IRC Logs

*******************************************************************
[03:02] deadchip ([email protected]) left irc: Remote closed the connection
[03:14] <Company> ds-work: i now know why wtay was opposed to my clocking changes
[03:14] mxpxpod ([email protected]) left irc: Read error: 60 (Operation timed out)
[03:14] <ds-work> ?
[03:14] <Company> ds-work: for wtay clocks were implicitly used for element synchronizations
[03:14] <ds-work> yes
[03:15] <Company> ds-work: synchronization is a hard problem though (wrt seeking)
[03:15] <Company> it is currently unsolved
[03:15] <Company> but clocks are not the way to do sync
[03:16] <Company> LeRoutier: i think i've fixed the warnings in asf, so thx for your help
[03:16] matrixise ([email protected]) left #gstreamer ("Leaving").
[03:16] <Company> LeRoutier: i dunno if that solves your problems though
[03:16] <LeRoutier> no prob, finding a bug exists is really little
[03:17] <Company> ds-work: currently synchronization is again done implicitly via "clock events"
[03:17] Zapek ([email protected]) joined #gstreamer.
[03:17] <Company> ds-work: and it works
[03:17] <LeRoutier> updating gst-plugins now to know if it fixed my prob
[03:18] Zapek ([email protected]) left #gstreamer ("blah").
[03:18] <ds-work> Company: presumably, synchronization needs to be two-way -- a clock may provide the time, but an element that is late needs to be able to tell the clock to stall
[03:19] <ds-work> although I'm not to keen on the stalling idea.  I'd prefer to just get it right every time
[03:19] <Company> a clock may not stall
[03:20] <Company> you can't stop gettimeofday
[03:20] <ds-work> but you can change the conversion between gettimeofday and GST_TIME
[03:20] <Company> yeah, but that sucks
[03:20] <Company> because you never know who else uses the clock
[03:21] <ds-work> how else do you cope with SIGSTOP?
[03:21] <ds-work> or sleeping a laptop
[03:21] <ds-work> or ntp updating the clock?
[03:21] <Company> you use a better clock
[03:22] <Company> but those are general cases that are relevant for every element depending on the clock
[03:23] <Company> anyway, we need to have a way to say "element1 and element2 are synchronized, both have the same element time"
[03:23] <Company> possibly an interface
[03:23] <Company> but we need to think that through because that can break when seeking
[03:24] <Company> discont events may arrive at different times in both elements
[03:25] <walters> Company: 
[03:25] <walters> --- gstid3tag.c.~1.10.~	2004-01-08 00:44:27.000000000 -0500
[03:25] <walters> +++ gstid3tag.c	2004-01-14 21:20:40.000000000 -0500
[03:25] <walters> @@ -763,7 +763,7 @@
[03:25] <walters>    }
[03:25] <walters>    if (!tag->srcpad)
[03:26] <walters>      gst_id3_tag_add_src_pad (tag);
[03:26] <walters> -  return gst_pad_renegotiate (tag->srcpad) != GST_PAD_LINK_REFUSED;
[03:26] <walters> +  return gst_pad_try_set_caps (tag->srcpad, tag->found_caps) != GST_PAD_LINK_REFUSED;
[03:26] <walters>  }
[03:26] <walters>  
[03:26] <walters>  static GstPadLinkReturn
[03:26] <walters> Company: is that ok to apply?
[03:27] <Company> walters: i prefer to use the correct function for that, dunno which one that is though :)
[03:27] <walters> isn't that the correct function?
[03:27] <walters> the negotiation docs only mention _try_set_caps for explicit negotiation
[03:27] <Company> walters: it works implicitly, the getcaps function does that
[03:30] <ds-work> what are you setting the caps to?
[03:30] <LeRoutier> 12 minutes to compile gst-plugins, argh
[03:31] <LeRoutier>  
[03:32] <LeRoutier> hum, i've got no more warning playing the .asf video. the problem is it does not play video either
[03:33] <Company> ds-work: hm?
[03:34] <ds-work> walters: that patch is wrong
[03:34] <walters> ds-work: yeah, i realized that now
[03:35] <Company> _renegotiate doesn't work though when the pad is not connected
[03:35] <Company> pads never negotiate when they're unconnected, right?
[03:35] <ds-work> correct
[03:36] <ds-work> renegotiate may complain about not being connected, too
[03:36] <LeRoutier>  
[03:36] <Company> very bad
[03:36] <LeRoutier> i still have optschel doing bad things : ERROR       scheduler(13741) gstoptimalscheduler.c(2074):gst_opt_scheduler_iterate: [GstOptScheduler@0x82b67a0] in error state
[03:36] <walters> LeRoutier: that doesn't mean it's a bug in opt
[03:36] <Company> or maybe not, i'm not sure
[03:37] <Company> we should make that a GST_WARNING
[03:37] <Company> so people don't blame opt :)
[03:37] <LeRoutier> walters, got that after it : element work_thread claimed state-change success,but state didn't change to PLAYING. State is PAUSED (NONE_PENDING pending), fix the element
[03:38] steve_b ([email protected]) left irc: Read error: 110 (Connection timed out)
[03:40] <LeRoutier>  
[03:40] <Company> man, we don't even have gst_pad_is_linked
[03:40] <LeRoutier> from what i read in changelog, i can now use spider & alsasink together without infinie loop ?!
[03:40] <ds-work> Company: GST_PAD_PEER()
[03:41] <Company> ds-work: i know, but i'd prefer plugins using a function for that
[03:41] NFusi0n ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[03:41] <Company> LeRoutier: yes
[03:42] <Company> (i hope you can)
[03:42] <ds-work> Company: of course, so do I
[03:42] <Company> ds-work: should I implement gst_pad_is_linked ?
[03:43] <ds-work> yes
[03:49] <Company> http://tranquillity.campus.luth.se/~banan/qdmc/qdmc2.pdf - gogogogo
[03:49] <Company> i need a motivation to fix qtdemux
[03:50] mxpxpod ([email protected]) joined #gstreamer.
[03:51] <LeRoutier>  
[03:51] <LeRoutier> has a gstreamer dev  a SMP machine around ?
[03:52] steve_b ([email protected]) joined #gstreamer.
[03:52] <Company> SMP is known to be broken, i hope you know that
[03:52] <ds-work> is QDM2 really not reverse engineered?
[03:53] <ds-work> really?
[03:53] <ds-work> I use SMP all the time
[03:53] <Company> i don't think it is
[03:53] <Company> SMP has ugly races in the core
[03:53] <Company> qdm is only supported via win32
[03:54] thaytan ([email protected]) left irc: Read error: 110 (Connection timed out)
[03:54] yoann ([email protected]) joined #gstreamer.
[03:55] <yoann> hi
[03:55] <steve_b> hey guys, your projecting on the screen at linuxconf
[03:55] <steve_b> don't swear ;)
[03:55] <Misirlou> Hi Mom!
[03:55] <LeRoutier> beware of this SMP user : yoann
[03:55] <ds-work> uh, that's supposed to _stop_ us?
[03:55] <yoann> so, gstreamer is known to be broken on SMP ?
[03:56] <yoann> LeRoutier: lol :)
[03:56] <ds-work> yoann: it's known to be non-SMP-clean, but it works
[03:56] <Misirlou> Clearly not working out.
[03:56] <yoann> ds-work: well, for me it crash ~ every minute :)
[03:57] DrZaius ([email protected]) joined #gstreamer.
[03:57] <yoann> ds-work: and when it work, it stop playing a song before the end of it 
[03:57] <yoann> lol
[03:57] <yoann> funily, it appear to work on UP
[03:57] <Misirlou> steve_b: So, what are we doing on the screen?
[03:57] <yoann> so I guess there are some re-entrancy problem :)
[03:57] <ds-work> yoann: what kernel are you using?
[03:58] <Company> UP works because the Linux scheduler is too nice when scheduling threads
[03:58] <yoann> ds-work: I'm using a very specific config... 2.6.1 on an NPTL enabled system
[03:58] <ds-work> yoann: use 2.4
[03:58] <yoann> ds-work: thought I have the same effect without NPTL
[03:58] Action: Company still thinks Linux needs echo 1 > /proc/evi_threading
[03:58] <yoann> ds-work: you kidding I hope ? :)
[03:58] <ds-work> yoann: no
[03:58] <Company> and echo 1 > /proc/evil_threading too
[03:59] <yoann> ds-work: I need 2.6 ability for my own work, which involve captuing network stream at gigabits speed :)
[03:59] <yoann> ds-work: 2.6 threading is not broken as soon as you disable NPTL, which is kinda easy
[04:00] <yoann> ds-work: so, what's the issue with 2.6 ?
[04:00] <ds-work> threading is "different"
[04:00] <yoann> it's not different unless you use NPTL
[04:00] <yoann> and to use NPTL you need an NPTL compiled glibc...
[04:01] <yoann> and I'm able to reproduce theses issue with or without NPTL
[04:01] <yoann>  
[04:01] <yoann> and I am kinda sure I could reproduce them under 2.4 too
[04:01] <yoann> just to know, anywhere took the time to trace gstreamer under helgrind ?
[04:02] <yoann> s/anywhere/anyone/
[04:03] <Company> no complaints (at least on UP ;))
[04:04] <yoann> Company: okay, I'll try to take some time to look at it
[04:04] <ds-work> yoann: your speculation is not useful, as I use gstreamer continuously without seeing SMP issues
[04:04] <yoann> Company: also, is the alsa support known to be working well, or could side effect come from it's usage ?
[04:04] <Company> yoann: don't try, SMP requires quite some redesign
[04:05] <yoann> ds-work: and what kind of machine do you use, then ?
[04:05] <ds-work> dual Athlon and quad P4, both running 2.4.23
[04:05] <Company> yoann: i'm using alsa as my output and it works well (aside from usual alsalib bugs ;))
[04:05] <yoann> ds-work: I'm not speculating. I'm tellign fact: fact is gstreamer is screwed on SMP machine with or without NPTL
[04:06] <Company> walters: fixed id3tag
[04:06] <Company> walters: requires core update
[04:06] <yoann> Company: ok, I guess when you say SMP require redesign, that a lot of access are not protected ?
[04:07] <Company> yoann: that's one thing and the other thing is that quite some pieces of code make assumptions about execution order
[04:07] <ds-work> yoann: you're speculating _why_ it's not working.  I'm telling you that there are known issues, and that your config is aggravating them
[04:07] <yoann> Company: is the lack of locking existing because of the fear of contention, or is it simply not done ?
[04:08] <yoann> ds-work: my interest is in fixing the issue, not in working them around. Thus my question.
[04:08] <Company> yoann: it's simply not done
[04:09] <Company> yoann: I plan to address all these issues for 0.10 when I hopefully got an SMP machine from somewhere...
[04:09] <yoann> Company: lol... I also got an SMP machine from somewhere for that kind of problem on another project :-D
[04:11] <yoann> Company: anyway, I'll try to do some tracing, and I'll come back to you if I can find some useful things
[04:11] <Company> d'you still need it?, ;)
[04:11] <yoann> Company: yes, unfortunaly... My stuff is about being able to analyze gigabits stream... 
[04:12] <yoann> Company: which is a lot of work, far from being finished :)
[04:12] <Company> heh
[04:12] <yoann> Company: (and require SMP machine in order to be done)
[04:12] <Company> well, i need to get an SMP from elsewhere then
[04:13] <Company> in the meantime it might be an interesting idea to use kernel calls to bind all gst threads to the same processor
[04:13] <yoann> Company: you can do so, but you have to be aware that theses call are not portable
[04:14] <yoann> Company: it's not even sure they work under Linux. They are define through.
[04:16] Misirlou ([email protected]) left irc: "Installing a new CD burner."
[04:18] <Company> yoann: it was a suggestion, so if you'd implement that on your box and it works, i'd include that in the core (at least as a command line option) to make life for SMPers easier
[04:23] thaytan ([email protected]) joined #gstreamer.
[04:24] <walters> Company: crashes on metadata load now
[04:25] <Company> that's at least a different bug
[04:42] thaytan ([email protected]) left irc: "See y'all"
[04:44] <LeRoutier> good night guys
[04:44] LeRoutier ([email protected]) left irc: "Leaving"
[04:56] LCA_Bruce ([email protected]) joined #gstreamer.
[05:03] <thomasvs> ah, the first person to have seen steveb talk it looks like ?
[05:03] <thomasvs> steve_b: did the recent slashdotting help to get you more people ?
[05:03] <steve_b> over half full 
[05:03] <LCA_Bruce> yeah i am in the talk
[05:04] <steve_b> your on the screen ;)
[05:04] <LCA_Bruce> anyway, bye
[05:04] LCA_Bruce ([email protected]) left #gstreamer.
[05:05] <thomasvs> steve_b: what does that mean, we're on the screen ?
[05:05] <thomasvs> our ugly mugs ?
[05:05] <thomasvs> send us a picture or something :)
[05:06] Action: thomasvs should really get some sleep instead of faffing about at 5 PM
[05:15] <walters> Company: committed a patch.
[05:16] NFusi0n ([email protected]) joined #gstreamer.
[05:22] hadley ([email protected]) joined #gstreamer.
[05:23] steve_b ([email protected]) left irc: Read error: 60 (Operation timed out)
[05:37] harshy ([email protected]) joined #gstreamer.
[05:42] Misirlou ([email protected]) joined #gstreamer.
[05:46] harshyWork ([email protected]) left irc: "I quit for now"
[05:47] Miko5881 ([email protected]) left irc: "ChatZilla 0.8.31 [Mozilla rv:1.4/20030928]"
[05:54] <walters> Company: still here?
[05:56] <thomasvs> walters: btw, congrats on the RH job, that's awesome news
[05:57] <walters> thanks :)
[05:59] <harshy> walters: you and the rest of us need to get drinks before you leave us
[05:59] <walters> harshy: sure, still two months out :)
[05:59] <harshy> ya I know
[06:00] <harshy> just makin sure
[06:07] trow ([email protected]) left irc: "Leaving"
[06:10] <taaz> walters: RH job?  nice.  going to try and convert them to Debian? ;)
[06:11] <walters> heh
[06:12] <walters> they're trying to do that themselves actually...
[06:12] <walters> (wrt Fedora)
[06:13] <dilinger> walters: where will you be working?
[06:13] <dilinger> (which rh office, i mean)
[06:13] <walters> in their Westford office, outside of Boston
[06:13] <dilinger> oh, cool
[06:14] <dilinger> the sistina people are going to be there in 2 weeks, i'm gonna meet up w/ them (boston's a 3 hr drive from here, i go there a lot to visit friends)
[06:15] <walters> dilinger: ah, is that where you work?
[06:15] <dilinger> walters: i work in ny (for voxel)
[06:16] <walters> ah, ok.
[06:16] <taaz> walters: what will you do at rh?  full time rb hacker? ;)
[06:17] <walters> taaz: desktop stuff, mostly GNOME development I think.
[06:34] ChrisHJW_log ([email protected]) left irc: Read error: 110 (Connection timed out)
[06:44] Action: ds needs to get a Red Hat job
[06:52] water ([email protected]) joined #gstreamer.
[06:55] <water> hi. short intro: i work on a smalltalk variant which happens to have a similar, but much less developed, streaming architecture
[06:55] thaytan ([email protected]) joined #gstreamer.
[06:58] <ds> cool
[06:59] <water> it actually is code that someone else wrote in 98 for an embedded media platform, but i've been extending it
[06:59] <water> and working with the author
[07:01] Action: water looks over the api's first
[07:06] <water> ok, i admit i am intrigued. there are clearly a lot of considerations in here that i haven't had the effort or ideas to work on
[07:07] <water> obviously this is a really good framework, or i wouldn't be here :)
[07:07] <water> well anyway, i will lurk, and maybe ask a few questions sometime
[07:08] <ds> there probably won't be much conversation for the next 3 hours
[07:08] <water> ok. this sounds like the american-european time gap
[07:08] <ds> sort of
[07:09] <thomasvs> yeah, us europeans have yet to wake up
[07:09] <ds> it's a bunch of europeans and me in California
[07:09] <water> hehe
[07:09] Action: water is in seattle
[07:09] <thaytan> hey!
[07:09] <ds> except for thomasvs who needs to be on a plane right now
[07:09] <thaytan> everyone forgets the anzacs 
[07:09] <thomasvs> ... and didn't get any sleep at all
[07:10] <thaytan> steve's gst talk went well earlier
[07:10] <ds> and the occasional ozzie
[07:10] <walters> thaytan: did he really show irc during the presentation?
[07:10] <thomasvs> thaytan: why did he show irc ?
[07:10] <thaytan> heh, at the end
[07:11] <thaytan> as he was closing the window, xchat was behind
[07:14] Action: taaz thinks he's turned into a gst groupie rather than developer.  sigh.
[07:14] <thomasvs> taaz: yeah, get going again dude !!!
[07:15] <thomasvs> taaz: people are writing dvd stuff with gst these days.  where are you ?
[07:16] <taaz> i'm one of the foolish sort of people that is trying for a phd.
[07:16] <ds> walters: did you finish?
[07:17] <taaz> plus i have various other programming interests in addition to gst
[07:17] <walters> ds: finish?
[07:17] <ds> walters: weren't you working on a degree?
[07:17] <taaz> and apps like totem reduce the urgency of working on vdv
[07:17] <walters> ds: oh, yeah.  i'm going to do that remotely
[07:17] <water> ack. you guys run your project like i do: from irc with logs :)
[07:18] <ds> we don't read the logs :)
[07:18] <ds> except when they're pasted into documentation
[07:18] <water> too bad. ours look like this: http://meme.b9.com/ ;)
[07:19] <water> (ok, so it's a third party)
[07:19] Action: water returns to lurking
[07:31] ChrisHJW ([email protected]) joined #gstreamer.
[07:43] thomasvs ([email protected]) left irc: Read error: 60 (Operation timed out)
[07:51] hadley ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[07:55] mathrick|sleep ([email protected]) left irc: Remote closed the connection
[08:07] kleppari ([email protected]) joined #gstreamer.
[08:08] <dilinger> hm, has anyone dealt w/ tta before?
[08:08] <dilinger> http://tta.iszf.irk.ru/index.htm
[08:08] herzi ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[08:08] <dilinger> (or even heard of it?  i just stumbled across it)
[08:10] herzi ([email protected]) joined #gstreamer.
[08:10] <dilinger> if those comparisons are valid, it's quite impressive looking
[08:11] <ds> looks pretty fast
[08:11] <dilinger> using normal encoding, 5% smaller than flac, and faster (looking at 2.0 comparisons)
[08:15] herzi ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[08:16] Action: taaz would suggest adding it to GstWishList wiki node but notices the wiki is totall fubar
[08:17] herzi ([email protected]) joined #gstreamer.
[08:19] ChrisHJW_log ([email protected]) joined #gstreamer.
[08:20] Nick change: rk|away -> rk
[08:21] herzi ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[08:22] herzi ([email protected]) joined #gstreamer.
[08:43] sublett ([email protected]) joined #gstreamer.
[08:44] <NFusi0n> taaz: Gstreamer support would be fantastic, especially with this transcode-but-keep-the-correct-tagging stuff i've heard about lately
[08:44] <NFusi0n> what tagging format does tta use
[08:46] Action: taaz knows nothing about tta
[08:47] <NFusi0n> Where did you find the 2.0 comparisons taaz 
[08:47] <NFusi0n> nevermind
[09:26] <ds> WOAH
[09:26] <ds> I appear to have gotten several hundred lines of code correct on the first try
[09:27] <ds> (aside from compile issues)
[09:29] swentel ([email protected]) joined #gstreamer.
[09:31] km_sleep ([email protected]) left irc: "Leaving"
[09:34] Markey- ([email protected]) joined #gstreamer.
[09:38] md` ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[09:41] Shoragan ([email protected]) joined #gstreamer.
[09:46] md` ([email protected]) joined #gstreamer.
[09:48] kleppari ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[09:53] sublett ([email protected]) left irc: "I like food, food is good!"
[10:00] walters ([email protected]) left irc: "out"
[10:00] harshy ([email protected]) left irc: "I quit for now"
[10:08] xcomp ([email protected]) joined #gstreamer.
[10:08] <xcomp> quick question
[10:09] <xcomp> i'm not too familiar with the details of multimedia programming and such
[10:10] <xcomp> is it feasible to use gstreamer to obtain YUV video image data from any supported video file format?
[10:10] <thaytan> yes
[10:10] BBB ([email protected]) joined #gstreamer.
[10:10] <thaytan> hi BBB 
[10:11] <BBB> mornin'
[10:11] <xcomp> thaytan: cool, thanks
[10:12] harshy ([email protected]) joined #gstreamer.
[10:19] Nick change: harshy -> harshyZ
[10:22] Markey- ([email protected]) left irc: "leaving"
[10:25] matrixise ([email protected]) joined #gstreamer.
[10:30] dolphy ([email protected]) joined #gstreamer.
[10:31] <dolphy> morning
[10:32] <thaytan> yo yo!
[10:33] <BBB> hi dolphy
[10:35] <dolphy> BBB: did you know we know have software colorbalance aswell in gstplay :)
[10:35] <dolphy> s/know/now
[10:36] <thaytan> dolphy: beautiful
[10:36] <thaytan> you're doing good work
[10:36] <thaytan> I'm attempting to have better DVD subtitle support for you RSN
[10:36] <thaytan> I have to learn about mpeg subpictures first
[10:36] <BBB> dolphy: no... how about software volume adjustment?
[10:37] <BBB> dolphy: I didn't check out gst-player lately... I will soon
[10:37] <BBB> but need to fix media formats first
[10:41] <BBB> ds/ds-work: are you awake?
[10:44] tlg ([email protected]) joined #gstreamer.
[10:45] <dolphy> BBB: i will implement mixer in volume soon
[10:45] <dolphy> BBB: adding the same type HARD/SOFT in the class
[10:46] <BBB> cool thing, thanks
[10:46] <BBB> oh, can you please add the type field to all implementations that are hw?
[10:46] <BBB> mixer: oss/alsa
[10:46] <BBB> colorbalance: v4l/v4l2
[10:46] <BBB> (saves me some fixing)
[10:48] <dolphy> i will
[10:48] <BBB> ty!
[10:48] <BBB> I'll focus on media format fixing
[10:49] <BBB> (and encoding)
[10:49] Action: Company is going to code up some broken files
[10:49] <Company> like setting the chunk size field on an avi to G_MAX_INT
[10:50] <BBB> actually...
[10:50] <BBB> I wanted to look at those things (security-related) before 1.0
[10:50] <BBB> just "not now"
[10:50] <Company> asfdemux and qtdemux are worse ofenders anyway
[10:50] <BBB> asfdemux needs a semi-rewrite
[10:51] <BBB> to become more like matroskademux and avidemux
[10:51] <BBB> "parsing and interpretting separated"
[10:51] <BBB> that allows for nice, clean code
[10:51] <BBB> where reading of chunks is centrlized in one place
[10:52] <BBB> I pretty much like the structure of matroskamux/demux
[10:52] pb_ ([email protected]) joined #gstreamer.
[10:53] <Company> and all encoders need to be fixed to output buffers in presentation order
[10:53] <Company> to get around issues with queue sizes
[10:53] <BBB> the matroska devs bugged me about that, too
[10:53] <BBB> avi expects 'coding order', matroska expects 'display order'
[10:54] <BBB> a bit ugly, we'll need to hack around stuff in avi to fix that
[10:54] Action: BBB doesn't like it
[10:54] <ChrisHJW> BBB : matroska is coding order also
[10:54] <ChrisHJW> but the timestamps on the block are presentation order
[10:54] <ChrisHJW> so, the timestamps on the blocks in the file may not be linear if b-frames are used
[10:55] <Company> gtg
[10:55] Action: BBB doesn't get it anymore
[10:55] <BBB> you guys go fix mpeg-4 stuff...
[10:55] <BBB> I'll focus on simpler stuff ;)
[10:55] <ChrisHJW> lol
[10:56] <ChrisHJW> there is a simple solution to overcome this, write AVI compatibility files, as i told you long ago ;)
[10:57] <ChrisHJW> even if we set the codec ID to V_MS/VFW/FOURCC, we actually call it AVI compatibility files
[10:57] <ChrisHJW> no complicated frame references, etc.
[10:58] <ChrisHJW> still, timestamps of b-frames can be non-linear in the file, as frames are stored in coding order in the file
[10:58] Company ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[10:59] gheet ([email protected]) joined #gstreamer.
[10:59] Action: BBB will uderstand this one day
[10:59] <BBB> I think...
[10:59] joanplanas ([email protected]) joined #gstreamer.
[11:01] somex1 ([email protected]) left #gstreamer.
[11:18] ChrisHJW ([email protected]) left irc: Read error: 60 (Operation timed out)
[11:21] <dolphy> BBB: when not implementing one method of an interface
[11:21] <dolphy> BBB: you just don't do klass->method = my_element_method; ?
[11:24] thaytan ([email protected]) left irc: Read error: 110 (Connection timed out)
[11:28] <BBB> yes
[11:28] <BBB> the interface sets it to NULL
[11:28] <BBB> and the wrapper functions check for it being NULL or not
[11:28] <BBB> so not setting it is enough
[11:34] <dolphy> ok
[11:34] <dolphy> BBB: set_record is quite useless in volume :)
[11:38] <BBB> yes, so don't implement it ;)
[11:38] <BBB> oh, you might want to add a third track type for 'volume': SOFTWARE
[11:39] <BBB> because 'volume' isn't really an input or an output
[11:41] <dolphy> ah so it is a software mixer
[11:41] <dolphy> and i should add a new track type aswell
[11:41] <dolphy> right ?
[11:43] <BBB> that's what I meant ;)
[11:48] danb ([email protected]) left irc: "leaving"
[11:48] danb ([email protected]) joined #gstreamer.
[11:53] <dolphy> ds ds-work: ping ?
[11:53] wheels ([email protected]) left irc: "using sirc version 2.211+KSIRC/1.3.9"
[11:54] <dolphy> BBB: how can you be sure of the caps you set in colorspace.c ?
[11:54] <dolphy> BBB: it might be xRGB or RGBx etc..
[11:54] <dolphy> BBB: and gsthermescolorspace.c needs the same fixes btw
[11:55] <BBB> I checked in the implementation
[11:55] <BBB> ah, /me notices
[11:56] teuf ([email protected]) joined #gstreamer.
[11:57] <teuf> hi
[11:57] <dolphy> yo teuf
[12:00] <BBB> working on this one too
[12:00] <teuf> who is this ximian hacker who's working on gstreamer/gnome integration ?
[12:00] <teuf> latest thomasvs blog entry made me curious ;)
[12:00] <BBB> fixed too
[12:01] <BBB> trow, I guess?
[12:02] <teuf> ah trow is a ximian dude ?  I didn't know that ;)
[12:02] <BBB> he is
[12:05] <BBB> gstsnapshot.h:28:31: ../hermes/yuv2rgb.h: No such file or directory
[12:05] <BBB> gstsnapshot.h:29:31: ../hermes/yuv2yuv.h: No such file or directory
[12:05] <BBB> eeeeeeeeeeeeeeeeeeeehw
[12:06] joanplanas ([email protected]) left irc: "Leaving"
[12:06] <BBB> gruesome
[12:07] <tlg> Hi, I was wondering how har it would be to do a simple use of gstreamer to decode video files with virtualy to simple callback : set_frame and read_yuvframe to get the yuv values of frames and then play the file by doing set_frame(frame++)
[12:08] <dolphy> BBB: ds made a big breakage
[12:08] Action: dolphy kicks ds :)
[12:11] <BBB> tlg: gstreamer doesn't work that way...
[12:11] <BBB> dolphy: /me 2
[12:12] <BBB> gr...
[12:17] Action: BBB fixes snapshot
[12:18] <BBB> how is snapshot being used?
[12:18] <BBB> let's do this differently
[12:18] <BBB> dolphy: tell thomas he may revert my next commit at any time
[12:20] <BBB> and CVS compiles again
[12:20] <BBB> tlg: gstreamer uses pipelines. pipelines stream data internally and the application doesn't need to control the frame flow
[12:20] <BBB> pipelines do that themselves
[12:20] <BBB> apps only set up pipelines
[12:24] <tlg> BBB: ok thx, I'll better have a look at thing like ffmpeg may be
[12:24] <BBB> depends on what you want to do...
[12:29] <tlg> BBB: have multiple video source support (divx,xvid,mpeg,....) and the simple use of libmpeg3
[12:31] <BBB> no, what you want to do, not what features you need
[12:31] <BBB> i.e., why do you need this specific set_frame() API?
[12:33] <tlg> I guess I'll be happy enough with the yuv stuff and have some next_frame() api
[12:33] <tlg> and a seek_frame()
[12:41] yoann ([email protected]) left irc: Read error: 110 (Connection timed out)
[12:42] <BBB> why next_frame()?
[12:42] <BBB> what's wrong with simply receiving YUV frames automatically?
[12:45] alley_cat ([email protected]) joined #gstreamer.
[12:47] sublett ([email protected]) joined #gstreamer.
[12:48] Nick change: rk -> rk|away
[12:52] <tlg> I might need to read the frame more than one time before a next_frame()
[12:55] Rotty ([email protected]) joined #gstreamer.
[12:56] Rotty ([email protected]) left irc: Client Quit
[12:56] Rotty ([email protected]) joined #gstreamer.
[13:02] <BBB> tlg: "might" is not a use case...if you want to repeat a frame because of a pause, then you simply pause the pipeline... if you want to slow down at the same framerate, we've got a plugin that does that for you... if you've got an exact reason for doing it, we might have a solution... "might" is not a use case (I "might" want to delete my whole harddisk, so why doesn't redhat provide a tool for that?)
[13:02] <BBB> but anyway, gstreamer is not oriented at low-level media access
[13:02] <BBB> ffmpeg is
[13:03] <BBB> gstreamer is targetted at desktop/application segment
[13:03] <BBB> you can go low-level, but that'll require more lines of code than the set_frame API you're lookign for
[13:05] <tlg> I did the wrong use of the word "might", I mean that the frame can be grab more than one times per frame
[13:05] <tlg> It is not a limitation
[13:05] <BBB> in gstreamer, you'd use a cache plugin for that
[13:05] <BBB> (or so0
[13:05] <BBB> s/0/)/
[13:06] <tlg> ok I'll have a look at that, thx
[13:26] <dolphy> BBB: wonders why people choose a range like 0.0 to 4.0 for volume element
[13:27] <BBB> float?
[13:27] <BBB> don't know...
[13:28] <dolphy> BBB: and do you why that :
[13:28] <dolphy> filter->volume_f       = g_value_get_float (value);
[13:28] <dolphy>   filter->volume_i       = filter->volume_f*8192;
[13:28] <dolphy> missing a "know"
[13:28] <dolphy> why 8192 ?
[13:30] <BBB> probably a random value
[13:31] <dolphy> lovely
[13:31] <dolphy> hmm i m fed up of doing range conversions 
[13:32] matrixise ([email protected]) left irc: Read error: 110 (Connection timed out)
[13:49] kleppari ([email protected]) joined #gstreamer.
[13:50] <aroedl> How can I watch my favourite AVIs with gst-launch?
[13:51] <aroedl> I have video or audio, but I don't know, how to have both at the same time.
[13:52] <alley_cat> aroedl: take a look at the gst-launch-ext script, it has all you need for the avi pipeline
[13:52] <alley_cat> just replace spider by your decoders, since spider doesn't work too well (at least here)
[13:53] <aroedl> I thought gst-launch-ext is a debugging versionf of gst-launch, but I'll check it out right now.
[13:55] <aroedl> Ah! Impressing!
[13:56] <aroedl> What is colorspace good for?
[13:58] <alley_cat> colorspace conversion
[13:58] <alley_cat> most elements only support some colospaces
[13:59] <aroedl> ic
[14:03] Nick change: rk|away -> rk
[14:04] LeRoutier ([email protected]) joined #gstreamer.
[14:05] <LeRoutier> hello
[14:05] <LeRoutier> BBB: i uploaded a sample .asf file to my server : http://bugzilla.gnome.org/show_bug.cgi?id=131502
[14:08] <BBB> ty
[14:10] <LeRoutier> no prob
[14:11] <LeRoutier> no more warning (because of yesterday night changes) but the "WARNING **: ffdec_wmav2: decoding error" repeted several time while it "plays" the audio file
[14:16] _246tNt_ ([email protected]) joined #gstreamer.
[14:17] <LeRoutier> BBB: not a ffmpeg bug but a bug in the way it is initialized ?
[14:17] <BBB> LeRoutier: probably, yes
[14:17] <LeRoutier> logical as mplayer plays it well using ffmpeg too
[14:18] <BBB> I'll have a look at it...
[14:18] <LeRoutier> well, got around 50 files like that
[14:18] <LeRoutier> only want audio files or also wmv ?
[14:24] iain ([email protected]) joined #gstreamer.
[14:24] <iain> yey 2 exams down one to go
[14:26] <BBB> LeRoutier: video is fine too
[14:28] <LeRoutier> k, uploading 1 more audio (longer) and 1 video (<3MB)
[14:28] <BBB> k, ty
[14:28] <BBB> I'll probably leave the audio for apoc
[14:28] <BBB> he fixed WMAV1 some time agof
[14:31] <LeRoutier> links to those files added to the bug report
[14:33] <LeRoutier> got to work a bit on my everyday project or my maintainer would kill me. bbl
[14:33] LeRoutier ([email protected]) left #gstreamer ("Leaving").
[14:41] thomasvs ([email protected]) joined #gstreamer.
[14:45] thomasvs ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[14:48] thomasvs ([email protected]) joined #gstreamer.
[14:49] thomasvs ([email protected]) left irc: Remote closed the connection
[14:52] matrixise ([email protected]) joined #gstreamer.
[14:52] <teuf> does gstreamer have sink elements to stream data on a network ?
[14:55] <dolphy> yup
[14:55] <dolphy> tcpsink
[14:55] <dolphy> updsink
[14:55] foser ([email protected]) joined #gstreamer.
[15:03] thomasvs ([email protected]) joined #gstreamer.
[15:16] dr88dr88 ([email protected]) joined #gstreamer.
[15:34] KoRnouille ([email protected]) joined #gstreamer.
[15:46] thomasvs ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[15:47] thomasvs ([email protected]) joined #gstreamer.
[15:51] KoRnouille ([email protected]) left irc: "Leaving"
[15:54] Zahir ([email protected]) joined #gstreamer.
[15:54] <Zahir> hi all
[15:58] <dolphy> hola
[16:08] je4d ([email protected]) left #gstreamer.
[16:23] rk ([email protected]) left irc: "leaving"
[16:23] trow ([email protected]) joined #gstreamer.
[16:27] mxpxpod ([email protected]) left irc: "Guns kill people as much as spoons made Rosie O'Donnell fat.."
[16:32] aldug ([email protected]) left irc: "I like core dumps"
[16:44] Action: BBB falls off-chair
[16:49] <dolphy> BBB: commited a big chunk of stuff
[16:49] <BBB> \o/
[16:49] <dolphy> BBB: mixer interface in volume
[16:50] <dolphy> BBB: interface types in v4l v4l2 oss alsa
[16:50] <dolphy> BBB: new flag in mixertrack object for SOFTWARE tracks
[16:51] <BBB> very nice stuff :)
[16:51] <dolphy> BBB: so now i have gst totem searching for a mixer prefering software one
[16:51] <dolphy> BBB: and searching for a colorbalance prefering hardware one
[16:51] <dolphy> BBB: those interfaces thing are very nice now :)
[16:52] Action: BBB cheers
[16:54] <dolphy> BBB: are you working on avidemux ?
[16:54] <dolphy> BBB: avi seeking is broken for some days now
[16:54] <BBB> it worked last time I tried?
[16:54] <BBB> I'll try to fix
[16:55] <BBB> gstplay/player seeks, right?
[16:55] <dolphy> hmm lemme try
[16:57] <dolphy> well my ppc is unable to play matrix.avi anymore with sync :)
[16:57] <dolphy> too hard for it
[16:57] <dolphy> i ll check without videobalance in the pipeline
[16:57] <BBB> videobalance takes a lot of CPU indeed
[16:58] <dolphy> yeah but i thought it was doing a passthrough if values were to default
[16:58] ska-fan ([email protected]) joined #gstreamer.
[16:58] <dolphy> at least that's what ds said
[17:00] <BBB> even then...
[17:00] <BBB> maybe it's not default
[17:13] <dolphy> i ll ask him tonight
[17:14] <BBB> note that it's several math functions per pixel
[17:14] <BBB> that's very expensive
[17:16] <dolphy> yeah i believe you :)
[17:16] <dolphy> but if all levels are at middle
[17:16] <dolphy> it should pass through
[17:16] <dolphy> dunno if there's a check for that
[17:20] <dolphy> BBB: k i checked by myself :)
[17:20] Action: thomasvs hates bbb for all of hte v4l errors to translate
[17:20] <dolphy> BBB: it s never passing through :)
[17:20] <dolphy> thomasvs: oh you here ? :)
[17:20] <thomasvs> yeah
[17:20] <thomasvs> fixing some old bug on the side
[17:22] <BBB> said so... ;)
[17:22] <BBB> thomasvs: at least it gives understandable errors
[17:22] <BBB> understandable for me...
[17:23] <thomasvs> BBB: hehe :)
[17:24] Action: dolphy tries to make videobalance being passthrough when settings are in the middle
[17:25] Uraeus ([email protected]) joined #gstreamer.
[17:25] <gheet> hi Uraeus :)
[17:25] <Uraeus> hi gheet
[17:26] <Uraeus> gheet: you never told me if you could maybe make a Sun Audio plugin for us :)
[17:34] <dolphy> BBB: hmm do you know if there's a way to make real pass through without copying from src to dest being a child of videofilter ?
[17:35] <BBB> say agian?
[17:35] <BBB> I guess if all settings are 0.0 or 1.0
[17:36] <dolphy> found it
[17:36] <dolphy> there's a 
[17:36] <dolphy> if(videofilter->passthru){
[17:36] <dolphy>     gst_pad_push(videofilter->srcpad, GST_DATA (buf));
[17:36] <dolphy>     return;
[17:36] <dolphy>   }
[17:36] <dolphy> in the chain
[17:36] <dolphy> so just setting passthru to TRUE when settings are 0
[17:36] <BBB> and passthru is always FALSE?
[17:37] <BBB> ah
[17:38] <dolphy> passthru is never initialized
[17:38] <dolphy> so i guess it s false by default
[17:38] <dolphy> and childs are responsible of managing it
[17:38] <BBB> it's 0 by default
[17:38] <BBB> objects are inited as g_new0, I think
[17:39] <dolphy> yup
[17:40] <dolphy> i ll add passthru handling in videobalance then
[17:40] <dolphy> and have my ppc breath again :)
[17:48] matrixise ([email protected]) left irc: "Leaving"
[17:50] <thomasvs> [gst-cvs] [thomas@otto gstreamer]$ wc /tmp/gstreamer.patch
[17:50] <thomasvs>    1588    5928   54432 /tmp/gstreamer.patch
[17:51] <thomasvs> [gst-cvs] [thomas@otto gst-plugins]$ wc /tmp/gst-plugins.patch
[17:51] <thomasvs>    3948   16267  163228 /tmp/gst-plugins.patch
[17:51] Action: thomasvs wonders about comitting
[17:51] <thomasvs> anyone want to try the patches ?
[17:52] <Uraeus> thomasvs: what do they do?
[17:52] <thomasvs> Uraeus: implement the new error system
[17:52] <thomasvs> Uraeus: so you can start translating core for example
[17:52] <thomasvs> Uraeus: and test it that way
[17:53] <thomasvs> by trying to trigger the errors
[17:53] <Uraeus> thomasvs: ok, mail em to me and I test
[17:54] Misirlou ([email protected]) left irc: Read error: 110 (Connection timed out)
[17:56] BBB ([email protected]) left irc: Read error: 110 (Connection timed out)
[17:58] swentel ([email protected]) left irc: 
[17:59] foolish ([email protected]) joined #gstreamer.
[18:02] thomasvs ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[18:05] ska-fan_ ([email protected]) joined #gstreamer.
[18:09] gheet ([email protected]) left irc: "Download Gaim [http://gaim.sourceforge.net/]"
[18:19] Nick change: harshyZ -> harshyOut
[18:19] ska-fan ([email protected]) left irc: Read error: 110 (Connection timed out)
[18:20] DrZaius ([email protected]) left irc: "Client exiting"
[18:31] aroedl ([email protected]) left irc: "using sirc version 2.211+KSIRC/1.3.8"
[18:31] <Uraeus> hmm, no gsterror.h in the patch
[18:37] Nick change: trow -> trow_lunch
[18:48] Markey- ([email protected]) joined #gstreamer.
[18:48] <Markey-> re
[19:07] Uraeus ([email protected]) left irc: Read error: 110 (Connection timed out)
[19:13] teuf ([email protected]) left irc: "Client exiting"
[19:19] matrixise ([email protected]) joined #gstreamer.
[19:30] pb_ ([email protected]) left irc: "Client exiting"
[19:44] dolphy ([email protected]) left irc: Remote closed the connection
[19:47] Nick change: trow_lunch -> trow
[20:18] pb_ ([email protected]) joined #gstreamer.
[20:21] mathrick ([email protected]) joined #gstreamer.
[20:21] <mathrick> re
[20:32] ChrisHJW ([email protected]) joined #gstreamer.
[20:41] matrixise ([email protected]) left irc: "Leaving"
[20:43] Misirlou ([email protected]) joined #gstreamer.
[20:45] <ds-work> moo
[20:47] Zeenix ([email protected]) joined #gstreamer.
[20:47] <Zeenix> hi
[20:48] <mathrick> hi Zeenix
[20:48] <mathrick> man, today was HUGE
[20:48] <mathrick> I had 3 classes, and 3 tests :)
[20:51] <mathrick> 6 topics in second try Electronics, Computer Programming, and Digital Circuits
[20:51] <Zeenix> the last three days had been terrible for me, i had been having HD access problems & the problem is further magnified by the fact that it is only with linux
[20:51] dolphy ([email protected]) joined #gstreamer.
[20:52] <foolish> Zeenix: hardware sucks
[20:52] <mathrick> y0 d01phy :)
[20:52] <mathrick> Zeenix: are they HW?
[20:52] Action: ds-work moos at dolphy
[20:53] <mathrick> also, what is the correct term for test you write after failing it first time?
[20:53] <Zeenix> mathrick: yeah it seems, the problem is evidently with my motherboard but it is hard to tell this to my HW vendor as he has no idea (like the rest of the people around here) about linux
[20:53] LeRoutier ([email protected]) joined #gstreamer.
[20:53] <LeRoutier> re
[20:53] <dolphy> ds-work: yo :)
[20:54] <Zeenix> mathrick: the kernel gives access errors the first time it mounts the root FS
[20:54] <dolphy> ds-work: dude videobalance was not passing through :)
[20:54] <mathrick> Zeenix: urk, sounds bad :\
[20:54] <mathrick> hi LeRoutier
[20:59] <Zeenix> anyone having RH please tell me the rpm that includes the wish command
[21:00] <LeRoutier> isn't that a python thing ?
[21:00] <mathrick> Zeenix: what does it do? something like 'wish new hdd'? ;)
[21:00] <ds-work> dolphy: yeah, I realized that after reading the backlog
[21:00] <ds-work> sorry about that
[21:01] <Zeenix> mathrick: no, it runs the amsn 
[21:02] <mathrick> Zeenix: what is amsn?
[21:02] <dolphy> ds-work: no pb
[21:02] Action: mathrick remembers Tk shell named wish
[21:02] <dolphy> ds-work: i need to improve that a little though
[21:02] <dolphy> ds-work: because values might be very close to median 
[21:02] <dolphy> ds-work: and it's not really worth calculating
[21:02] <dolphy> ds-work: it's hard to get the exact median with a slider
[21:02] <Zeenix> mathrick: yeah thats it
[21:03] <mathrick> Zeenix: then you should try Tk IIRC
[21:03] <ds-work> dolphy: that sounds like a UI problem :)
[21:04] <dolphy> ds-work: hehe true
[21:04] water ([email protected]) left #gstreamer ("The water has parted.").
[21:09] <trow> dolphy: I've been trying to figure out why seeking doesn't work in GstPlay, and part of the problem seems to be that the clock on the pipeline is weird.
[21:10] <trow> dolphy: It takes on very large values (i.e. 1074197097518163727) and seems to be in different units than the seek times we are requesting.
[21:11] Action: mathrick thinks there shoud be no sliders w/o assciated spinbuttons in UI
[21:12] <dolphy> trow: i ve seen strange tick sometimes that's true
[21:12] <dolphy> trow: dunno why yet though
[21:12] <dolphy> trow: haven't been able to look into that yet
[21:15] <dolphy> ds-work: have you found anything about switch ?
[21:16] <ds-work> dolphy: I still don't really understand the problem
[21:20] <dolphy> ds-work: if i connect 2 videotestsrc it works just fine
[21:20] <dolphy> ds-work: if i connect spider to it 
[21:20] <dolphy> ds-work: it breaks the picture
[21:20] <dolphy> ds-work: even with only spider
[21:20] <dolphy> ds-work: the picture is green
[21:21] <mathrick> dolphy: wrong colorspace negotiated?
[21:22] <dolphy> mathrick: probably
[21:23] <mathrick> dolphy: what exactly is negoed? what do you use - xvimagesink or ximagesink?
[21:23] <dolphy> mathrick: the same happens for both
[21:25] <mathrick> dolphy: hmm, are you able to link videotestsrc ! switch ! ximagesink ?
[21:25] <mathrick> dolphy: i.e. w/o using colorspace in between?
[21:26] <dolphy> mathrick: yes that works
[21:27] <mathrick> dolphy: hmm, right, it works. For me adding spider gives black window in case of ximagesink tho
[21:27] <mathrick> and xvimagesink is all green indeed
[21:28] <dolphy> mathrick: filesrc location="/home/dolphy/south.mp3" ! spider ! goom ! switch ! ffcolorspace ! ximagesink
[21:28] <dolphy> mathrick: try thaat
[21:28] <dolphy> with and without switch
[21:28] <dolphy> you will see
[21:29] walters ([email protected]) joined #gstreamer.
[21:32] <mathrick> mathrick@megumi:~$ gst-launch-0.7 filesrc location="/home/mathrick/Desktop/Stellvia OST/Uchu no Stellvia Original Soundtrack - 01 - Asu e no Brilliant Road -TV Size Version-.mp3" ! spider ! goom ! switch ! ffcolorspace ! ximagesink
[21:32] <mathrick> RUNNING pipeline
[21:32] <mathrick>  
[21:32] <mathrick> (process:1708): GStreamer-CRITICAL **: Could not find a compatible pad on element id3tag0 to link to src_0:sink
[21:32] <mathrick> this is for tagged ones
[21:32] <mathrick> mathrick@megumi:~$ gst-launch-0.7 filesrc location="/mnt/win/Misc/Muza/ost/Vampire Princess Miyu TV/Sub title.mp3" ! spider ! goom ! switch ! ffcolorspace ! ximagesink
[21:32] <mathrick> RUNNING pipeline
[21:32] <mathrick>  
[21:32] <mathrick> (process:1739): GStreamer-CRITICAL **: file gstdata.c: line 236 (gst_data_unref): assertion `GST_DATA_REFCOUNT_VALUE (data) > 0' failed
[21:32] <mathrick> and this for untagged
[21:32] Action: mathrick should probably update checkout
[21:33] <mathrick> yo walters
[21:34] <trow> mathrick: I get that warning all of the time when I try to do relative seeks w/ totem.
[21:35] <trow> (the refcounting warning, not the 'no compatible pad' warning)
[21:37] <walters> hey.
[21:40] <mathrick> trow: maybe it's just some forgotten _ref somewhere. IIRC, there are some options to disable GLib's cached malloc's, this could make it nice crasher easier to debug
[21:41] <mathrick> trow: also, poisoning seems like good idea
[21:43] desrt ([email protected]) joined #gstreamer.
[21:44] <desrt> does anyone know what's up with spider in 0.6?
[21:44] <mathrick> desrt: AFAIK it's to be considered broken, or at least heavily suboptimally working
[21:45] <desrt> that's sort of the impression i'm getting, too :/
[21:46] <mathrick> desrt: are you putting typefind in front of spider? it may help sometimes (at least in 0.7 it does)
[21:46] <desrt> ! typefind ! spider?
[21:47] <mathrick> desrt: yep
[21:47] <desrt> ERROR: pipeline could not be constructed: Could not link element typefind0 to spider0
[21:47] <mathrick> desrt: full pipeline?
[21:47] <desrt>  gst-launch-0.6 gnomevfssrc location=file:///mnt/media/music/albums/sleater-kinney/one\ beat/sleater-kinney\ -\ one\ beat\ -\ 01\ -\ one\ beat.ogg ! typefind ! spider ! osssink
[21:48] <mathrick> gst-launch-0.6 -v -i1 gnomevfssrc location=file:///mnt/media/music/albums/sleater-kinney/one\ beat/sleater-kinney\ -\ one\ beat\ -\ 01\ -\ one\ beat.ogg ! typefind ?
[21:48] <desrt> Error on option -i1: unknown option.
[21:49] <mathrick> desrt: hmm, seems it's not there in 0.6, just omit it
[21:49] <desrt> ERROR: /pipeline0/typefind0: typefind could not determine type after 1 buffers
[21:49] <desrt> is that what you wanted to see?
[21:49] <mathrick> desrt: tagged?
[21:49] <desrt> mathrick; yes.
[21:49] <mathrick> desrt: kind of
[21:49] <desrt> encoded using xiph.org oggend
[21:49] <desrt> *oggenc
[21:50] <mathrick> desrt: then it's probably the cause. Tagging and typefind in 0.6 don't go along very well
[21:50] <desrt> perhaps i should just go with 0.7?
[21:50] <ds-work> heh
[21:50] <mathrick> desrt: try gst-launch-0.6 -v gnomevfssrc location=file:///mnt/media/music/albums/sleater-kinney/one\ beat/sleater-kinney\ -\ one\ beat\ -\ 01\ -\ one\ beat.ogg ! typefind max-buffers=0
[21:51] <ds-work> I just spent 30 minutes trying to figure out why 'videotestsrc ! xvimagesink' didn't work
[21:51] <desrt> 0 litreally does 0 buffers
[21:51] <ds-work> and it was because DISPLAY=:1
[21:51] <desrt> *literally
[21:51] <desrt> :)
[21:51] <mathrick> desrt: no, it means unlimited
[21:51] <desrt> i think you're wrong...
[21:52] <desrt> ERROR: /pipeline0/typefind0: typefind could not determine type after 0 buffers
[21:52] <desrt> (really quick return.. almost instant)
[21:52] <mathrick> ds-work: yesterday i wasted all day debugging valid code, b/c gdb gone crazy
[21:52] <desrt> whereas if i put 200 it takes a while before bombing and prints this out a lot of times: ** (process:23179): WARNING **: factory->typefindfunc set to _dummy, not recursing, FIXME
[21:53] <mathrick> desrt: personally, i'd just recommend CVS HEAD, it's really usable now, more than 0.6 ever was
[21:53] <desrt> honestly, i'm using 0.6 for 2 reasons:
[21:53] <desrt> (1) it's in portage stable
[21:53] <desrt> (2) it prints out pretty colours on startup
[21:54] <desrt> so i think i'll give 0.7 a go
[21:54] <mathrick> desrt: you will quickly get used to no colors, and it's easier when dumping into file anyway
[21:55] <desrt> i was sort of kidding about the colours :)
[21:55] <mathrick> desrt: and I believe CVS HEAD isn't in portage ;)
[21:55] <mathrick> desrt: yeah, i know :)
[21:55] <desrt> i can put it in portage :)
[21:55] trow ([email protected]) left irc: "Leaving"
[21:56] <mathrick> desrt: hmm, why not, it breaks rather rarely now ;)
[21:56] <alley_cat> depends what you define as breaking ;)
[21:58] <mathrick> alley_cat: works way more often it did month ago ;)
[21:58] <alley_cat> gst-launch is great in HEAD, but gst-player for 0.6 is much more usable as player than current one
[21:59] <mathrick> alley_cat: I use gst-launch as my player ;)
[21:59] Action: alley_cat needs seeking to test some stuff right now
[21:59] <mathrick> alley_cat: and honestly, i didn't try new gst-player yet
[21:59] <alley_cat> i hope updating core and plugins will restore avi seeking
[21:59] <mathrick> alley_cat: but dolphy is working hard on it, it's good omen :)
[22:00] <alley_cat> since i don't want to wait 38 minutes just to test if the file plays correctly ;)
[22:01] <desrt> i'm using gstreamer in my own media player
[22:01] <desrt> so 0.7 is probably the way to go, anyway
[22:01] <desrt> i stopped using it a while ago because spider was totally broken
[22:02] somex1 ([email protected]) joined #gstreamer.
[22:02] <mathrick> desrt: Company has writing new autoplugger scheduled :)
[22:02] <desrt> :)
[22:03] <desrt> so spider will be broken in 0.7 too?
[22:03] <mathrick> desrt: current spider sucks, too, but typefind can help. so it mostly works ok
[22:03] <desrt> i see
[22:03] <desrt> can i feed it a mime type somehow?
[22:03] <alley_cat> spider alone doesn't work at all here
[22:03] <desrt> like...
[22:03] <desrt> gnomevfssrc location=file:/// mime_type=[whatever]
[22:03] <desrt> or actually
[22:03] <mathrick> desrt: who knows :), depends on Company's motivation to write it
[22:03] <desrt> why doesn't gnomevfs tell it the mimetype?
[22:04] <mathrick> desrt: yep, you can specify filtered caps
[22:04] <desrt> gnomevfs supports mime type detection (from filecntents and extension)
[22:04] <mathrick> desrt: it's useless for us anyway
[22:04] <desrt> well...
[22:05] <desrt> vorbisfile knows that it can take a application/ogg and convert it to a raw stream, right?
[22:06] <ds-work> yes
[22:06] <desrt> i suspect i don't understand enough of what's going on inside :)
[22:06] <desrt> but i do know that gnomevfs will tell you the mime type of the file that you're reading
[22:06] <desrt> (and do so rather reliably)
[22:06] <walters> desrt: yeah, we don't use gnome-vfs mimetypes though at the moment.
[22:07] <walters> desrt: it was tried, but there were incompatibilites
[22:07] <alley_cat> and mimetype only tells you about the container, not the content
[22:08] <desrt> alley_cat; true.
[22:08] <mathrick> desrt: fyi, filtered caps look like elem ! mime/type,property1=val,prop2=val2 ! elem
[22:10] <desrt> can you give a concrete example?
[22:10] <desrt> like say i wanted to force to mime type application/ogg
[22:10] <alley_cat> audioscale ! audio/x-raw-int,rate=48000 ! alsasink device=dsp1
[22:10] <alley_cat> that's my default audiosink :)
[22:12] Uraeus ([email protected]) joined #gstreamer.
[22:12] <Uraeus> g'evening
[22:12] <mathrick> hi Uraeus :)
[22:12] <desrt> i need to start using alsa :)
[22:12] <desrt> s/alsa/2.6/
[22:13] Action: ds-work gets annoyed at GStreamer's core macros
[22:14] herzi ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[22:14] matrixise ([email protected]) joined #gstreamer.
[22:14] Action: mathrick will probably wait till 2.6.2
[22:16] Action: alley_cat switched with test10
[22:18] lypanov ([email protected]) joined #gstreamer.
[22:18] <lypanov> ugh
[22:18] <lypanov> wheels didn't get the story done or what?
[22:18] <lypanov> i should have just done it...
[22:18] tlg ([email protected]) left irc: "Client exiting"
[22:18] <Uraeus> lypanov: doesn't seem like it :(
[22:19] <lypanov> that really sucks
[22:19] m_wheels ([email protected]) joined #gstreamer.
[22:19] <lypanov> m_wheels: story?
[22:19] Nick change: m_wheels -> wheels
[22:20] Action: mathrick wonders if it's worthy trying Mono/C#
[22:20] <mathrick> Hallski seems pretty impressed
[22:20] <Uraeus> mathrick: if you do you can try out Muine and tell us what you think :)
[22:20] <Markey-> heyho lypie & wheelie
[22:21] <LeRoutier> bbl
[22:21] <dolphy> Uraeus: yo chris :)
[22:21] LeRoutier ([email protected]) left irc: "Leaving"
[22:22] <lypanov> hey Uraeus 
[22:22] <lypanov> mathrick: mono is funky
[22:22] <mathrick> lypanov: what exactly is funky about it?
[22:23] <Uraeus> dolphy: just sent you the latest logo version :)
[22:23] <lypanov> mathrick: it works
[22:23] <lypanov> mathrick: thats pretty darn amazing considering what it does
[22:23] Action: dolphy check mails
[22:23] <lypanov> hey dolphy 
[22:23] <mathrick> lypanov: ahh, I thought you mean funky as in 'having lotsa quirks' :)
[22:24] <lypanov> heh. no
[22:24] <lypanov> i like the project
[22:24] Action: Uraeus has a really bad cold, I think I crawl back to bed again
[22:24] Uraeus ([email protected]) left irc: "Client exiting"
[22:24] ska-fan_ ([email protected]) left #gstreamer.
[22:25] <dolphy> lypanov: hi :)
[22:26] <mathrick> lypanov: all that .NET things is kinda strange for me... I sort of know what it's meant to do, but can't really feel it :)
[22:26] foolish ([email protected]) left irc: Remote closed the connection
[22:27] <mathrick> lypanov: it's probably b/c i missed it's actual adoption after all that initial hype... so for me it's still one more M$ revolutionary vaporware ;)
[22:28] <lypanov> hehe :)
[22:30] Rotty ([email protected]) left irc: "Bye"
[22:37] Company ([email protected]) joined #gstreamer.
[22:37] <Company> good morning angels
[22:37] <mathrick> yo Company
[22:38] <dolphy> Company: hi
[22:38] <dolphy> Company:  i have some small change to alsamixer.c
[22:38] <dolphy> Company: to set the interface type being HARDWARE 
[22:38] <Company> dolphy: go ahead
[22:38] <dolphy> Company: it s in already :)
[22:38] <Company> i haven'T written alsamixer anyway, that was leif :)
[22:41] LeRoutier ([email protected]) joined #gstreamer.
[22:42] <dolphy> LeRoutier: re
[22:42] <LeRoutier> s'lut dolphy
[22:44] <LeRoutier> and bye
[22:44] LeRoutier ([email protected]) left irc: Client Quit
[22:48] matrixise ([email protected]) left irc: "Leaving"
[23:12] <Company> "Do you think it's possible to take the code of GStreamer/GLib and reassemble it in this way ? Instead of starting from scratch ?"
[23:12] <Company> muahahaha
[23:13] Action: Company unsubscribes from media-api
[23:14] <ds-work> ?
[23:14] <Company> are you subsribed to media-api list?
[23:14] <ds-work> no
[23:15] <Company> http://lists.matroska.org/pipermail/media-api/2004-January/000201.html
[23:17] <ChrisHJW> Company : this guy is not representing our opinion
[23:17] <Company> i know
[23:18] <ChrisHJW> robux4 likes the idea more and more to port gstreamer to win32
[23:18] <ChrisHJW> we love you guys :)
[23:18] <Company> but i've had enough of those guys arguing about that stuff and cluttering my inbox :)
[23:18] <ChrisHJW> lol
[23:18] <ds-work> er, uh, "don't use GObject"
[23:18] <ChrisHJW> Company : i use gmane.org and my newsreader since ages
[23:19] <ChrisHJW> news.gmane.org gmane.comp.multimedia.media-api
[23:20] dolphy ([email protected]) left irc: "Network down, IP Packets delivered via UPS"
[23:23] <ds-work> what a turd
[23:28] <NFusi0n> turd.
[23:32] <taaz> ds-work: i see you added the current-level-buffers notify code.  should that be in the _get func too?  or fire when reset to 0?
[23:32] <ds-work> oops, that should have been removed
[23:33] <taaz> that was my other guess ;)
[23:34] <ds-work> I think we've decided not to call g_object_notify() for every buffer
[23:34] <taaz> why the [1.0 100.0] framerate limits?
[23:34] <ds-work> because we can't really display framerates outside that range
[23:35] Nick change: Markey- -> markov
[23:35] <ds-work> so there's no need to advertise it
[23:39] <taaz> you know someone will end up writing a slow-mo or speed up element just to prove you wrong ;)
[23:40] mxpxpod ([email protected]) joined #gstreamer.
[23:40] thaytan ([email protected]) joined #gstreamer.
[23:41] trow ([email protected]) joined #gstreamer.
[23:42] ChrisHJW ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[23:47] dr88dr88 ([email protected]) left #gstreamer.
[23:48] sublett ([email protected]) left irc: "I like food, food is good!"
[23:59] lypanov ([email protected]) left #gstreamer ("ciao all").
[00:00] --- Fri Jan 16 2004
[00:06] lmjohns3 ([email protected]) joined #gstreamer.
[00:06] <lmjohns3> hi hi
[00:06] <lmjohns3> Company: ping
[00:06] <Company> lmjohns3: pong
[00:06] <lmjohns3> Company: excellent ! i was just wondering how you all generate the changelog entries for your cvs commits.
[00:08] <Company> lmjohns3: www.gnome.org/tools/scripts
[00:08] <Company> there's a preparechangelog.pl script
[00:11] <lmjohns3> excellent --- thanks
[00:13] <lmjohns3> do i have to run this in the root source directory ?
[00:14] <iain> Company: is that what we're supposed to be running? I've been making my changelogs by hand
[00:15] <lmjohns3> hm, also, does it add the entry to ./ChangeLog, and then i have to copy/paste it into my commit message ?
[00:17] <Company> lmjohns3: yes, tree root dir
[00:17] <Company> iain: no, do it as you wish, it's just nice to do it that way
[00:17] walters ([email protected]) left irc: "out"
[00:18] <Company> lmjohns3: yes, use cvs commit -m "`head -n10 ChangeLog`" (or whatever number of lines you added)
[00:18] <lmjohns3> Company: ok, thanks
[00:19] <ds-work> I have a script that commits the files in the first changelog entry, with the entry as the message
[00:19] <ds-work> I'll post it to the list in the next few days
[00:19] <lmjohns3> have any of you tried using the alsa mixer interface ? uraeus wrote me and said there have been reports of problems.
[00:20] <lmjohns3> i can't test atm though because i'm running through putty on a windoze box. :(
[00:20] <Company> it doesn't work at all here (gst-player and gst-mixer don't find anything), but i haven't looked
[00:21] <iain> yeah, it doesn't work here either, but I've got an extigy
[00:21] <iain> so I never expect alsa to work
[00:21] <lmjohns3> ok
[00:22] <lmjohns3> iain: why doesn't the extigy work ?
[00:22] <iain> oh, all sorts of reasons
[00:22] <iain> funny usb problems, the fact that it can only output in 48000...
[00:22] jdahlin ([email protected]) joined #gstreamer.
[00:23] <lmjohns3> hm. does the alsa project claim to support it ?
[00:23] <iain> yup
[00:23] <lmjohns3> iain: have you messed around with the .asoundrc stuff any ?
[00:24] <iain> yes, but I've never understood it
[00:24] <lmjohns3> me either :/
[00:25] jimmy_dean ([email protected]) joined #gstreamer.
[00:25] <iain> I now have some random assed thing that might use dmix and the plug interface to do everything, but I have no idea what its about
[00:26] <jimmy_dean> I'm trying to follow the CVS-HEAD version of the plugin tutorial...however I don't see the gst-template/tools directory which is support to contain the tools for making the boilerplate...what happened to these tools?
[00:26] <jimmy_dean> s/support/supposed
[00:27] <lmjohns3> jimmy_dean: they're still under development ... i worked on them a while back but never got anything done with them. sorry ...
[00:27] <jimmy_dean> lmjohns3: oh, so should I just skip that section in the tutorial?
[00:27] <lmjohns3> jimmy_dean: i could send you the plugin stamp script, but it might be easier to copy an existing plugin and do global search/replace.
[00:28] <jimmy_dean> lmjohns3: ok, does the script just copy the template code and set some macros equal to the boilerplate values?
[00:28] <ds-work> jimmy_dean: I just added a make_element script that I've used locally for months
[00:29] <lmjohns3> jimmy_dean: yes, it's basically a perl script that does search/replace
[00:29] <lmjohns3> ds-work: did you write a canonical audio element somewhere in gst-libs ?
[00:29] <jimmy_dean> lmjohns3: does the template contain the macros defined already?  Will I see these as obvious if I look at the source code?
[00:29] <jimmy_dean> ds-work: where's that script at?
[00:29] <lmjohns3> jimmy_dean: yes, they're pretty obvious, and seem to be standard for all glib type projects.
[00:30] <ds-work> lmjohns3: you mean for audiofilter?
[00:30] <ds-work> jimmy_dean: gst-plugin/tools/
[00:30] <jimmy_dean> lmjohns3: ok, well I'll compare the template to a working plugin then...or maybe even the filter plugin example code
[00:31] <jimmy_dean> ds-work: anonymous cvs must be slow to update then, there is no tools directory
[00:32] <lmjohns3> jimmy_dean: your best bet is to look at a simple existing plugin, probably passthrough is the least complex right now, but i don't know if it's a great example at this point
[00:32] <ds-work> yes, it is
[00:32] <jimmy_dean> lmjohns3: ok, well I'm trying my best attempt at making an audio equalizer plugin
[00:32] <jimmy_dean> lmjohns3: so I was thinking of looking at the volume plugin
[00:32] <lmjohns3> jimmy_dean: yeah, or the level plugin might be good too
[00:33] <jimmy_dean> lmjohns3: what does the level plugin do exactly?
[00:33] <jimmy_dean> lmjohns3: does it adjust the gain?
[00:36] <lmjohns3> jimmy_dean: no, i think it's just a passive sensor analysis element
[00:36] kleppari ([email protected]) left irc: "Leaving"
[00:36] <lmjohns3> ds-work: yeah, is the gstaudiofilter stuff up to date ?
[00:37] <ds-work> lmjohns3: should be.  there's no subclass example, iirc
[00:37] <jimmy_dean> lmjohns3: ok, so in working with gstreamer, what form will I be presented the data in from gstreamer...will it be in the discrete time domain chunks of data or already in discrete frequency chunks?
[00:37] <ds-work> ah, there is
[00:38] <lmjohns3> jimmy_dean: normally audio comes in time-domain chunks
[00:38] <jimmy_dean> lmjohns3: ok, excellent :)
[00:38] <jimmy_dean> lmjohns3: that's what I was hoping for
[00:39] <jimmy_dean> lmjohns3: do you think there'd be a use in making a plugin that has only one purpose...to take the audio in the time domain and transform it to the frequency domain?
[00:39] <jimmy_dean> lmjohns3: other plugins might find that useful
[00:39] <lmjohns3> ds-work: would it make sense to make plugins like passthrough a subclass of the audiofilter ?
[00:40] <ds-work> passthrough should be killed
[00:40] <lmjohns3> jimmy_dean: sure, you'd have to make a different set of caps to represent that ... check out gst-libs/gst/audio/audio.h for the current normal audio caps
[00:40] <lmjohns3> ds-work: well, yeah. maybe volume or speed then ?
[00:41] <jimmy_dean> lmjohns3: ok, I think I might try writing the basic equalizer first but then maybe separate out that functionality later...gstreamer is a whole new beast for me to learn about :)
[00:41] <ds-work> not speed
[00:41] <ds-work> but volume, yes
[00:41] <Company> what are we gonna do with dparams btw?
[00:41] <Company> especially because they need to be threadsafe probably
[00:41] <lmjohns3> so i made rather large changes to the gstreamer_and_midi file ... do i need to do a normal changelog thing for that ?
[00:42] <ds-work> Company: pretend they don't exist :)
[00:42] <Company> (that was a long term question)
[00:42] <mathrick> ds-work: what's wrong w/ dparams?
[00:42] <lmjohns3> Company: it would be nice to get some MIDI<->dparams integration
[00:44] markov ([email protected]) left irc: "nity"
[00:46] <Company> what would you use that for?
[00:47] <ds-work> mathrick: nothing
[00:47] Zeenix ([email protected]) left irc: "Client Exiting"
[00:49] <mathrick> ds-work: hmm, seems like i didn't understand Company's q then :)
[00:50] <ds-work> mathrick: there are threading issues with dparams
[00:51] <ds-work> it's sort of a catch-22.  GStreamer would have never gotten started if it had to be thread-safe from the ground up
[00:51] <ds-work> but now, it's hard to make it thread-safe
[00:52] <ds-work> we should throw it all away and take bits and pieces of GStreamer/Glib and create a new project called media-api
[00:52] <ds-work> (that was a joke, if you didn't get it)
[00:55] <mathrick> ds-work: heh, I skimmed through media-api archives, these guy(s) is clinical case of heavy NIH syndrome...
[00:56] <ds-work> Log message:
[00:56] <ds-work> `head -5 ChangeLog `
[00:56] <ds-work> cute
[00:56] <mathrick> and MDAI syndrome too
[00:56] walters ([email protected]) joined #gstreamer.
[00:56] <lmjohns3> doh ! is that my commit message ?
[00:56] <mathrick> ds-work: yep, I wonder how did he manage to put it in ;)
[00:57] <lmjohns3> *sigh*
[00:59] <Company> you could try to use echo ?`head -5 ChangeLog " first ;)
[00:59] <Company> echo "`head -5 ChangeLog " too
[01:00] <mathrick> Company: echo "`head -5 ChangeLog `" ? ;)
[01:00] <lmjohns3> Company: i did, i did ... but then i used single quotes :-/
[01:01] <mathrick> lmjohns3: commit message is with backticks... that's what's intriguing :)
[01:01] <mathrick> lmjohns3: er, nvm
[01:01] <lmjohns3> argh :)
[01:03] steve_b ([email protected]) joined #gstreamer.
[01:04] <mathrick> btw, did guys on media-api agreed on creating codec api over full-blown framework yet? They seemed pretty much indecisive, and I don't feel like reading all that flamage now ;)
[01:05] <steve_b> thaytan, hey - do you want the mpeg of my preso to see whats wrong with it
[01:06] <Company> mathrick: media-api is essentially useless, the people on there are not the important ones
[01:07] <mathrick> Company: yeah, I know, but it's kinda funny to see attempts to reimplement Glib/Gst as a side effect of writing codecs :)
[01:08] <iain> and the ones that are on there say things like "Well, everyone's going to have to use our API cos we're not changing"
[01:08] <ds-work> I'm still trying to avoid using glib in liboil, but it's getting increasingly hard
[01:08] <Company> ds-work: do it like pkg-config, keep a local copy in your tree ;)
[01:09] <ds-work> Company: I think I may settle for using glib internally, but no exported interface should use glib
[01:09] <walters> yeah, glib's actually pretty minimal, realistically speaking
[01:09] <mathrick> and basically, most of them seem not to differentiate between codec api and framework, and that's rather fundamental mistake
[01:09] <walters> i mean, when you start out, you don't need all of it - but by the time you get to the end, you've basically reimplemented glib
[01:10] <mathrick> yeah, glib's all that libc lacks for real world apps
[01:10] <Company> na, glib is much to big, embedded developers despise it and it should be split up into chunks
[01:11] <mathrick> ds-work: is there any info on what liboil is?
[01:11] <ds-work> glib totally rocks for embedded developers
[01:11] <Company> like glib-hashtable and glib-queue and glib-list probably
[01:11] <ds-work> mathrick: no
[01:11] <mathrick> ds-work: so, what is it? :)
[01:11] <Company> mathrick: it's a top secret project and the first one i named ;)
[01:11] <Company> assembler optimizations
[01:11] <ds-work> glib puts all of that junk that normally gets repeated 100 times in the code into one place
[01:11] <mathrick> ds-work: library of optimised instructions, right?
[01:11] <Company> optimized inner loops
[01:11] <Company> oil
[01:12] <mathrick> Company: and what does that mean?
[01:12] <iain> oh is that what liboil is
[01:12] <Company> mathrick: that it only optimizes inner loops, not outer ones :)
[01:12] <iain> I was expecting it to be liboil to fry libegg and libbacon in
[01:12] <lmjohns3> mmm, bacon
[01:12] <mathrick> Company: ok, but I still don't know what it's supposed to do :)
[01:13] <mathrick> iain: ROTFL
[01:13] <Company> mathrick: it's just a collection of functions/loops that are often used
[01:13] <Company> mathrick: and they are heavily assembler optimized
[01:13] <mathrick> iain: it's the second funniest thing today, after Hallski's Blam! announce :)
[01:14] <mathrick> Company: example?
[01:14] <iain> mathrick: really? thats what I seriously thought it meant
[01:14] <iain> mathrick: that announcement was good mind :)
[01:14] <Company> mathrick: conversion (int16 to float), idct, colorspace conversions, ...
[01:14] <mathrick> Company: ahh, things like that
[01:14] <lmjohns3> ds-work: in the gstaudiofilter, who calls gst_audiofilter_class_add_pad_templates ? is that supposed to be called by subclasses ?
[01:15] <ds-work> lmjohns3: hold on, I'm writing an example
[01:15] <lmjohns3> ds-work: nice :)
[01:18] <Company> does anyone of you remember how that book was called that omega was reading in Norway?
[01:19] <lmjohns3> Company: do you remember what it was about ?
[01:19] <Company> lmjohns3: it was about hacker tricks
[01:20] <ds-work> Company: there's a books section on the gstreamer web page somewhere
[01:21] <Company> ds-work: cool, thx
[01:24] Action: Company just rewrote state changes to be more threadsafe
[01:24] <Company> s/rewrote/fixed
[01:24] <Company> phew, BBB's not here
[01:24] <iain> you love the word rewrote :)
[01:25] <alley_cat> hmm, is it normal that gstreamer apps don't show in top, or is that a "feature" of linux 2.6/NPTL glibc?
[01:25] <Company> fixing is what you do with opt
[01:25] <ds-work> heh.  I actually rewrote much of colorspace last night
[01:28] <steve_b> hey, i think i've convinced libsndfile author to virtualise his io
[01:28] <steve_b> so we can use it for autoplugging
[01:29] <ds-work> that reminds me (somehow) -- I need to write a perl script replacement for sox
[01:34] pb_ ([email protected]) left irc: "Client exiting"
[01:35] <thaytan> hi, all
[01:43] <lmjohns3>  Company: ok, but I still don't know what it's supposed to do :)
[01:43] <lmjohns3> 16:12 <mathrick>  Company: ok, but I still don't know what it's supposed to do :)
[01:43] <lmjohns3> 16:12 <mathrick> oops, sorry
[01:44] <lmjohns3> cheers all, gotta run
[01:44] lmjohns3 ([email protected]) left irc: "leaving"
[01:46] steve_b ([email protected]) left irc: Read error: 110 (Connection timed out)
[01:53] <mathrick> nite
[01:54] Nick change: mathrick -> mathrick|sleep
[02:16] jimmy_dean ([email protected]) left irc: Read error: 110 (Connection timed out)
[02:26] jimmy_dean ([email protected]) joined #gstreamer.
[02:28] iain ([email protected]) left irc: "and we sat on our own star and dreamed of the way that we were and the way that we wanted to be"
[02:28] foser ([email protected]) left irc: "[ I want to believe ]"
[02:32] sub_pop ([email protected]) left irc: "Client exiting"
[02:39] water ([email protected]) joined #gstreamer.
[02:42] water ([email protected]) left irc: Client Quit
[02:43] water ([email protected]) joined #gstreamer.
[02:45] Zahir ([email protected]) left irc: "adeu"


-------------------------------------------------------
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