From: IRC <wt...@us...> - 2003-11-08 06:41:54
|
******************************************************************* [03:00] <taaz> rewrite! [03:00] <ds-work> yup [03:00] <ds-work> not today [03:00] <ds-work> caps first [03:00] <Company> do caps first [03:00] <Company> :) [03:01] <Company> gst_buffer_new_and_alloc should be killed [03:01] <Company> gst_buffer_new_from_pad is the right function to use [03:01] <taaz> so the videosink frame_time... is it supposed to be uint64, int64 or int? since it's used all three ways and throwing errors with gst-inspect [03:01] <ds-work> bufferpools is one of those things that we can't even theoretically get correct with the current implementation [03:01] <ds-work> is it a absolute or relative time? [03:02] <taaz> i have no idea. [03:02] <Company> it's probably just the timestamp of the last buffer displayed [03:02] <Company> in which case it'd be guint64 [03:02] <ds-work> "The interval between frames" [03:03] <Company> oh [03:03] <ds-work> should be gint64 [03:03] <Company> because it's a GstClockTimeDiff :o [03:03] <ds-work> even though it doesn't make sense negative [03:03] <ds-work> yep [03:03] <taaz> lovely... aasink and v4lmjpegsink are confused the same way too [03:03] <Company> i need to rewrite clocking [03:04] <ds-work> we could realistically make GstClockTime[Diff] into GValues and use those directly [03:04] <ds-work> taaz: please fix [03:04] <taaz> i'll fix... just didn't know what to fix to ;) [03:04] <ds-work> but why isn't it part of videosink? [03:04] <Company> i want GstClockTime gst_element_get_time (GstElement *element); [03:05] <taaz> why isn't what part of videosink? [03:05] <taaz> glsink gets it wrong too [03:05] <ds-work> the object property [03:05] <taaz> go go copy & paste! [03:06] <taaz> i guess not every element is a subclass of that? [03:06] <taaz> that should certainly be fixed as well [03:07] Action: taaz decides to head home first... [03:08] walters (wa...@ve...) left irc: "brb" [03:15] walters (wa...@ve...) joined #gstreamer. [03:19] <Company> ds-work: you haven't committed "my" mpeg fix yet, have you? [03:32] <Company> being ill sucks [03:32] <Company> you can't concentrate [03:32] Action: Company goes to bed [03:32] <Company> gnight [03:53] abraar (~goo@AToulouse-105-1-30-248.w81-53.abo.wanadoo.fr) left irc: Read error: 60 (Operation timed out) [04:00] Francis (~francis_l@MTL-ppp-157336.qc.sympatico.ca) joined #gstreamer. [04:04] The_Company (~Co...@p5...) joined #gstreamer. [04:09] abraar (~goo@AToulouse-105-1-24-200.w81-50.abo.wanadoo.fr) joined #gstreamer. [04:22] Company (~Co...@p5...) left irc: Read error: 110 (Connection timed out) [04:39] Francis (~francis_l@MTL-ppp-157336.qc.sympatico.ca) left irc: [04:53] TheCompany (~Company@pD958B3BF.dip.t-dialin.net) joined #gstreamer. [04:57] <taaz> i'm just getting confused trying to fix up aasink [05:03] <ds-work> howso? [05:13] The_Company (~Co...@p5...) left irc: Read error: 110 (Connection timed out) [05:16] <ds-work> wow, I think I just found a longstanding bug [05:16] <ds-work> that stupid green flickering as a movie starts [05:17] <ds-work> now it's a blue flickering :) [05:59] thaytoo (ja...@ad...) left irc: "leaving" [06:20] <taaz> hmm.. yeah stuff like knowing where to set video size [06:20] <taaz> which you just changed ;) [06:21] <taaz> i noticed xvideosink sets the w/h fields direct then calls that got_size thing [06:21] <taaz> is that the prefered way to do it? [06:22] <taaz> theres also set_geometry but no get_geometry [06:22] <taaz> in videosink [06:22] <taaz> not sure how all this fits with aasink [06:22] <taaz> since it's get/set_property query some aasink fields directly [06:23] <taaz> kind of seems like videosink get/set_geometry might be the way to query that [06:23] <taaz> have to add get_ [06:47] Nick change: harshy -> harshyZ [07:34] mrkidd (~mr...@12...) joined #gstreamer. [07:37] walters (wa...@ve...) left irc: "z" [08:21] Rotty (~an...@ch...) joined #gstreamer. [08:37] thaytan (~ja...@ad...) joined #gstreamer. [08:57] Rotty (~an...@ch...) left irc: Remote closed the connection [09:18] kwm (~kw...@c7...) joined #gstreamer. [09:25] swentel (sw...@d5...) joined #gstreamer. [09:39] <sxpert> gst-player from debian packages doesn't work [10:03] athom (~the...@so...) joined #gstreamer. [10:06] herzi (~herzi@158.64.76.5) joined #gstreamer. [10:31] wheels (~sc...@ds...) left irc: "using sirc version 2.211+KSIRC/1.3.8" [10:38] md` (ill...@md...) left irc: Read error: 104 (Connection reset by peer) [10:44] md` (ill...@md...) joined #gstreamer. [10:44] dolphy (~do...@po...) joined #gstreamer. [11:02] TheCompany (~Company@pD958B3BF.dip.t-dialin.net) left irc: Remote closed the connection [11:12] apoc (~ap...@dy...) joined #gstreamer. [11:12] <apoc> yo guys [22:29] _gst_newt_ joined #gstreamer. [22:29] Rotty (~an...@ch...) left irc: Remote closed the connection [22:29] thaytan (~ja...@ad...) left irc: "See y'all" [22:47] murrayc (~murrayc@80.121.222.26) joined #gstreamer. [22:51] <murrayc> hi. Is gst-plugins HEAD meant to build now? [22:51] sjoerd (~sj...@si...) joined #gstreamer. [22:52] Nick change: harshyWork -> harshy [22:56] Ridcully (~rid...@d0...) left irc: "Leaving" [22:56] <taaz> murrayc: sure. does it? ;) [23:06] alley_cat (AlleyCat@pD9519741.dip.t-dialin.net) left irc: Read error: 60 (Operation timed out) [23:10] murrayc (~murrayc@80.121.222.26) left irc: Read error: 104 (Connection reset by peer) [23:13] murrayc (~murrayc@80.121.222.26) joined #gstreamer. [23:13] <murrayc> Sorry if missed a reply before. Should gst-plugins on HEAD build? [23:14] <taaz> i said: [23:14] <taaz> murrayc: sure. does it? ;) [23:14] alley_cat (AlleyCat@pD9E6148D.dip.t-dialin.net) joined #gstreamer. [23:15] <murrayc> Nope. One moment. [23:15] <murrayc> A few more moments. [23:21] <murrayc> So far it's building. Maybe somebody checked something in in the last 10 minutes. [23:28] <murrayc> ximagesink.c: In function `gst_ximagesink_chain': [23:28] <murrayc> ximagesink.c:633: `GST_EVENT_RENEGOTIATE' undeclared (first use in this function) [23:28] <murrayc> ximagesink.c:633: (Each undeclared identifier is reported only once [23:28] <murrayc> ximagesink.c:633: for each function it appears in.) [23:28] <murrayc> make[3]: *** [libgstximagesink_la-ximagesink.lo] Error 1 [23:29] <murrayc> taaz: Nope, doesn't build. [23:29] <taaz> that's due to changes from today [23:30] <murrayc> In gstreamer? [23:30] <taaz> update your core [23:30] <murrayc> core? [23:30] Action: taaz doesn't quite understand why that change is even needed... [23:30] <taaz> yeah, it's in gst/gstevent.h or something [23:30] <murrayc> Are you saying that I need to update gstreamer? [23:30] <taaz> yeah [23:30] <murrayc> OK, trying now, but I think I'm very up-to-date already. [23:32] <murrayc> Yeah, I think I was already fully up-to-date. I think it's broken. This is jhbuild's gnome26 moduleset. [23:33] <taaz> or you could just update to ds's fix for that problem [23:33] <taaz> which was just checked in [23:34] <murrayc> just? [23:34] <murrayc> minutes ago? [23:34] <murrayc> In gst-plugins, or gstreamer? [23:35] <ds-work> two checkins, one lunch apart. If you got one but not the other, it will fail to build [23:37] <murrayc> I some changes now in ximagesink.c and the Makefile.am, so I guess that's it. Thanks. The gnome 2.5 build is now looking a lot better than last week. [23:38] <murrayc> There's an .LoT file checked in too. Isn't that a libtool thingy? [23:38] <murrayc> Oh, nonsense, ignore that. [23:45] <ds-work> murrayc: do you happen to know if gnome-media compiles against gstreamer-HEAD? [23:45] <murrayc> ds-work: I'm building gstreamer+gst-plugins now, so I guess I'll get to gnome-media soon. [23:46] hk (~SpyVB@200.182.62.135) joined #gstreamer. [23:53] yippi (~br...@nw...) left irc: "Client exiting" [23:55] <murrayc> ds-work: gnome-media builds. nautilus-media doesn't. [00:00] --- Sat Nov 8 2003 [00:26] m_wheels (~sc...@ds...) joined #gstreamer. [00:27] Nick change: m_wheels -> wheels [00:30] md` (ill...@md...) left irc: Read error: 54 (Connection reset by peer) [00:31] kwm (~kw...@c7...) left irc: "bed looks nice" [00:32] <wheels> Company: ping? [00:34] <Company> wheels: unknown host Company [00:34] <wheels> Company: interesting in the "make RB compile with 2.95" patch of the week? :-) [00:35] <wheels> err, "interested" [00:35] <wheels> (walters isn't around for me to bug) ;-) [00:36] md` (~ill...@md...) joined #gstreamer. [00:36] <Company> wheels: rhy...@gn... is better - my current rb build is quite old already [00:37] <ds-work> Company: do you have any idea why we have GST_PAD_TEMPLATE_GET()? [00:37] <wheels> Company: Bah, I'll just wait for walters to appear and dcc it. :-) [00:38] <Company> wheels: that might work, too [00:38] <Company> ds-work: I assume because someone wanted to make padtemplates static structs instead of functions some day [00:39] <Company> ds-work: and then you'd only have to change that macro [00:43] <ds-work> yeah, I'd like padtemplates to be static structs myself [00:44] <ds-work> we can do that after we get static caps [00:48] Rotty (~an...@ch...) joined #gstreamer. [00:49] sublett (~rv...@21...) joined #gstreamer. [00:49] Ridcully (~rid...@d0...) joined #gstreamer. [00:55] <ds-work> one can delete a bunch of lines by simply creating the pad templates in the base_init() function now [00:56] <Company> as long as you don't need the same padtemplate elsewhere [00:56] <Company> like when creating pads [00:57] <ds-work> the plugin_init reorg certainly made things a lot more sane [00:57] <ds-work> you can create pads from the templates in the class [00:57] <ds-work> I think that's really clean, actually -- gst_element_add_pad_from_template(element, "src"); [00:58] <Company> it was a shitty thing though [00:58] <Company> not fun to hack on [00:58] <Company> i'm glad it's done [00:58] <ds-work> thanks for doing it [00:58] <ds-work> I think it will make elements a lot simpler in the long run [00:59] <Company> i should probably give out a beer to BBB, ain, thaytan and whoever helped me convert the plugins [00:59] <ds-work> and sorry about giving you so much shit about breaking the build [00:59] <Company> you were quite right though [00:59] <Company> next time i'll collect volunteers in advance [01:00] <Company> branch on a weekend and do it then [01:03] <ds-work> so you volunteer for the caps migration? [01:03] <steveb_> group hug! [01:03] <ds-work> steveb_: fuck off [01:03] <ds-work> :) [01:03] <steveb_> hoho [01:03] <ds-work> :) [01:04] <Company> i volunteer for helping in the caps migration, sure [01:05] <ds-work> I'm almost to the point of doing a branch to migrate the gstreamer/gst/ code [01:05] <ds-work> intersection started working last night [01:05] Rotty (~an...@ch...) left irc: "Bye" [01:06] <ds-work> the code is _so_ much cleaner than gstprops.c [01:06] <Company> that's not hard [01:09] Rotty (~an...@ch...) joined #gstreamer. [01:09] Ridcully (~rid...@d0...) left irc: Remote closed the connection [01:10] Ridcully (~rid...@d0...) joined #gstreamer. [01:13] Rotty (~an...@ch...) left irc: Client Quit [01:16] Rotty (~an...@ch...) joined #gstreamer. [01:20] Nick change: TD -> TD[zZz] [01:21] sublett (~rv...@21...) left irc: "I like food, food is good!" [01:24] <ds-work> shouldn't g_object_class_install_property() be called from base_init()? [01:28] <apoc> night all [01:28] apoc (~ap...@dy...) left irc: "Client exiting" [01:28] <Company> ds-work: no, because properties are only installed in the class they belong to, not subclasses again [01:29] <ds-work> ah, ok [01:29] <ds-work> that makes sense [01:29] <Company> it's that evil gobject "some stuff belongs in subclasses, some doesn't" stuff ;) [01:30] <ds-work> just need to remember what searches all parent classes [01:30] <ds-work> (just) [01:33] somex1 (~kn...@d6...) joined #gstreamer. [01:33] <ds-work> Company: did you mention that we shouldn't be using g_type_class_ref()? [01:34] <Company> ds-work: yes [01:34] <Company> ds-work: see docs of g_type_class_peek_parent for reasons [01:34] Ridcully (~rid...@d0...) left irc: "Leaving" [01:36] <ds-work> I see that I already added a check to gst-lint, but with a typo [01:42] <taaz> blah... latest debian bug has xvideosink crap messages still being printed out. grr. if anyone knows how to fix that please do so. [01:43] <ds-work> what changed? [01:43] <taaz> plus i have no idea what that bug is refering to [01:43] <taaz> maybe nothing changed... i have no idea. i'm not very good at recreating those x problems. so therefore i don't bother to debug them. [01:45] Action: Company doesn't care about bugs atm [01:46] <Company> i'll care only if it is affecting me [01:46] <Company> i'll care again when 0.8 is out [01:46] <ds-work> I care if it affects an app that people use to test, like rb or gst-player [01:46] Misirlou (~ae...@c-...) left irc: [01:46] <Company> i use them ;) [01:47] <Company> we need a policy for commits when 0.8 is out btw [01:48] <ds-work> IMO, we just need a gatekeeper/maintainer that keeps track of things [01:48] <ds-work> we didn't/don't have anyone like that for 0.6 [01:48] <Company> i want to be allowed to commit what i think is ok [01:48] <Company> like simple bugfixes [01:49] <ds-work> thomasvs was doing it mostly, but I don't think he felt "empowered" [01:49] <Company> even more advanced stuff when i feel it's needed [01:50] <Company> this policing always makes me feel like someone doesn't trust me [01:50] <Company> "you're not worthy to decide" [01:50] <ds-work> hopefully, we'll have more testing available for 0.8 [01:50] <ds-work> Company: it's not policing, it reviewing [01:51] <Company> ds-work: reviewing works best with stuff checked in ;) [01:51] <ds-work> if it's checked in, it never gets reviewed :) [01:51] <Company> i'm all for a no-commit period before releases [01:51] <Company> but not a no-commit-ever period [01:51] <Company> that makes me want to hack on a branch where i may commit [01:55] Rotty (~an...@ch...) left irc: "Bye" [01:57] <ds-work> there's no way to remove pad_templates from an element [01:58] <Company> there never was :) [01:58] <Company> why should there be? [01:59] <ds-work> In a subclass, I'd like to potentially modify a pad_template [01:59] <ds-work> anyway, it's not that important [02:06] sublett (~rv...@21...) joined #gstreamer. [02:11] <jaiserca> is supossed id3types can detect id3v1 tags? [02:11] <Company> not yet [02:12] Action: Company mumbles about typefinding and that that includes a plugin that does that [02:15] <jaiserca> well, gsttypefindfunctions detect id3v1 tag but not parses it [02:15] <Company> yup [02:16] <Company> some part of gstreamer are better than others ;) [02:18] hk (~SpyVB@200.182.62.135) left irc: "Leaving" [02:18] <jaiserca> ok, i'll check id3types tomorrow for a working MediaInfo [02:22] <Company> it'll take some time before it's merged [02:23] <jaiserca> but i need it for my program (: [02:24] <Company> what program? [02:26] <jaiserca> its not released. It's a id3, ogg tagger but something completely different [02:26] <jaiserca> *completely* different [02:27] <ds-work> er, wouldn't that make it _not_ an id3, ogg tagger? [02:27] <jaiserca> w/ nautilus intregation and more [02:27] <Company> it's different as in gstreamer [02:28] <Company> gstreamer is a media framework, but different [02:28] <Company> ;) [02:28] <Company> actually it's a tagging framework soon ;) [02:29] Action: ds-work wonders when [02:29] Action: Company wonders, too [02:29] <jaiserca> basically you make groups and assign rules to the group and rules are applied when you move files even in nautilus [02:30] <Company> i don't like the api yet and need to port it to gst-structure anyway [02:30] <ds-work> the videofilter example is down to 225 lines [02:30] <ds-work> writing GstElements is hard [02:31] <Company> i want an API that assumes 1 tag => 1 GValue by default [02:31] <Company> ds-work: i know, it needs to become a lot simpler [02:31] <Company> ds-work: that's why i want loads of stuff simplified (event mask functions, qwueries, ...) [02:32] Rotty (~an...@ch...) joined #gstreamer. [02:32] Rotty (~an...@ch...) left irc: Read error: 104 (Connection reset by peer) [02:33] <Company> and that's why we need subclassing badly [02:34] <taaz> there is an example? [02:34] <ds-work> videofilter [02:35] <taaz> oh... i read videosink ;) nevermind [02:35] <ds-work> videosink, too [02:35] pb_ (~pb...@cp...) left irc: "Client exiting" [02:48] thaytan (~ja...@no...) joined #gstreamer. [02:49] thaytan (~ja...@no...) left irc: Read error: 54 (Connection reset by peer) [02:50] Rotty (~an...@ch...) joined #gstreamer. [02:51] thaytan (~ja...@no...) joined #gstreamer. [02:51] thaytan (~ja...@no...) left irc: Read error: 104 (Connection reset by peer) [02:52] Rotty (~an...@ch...) left irc: Remote closed the connection [02:52] <ds-work> ugh [02:53] thaytan (~ja...@no...) joined #gstreamer. [02:53] thaytan (~ja...@no...) left irc: Read error: 104 (Connection reset by peer) [02:54] Rotty (~an...@ch...) joined #gstreamer. [02:56] Rotty (~an...@ch...) left irc: Remote closed the connection |