <BBB> same for properties (notify) [16:27] <BBB> etc. [16:28] <BBB> if you want, add pieces [16:29] <dolphy> BBB: i ll finnish xoverlay first :) [16:30] <BBB> k [16:30] Action: BBB continues tagging [16:37] Action: thomasvs hates underscores [16:37] <thomasvs> BBB: I'm going to rename those silly underscore files [16:41] ct_ ([email protected]) left irc: "Client exiting" [16:42] <BBB> nooooooooooooooooooooooooooooooooooooooo [16:42] Action: BBB kicks thomasvs [16:42] <BBB> I'm working on them! [16:43] <thomasvs> I know, that's why I asked you to wait for a sec [16:43] <jdahlin> dolphy, ping? [16:43] <thomasvs> you just commited all your changes you said ! [16:44] <BBB> yes, but I continued after that [16:45] Action: BBB whistles [16:45] <BBB> I'll run a diff and reapply to the new file [16:45] <BBB> grmbl [16:45] <BBB> why do you want to rename them anyway? [16:46] <thomasvs> BBB: so they are the same for all our docs [16:46] <thomasvs> why have _ when all hte others have - ? [16:46] <thomasvs> BBB: give me a few more minutes, almost done [16:46] <thomasvs> BBB: but I moved files in the cvsroot, don't know if you can diff then [16:47] <BBB> I'll just overwrite yours [16:47] <thomasvs> overwrite what ? [16:47] <thomasvs> the files have moved [16:47] <thomasvs> btw, it's your mistake [16:47] <BBB> cp my_foo.xml myfoo.xml; cvs update -dP; cp myfoo.xml my-foo.xml [16:47] <thomasvs> I asked you to wait :) [16:47] <thomasvs> yeah, but I just fixed up id's [16:47] <thomasvs> argh [16:47] <thomasvs> you suck :) [16:47] <thomasvs> well you don't [16:47] <thomasvs> but you're too impatient :) [16:48] <BBB> you take too long [16:48] <BBB> :p [16:48] <thomasvs> you're still my hero [16:48] <thomasvs> but communication is ipmortant :) [16:48] <BBB> thank tyou ;) [16:48] <BBB> I'll fix the Ids [16:48] <BBB> can I update? [16:48] <thomasvs> how much files did you change ? [16:48] <thomasvs> hang on, almost done [16:48] <BBB> one [16:49] <BBB> it only has three IDs [16:49] <thomasvs> ok, first make a backup copy of your dir to make sure you don't lose anything [16:49] <BBB> I only have one file, I'll only backup that one [16:49] <thomasvs> ok [16:49] <BBB> the rest is in CVS :) [16:50] <thomasvs> it's going to complain a little when you update, don't be scared [16:50] <thomasvs> easiest is to remove your whole pwg dir and do a fresh checkout [16:50] <thomasvs> but keep your Makefile and your changed file [16:50] <thomasvs> or wait [16:50] <thomasvs> no [16:50] <thomasvs> do as I said :) [16:50] <thomasvs> cp Makefile /tmp [16:50] <thomasvs> cp myfile.xml /tmp [16:50] <thomasvs> rm -rf * [16:50] <thomasvs> cd .. [16:50] <thomasvs> cvs update -dP [16:50] <thomasvs> mv /tmp/Makefile pwg [16:50] <thomasvs> mv /tmp/myfile.xml pwg [16:50] <thomasvs> cd pwg [16:51] <thomasvs> *hope everything is ok* [16:52] <BBB> seems to build fine [16:52] <thomasvs> strange :) [16:53] trow ([email protected]) joined #gstreamer. [16:55] <thomasvs> BBB: ok, so docs ok now for you ? [16:55] <thomasvs> hi jon [16:56] Action: thomasvs is going to be very happy when he can put the website down again [16:57] <BBB> thomasvs: seems so, yes [16:58] <BBB> why is the first page of a section always stored on the same page as the chapter content itself, but the second section not? [16:59] mathrick ([email protected]) joined #gstreamer. [16:59] <thomasvs> I have wondered myself too [17:00] <mathrick> lo [17:01] trow ([email protected]) left irc: "Leaving" [17:02] trow ([email protected]) joined #gstreamer. [17:03] <thomasvs> mathrick: ready for some more website checking ? [17:04] <mathrick> thomasvs: k, gimme :) [17:07] <thomasvs> mathrick: same url [17:07] <thomasvs> mathrick: also, feel free to read the pwg, bbb is updating it as we go [17:07] <mathrick> thomasvs: k [17:08] yippi ([email protected]) joined #gstreamer. [17:10] Company ([email protected]) joined #gstreamer. [17:11] Action: Company implemented seeking querying in oggdemux [17:11] <Company> that was hard [17:13] KoRnouille ([email protected]) left irc: Read error: 104 (Connection reset by peer) [17:14] KoRnouille ([email protected]) joined #gstreamer. [17:18] dilinger ([email protected]) left irc: Read error: 113 (No route to host) [17:20] swentel ([email protected]) left irc: [17:30] <Company> you know what GStreamer needs that we should steal from gtk? [17:30] <Company> child properties [17:31] <Company> to allow elements to attach properties to pads [17:34] <thomasvs> anyone have philosophical/religious problems with the layout of [17:34] <thomasvs> http://freedesktop.org/~gstreamer/src/ [17:34] <thomasvs> ? [17:34] <teuf> my religion forbids directory names starting with gst, if I look at such a layout, I'll burn in hell forever [17:35] <thomasvs> teuf: first fix tagging everywhere before you look then :) [17:38] Action: Company renames gst-autoplug autoplug-gst [17:39] Action: Company then renames teufs account gsteuf [17:39] <thomasvs> heh :) [17:40] <dolphy> Company: any news about my time issue ? [17:40] Action: sxpert_work is away: goes home [17:41] <Company> dolphy: nope [17:43] <thomasvs> BBB: I want to rename gst-record to gst-recorder - opinions ? [17:45] <BBB> that'll break my CVS, won't it? [17:45] <thomasvs> BBB: well, if you commit first, no [17:45] <teuf> you should all be using arch ;) [17:46] <thomasvs> BBB: which is why I'm asking you first :) [17:46] <BBB> [rbultje@shrek gst-record]$ cvs -Q diff -u [17:46] <BBB> ? src/gst/unused-emulator.c [17:46] <BBB> ? src/gst/unused-emulator.h [17:46] <BBB> [rbultje@shrek gst-record]$ [17:46] <BBB> do your thing [17:46] <thomasvs> rock :) [17:48] <thomasvs> BBB: execute this command in your gst-record directory after you mv'd it to gst-recorder: [17:48] <thomasvs> find . -name Repository -exec perl -i -p -e 's@record@recorder@g' {} \; [17:48] <thomasvs> BBB: that'll fix the locations [17:54] <BBB> cvs update -dP seems to work fine [17:54] <BBB> does that mean it works? [17:55] <thomasvs> yep [17:55] sub_away ([email protected]) left irc: "Client exiting" [18:02] <dolphy> hmm <programlisting> tag is refusing my g_blah (&width) [18:02] <BBB> &width) [18:02] <BBB> escape :) [18:02] <dolphy> i think there's a trick [18:02] <thomasvs> dolphy: three chars you can't use as-is [18:02] <dolphy> better than that [18:02] <thomasvs> dolphy: < > & [18:02] <thomasvs> dolphy: change them to < > & [18:03] <BBB> long live XML [18:04] <dolphy> <programlisting><![CDATA[ [18:04] <dolphy> blah [18:04] <dolphy> ]]></programlisting> [18:04] <dolphy> works better [18:04] <dolphy> &blah even [18:06] <dolphy> BBB: that CDATA tag let you copy paste what you want [18:06] <dolphy> BBB: much better imho [18:06] <BBB> oh [18:06] <BBB> hm [18:06] <BBB> ok [18:06] <dolphy> no escaping needed [18:06] <BBB> will use that in future [18:06] <mathrick> dolphy: everything except "]]>" :) [18:07] <dolphy> mathrick: true :) [18:07] <dolphy> mathrick: if you have that in your code i m curious to read it :) [18:09] <dolphy> the doc building system works just fine now [18:09] <dolphy> thanks thomasvs :) [18:09] <BBB> it's perfect ): [18:09] <BBB> :) [18:10] <thomasvs> ok, silly question [18:10] <thomasvs> I want a short term describing "the apps we as a project maintain" [18:10] <thomasvs> ie, gst-recorder, gst-player, gst-editor [18:10] <thomasvs> what term ? [18:12] <BBB> internally developed applications [18:13] <mathrick> thomasvs: in-house developed apps? [18:14] <thomasvs> hm, both are as long as what I already have though :) [18:14] <thomasvs> "our apps" ? [18:14] <mathrick> thomasvs: or auxiliary apps maybe [18:14] <dolphy> homemade apps [18:14] <dolphy> homegrown apps [18:15] <dolphy> my precious [18:15] Action: mathrick votes for "my preciousss" [18:15] <thomasvs> they're not exactly carrots :) [18:16] <thomasvs> argh, naming sucks :) [18:16] <BBB> I like homemade applications [18:16] <mathrick> thomasvs: is 'auxiliary' no good? [18:16] <BBB> even though my preciousssssssssss sounds better [18:16] <dolphy> yeahh [18:17] <dolphy> our chest [18:17] <dolphy> Jewel room [18:17] Action: BBB goes home [18:17] <BBB> bye [18:17] <dolphy> cius [18:17] BBB ([email protected]) left irc: "Client exiting" [18:17] <thomasvs> dolphy: care to write the gst-player page ? :) [18:17] <mathrick> argh, BBB left :\ [18:17] <iain> oh fedora has fancy coloured animated cursors during the installation... [18:18] <iain> that shows hopw that my laptop might get a decent XServer :D [18:18] <mathrick> anyone knows if he revised 'II. Building a Filter' yet? [18:18] <thomasvs> mathrick: cvs.freedesktop.org will tell you [18:19] <mathrick> thomasvs: hmm, right, it has that fancy annotated view [18:20] Action: mathrick notices Helix article of one of our Linux sites [18:20] <mathrick> gotta write GStreamer one :) [18:20] <sxpert> here's a good one : http://home.elka.pw.edu.pl/~mstolars/ftp/mars/Mars.avi [18:22] apoc ([email protected]) joined #gstreamer. [18:26] markey ([email protected]) joined #gstreamer. [18:26] <dolphy> thomasvs: hmm [18:26] <apoc> yo [18:26] <mathrick> yo apoc [18:27] teuf ([email protected]) left irc: "Client exiting" [18:27] <thomasvs> dolphy: ? [18:32] <dolphy> thomasvs: about writing the gst-player page [18:34] jdahlin ([email protected]) left irc: "Leaving" [18:42] alley_cat ([email protected]) left irc: Read error: 60 (Operation timed out) [18:44] pb_ ([email protected]) joined #gstreamer. [18:45] ilinger ([email protected]) joined #gstreamer. [18:48] Nick change: ilinger -> dilinger [18:54] alley_cat ([email protected]) joined #gstreamer. [18:56] ChrisHJW ([email protected]) joined #gstreamer. [18:58] dilinger ([email protected]) left irc: Read error: 60 (Operation timed out) [19:03] dolphy ([email protected]) left irc: "Network down, IP Packets delivered via UPS" [19:07] Kurosu_ ([email protected]) joined #gstreamer. [19:11] kmaraas ([email protected]) joined #gstreamer. [19:26] Kurosu ([email protected]) left irc: Read error: 110 (Connection timed out) [19:27] <thomasvs> http://www.freedesktop.org/~gstreamer/modules/gst-player.html [19:27] Nick change: Kurosu_ -> Kurosu [19:27] <thomasvs> ah, hm, dolphy left [19:29] <taaz> the lists page should probably also point people to gmane [19:30] hadley ([email protected]) joined #gstreamer. [19:32] <thomasvs> taaz: yeah, I'm going to do that [19:32] <thomasvs> taaz: though gmane still doesn't have the full archive up yet [19:32] <thomasvs> because they have a very painful rehashing process that takes four hours of downtime :) [19:33] <taaz> i thought it did? should ping the admin guy about that [19:33] <thomasvs> I already did [19:33] <thomasvs> he explained the pain to me :) [19:33] <thomasvs> I gave htem our complete archives, and through nttp it's completely accessible [19:33] <taaz> heh ;) [19:33] <thomasvs> but not yet in the web interface [19:33] <thomasvs> but it's nice to read the first few mails in pan [19:33] <taaz> oh... i just use the nntp access [19:35] <iain> mmmm fedora hangs halfway though boot [19:35] hallibaby ([email protected]) joined #gstreamer. [19:36] <iain> oh, only if I have my usb card reader plugged in [19:43] harshy ([email protected]) left irc: Read error: 110 (Connection timed out) [19:43] Zeenix ([email protected]) joined #gstreamer. [19:44] <Zeenix> hi [19:50] iain ([email protected]) left irc: "Turning mains off" [19:55] iain ([email protected]) joined #gstreamer. [20:05] <thomasvs> woah [20:05] Action: thomasvs just received a 203 page .doc file from someone with, get this, how to write gstreamer plugins [20:05] <thomasvs> I must be dreaming [20:05] <thomasvs> I really don't know what to do with this [20:05] Action: thomasvs is going to read it at home [20:17] thomasvs ([email protected]) left irc: Read error: 60 (Operation timed out) [20:24] <ds-work> the buffer pool stuff needs to be rewritten [20:25] somex1 ([email protected]) joined #gstreamer. [20:26] hallibaby ([email protected]) left irc: "Bye bye" [20:31] hadley ([email protected]) left irc: "Leaving" [20:36] KA ([email protected]) joined #gstreamer. [20:36] <KA> hi [20:36] <mathrick> lo [20:37] kmaraas ([email protected]) left irc: "Leaving" [20:41] Shoragan ([email protected]) left irc: Read error: 113 (No route to host) [20:41] shanatk ([email protected]) joined #gstreamer. [20:51] dolphy ([email protected]) joined #gstreamer. [20:52] <dolphy> re [20:54] <taaz> boo [20:55] <taaz> so, why does resize cause segfaults? ;) [20:57] <dolphy> caps renego [20:57] <dolphy> it's segfaulting all the time for a single resize ? [20:58] <ds-work> it works here [20:58] <taaz> no, just wild resizing intentionally trying to make it segfault ;) [20:58] <dolphy> yeah works here too [20:58] <dolphy> ds-work: but if you stress the resizing of the window [20:58] <dolphy> ds-work: then sometimes it crashes [20:58] <taaz> videotestsrc also does some diagnal bar thing [20:58] <ds-work> taaz: most elements don't get odd-sized video correct [20:59] <ds-work> especially ffcolorspace [20:59] <ds-work> videotestsrc and x[v]imagesink get it correct [21:01] Action: dolphy installs synergy [21:01] mxpxpod ([email protected]) joined #gstreamer. [21:01] <Company> we need ranges with stepping ;) [21:01] <ds-work> no, we need to fix broken elements [21:01] <taaz> is it easy to make cacasink handle resize? i haven't looked at code ;) [21:02] <Company> well, ranges with stepping would be the easiest fix :p [21:02] <taaz> fwiw, sdlvideosink doesn't work over remote X due to shm issues [21:02] <Company> isn't that an SDL bug? [21:03] <taaz> most likely [21:08] somex1 ([email protected]) left #gstreamer. [21:08] dilinger ([email protected]) joined #gstreamer. [21:11] shanatk ([email protected]) left #gstreamer. [21:15] mxpxpod ([email protected]) left irc: "Guns kill people as much as spoons made Rosie O'Donnell fat.." [21:18] thomasvs ([email protected]) joined #gstreamer. [21:21] dolphy_ ([email protected]) joined #gstreamer. [21:21] dolphy_ ([email protected]) left irc: Client Quit [21:31] Shoragan ([email protected]) joined #gstreamer. [21:40] robUx4 ([email protected]) joined #gstreamer. [21:40] <robUx4> lo [21:43] <mathrick> yo robUx4 [21:46] Action: dolphy looks at seeking [21:49] Action: thomasvs wonders what intel is doing with gstreamer [21:50] Shoragan ([email protected]) left irc: "Leaving" [21:50] <mathrick> thomasvs: huh? [21:51] <thomasvs> mathrick: I just received a 207 page word document with a gstreamer + plugin manual [21:52] <iain> thomasvs: ask sri? [21:52] <thomasvs> does sri work at intel ? [21:53] <ds-work> word? [21:53] <thomasvs> yeah, word [21:53] <iain> thomasvs: yeah, I think so [21:53] <ds-work> send it back and ask for XML :) [21:53] <thomasvs> sadly it seems he copied huge chunks [21:53] <iain> thomasvs: someone does I think [21:53] <thomasvs> out of the api docs and so on [21:53] <thomasvs> and it almost looks like someone copied huge chunks of html out of the api docs too [21:53] <thomasvs> strange [21:54] <thomasvs> what a world :) [21:54] <thomasvs> ds-work: ok so for the error stuff - you want uppercase macro right ? [21:54] <mathrick> thomasvs: and what does it have to do with intel? [21:54] <ds-work> thomasvs: it doesn't matter [21:54] <thomasvs> ds-work: anything special we need to make sure gets changed before release ? [21:54] <ds-work> thomasvs: we'll figure it out in 0.9 [21:55] <thomasvs> mathrick: it was someone from intel [21:55] <dolphy> hmm avidemux is not handling seeking correctly [21:55] <dolphy> grmbl [21:55] <thomasvs> dolphy: seeking ? gstreamer does seeking ??? [21:55] <mathrick> thomasvs: how do you know that? was it signed as "Anonymous Coward from Intel" ? ;) [21:56] <taaz> thomasvs: i've got a small patch and related question for the pkgconfig stuff [21:56] <thomasvs> taaz: shoot [21:56] <dolphy> thomasvs: well not exactly :) [21:56] <thomasvs> mathrick: no, his email address was from intel, and the document title wasn't changed from the template and also contained some intel info [21:56] <thomasvs> dolphy: http://www.freedesktop.org/~gstreamer/modules/gst-player.html [21:56] <dolphy> thomasvs: it is time persistent :) [21:56] <thomasvs> dolphy: feel free to change that page [21:56] <ds-work> thomasvs: I'd like to apply #130439 [21:56] <dolphy> thomasvs: means you can resume from where you were playing the previous media [21:56] <dolphy> thomasvs: and that's automagic [21:57] <robUx4> I have a question about GLib [21:57] <thomasvs> dolphy: nice feature no ? i'm sure we're the first to have that [21:57] BBB ([email protected]) joined #gstreamer. [21:57] <mathrick> y0 BBB [21:57] <Company> ds-work: i thought you had already [21:57] <robUx4> is it meant to be a static link or a dynamic link ? [21:57] <robUx4> lo BBB :) [21:57] <dolphy> thomasvs: yup totally l33t feature [21:57] Action: BBB kicks dolphy [21:57] <taaz> thomasvs: i'm still confused by the gstreamer-gconf*pc* files in both gst-libs/gst/gconf/ and pkgconfig/ [21:57] <BBB> dolphy: do you have a webcam? [21:57] <ds-work> BBB: the buffer pool stuff in the pwg is old [21:57] <dolphy> BBB: yeah but it's broken :) [21:57] <ds-work> BBB: btw, nice work [21:57] <dolphy> BBB: well the driver:) [21:57] Action: iain es how Marlin can work one day and stop working after an update of GStreamer [21:58] <thomasvs> ds-work: you're asking me ? you're the maintainer :) [21:58] <BBB> dolphy: try resizing a window [21:58] <thomasvs> ds-work: anyway, what's the catch of the patch ? [21:58] <taaz> thomasvs: all the other .pc are only in pkgconfig [21:58] <BBB> ds-work: I know, I didn't upodate that yet [21:58] Action: ds-work kicks iain for gratuitous use of UTF-8 [21:58] <thomasvs> taaz: check closely [21:58] <dolphy> BBB: with v4lsrc ! ximagesink you mean ? [21:58] <BBB> ds-work: please put it all together in a bugzilla entry or so [21:58] <mathrick> robUx4: dynamic [21:58] <BBB> so I don't forget [21:58] <BBB> dolphy: yes [21:59] <ds-work> BBB: actually, I'll just rewrite it [21:59] <dolphy> BBB: well my webcam is not working [21:59] Action: iain ds ds-work [21:59] <ds-work> if you want [21:59] <dolphy> BBB: can't try [21:59] <thomasvs> taaz: gst-libs/gst/gconf contain the non-versioned .pc files [21:59] <BBB> dolphy: it takes *TEN* seconds for image to come back [21:59] <BBB> and that's ximagesink's fault [21:59] <BBB> it renegotiates for each x event [21:59] <mathrick> robUx4: at least in normal usage, I don't know if it's possible to do static [21:59] <thomasvs> taaz: ie, the ones that should never ever be used [21:59] <iain> oh wait, thats a spade [21:59] <BBB> it should flush all resize events except that last if there's more than once queued [21:59] <dolphy> BBB: true :) [21:59] <thomasvs> taaz: they get copied to the locatoin that *is* important [21:59] <thaytan> morning all [21:59] <thomasvs> namely, pkgconfig/ [21:59] <thomasvs> with proper version [21:59] <robUx4> mmm, ok [21:59] <dolphy> BBB: i was planning to do that [21:59] <thaytan> gotta go to work, see you from there [21:59] thaytan ([email protected]) left irc: "See y'all" [21:59] <robUx4> I'm facing a problem with DLLs [21:59] <BBB> ds-work: even better! :) [21:59] <BBB> ds-work: yes please! [21:59] <thomasvs> BBB: no, it's your fault [21:59] <taaz> thomasvs: that's confusing. people, like me, just see the ones in pkgconfig/ and would edit those [21:59] <dolphy> BBB: could you loook at avidemux seeking for me please ? [21:59] <BBB> I love making others do work for me ;) [21:59] <BBB> dolphy: so, fix it :) [21:59] <robUx4> it seems that all functions has to be specified as import to be usable from the DLL [22:00] <robUx4> but I'm no DLL expert [22:00] <ds-work> thomasvs: I was just telling you so you know what I want to do before releasing [22:00] <thomasvs> ds-work: well, so what is the downside of the patch ? people are forced to have glibc ? [22:00] <robUx4> apart from that I have a native Win32 DLL with GLib built with MSVC 7.1 :) [22:00] <BBB> dolphy: I'll work on it, ok? really, asap [22:00] <BBB> before monday [22:00] <dolphy> BBB: i think i found an issue [22:00] <dolphy> BBB: and i would need your help [22:00] <thomasvs> robUx4: you really should discuss this with some glib people doing tests on win [22:01] <robUx4> k [22:01] <ds-work> thomasvs: non-glibc users will get 0x12345678 instead of a caps string [22:01] <robUx4> they have an IRC channel ? [22:01] <robUx4> I have a few patches here and there [22:01] <thomasvs> ds-work: only for debugging, or also for caps serialization as a string ? [22:01] <robUx4> idem for GStreamer (that just needs to be linked to GLib and I'm done) [22:01] <ds-work> thomasvs: only debugging [22:02] <thomasvs> ds-work: then that's fine. as long as I can still compile it with uclibc on my ia1 machines, i guess it's ok [22:02] <BBB> robUx4: put them in bugzilla [22:02] <BBB> robUx4: we handle that quickest [22:02] <thomasvs> is there anyone who knows how I can fix keymaps so that a key that gives the wrong symbol gives the right one instead ? [22:02] <thomasvs> where do I begin ? [22:02] <mathrick> robUx4: yep, DLLs need some export specs to be usable [22:03] <robUx4> well, I don't have patches in the UNIX sense [22:03] <robUx4> just some modified code [22:03] <mathrick> thomasvs: somewhere in /usr/{share,lib}/X11, but it's scary to look in there ;) [22:03] <robUx4> mathrick: that means I have to patch all the GLib and GStreamer sources [22:03] <robUx4> :/ [22:04] <mathrick> robUx4: don't they release MSVC compilable version? [22:04] <robUx4> of GLib ? [22:04] <mathrick> robUx4: i'd swear there's native win32 version [22:04] <robUx4> I haven't found them [22:04] <robUx4> and from the code I've seen no [22:04] <mathrick> robUx4: so, Thor prepares mingw32 only? [22:04] <robUx4> they might do it with MinGW but if they produce DLL they don't export anything anyway [22:04] <taaz> thomasvs: there was talk (and bug 126999) about renaming gstreamer-libs.pc. that's a build sys thing that should happen sooner rather than later. [22:05] <mathrick> robUx4: but I think it still needs export for DLLs [22:05] <robUx4> or maybe they have static libraries [22:05] <mathrick> robUx4: it may be, I remember hearing about GTK on Win32 being suboptimal in terms of shared libs [22:05] <ds-work> thomasvs: how do I check if a library function is usable in automake? [22:06] <robUx4> mmm, wait I found something [22:07] <thomasvs> ds-work: first AC_CHECK_LIB on the lib, then AC_CHECK_FUNC on the function you want [22:07] <Company> AC_CHECK_LIB (libc) :) [22:08] <Company> ds-work: do you think it'd be good if we only enable %P when we do --enable-gst_debug ? [22:08] <ds-work> Company: yes [22:08] <Company> ok [22:08] <Company> so you'll ifdef it :) [22:08] <thomasvs> taaz: not really sure I agree - maybe plugin-libs, but ... [22:08] <thomasvs> taaz: or separate ones per functionality [22:09] <taaz> thomasvs: i'm not really for or against it, just pointing out the bug report since i can see how it is a bit confusing the way we have things now [22:10] Action: Company votes plugin-libs (but he wrote that bug) [22:11] <Company> i think there's 1 person per week in #gstreamer and 2 per week in #rhythmbox asking why they don't have gstreamer-libs.pc [22:14] Action: robUx4 found the glib.def file... [22:15] Action: Company needs to get all the new names into his head [22:16] <taaz> thomasvs: why are the gconf files in pkgconfig/ in CVS if they are 'generated' via cp from the other dir? [22:19] <thomasvs> are they ? [22:20] <taaz> they are, and they are different i think [22:20] somex1 ([email protected]) joined #gstreamer. [22:21] <mathrick> taaz: gconf? or do you mean pkgconfig? [22:21] <taaz> thomasvs: want to see my uninstalled pc patch? it adjusts includedir and libdir to be useful vars. something i'm using for gst-python to find headers [22:22] <BBB> robUx4: how is the porting going? ;) [22:22] <taaz> mathrick: the pkgconfig/ dir copies the gconf files from the gconf src dir but they are also in cvs in pkgconfig/. i don't think they should be [22:23] <thomasvs> taaz: yeah I just noticed - the ones in pkgconfig/ are ones steveb put in and serve no purpose, removing them [22:23] <mathrick> taaz: gconf files are in pkgconfig/ ? strange, out of place i'd say [22:24] <Company> it copies the gstreamer-gconf pkgconfig file [22:24] <mathrick> Company: ah, l [22:24] <Company> so that all .pc files are in pkgconfig/ [22:24] <mathrick> s/l/k/ [22:24] <thomasvs> taaz: ok, they're gone [22:26] eazel7 ([email protected]) joined #gstreamer. [22:26] <eazel7> hi [22:26] <BBB> ds: will you write the PWG part on the element checklist? [22:26] <BBB> or ds-work [22:26] <BBB> hi eazel [22:27] <thomasvs> ar ? argentina ? [22:27] <ds-work> BBB: sure. remind me if I don't [22:27] <eazel7> yes [22:27] <BBB> http://kdedevelopers.org/node/view/325 ? [22:27] <BBB> ds-work: will do :) [22:27] <BBB> ty [22:27] <eazel7> I'm from Argentina [22:27] <eazel7> where are you from? [22:28] <BBB> the country from the prince that maxima zorreguieta married [22:28] <eazel7> yes [22:28] <BBB> guess ;) [22:28] <dolphy> grmbl [22:28] Action: dolphy finds a some more bugs.. [22:28] <dolphy> s/a// [22:28] <mathrick> BBB: btw, make sure to s/GNOME/GObject/ in "II. Building a Filter" intro :) [22:29] <eazel7> m/19/ [22:29] <eazel7> I need to ask [22:29] <eazel7> when I play videos, the screen is blue [22:29] <eazel7> why? [22:29] <BBB> mathrick: there's a PWG bug in bugzilla, plese add it there [22:29] <BBB> mathrick: I will forget, I assure you [22:30] <mathrick> BBB: k [22:30] <mathrick> BBB: have its number handy? [22:30] <BBB> nope :p [22:31] <BBB> #125890 [22:31] <BBB> http://bugzilla.gnome.org/show_bug.cgi?id=125890 [22:32] <BBB> eazel7: xv output? what version of gst? [22:32] <thomasvs> eazel7: do you have xinerama ? [22:32] <BBB> eazel7: could it be that there's another xv app (mplayer, xine) active at the same time? [22:32] <eazel7> BBB, no idea [22:32] <eazel7> how do I know? [22:33] <thomasvs> eazel7: do you have more than one monitor [22:33] <thomasvs> ? [22:33] <BBB> version... pkg-config --modversion gstreamer [22:33] <BBB> or pkg-config --modversion gstreamer-0.7 [22:33] <BBB> or pkg-config --modversion gstreamer-0.6 [22:33] <BBB> one of those three [22:33] <eazel7> no [22:33] <eazel7> I have only one [22:33] <thomasvs> BBB: easy, isn't it ? :) [22:33] <eazel7> 0.6.4 [22:33] <BBB> there must be an easier way... [22:34] <eazel7> mm [22:34] <eazel7> gstreamer-properties [22:35] <BBB> what's the output set to? xvideosink? sdlvideosink? [22:35] <eazel7> there it says: Output: XWindows (X11/XShm/Xv) [22:35] <BBB> does it change if you set video output to "xvideosink disable-xv=true"? [22:35] <mathrick> BBB: should I add new bug as a child there? or just throw it to comments? [22:35] <BBB> mathrick: throw comments in there [22:35] <eazel7> yes [22:35] <eazel7> it works now [22:35] <eazel7> =) [22:36] <BBB> :) [22:36] <BBB> your xv setup might be slightly troublesome... if you see it in version 0.7.x too, please come back so we can have a closer look, ok? [22:37] <Company> ds-work: is a return value of gst_caps_new_empty ok for a getcaps function? [22:37] <eazel7> mmm [22:37] <eazel7> but now the video does not resize [22:37] <eazel7> =( [22:37] <BBB> true [22:37] <BBB> that's one of the missing features in 0.6.x... [22:37] <BBB> it's added in 0.7.x [22:38] <BBB> but 0.7.x is still early beta... :/ [22:38] <ds-work> Company: yes, although it doesn't play well with linking [22:38] <ds-work> Company: that was the original plan for explicit_caps() [22:38] <Company> ds-work: that pad is not supposed to link as it's an unidentified stream [22:38] <mathrick> eazel7: what does xvinfo return? [22:39] <dolphy> what is the default event handler of a pad ? [22:39] <eazel7> it's too large [22:39] <eazel7> do I paste it in #flood? [22:39] <Company> ds-work: though I'd like to have a way to make it connect to ANY caps [22:39] <dolphy> and where is it set ? [22:39] <mathrick> eazel7: k [22:39] <BBB> dolphy: gst_pad_default_event_handler? [22:39] <BBB> dolphy: in gstpad.c [22:39] <ds-work> Company: for now, make the link function return DELAYED [22:39] <ds-work> Company: and the getcaps function return the pad template caps [22:39] <ds-work> Company: because that's how explicit_caps() works [22:40] <ds-work> Company: unless it is specifically different than explicit caps [22:40] <Company> ds-work: i'll just say use explicit caps and not set any [22:40] <robUx4> <BBB> robUx4: how is the porting going? ;) <- getting closer everyday [22:40] <ds-work> Company: what kind of pad is this? [22:40] <robUx4> I now have GLib.dll done with the MS compiler [22:40] <robUx4> now I just figure out I need a gobject.dll [22:40] <Company> ds-work: unidentified ogg stream [22:40] <robUx4> and probably a gmodule.dll and gthread.dll too... [22:41] <BBB> yes [22:41] <dolphy> BBB: dolphy@dell:~/workdir/gstreamer-0.7/gstreamer/gst$ grep gst_pad_default_event_handler * [22:41] <dolphy> dolphy@dell:~/workdir/gstreamer-0.7/gstreamer/gst$ [22:41] <robUx4> but in the end it should work [22:41] <eazel7> mathrick, done [22:41] Nick change: KA -> KA_afk [22:41] <BBB> dolphy: or something like that...? [22:41] <ds-work> Company: you could make it application/octet-stream [22:41] <Company> ds-work: you mean any? [22:41] <Company> ds-work: any is fixed? [22:42] <mathrick> eazel7: your xv looks ok, so probably something else is blocking xv port. You sure no mplayer or xine fired up? [22:42] <ds-work> no, application/octet-stream is the fixed version of ANY, sort of [22:42] <eazel7> I have no mplayer, neither xine [22:42] <dolphy> gst_pad_event_default [22:42] <Company> ok, i'll try that - i hope that doesn'tr break autoplugging in any way now or in the future [22:43] <BBB> dolphy: yes, that [22:43] <mathrick> eazel7: hmm, then it's really hard to say, i don't know if there's any way to check if something uses Xv [22:43] ChrisHJW ([email protected]) left irc: Read error: 60 (Operation timed out) [22:43] <mathrick> dolphy: do you know of any? [22:43] <Company> mathrick: running an app that's known to work and uses xv :) [22:43] <taaz> thomasvs: are you the as-*.m4 maintainer? [22:44] <mathrick> Company: yep, but it doesn't print PID or anything, does it? ;) [22:44] ChrisHJW ([email protected]) joined #gstreamer. [22:44] <mathrick> Company: PID of the supposed prior user, that is [22:44] <taaz> ChrisHJW: ping? ;) [22:44] <eazel7> mathrick, ok [22:45] <eazel7> mathrick, thanks anywa [22:45] <dolphy> mathrick: any what [22:45] <dolphy> ? [22:45] <mathrick> dolphy: way of determining if and what uses Xv already [22:45] thaytoo ([email protected]) joined #gstreamer. [22:46] <BBB> thaytoo: YES! I have a Q for you [22:46] <mathrick> eazel7: come back with 0.7 :), and you can try running xine/mplayer indeed, if they don't work, then it's your Xv b0rked probably [22:46] <mathrick> yp thaytoo [22:46] <mathrick> s/yp/yo/ :) [22:46] <thaytoo> hiyas [22:46] <thaytoo> BBB: wassup? [22:46] <thomasvs> taaz: yeah [22:46] <eazel7> mathrick, thanks [22:47] <BBB> thaytoo: could you please write me some PWG documentation in gstreamer/docs/pwg/advanced_interfaces.xml (placeholder is already there, but I need content) about the navigation interface? [22:47] <mathrick> eazel7: np [22:48] <Company> YES [22:48] Action: Company listens to the audio stream of a theora file [22:49] <thaytoo> Company: excellent :) [22:49] <taaz> thomasvs: want to fix a bug? in as-docbook.m4 it defines VERSION=4.1.2 which when stuffed into configure redefines VERSION like that and is then put into some of the pkgconfig/ files during AC_OUTPUT. could either use another var name or save/restore to fix. [22:49] <mathrick> Company: theora is container or codec? [22:49] <thaytoo> theora is the video codec [22:49] <Misirlou> theora's a codec [22:49] <Misirlou> Ogg's the container [22:49] <Company> mathrick: theora is the video codec of xiph [22:50] <Misirlou> join #theora! join #theora! omg omgomg [22:50] <Company> it just means thatr oggdemux demuxes correctly :) [22:50] <mathrick> Company: * Company listens to the audio stream of a theora file <-- audio? [22:50] <Company> #theora is boring :p [22:50] <mathrick> Company: ah, ic :) [22:50] <thaytoo> Company: it also means you somewhere have a vorbisdec? [22:50] <eazel7> ok [22:51] <eazel7> well, I'll try other things bbl [22:51] <Company> thaytoo: yes, it's not committed because the raw float mimetype is pending [22:51] <eazel7> thanks [22:51] eazel7 ([email protected]) left irc: "Leaving" [22:51] <BBB> Company: I think it's good to make it output float caps [22:51] <BBB> it'll make gst-player aware ;) [22:52] <Company> BBB: that's not the problem [22:52] <Company> BBB: the problem is that I have a float** [22:52] <thomasvs> taaz: oops, that's braindead of me [22:52] <thomasvs> cvs -z3 -d:pserver:[email protected]:/cvs/gstreamer co gst-editor [22:52] <thomasvs> can someone run this and tell me if it works ? [22:52] <thomasvs> it's locked here [22:52] <BBB> Company: oh, wait, is that the ">1 chan/pad" issue? [22:53] <BBB> thomasvs: /home/cvs/gstreamer [22:53] <thomasvs> BBB: no, that's wrong [22:53] <Company> thomasvs: [22:53] <Company> cvs checkout: warning: cannot write to history file /cvs/gstreamer/CVSROOT/history: Permission denied [22:53] <Company> cvs checkout: Updating gst-editor [22:53] <Company> cvs checkout: failed to create lock directory for `/cvs/gstreamer/gst-editor' (/cvs/gstreamer/gst-editor/#cvs.lock): Permission denied [22:53] <Company> cvs checkout: failed to obtain dir lock in repository `/cvs/gstreamer/gst-editor' [22:53] <Company> cvs [checkout aborted]: read lock failed - giving up [22:53] <thomasvs> ok, so I need to check with some fdo admin then [22:53] <thomasvs> strange [22:53] <Company> BBB: yeah, the >1 channel problem [22:54] <Company> BBB: and the "float2int sucks" problem [22:54] <BBB> *grin* [22:54] <BBB> so fix it [22:54] <Company> but i'll just fix this in a totally compatible way :) [22:54] <BBB> or, better yet, integrate it in audioconvert [22:54] <Company> nah [22:54] <Company> float is buffer-frames aware, audioconvert is not :) [22:55] <BBB> oh [22:55] <BBB> right [22:55] <BBB> ohwell [22:55] <BBB> anyway [22:55] <BBB> (buffer-frames is evil [/rant]) [22:56] Action: BBB will go offline and finish docs on tagging [22:56] <BBB> Company: please write clocking docs some day ;) [22:56] <Company> ah, reminds me [22:56] <dolphy> so many things broken [22:56] <dolphy> videoscale src event handler is broken [22:56] <Company> should i write ChangeLog entries for docs/random btw? [22:56] somex1 ([email protected]) left #gstreamer. [22:57] <BBB> Company: yes please [22:57] <mathrick> Company: what is buffer-frames? [22:57] <Company> mathrick: number of frames in a buffer [22:57] <Company> mathrick: float audio specific caps property [22:57] <ChrisHJW> BBB : once robUx4 has gstreamer ported, how can i start testing it ? any recommendations ? [22:57] <BBB> ChrisHJW: uhm... gst-launch videotestsrc ! ximagesink [22:58] <BBB> ChrisHJW: if that displays a test image (like on TV), you're the man! [22:58] yippi ([email protected]) left irc: "Client exiting" [22:58] <BBB> but gst-launch fakesrc ! fakesink is also a good testcase [22:58] <ChrisHJW> is this a CLI command ? :) [22:58] <BBB> yes [22:58] <mathrick> Company: ah, so audioconvert would lose that property? [22:58] Action: ChrisHJW hates CLI [22:58] <BBB> I'm affraid the gnome utilities won't work under win32 [22:59] <ChrisHJW> no wxwindows ? [22:59] <BBB> no [22:59] <ChrisHJW> :sigh: [22:59] <mathrick> ChrisHJW: CLI under win32 is painful indeed, but in general, CLI rocks ;P [22:59] <Company> mathrick: well, it could just not implement it [22:59] <ChrisHJW> so, i better wait until some guys have made some nice player with a GUI for it :D .... [23:00] <Company> mathrick: but that's a bad idea for a float converter element ;) [23:00] <mathrick> although, on win2k there is *AUTOCOMPLETION* in CLI! Now that's w00t ;) [23:00] <mathrick> pretty lame one, but it is ;) [23:01] <Company> BBB: check commit list :) [23:01] Action: BBB goes offline [23:01] <BBB> seeya tomorrow [23:01] <BBB> oh [23:01] <BBB> crap [23:01] Action: BBB checks mail [23:01] <mathrick> cya BBB [23:01] <Company> or cvs up docs/random :) [23:02] Action: BBB waits for email one more minute [23:03] <robUx4> where is 'gconstpointer' defined ? [23:04] <robUx4> I can't find it anywhere [23:04] <robUx4> and so gstindex.c doesn't compile [23:04] <alley_cat> gtypes.h in glib [23:05] <robUx4> oh [23:07] <BBB> Company: not appearing [23:08] Action: BBB checks CVS [23:08] <BBB> what file? [23:09] <BBB> Company: time? [23:09] <BBB> well, looks fine, but I need it in the xml [23:09] <BBB> advanced_clock.xml [23:09] <BBB> ;) [23:10] <BBB> ok, so SF mail is dead [23:10] <BBB> ohwell [23:10] <BBB> bedtime [23:10] <BBB> bye! [23:11] BBB ([email protected]) left irc: "Client exiting" [23:12] <dolphy> bedtime too [23:12] <dolphy> i ll continue fixing seeking tomorrow [23:12] <dolphy> ds-work: i just committed a fixed videoscale handle src method [23:13] <taaz> ChrisHJW: ping ping? did anyone reply to your win32 port message that you said to send to matroska list too and didn't give an address for that list? [23:13] <dolphy> ds-work: the actual one was leaking an event and not forwarding events != NAVIGATION [23:13] <dolphy> :-( [23:13] <ds-work> dolphy: ah, cool. is that the seeking problem? [23:14] <KyleB> Company, remember talking to me about "gstreamer's worst bug" the other day? and you said it would go away? [23:14] <dolphy> ds-work: that's a part of it [23:14] <dolphy> ds-work: i will fix the whole thing tomorrow [23:14] apoc ([email protected]) left irc: "Leaving" [23:14] <ChrisHJW> taaz : ? [23:14] <dolphy> ds-work: i m discovering tons of broken stuff tracing events road [23:15] <ChrisHJW> i am almost asleep, so that question was too difficult for me :) [23:15] <dolphy> ds-work: asfdemux is broken too [23:15] <dolphy> ds-work: i ll fix them [23:15] <dolphy> night [23:15] <taaz> ChrisHJW: you send message about win32 port of gst. said to reply to matroska list too. i have no idea what that address is. gmane might have stripped it from the headers. [23:16] <dolphy> ds-work: i will also add something in the PWG doc talking about how src events should be handled [23:16] <taaz> ChrisHJW: i was wondering if anyone bothered to reply. i didn't yet because i didn't know that other mailing address. ;) [23:16] <dolphy> ds-work: so that people stop breaking the whole event pipeline [23:17] <KyleB> is it kosher to post a gstreamer error message in here? [23:17] <ds-work> dolphy: write some notes at the end of docs/random/ds/element-checklist [23:17] <dolphy> i will [23:17] dolphy ([email protected]) left irc: "Network down, IP Packets delivered via UPS" [23:17] <ChrisHJW> taaz : [email protected] [23:18] <ChrisHJW> there was one reply to me, asking me why we wouldnt use gstreamer CVS for that [23:18] Action: taaz can't send/receive mail now anyway... windows virus took out entire vt.edu mail system since monday :( [23:18] <ChrisHJW> which of course is the much better solution [23:18] <taaz> ChrisHJW: yeah, that's about what i was going to say too ;) [23:19] <ChrisHJW> :) ... lol .... we didnt know if you guys wnated to have windows specific code on your freebsd server ;) [23:20] <mathrick> KyleB: if they're short, then go ahead [23:20] <ds-work> isn't windows half BSD anyway? [23:20] <KyleB> ERROR scheduler(30310) gstoptimalscheduler.c(2074):gst_opt_scheduler_iterate: [GstOptScheduler@0x8191630] in error state [23:20] <KyleB> Execution ended after 1 iterations (sum 337016000 ns, average 337016000 ns, min 337016000 ns, max 337016000 ns). [23:20] Action: ds-work trolls and wanders off to lunch [23:21] <ChrisHJW> robUx4 : stupid question of the day .... if you were really succesfully in porting gstreamer, does that mean the decision to build TCME on it is done with that ? [23:21] <ChrisHJW> lol@ ds-work [23:21] <KyleB> anyone seen that error before? [23:21] <mathrick> ChrisHJW: will you build TCME on gst? rock :) [23:21] <ChrisHJW> mathrick : i have nothing to say here :) [23:22] <robUx4> ChrisHJW: no, no decision will be done [23:22] <mathrick> KyleB: yep, it's standard "something is nasty broken" type of message [23:22] <ChrisHJW> this decision will be done by the devs [23:22] <robUx4> but that will be a major point for GST [23:22] <mathrick> ChrisHJW: i mean, did you consider it? [23:22] <taaz> ChrisHJW: it would be nice to have all the code in one place. otherwise you have syncing issues and so on. [23:22] <KyleB> mathrick: oh. I can play oggs but not mp3s. Also, I did a cvs update, rebuilt and still the same [23:22] <ChrisHJW> taaz : no question, its the very best option [23:23] <KyleB> Company was saying it was a thread racing issue and it would go away on its own [23:23] <mathrick> KyleB: lemme grep for recents bugs with that symptoms [23:23] <thomasvs> ChrisHJW, robUx4: we don't mind at all having it integrated with our cvs if it is possible [23:23] <mathrick> KyleB: ahh, if he said so, then it's prolly true [23:23] <KyleB> but it hasn't gone away :( [23:23] <KyleB> im just trying to do rhythmbox development damnit! ;-) [23:23] Action: ChrisHJW shivers by the idea to have a nice, feature packed player based on gstreamer, for both windows and Linux [23:24] <robUx4> I'm off to bed now [23:24] <robUx4> I'll need GThread.dll tomorrow [23:24] <robUx4> and maybe GModule.dll too... [23:24] <robUx4> then all the remaining pbs will be in GStreamer [23:24] <mathrick> cya robUx4 [23:24] <robUx4> especially the macro using the ellipsis arg [23:25] <robUx4> #define MACRO(x,y,z,...) [23:25] <robUx4> which is only supported by C99 compilers [23:25] <thomasvs> robUx4: we should have defines for those anyway to make them different [23:25] <robUx4> which is very few [23:25] <thomasvs> KyleB: what exactly is your bug ? [23:25] <robUx4> so if you intend to have GST work on hardware devices, you'd better be less restrictive with compilers [23:26] <robUx4> unless you only allow GCC 3 [23:26] <KyleB> thomasvs, what's the best way to describe it.. I can't play from rb, so I tried using the same command line to gst-launch that rb uses. the lines I pasted are from the error it spit out [23:26] <thomasvs> robUx4: well, we need people on our team who actually DO that. someone, like, say, you [23:26] Uraeus ([email protected]) joined #gstreamer. [23:26] <robUx4> mmm, heh [23:26] <thomasvs> KyleB: can you paste the complete output to me privately ? [23:26] <mathrick> y0 Uraeus [23:26] <KyleB> sure [23:26] <thomasvs> Uraeus: so, are you going to fix the tasks now ? [23:26] <Uraeus> good evening :) [23:27] <robUx4> well for the moment I'll just patch the code to do nothing [23:27] <ChrisHJW> thomasvs : dont grab our main developer :P [23:27] <robUx4> as it's only logs [23:27] <Uraeus> thomasvs: I will :) I did update them not long ago actually [23:27] <robUx4> but later something cleaner will need to be done [23:27] <Uraeus> thomasvs: but with wingo's mail I think it can at least make that one better [23:28] <robUx4> bye [23:28] robUx4 ([email protected]) left irc: "Baby wants to bleep" [23:28] <Uraeus> problems with our Matroska plugin? [23:29] <ChrisHJW> Uraeus : nope [23:29] <ChrisHJW> but your software refuses to run on windows :P [23:29] <ChrisHJW> and robux4 is trying to convince it to do exactly that [23:29] <ChrisHJW> :) [23:31] <Uraeus> ah :) [23:31] <thomasvs> KyleB: can you try without typefind in that pipeline ? [23:31] <KyleB> sure [23:31] <thomasvs> Uraeus: no, I mean, make sure that the code locations you refer to are actually defined [23:31] <thomasvs> Uraeus: also, provide code locations for the ones you didn't give [23:31] <Uraeus> thomasvs: ah ok, yeah I look into that tommorow after work [23:31] <KyleB> thomasvs, same [23:32] <thomasvs> KyleB: what version of gst ? [23:32] <KyleB> cvs [23:32] <KyleB> as of yesterday afternoon sometime [23:33] <thomasvs> KyleB: something got fixed re: mp3's, can you update and try again [23:33] <thomasvs> KyleB: is the file available somewhere so I can test ? [23:33] <KyleB> thomasvs, I can make it available, but it happened for all my mp3s (that I tested anyway) [23:34] <Uraeus> aaggg, I forgot to list Jan in my list of status reporters [23:35] markey ([email protected]) left irc: "leaving" [23:35] <thaytoo> Uraeus: it's alright [23:35] <thaytoo> I had a cry, but I'm over it ;) [23:37] <KyleB> thomasvs, thanks for looking at it, I have to run for a bit, but I will update and then build when I get back [23:37] <Uraeus> thaytoo: it was not meant as a slight, I think your work on gstreamer over the last months has been essential and fantastic [23:41] <thaytoo> Uraeus: ta :) [23:44] wheels ([email protected]) joined #gstreamer. [23:46] <mathrick> lo wheels [23:46] <wheels> hiya [23:50] <Zeenix> hi Uraeus [23:51] walters ([email protected]) left irc: "out" [23:51] ahern ([email protected]) joined #gstreamer. [23:52] <ahern> Howdy. Do you guys know of any problems with anonymous CVS currently? I can't seem to check out the gstreamer sources. [23:52] <Uraeus> ahern: from freedesktop CVS? [23:52] <Uraeus> hi Zeenix [23:53] <thaytoo> ahern: lock problems? [23:54] <thaytoo> thomasvs was contacting the admins about that [23:59] Action: sxpert is away: bed time [00:00] --- Thu Jan 29 2004 [00:04] ChrisHJW ([email protected]) left irc: Connection timed out [00:07] <Uraeus> night guys [00:07] Uraeus ([email protected]) left irc: "Client exiting" [00:12] <thomasvs> ahern: should be fixed now [00:14] bilboe1 ([email protected]) joined #gstreamer. [00:14] bilboe1 ([email protected]) left #gstreamer. [00:18] smoke_ ([email protected]) left irc: Remote closed the connection [00:28] <ahern> sorry - afk. Roof inspector. [00:29] <Company> ds-work: srcpads of filters should return DELAYED until the sinkpads have a format? [00:29] <ahern> Yeah, it was lock issues. Lemme go check. [00:29] <ds-work> Company: no [00:29] <ahern> Works perfectly now. Thanks much for your help. [00:29] <ds-work> Company: they call try_set_caps() and return that value [00:29] <ahern> thomasvs: Thanks for the help. [00:29] <Company> ds-work: right [00:32] <ds-work> GST_ERROR ("This should print caps: %" GST_PTR_FORMAT, caps); [00:32] <Company> there's no way to make that runtime defineble, right? [00:33] <ds-work> GST_PTR_FORMAT? [00:33] <ds-work> no [00:34] ahern ([email protected]) left irc: "Leaving" [00:38] <ds-work> nice. we have potential for recursion if code used by the printf extension calls printf with %P [00:39] <Company> that shouldn't happen [00:39] <ds-work> gst_caps_to_string(), gst_structure_to_string()? [00:39] <ds-work> I agree [00:39] <Company> oh, but it could if you have debugging stuff in there :) [00:39] <Company> no, it couldn't [00:39] <Company> that would have looped anyway [00:40] <ds-work> yes it would, since now gst_caps_to_string() is called from inside printf [00:40] <ds-work> or vnsprintf() or whatever is actually used [00:41] <Company> could you add a comment in caps_to_string saying exactly that? [00:41] <ds-work> yes [00:41] <ds-work> do you know of any threadsafe way of checking for recursion? [00:41] <ds-work> other than an atomic variable [00:42] <Company> nope [00:42] <ds-work> should I add a check? [00:42] <ds-work> seems a little bit heavy for something that isn't supposed to happen [00:43] Kurosu_ ([email protected]) joined #gstreamer. [00:43] <Company> that'll probably kill performace [00:43] <ds-work> an atomic variable shouldn't on normal architectures [00:43] <Company> and infinite loops are easy to debug [00:44] <ds-work> mmm... good point [00:44] <Company> crap [00:44] <Company> i have a problem [00:45] <Company> oggdemux parses the pipeline 2 times, once for stream detection, once for actual playback [00:45] <Company> i need to detect eos before i get an eos event [00:45] <Company> so i can seek back to the beginning without all the elements going to paused [00:46] <ds-work> qtdemux needed that, too [00:46] <ds-work> I forget what I did [00:46] <Company> i could query the length... [00:48] <ds-work> gst_bytestream_length(qtdemux->bs) [00:49] <ds-work> presumably that's what bytestream does [00:49] <Company> hmhm [00:49] <ds-work> qtdemux ends up at the end of the file at least once, but doesn't get an EOS [00:49] <ds-work> however, that could be the root of several qtdemux bugs [00:49] <Company> yeah, I'll implement something like that [00:50] <Company> in the future we want EOS not change the state - that must be done in 0.9 :) [00:50] <Company> so elements can still seek when they hit eos [00:51] <ds-work> gstreamer needs to split off its own glib [00:51] <ds-work> for structures and debug [00:51] <Company> and values [00:51] <ds-work> yeah [00:51] <Company> yeah [00:52] <Company> gstlib :) [00:52] <ds-work> perhaps buffers. I've wanted buffers for other projects [00:52] <ds-work> it could be the base of the elusive codec project [00:53] <Company> we could make the glib guys have a look, but i don't think they want that [00:53] <Company> serialization is another part that would belong in gstlib [00:53] <Company> i want to be able to register serialization functions for GTypes [00:54] <ds-work> gst_caps_from_string() now uses the GstValue deserialize functions [00:54] <Company> yeah [00:54] <Company> i need to port pipeline parsing to do this, too [00:55] <Company> the property=value code i mean [00:55] <ds-work> good idea [00:55] <Company> it should be dead easy to do right now [00:56] <Company> you'd only need a (de)serializer for flags [00:56] <Company> and enums [00:56] <ds-work> wouldn't be too hard to write [00:56] <ds-work> it would be special cased in gst_value_deserialize() [00:56] <Company> the code is already there, it just needs to be moved out of grammar.y [00:58] <ds-work> ERROR default(30618) printf_extension.c(34):: This should print an object: <GstIdentity:0x8237ee0> [00:58] <ds-work> sweet [00:59] <Company> ds-work: you should make gst_debug_log_default use this for the first property [00:59] <Company> s/property/argument/ [00:59] Kurosu ([email protected]) left irc: Read error: 110 (Connection timed out) [00:59] Nick change: Kurosu_ -> Kurosu [00:59] <ds-work> ok [01:00] <Company> btw: my goal is to abstract (de)serialization so much, that _to_string and _to_xml are possible [01:01] <Company> that way you just register (de)serializations for GST_TYPE_ELEMENT, GST_TYPE_ELEMENT_FACTORY, GST_TYPE_REGISTRY and you have a registry in plain text or xml - as you wish [01:01] <ds-work> yeah [01:01] <Company> i dunno how easy that would be though [01:03] Kurosu ([email protected]) left irc: [01:03] <ds-work> I wouldn't try to hard with the current architecture [01:03] <ds-work> it needs a few barriers to avoid recursion [01:04] <ds-work> Houston, we have a problem [01:04] <ds-work> %P doesn't work with bogus pointers [01:07] <ds-work> how do I test if memory is readable? [01:08] <Company> bogus pointers? [01:09] <ds-work> (void *)1 [01:09] <ds-work> segfaults [01:09] <Company> that's like using %s with that [01:09] <Company> that needs to be punished [01:09] <Company> %P only works with Objects and caps [01:10] <ds-work> I'd prefer to write a decent error message and then call abort() [01:11] <ds-work> why isn't there a function similar to mprotect() that allows you to _read_ the access permissions on a page? [01:13] <Company> what does glibc do when you do printf ("%s", (void*) 1); ? [01:13] <ds-work> segfault [01:15] <Company> hm, being better then glibc sounds hard [01:16] <ds-work> I could do it with madvise [01:17] <ds-work> that doesn't seem wise [01:17] Action: Company has no clue about such lowlevel stuff [01:17] <ds-work> it still touches things [01:17] <Company> you could also install a custom segfault handler :p [01:17] <ds-work> ah [01:17] <ds-work> mincore() [01:17] <ds-work> other times, I've used a segfault handler [01:18] <ds-work> but not in a library, of course [01:22] <thaytoo> can you test for ptr >= PAGE_SIZE && ptr < sbrk(0) ? [01:22] <thaytoo> are there allowed to be holes in the data segment? [01:22] <ds-work> it could be a mmaped pointer [01:23] <thaytoo> is %P some magic thing that prints any GST object? [01:24] <ds-work> yup [01:24] <Company> it's supposed to [01:25] <Company> ma, spider doesn't work [01:25] <ds-work> except it's ...%" GST_PTR_FORMAT "... [01:25] <Company> it plugs correctly but never plays [01:26] <ds-work> well, my madvise() trick work, but it's gross, so I'm removing it [01:26] <ds-work> works [01:26] <Company> ah, found it [01:32] <thaytoo> gst-launch filesrc ....mp3 ! mad ! esdsink isn't working for me now [01:32] <thaytoo> nor with audioscale in there [01:32] <ds-work> audioconvert [01:32] <Company> try audioconvert [01:32] <thaytoo> doh :) [01:33] <thaytoo> same [01:33] <thaytoo> mad gets REFUSED from try_set_caps [01:33] <ds-work> perhaps both [01:33] <Company> what caps is it trying to set? [01:33] <thaytoo> ds-work: yep [01:33] <ds-work> I changed esdsink to only rate=44100,width=16,channels=2 [01:33] <ds-work> I'm tired of esdsink bugs [01:35] <thaytoo> perhaps I should make the gstreamer-properties pipeline for esd be audioscale ! audioconvert ! esdsink then [01:35] <thaytoo> or, better, put it in rhythmbox [01:36] <Company> that reminds me that i wanted to "fix" sinesrc [01:36] <Company> so that it supports all audio formats, just like videotestsrc [01:36] <ds-work> maybe call it audiotestsrc then [01:36] <Company> good idea [01:37] <Company> is there another sound that's simple to create but not so unnerving? [01:37] <ds-work> noise [01:37] <ds-work> white noise [01:37] <thaytoo> white noise amplitude modulated by low-level sine [01:37] <thaytoo> beachsrc ;) [01:37] <ds-work> ideally, audiotestsrc should default to a 0dB 1000 Hz sine wave [01:38] <iain> I'm going to write clicksrc [01:38] <ds-work> since that's an actual standard test frequency [01:38] <Company> that's an evil beeping [01:38] <iain> for everyone who wants a metronome to add to their pipeline [01:39] <Company> somebody needs to write videotestsink btw [01:39] <ds-work> iain: replace gtick [01:39] <Misirlou> 0 dB? [01:39] <ds-work> Misirlou: line level [01:39] <iain> Misirlou: yeah, LOUD [01:39] <Misirlou> oh [01:39] <Misirlou> I thought 0 dB couldn't be heard. [01:40] <iain> no dB is negative scale [01:40] <ds-work> Misirlou: that's dBA [01:40] <Company> ds-work: something against me adding gst_structure_remove_fields ? [01:40] <Misirlou> ahhhh [01:41] <Misirlou> Thank you. [01:41] <ds-work> Company: remove a list of fields, or all fields? [01:41] <Company> ds-work: a list of [01:41] <ds-work> Company: I suppose not [01:42] harshy ([email protected]) joined #gstreamer. [01:42] <ds-work> harshy: what's the reasoning behind libburn being GPL instead of LGPL? [01:43] <harshy> well [01:43] <harshy> that would be a question for the libburn guys [01:43] <harshy> iirc they just did it under gpl and may go lgpl [01:43] <ds-work> harshy: I thought you were a libburn guy [01:43] <harshy> nah [01:43] <harshy> to dumb to be [01:44] <harshy> libburn tho is developed along side coaster [01:46] <Misirlou> Is it even possible to go GPL -> LGPL without the agreement of all the contributors? [01:46] <ds-work> no [01:47] <Misirlou> That's what I thought . . . [01:47] Rotty ([email protected]) left irc: Remote closed the connection [01:48] Action: mathrick goes to bed [01:48] <mathrick> nite guys [01:48] pb_ ([email protected]) left irc: "Client exiting" [01:48] Zeenix ([email protected]) left irc: Read error: 60 (Operation timed out) [01:49] Nick change: mathrick -> mathrick|sleep [01:52] <Company> the AUTHORS file needs updating [01:55] <ds-work> a bit [01:56] Action: ds-work wonders if himself and Company have done enough work to be considered AUTHORS [01:59] <Company> probably not [01:59] <Company> we need to go into MAINTAINERS [01:59] <Company> :p [02:00] AlexCONRAD ([email protected]) joined #gstreamer. [02:02] <ds-work> <pipeline0(GstPipeline)@0x823ce68> [02:02] <ds-work> is that too ugly for a named object? [02:02] <Company> dunno [02:02] <Company> it's quite long for debugging output [02:03] <Company> pipeline0 is pretty much ok [02:03] Action: ds-work doesn't like that element names are different than gtype names [02:05] <ds-work> printing the object name in gst_debug_log_default() has a slightly different goal that %P, imo [02:06] <Company> hm, yeah, probably [02:06] <ds-work> but I'm not going to worry about it [02:06] <ds-work> they're close enough [02:07] <ds-work> they use the same function now [02:07] <ds-work> gst_debug_print_object() [02:16] <Company> yep [02:16] <Company> spider is picky about autoplugging [02:16] KoRnouille ([email protected]) left irc: Read error: 110 (Connection timed out) [02:16] <ds-work> naw, really? [02:17] alley_cat ([email protected]) left irc: "May the Source be with you!" [02:17] <Company> s/autoplugging/caps/ [02:19] <ds-work> %P is in CVS [02:38] <Company> hm, saeeking is really funky [02:41] iain ([email protected]) left irc: "leaving" [02:48] sublett ([email protected]) left irc: Remote closed the connection [02:56] C ([email protected]) joined #gstreamer. ------------------------------------------------------- The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA. http://www.eclipsecon.org/osdn