IRC Logs

*******************************************************************
[03:06] jonathang ([email protected]) left irc: "Leaving"
[03:24] Nick change: jcsston` -> jcsston
[03:36] sub_pop ([email protected]) joined #gstreamer.
[03:45] Zeenix ([email protected]) left irc: "The whole problem with the world is that fools and fanatics are always so certain of themselves, but wiser people so full of 
[04:20] Marsupilami23 ([email protected]) joined #gstreamer.
[04:21] joem ([email protected]) left irc: "Leaving"
[04:27] joem ([email protected]) joined #gstreamer.
[05:25] harshy ([email protected]) joined #gstreamer.
[05:31] mxpxpod ([email protected]) left irc: "To quit or not to quit.. wait... was that a question?"
[05:43] jcsston ([email protected]) left irc: Read error: 60 (Operation timed out)
[05:50] jcsston ([email protected]) joined #gstreamer.
[06:04] trow ([email protected]) left irc: "Leaving"
[06:12] joem ([email protected]) left irc: Remote closed the connection
[06:19] joem ([email protected]) joined #gstreamer.
[07:03] chreekat ([email protected]) joined #gstreamer.
[07:03] chreekat ([email protected]) left #gstreamer.
[07:13] Rotty ([email protected]) joined #gstreamer.
[07:24] Rotty ([email protected]) left irc: Remote closed the connection
[07:36] markey ([email protected]) joined #gstreamer.
[08:13] mathrick|sleep ([email protected]) left irc: Remote closed the connection
[08:38] Uraeus ([email protected]) joined #gstreamer.
[08:38] <Uraeus> morning
[08:39] <thaytan> hi :)
[08:40] <Uraeus> thaytan: have you seen Martin's DVD code?
[08:47] Action: Uraeus watch Quicktime movie with gst-player :)
[08:50] <thaytan> some of it
[08:50] <thaytan> dvdnavsrc changes only
[08:50] <thaytan> I want to see more so I can get mpeg2subt working with it
[08:51] <Uraeus> thaytan: I was just thinking that since he still haven't CVS access afaik maybe you should mail asking him for his diff and merge what you think should be merged etc.?
[08:51] <markey> morning
[08:52] <thaytan> I think I'd rather see him get CVS access so we've got a better chance of keeping him actively developing :)
[08:52] Nick change: apoc_ -> apoc
[08:52] <apoc> morning
[08:53] <thaytan> yo
[08:54] <Uraeus> morning apoc, i just played the Shrek.mov file in gst-player successfully :)
[08:54] <Uraeus> thaytan: yeah, I try and remember to nag thomas about it when he comes online
[08:54] <Uraeus> morning markey
[08:55] <apoc> Uraeus: cool
[08:55] <apoc> Uraeus: does qtdemux support seek event now ?
[08:56] <Uraeus> apoc: no
[09:01] <apoc> monkeysaudio plugin is almost ready here, I can encode an ape file to flac format with tag tanscoding 
[09:02] <Uraeus> cool :)
[09:06] Action: Uraeus is in QA mood, filing tons of bugs on things not working 100% :)
[09:11] <Uraeus> apoc: I have a non-drm wmv file if you want a testcase for that
[09:12] <dolphy> morning
[09:14] <Uraeus> morning dolphy
[09:14] md` ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[09:14] <Uraeus> dolphy: I have some Totem problems. It starts then (as it seems to be initalizing the video) is segfaults for me
[09:14] md` ([email protected]) joined #gstreamer.
[09:14] <dolphy> bt
[09:14] <Uraeus> dolphy: gst-player works however
[09:15] <Uraeus> dolphy: bug-buddy refuse to 'see' totem crashing, how do I get a bt without using bug-buddy? running totem in gdb?
[09:15] <dolphy> yep
[09:15] <dolphy> gdb totem
[09:15] <dolphy> run
[09:16] <dolphy> and then when it crashed
[09:16] <dolphy> bt
[09:16] <thaytan> or thread apply all bt
[09:16] <thaytan> yo dolphy
[09:16] <dolphy> thaytan: heya
[09:17] <thaytan> what do you think is the best way to get libgstplay to build dvdnavsrc ! mpegdemux name=demux ! mpeg2dec ! mpeg2subt name=sub demux. ! sub. when playing dvd?
[09:17] <thaytan> just hardcode the pipeline?
[09:18] <thaytan> perhaps replacing mpeg2dec with spider
[09:18] <thaytan> (working on mpeg2subt atm)
[09:20] mxpxpod ([email protected]) joined #gstreamer.
[09:21] <dolphy> hmm
[09:22] <dolphy> no way to autoplug that?
[09:22] <thaytan> it builds most of the pipeline automatically
[09:22] <thaytan> but won't feed the video and subtitles through mpeg2subt to do subtitle overlay
[09:22] camh ([email protected]) left irc: "Client exiting"
[09:23] <apoc> Uraeus: sure 
[09:24] <Uraeus> apoc: http://www.troll-mountain.no/video/bred_1.wmv
[09:24] <apoc> Uraeus: ok I'll have a look
[09:25] <Uraeus> apoc: for me it gst-player just plays sound (and much delayed)
[09:25] <dolphy> thaytan: mpeg2subt is reading text files from hd ?
[09:25] <thaytan> no, it reads the subtitle overlay from mpegdemux
[09:25] <dolphy> ah right
[09:25] <thaytan> which is a run length encoded bitmap to be blended with the video
[09:25] <dolphy> then there should be a way autoplug it
[09:26] <dolphy> it's much harder to handle divx subtitles from text files though :)
[09:27] <thaytan> I just don't know how we inform the autoplugger that it needs to feed the video through the overlay
[09:27] <thaytan> as far as caps is concerned, the video is already nicely decoded and needs no more work
[09:28] <thaytan> I guess I'll ask the list after I get mpeg2subt in better shape
[09:28] <thaytan> it's still completely broken in CVS
[09:36] <Uraeus> dolphy: http://bugzilla.gnome.org/show_bug.cgi?id=137749
[09:41] <dolphy> Uraeus: something's wrong with the logo it seems
[09:41] <dolphy> Uraeus: try to run totem your_file_name 
[09:41] <dolphy> from command line
[09:41] <dolphy> to see if that solves
[09:42] foser ([email protected]) joined #gstreamer.
[09:43] <Uraeus> dolphy: that kinda works, but Totem starts up with a 0x0 video window, so I need to resize totem to actually see the video
[09:46] <dolphy> Uraeus: ??
[09:46] <dolphy> Uraeus: so your gtk is buggy
[09:47] <Uraeus> dolphy: no, the gtk widgets displays nicely, but it seems the missing logo at the begining makes the video window just use 0x0 as the size until you resize
[09:47] <Uraeus> dolphy: also goom does not seem to work in either totem or gst-player
[09:48] <dolphy> no visualization support for the moment
[09:49] mxpxpod ([email protected]) left irc: "To quit or not to quit.. wait... was that a question?"
[09:52] Action: dolphy is off to shower
[09:53] mxpxpod ([email protected]) joined #gstreamer.
[09:58] ChrisHJW_log ([email protected]) left irc: Read error: 110 (Connection timed out)
[09:59] <NFusi0n> Anybody down for some crack attack?
[10:00] apoc_ ([email protected]) joined #gstreamer.
[10:00] <markey> pass the pipe, dude :)
[10:00] <NFusi0n> wrong crack attack
[10:00] <NFusi0n> though i wouldn't be adverse under the right circumstances
[10:00] wingo-z ([email protected]) left irc: Read error: 60 (Operation timed out)
[10:00] <markey> hehe
[10:03] markey ([email protected]) left irc: "leaving"
[10:04] jcsston ([email protected]) left irc: Read error: 60 (Operation timed out)
[10:10] <Uraeus> NFusi0n: I think crack attacks fits better in bugzilla :)
[10:17] apoc ([email protected]) left irc: Connection timed out
[10:23] Shoragan ([email protected]) joined #gstreamer.
[10:35] markey ([email protected]) joined #gstreamer.
[10:40] Nick change: mxpxpod -> mx|sleep
[10:42] Nick change: Uraeus -> Ura_out
[10:58] ChrisHJW ([email protected]) joined #gstreamer.
[11:04] david_ ([email protected]) left irc: "Leaving"
[11:18] sublett ([email protected]) joined #gstreamer.
[11:21] Nick change: markey -> afk
[11:22] ChrisHJW ([email protected]) left irc: Client Quit
[11:24] Nick change: Ura_out -> Uraeus
[11:41] ChrisHJW_log ([email protected]) joined #gstreamer.
[11:50] smoke ([email protected]) joined #gstreamer.
[11:58] guize ([email protected]) joined #gstreamer.
[12:08] Nick change: afk -> markey
[12:35] Uraeus ([email protected]) left irc: "Client exiting"
[12:39] TimeRever ([email protected]) joined #gstreamer.
[12:39] Uraeus ([email protected]) joined #gstreamer.
[12:40] <TimeRever> is that bug or so whatever thing that i can't get esdsink and gnome-sound-events at same time fixed?
[12:40] <Uraeus> they are always broken every other time? :)
[12:40] <TimeRever> humm?
[12:41] <TimeRever> still not fixed?
[12:42] <Uraeus> TimeRever: my guess is that you are having the issue of esdsink being 'locked' or something when gstreamer is outputting sound to it, that is afaik still not fixed
[12:42] <TimeRever> its that nasty thing that i can't use rhythmbox and gnome-sound-events
[12:44] <TimeRever> or is it a rhythmbox only error?
[12:45] <TimeRever> Uraeus no ideia?
[12:45] <thaytan> have you set up rhythmbox to use esdsink?
[12:45] <thaytan> if so, gnome-sound-events and rhythmbox should both work ok
[12:45] <TimeRever> no cause rhythmbox doesn't have such config
[12:45] <Uraeus> TimeRever: not it is a issue with GStreamer, we are aware of the bug, but I don't think anyone has managed to get around to fixing it yet, but it is definetly high on our priority list
[12:46] <TimeRever> so it is a rhythmbox bug?
[12:46] <Uraeus> TimeRever: try running gstreamer-properties (it is a gui tool shipping with gnome media)
[12:46] <Uraeus> TimeRever: no, it is a GStreamer bug
[12:47] <thaytan> it's a more general problem than that
[12:47] <thaytan> we need a new sound server architecture
[12:47] <TimeRever> Uraeus, gstreamer-properties work with gnome-sound-events (at least it makes that noisy sound)
[12:49] <TimeRever> is there some way to get rhythmbox and gnome-sound-events?
[12:50] <thaytan> using esdsink as your audio output
[12:50] <Uraeus> TimeRever: if setting the sink to esound in gstreamer-properties doesn't work then the only way is to have a sound card that support multiple connections
[12:51] pb_ (~pb@2002:5169:e72f:1:240:95ff:fe30:aa84) joined #gstreamer.
[12:52] <TimeRever> its not that i think... rhythmbox popup a messagebox saying it won't worm with esdsink and gnome-sound-events
[12:53] <TimeRever> don't think a nise sound card changes rhythmbox beaviour
[12:53] <Uraeus> ok, then you need a new soundcard to use the together until we have managed to sort of all issues
[12:53] <Uraeus> TimeRever: it does, my soundblaster card can take multiple connections so that both RB and g-s-e can write to the soundcard at the same time
[12:54] <TimeRever> mine is a ac'97 chip :S
[12:54] hyriand ([email protected]) left irc: "boot"
[12:54] <Uraeus> TimeRever: yeah, have one of those and they are not nice, all I can say is that we are working hard at trying to solve all the issues involved
[12:55] <TimeRever> ok...
[12:56] <TimeRever> what about using other sink? osssink makes the songs go crappy after a while but maybe alsa can work with esd?
[12:56] <Uraeus> try :)
[12:58] ChrisHJW ([email protected]) joined #gstreamer.
[13:00] <TimeRever> only 0.7 and up have alsasink?
[13:00] <Uraeus> yes
[13:00] <Uraeus> or rather they are the only ones with a working alsasink
[13:02] <TimeRever> i would like to give a try to 0.8 but those a$$h0£es at mandrake take forever to build a damn package! fedora already have packages for 0.8 and mdk is still at 0.6.4
[13:02] hyriand ([email protected]) joined #gstreamer.
[13:03] <thaytan> TimeRever: the other thing you can try is looking at setting up dmix in ALSA
[13:03] <thaytan> then ALSA does a software mixing
[13:03] <TimeRever> is dmix a command?
[13:03] <thaytan> I've never figured it out on my AC97 though
[13:03] <thaytan> no, it's set up in /etc/asound.conf 
[13:04] <thaytan> and then when you use alsa, it automatically does software mixing of streams
[13:04] <thaytan> http://alsa.opensrc.org/index.php?page=DmixPlugin
[13:07] <TimeRever> that's a big mess up...
[13:07] <thaytan> yeah, none of the solutions are nice right now
[13:07] <TimeRever> this would be a prob if there has a standart for linux sound
[13:07] <TimeRever> wouldn't*
[13:08] <TimeRever> also and offtopic doesn't kernel 2.6 droped oss?
[13:08] <TimeRever> and now only uses alsa
[13:08] <thaytan> it's still there, but deprecated
[13:08] <TimeRever> :)
[13:09] <thaytan> soon, yes, it will disappear
[13:09] <TimeRever> :D
[13:09] <TimeRever> i see some progress!
[13:10] <TimeRever> well thanks for your help
[13:10] <TimeRever> now i'm going
[13:10] <TimeRever> c'ya!
[13:10] TimeRever ([email protected]) left irc: "Programmers do it bit by bit."
[13:14] s1 ([email protected]) joined #gstreamer.
[13:14] <s1> hi, I'm looking for a channel for rhythmbox music player
[13:15] <s1> is there one?
[13:16] Company ([email protected]) joined #gstreamer.
[13:16] <Company> good morning angels
[13:17] <thaytan> s1: not sure
[13:17] <thaytan> Company: hi!
[13:17] <thaytan> Does rhythmbox have a channel?
[13:18] <Company> irc.gimp.org #rhythmbox
[13:18] <s1> Company: thanks :D
[13:19] Action: Company bets that could have been found on www.rhythmbox.org, too ;)
[13:20] <s1> do'h 
[13:20] s1 ([email protected]) left #gstreamer.
[13:22] Rotty ([email protected]) joined #gstreamer.
[13:22] <Company> weee, we're 702nd
[13:23] <thaytan> mm?
[13:24] <Company> http://popcon.debian.org/by_inst
[13:28] <Company> wow, there's more people with vorbis plugins than mp3 plugins
[13:30] <Company> there's even 9 people with gstreamer 0.4.2 and one with 0.3.3
[13:30] Action: Company shouldn't been given statistics
[13:31] <Uraeus> hehe :)
[13:32] jwal ([email protected]) joined #gstreamer.
[13:32] <Company> wow, there's even 18 popcon entries from people without popcon installed
[13:38] <bitshifter> or they have installed it from source code instead...
[13:39] <Company> hm, there's more people with gnome than kde
[13:40] <jwal> For me, gstreamer-properties' test works for OSS but not ALSA using ALSA's OSS emulation.  What can I try to fix this?
[13:41] <thaytan> any error output?
[13:42] <jwal> thaytan: Plenty.  #flood or here?
[13:42] <thaytan> #flood
[13:43] <thaytan> that's using alsasink, right?
[13:43] <thaytan> try 'custom pipeline' and 'alsasink period-size=4096' for me?
[13:44] <Uraeus> hmm, think I should add NAS to gstreamer-properties soon 
[13:44] <thaytan> Uraeus: go for it
[13:45] <Uraeus> thaytan: a bit unsure about freeze issues
[13:45] <thaytan> yeah, after the thaw
[13:46] <jwal> thaytan: An error message window "Failed to construct test pipeline for 'Custom'".  (Before it was silently failing - literally)
[13:47] <thaytan> hang on, whichgstreamre version?
[13:47] foser ([email protected]) left irc: "[ I want to believe ]"
[13:48] <jwal> 0.6.4, perhaps
[13:48] <thaytan> ahr, right
[13:48] <thaytan> no alsasink is just broken in 0.6
[13:50] <Company> there's even 8 people with mas installed
[13:51] <Company> that's 0.3%!
[13:52] <thaytan> gah!
[13:52] <thaytan> where the hell are my subtitles going?
[13:52] <jwal> thaytan: OSS seems to work fine anyway.  Thanks.
[13:53] <thaytan> aww damn
[13:54] <thaytan> jwal: good :)
[13:54] <thaytan> (damn = just found my subtitles)
[13:54] Action: Company still fears our dvd guys do a lot of hacks that shouldn'T work in an ideal world...
[13:55] <Company> someone get me a dvd drive :)
[13:56] <thaytan> hacks? I am offended
[13:56] <thaytan> ;)
[14:02] <Uraeus> Company: if you are not picky so do I have a Creative 5x DVD drive I can send you
[14:04] <Company> Uraeus: i'm not at all picky as i don'T have a dvd player
[14:05] <thaytan> hrmn. I have 4, but I think I'd get in trouble for sending any of them
[14:06] <Uraeus> Company: ok, mail me your snail mail address and I send it to you
[14:07] <thaytan> Company: how do I make this subtitle element work 'properly'?
[14:07] <thaytan> it needs to read from 2 sink pads
[14:07] <thaytan> but subtitle buffer arrive irregularly
[14:08] <Company> gst_pad_select doesn't work with opt, right?
[14:08] <thaytan> at the moment, I'm leaving it as a chain based element, which works 'well enough'
[14:08] <thaytan> gst_pad_select still has no guarantee, does it?
[14:08] <Company> guarantee for what?
[14:08] <thaytan> since you want to process the subtitle buffer for time 'x' before you process the video with the same timestamp
[14:08] <Company> ah
[14:08] <thaytan> but you don't want to hold onto video 'in case' a subtitle arrives
[14:08] <Company> n
[14:09] <Company> no
[14:09] <Company> with current gst you need to use queues
[14:09] <Company> which probably doesn't work when subtitles are very rare because the queue fills up
[14:09] <thaytan> exactly
[14:10] <thaytan> it's a problem that needs solving eventually
[14:10] <Company> hm
[14:10] <Company> can't you send filler events?
[14:11] <thaytan> sending those from mpegdemux might work for now
[14:11] <Company> filler events were supposed to say "don't expect a buffer with a timestamp < $filler_time"
[14:12] <thaytan> *nod*
[14:12] <thaytan> that sounds good
[14:13] <thaytan> I'll try that later when I go to convert mpeg2subt to loop-basd
[14:15] jwal ([email protected]) left #gstreamer.
[14:25] coqRe ([email protected]) joined #gstreamer.
[14:47] md` ([email protected]) left irc: Read error: 110 (Connection timed out)
[14:49] smoke ([email protected]) got netsplit.
[14:55] smoke ([email protected]) got lost in the net-split.
[15:24] bitshifter ([email protected]) left irc: "bye"
[15:33] snikkt ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[15:51] foser ([email protected]) joined #gstreamer.
[15:52] pb___ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) joined #gstreamer.
[15:58] bitshifter ([email protected]) joined #gstreamer.
[16:02] Marsupilami23 ([email protected]) left irc: "I don't wanna grow up, I'm a Toys R Us Kid. There's a million toys at Toys R Us that I can play with!"
[16:03] Marsupilami23 ([email protected]) joined #gstreamer.
[16:08] alley_cat ([email protected]) left irc: Read error: 110 (Connection timed out)
[16:09] Kurosu ([email protected]) joined #gstreamer.
[16:11] alley_cat ([email protected]) joined #gstreamer.
[16:14] Marsupilami23 ([email protected]) left irc: Read error: 54 (Connection reset by peer)
[16:23] pb___ (~pb@2002:3e03:42cc:1:20b:6aff:fe19:f290) left irc: "Client exiting"
[16:32] iain ([email protected]) joined #gstreamer.
[16:40] coqRe ([email protected]) left irc: "Leaving"
[16:42] md` ([email protected]) joined #gstreamer.
[16:48] kmaraas ([email protected]) left irc: Client Quit
[16:51] coqRe ([email protected]) joined #gstreamer.
[16:55] Misirlou ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[16:57] Misirlou ([email protected]) joined #gstreamer.
[17:03] spyder482 ([email protected]) joined #gstreamer.
[17:22] guize ([email protected]) left irc: Remote closed the connection
[17:32] snikkt ([email protected]) joined #gstreamer.
[17:56] jdahlin ([email protected]) joined #gstreamer.
[17:56] kingsley ([email protected]) joined #gstreamer.
[17:57] <kingsley> Does anyone here happen to know how to extract the frames from within an mpeg movie?
[18:14] kingsley ([email protected]) left irc: "[x]chat"
[18:14] kmaraas ([email protected]) joined #gstreamer.
[18:20] walters ([email protected]) joined #gstreamer.
[18:20] jdahlin ([email protected]) left irc: Read error: 104 (Connection reset by peer)
[18:31] <Company> yay, we have elements that eat DICONTs...
[18:31] <Company> DISCONTs, too
[18:32] coqRe ([email protected]) left irc: "Leaving"
[18:35] <walters> can someone here approve a patch to gstxmlregistry.c?
[18:36] <walters> http://web.verbum.org/files/gst-registry.patch
[18:36] jcsston ([email protected]) joined #gstreamer.
[18:36] <walters> basically it's avoiding the evil hack of setting the mod time on the registry file
[18:41] jcsston` ([email protected]) joined #gstreamer.
[18:42] <walters> ds: here?
[18:45] <walters> Company: can you say whether you think it's sane or not?
[18:52] jdahlin ([email protected]) joined #gstreamer.
[18:53] <walters> i'd like to commit this patch so i can fix something else in gstxmlregistry.c :)
[18:53] Action: walters notes if gst was using arch he could have already committed this one in his local repo and went onto the next thing...
[18:54] <Company> walters: shouldn't the READABLE flag check the file and not the dir?
[18:55] kmaraas ([email protected]) left irc: Read error: 110 (Connection timed out)
[18:55] <Company> and the WRITABLE check should probably check the file, too if it exists
[18:55] <Company> though i'm not really sure what those flags mean exactly anyway
[18:56] <walters> Company: well, it seems unlikely to me that the dir would be readable but the file not
[18:56] <walters> Company: as for writable - no, because we only need write access to the dir, after my next patch
[18:56] <taaz> walters: i've started using arch for debian packages.
[18:56] <walters> taaz: cool, how is it working?
[18:56] <walters> taaz: i am probably going to maintain my rh packages with it
[18:57] <taaz> walters: ok i suppose.  it's a very painful learning curve.  i really don't think mortals will be able to figure it out at this point in time
[18:57] <Company> taaz: you know about the configure flags --witrh-package-name and --with-package-origin, right?
[18:57] <taaz> Company: um. err. should i? ;)
[18:57] <Company> taaz: you suggested them ;)
[18:58] <taaz> i did?
[18:58] <jdahlin> taaz: how far did you come wrt threads when you played around with gst-python?
[18:58] <walters> taaz: the curve is definitely high, i agree.  but it's been getting better.  i think it's mostly just the initial import that's confusing for cvs refugees.  after that it's mostly the same update/commit cycle
[18:58] <jdahlin> taaz: I've had some small successfull attempts with GThread
[18:58] <Company> taaz: i think so - anyway, use them for the Debian packages :)
[18:59] <taaz> walters: the tla/arch docs really need more tutorials that just explicitly spell out what to type.  seems some people assume that you know you need to run certain setup commands before other will work.  and the error messages are crap like "I/O Panic" rather than "archive doesn't exist" or whatnot.
[18:59] <walters> taaz: yeah
[18:59] <walters> taaz: totally agree there
[19:01] <taaz> walters: and there are various ways to manage debian packages and no real comparison or explaination why any method is better than the others.  manojs arch.d.o discussion page is quite nice though.  i did a variant of that and what is described in the arch-buildpackage README
[19:01] <taaz> Company: what do i put in them?
[19:03] <taaz> walters: i'm using 'configs' stuff but i don't even really know why. ;)  got a debian tree branch of upstream but also merge in the debian/ dir via the configs.  which seems a bit silly to have both other than making it easy to see what non-debian-dir changes i've made.  but that never was an issue in the past since grepdiff was born
[19:04] <Company> taaz: default values are "GStreamer" and "http://gstreamer.net"
[19:04] bitshifter ([email protected]) left irc: "bye"
[19:05] <taaz> walters: are you still going to do debian work are did the dark^wredside capture all your skillz?
[19:05] <walters> taaz: there's definitely multiple ways to do it.
[19:05] <walters> taaz: ah, i think i'll still do cdbs
[19:05] <walters> taaz: or try anyways :)
[19:05] bitshifter ([email protected]) joined #gstreamer.
[19:05] <walters> bbiab
[19:05] <Company> taaz: so something like "GStreamer (Debian Package v.0.8.0-1)" and "http://debian.org/gstreamer" or something
[19:07] <Company> taaz: it's visible when you do gst-inspect gnomevfs
[19:07] <taaz> it is?
[19:08] jdahlin ([email protected]) left irc: "Leaving"
[19:08] <taaz> oops.  was just typing response to jdahlin.. oh well
[19:09] <Uraeus> walters: do you know if it is is possible to tell gtkImage to use a specific colour as its background (as opposed to the default grey?)
[19:09] jcsston ([email protected]) left irc: Read error: 110 (Connection timed out)
[19:12] <taaz> oh.. i see.  gst-inspect-0.8 pluginname vs elementname.  i see that info now
[19:13] smoke ([email protected]) joined #gstreamer.
[19:14] <Company> Uraeus: it's transparent
[19:17] <Company> (at least I think it is)
[19:17] <Company> but you can change its attached GtkStyle
[19:17] <Company> the video widget does that
[19:19] <Uraeus> Company: the issue is that we have a mozilla plugin, which due to the transparancy allways ends up grey, and I think white would be better soltution
[19:19] <Misirlou> Company: There were a few Theora bit stream changes lately. Do you know if anyone has been catching up?
[19:19] <Uraeus> Company: ok, I look into changing the GtkStyle then
[19:19] <taaz> should the package name for the plugins be GStreamer Plugins rather than just "GStreamer"?
[19:19] Action: Uraeus tries hacking a bit again
[19:20] <Company> taaz: right
[19:20] kmaraas ([email protected]) joined #gstreamer.
[19:21] Nick change: khalek -> kh_zZz
[19:21] <walters> Uraeus: not sure offhand
[19:22] <walters> Uraeus: btw...committing non-translation stuff to cvs for rb is kind of deprecated atm
[19:22] <walters> Uraeus: the fixes are appreciated though :)
[19:22] <Uraeus> walters: I don't think the spec file is part of the freeze as it is not really part of the applications
[19:22] <walters> Uraeus: i can hook you up with direct arch commit access if you want
[19:23] <walters> Uraeus: right...well, you also did modify autogen.sh
[19:23] <Uraeus> walters: oh, sorry :)
[19:23] <Uraeus> not sure what I changed in autogen aymore :)
[19:23] <walters> Uraeus: np, just mentioning :)
[19:23] <Uraeus> I don't have any arch software installed
[19:24] <walters> Uraeus: i have rpms for fc1 lying around
[19:24] <walters> need to update, mm.
[19:24] Action: walters does that now
[19:25] spyder482 ([email protected]) left irc: Read error: 113 (No route to host)
[19:26] <Uraeus> walters: do using arch really give me anything? considering how simple my needs are I have to admit learning a new system doesn't seem tempting
[19:27] <walters> Uraeus: for what you're doing, you won't have to know anything more than "tla update" and "tla commit"
[19:27] <walters> Uraeus: basically i want to have cvs only used for translations, so that i can have it automatically sync
[19:28] <walters> that way i know there'll never be conflicts
[19:28] <Uraeus> ugh :)
[19:32] <bitshifter> Uraeus: gtk_widget_modify_bg(), but you'll need to modify the parent's background because GtkImage doesn't have its own GdkWindow IIRC. 
[19:32] <Uraeus> bitshifter: the parent in this case being mozilla or the plugin?
[19:33] <bitshifter> parent widget
[19:33] <taaz> Company: i ask because gst-plugins tarball just uses "GStreamer"
[19:36] Company ([email protected]) left irc: Remote closed the connection
[19:36] Company ([email protected]) joined #gstreamer.
[19:37] <Company> taaz: yeah, can you fix that so it says "GStreamer Plugins"
[19:37] Action: Company has a heavily modified configure.ac
[19:37] Nick change: mx|sleep -> mxpxpod
[19:39] <Company> Misirlou: that shouldn't matter since we use libtheora
[19:39] <Misirlou> Company: ooh, wheeeee!
[19:40] <Company> yep, still compiles fine
[19:41] <Company> doesn't play old files though...
[19:43] <Company> Misirlou: should it still play old files?
[19:43] <Misirlou> Company: Nope, don't think so. There were a few changes that broke backward compatibility I believe.
[19:44] <Company> Misirlou: you know of any new files i could use to test?
[19:45] <Misirlou> Company: nope
[19:46] <taaz> how do i get utf-8 to show up in vim over ssh in a gnome terminal?
[19:46] <taaz> or show up at all really.
[19:46] <taaz> i just see "???" which is useless and doesn't edit well
[19:48] kmaraas ([email protected]) left irc: Client Quit
[19:52] Nick change: Uraeus -> Ura_out
[19:57] kmaraas ([email protected]) joined #gstreamer.
[20:00] harshy ([email protected]) left #gstreamer ("PART!").
[20:48] foser_ ([email protected]) joined #gstreamer.
[20:48] foser_ ([email protected]) left irc: Client Quit
[21:34] benow ([email protected]) joined #gstreamer.
[21:35] <benow> hello all...
[21:35] <NFusi0n> hi
[21:37] <benow> so, narrowing down the JNI problems... having the newest version helps (was using old sf cvs ver)... have now found that:
[21:37] <benow>  DEBUG GST_PLUGIN_LOADING(14384) gstplugin.c(395):gst_plugin_load_file: Error loading plugin /usr/lib/gstreamer-0.8/libgstoptscheduler.so, reason: /usr/lib/gstreamer-0.8/libgstoptscheduler.so: undefined symbol: _gst_bin_type
[21:37] BBB ([email protected]) joined #gstreamer.
[21:38] <benow> perhaps I'm not linking the native library correctly?
[21:38] <benow> heya BBB
[21:39] <BBB> hi
[21:41] <Company> _gst_bin_type is exported by libgstreamer.so
[21:41] spyder482 ([email protected]) joined #gstreamer.
[21:42] jcsston ([email protected]) joined #gstreamer.
[21:46] jcsston` ([email protected]) left irc: Read error: 60 (Operation timed out)
[21:54] jcsston ([email protected]) left irc: Read error: 60 (Operation timed out)
[21:56] <NFusi0n> Has anybody done any gapless stuff with gstreamer?
[22:00] <BBB> gapless?
[22:02] <spyder482> gapless playback
[22:02] <spyder482> like playing 2 audio streams without a gap in samples at the transition
[22:05] psm ([email protected]) joined #gstreamer.
[22:05] <psm> hi
[22:11] <BBB> OH
[22:11] <BBB> THAT
[22:11] <BBB> oops
[22:11] <BBB> dunno
[22:13] steve_b ([email protected]) left irc: "I don't think I'm going to use hamster style any more."
[22:18] pel ([email protected]) joined #gstreamer.
[22:20] <pel> hi, anyone alive?
[22:20] <BBB> yes
[22:20] <taaz> no
[22:20] <pel> taaz, =)
[22:21] <pel> I've allways liked the undead... ;-)
[22:21] <pel> gstplay just told me assertion `GST_IS_TAG_LIST (list)' failed
[22:21] <pel> what does that mean?
[22:22] <pel> (just compiled this with jhbuild)
[22:22] <BBB> that it received metadata but the received metadata was corrupt
[22:23] <pel> hm.. totem with libxine seems to have no problem with the file
[22:24] <pel> well, basicly - that's the most informative answer I've gotten so far from anything gstreamer in this compile run.
[22:25] <pel> trying to find out what went wrong.
[22:26] <BBB> do you get that with specific media files?
[22:26] <BBB> if so, please upload the file to bugzilla so we can fix the bug :)
[22:27] <pel> I fear it does not matter what I throw at it
[22:27] <pel> otherwise I'd do that
[22:27] <pel> hmm... or post a link, perhaps.
[22:29] <BBB> in any case, make this a bug report
[22:29] <BBB> so we know something's wrong
[22:29] <BBB> ok? :)
[22:29] <pel> =)
[22:29] <pel> i probably should try to narrow it down from 'nothing works' ;-)
[22:29] <pel> well, it compiles at least.
[22:30] <pel> and the testsuite does not complain.
[22:30] <BBB> nothing works is sort of crude, yes
[22:30] <BBB> it works for us
[22:30] <BBB> how did you install gst?
[22:30] <BBB> manual compile?
[22:30] <pel> jhbuild
[22:30] <BBB> did you run gst-register?
[22:30] <pel> yup
[22:30] <BBB> both as user and as root?
[22:30] <BBB> (some people have two registries, sometimes it keeps using the root one, even as user...)
[22:30] <pel> is root user really necessary?
[22:30] <BBB> (that's a bug in gst, btw, but ... ;) )
[22:31] <pel> but yes, I have.
[22:32] <BBB> hm
[22:32] <BBB> odd
[22:32] <BBB> totem? or gst-player?
[22:33] <pel> gst-player is what I'm trying ATM
[22:34] <BBB> hm
[22:34] <BBB> ok
[22:34] <BBB> I mostly test with totem
[22:34] <BBB> maybe that's related?
[22:34] <pel> totem didn't work, so I tried gst-player.
[22:35] <BBB> "didn't work"?
[22:35] <BBB> :)
[22:35] <pel> well, it froze and then stoped. pressing play repeats procedure.
[22:36] <BBB> that's odd, really odd
[22:36] <pel> <- beeing very informative
[22:36] <BBB> :P
[22:37] <pel> *recompiling*
[22:37] <pel> hmm.. what is the simplest way to test mp3 sound?
[22:37] <pel> gst-launch?
[22:40] <BBB> yes
[22:40] <BBB> gst-launch filesrc location=file.mp3 ! mad ! osssink
[22:40] <BBB> or s/mad/spider/ to test autoplugging
[22:42] <pel> hmm
[22:42] <pel> WARNING: erroneous pipeline: no element "mad"
[22:42] <pel> does that mean that the registry is the problem anyways?
[22:43] <BBB> no
[22:43] <BBB> you probably miss a lot of media libs
[22:43] <BBB> which is why media playback doesn't work
[22:43] <BBB> did you compile gst-plugins and gst-ffmpeg?
[22:43] <pel> gst-plugins is compiled
[22:43] <BBB> running gst-inspect should show you a list of plugins
[22:44] <BBB> are plugins like mad, mpeg2dec in there?
[22:44] <BBB> if not, then you can barely playback anything
[22:44] <pel> http://www.update.uu.se/~peterl/gst-inspect.txt
[22:46] BBB ([email protected]) left irc: "Download Gaim: http://gaim.sourceforge.net/"
[22:47] <benow> hmmm, still having link issues, can anyone solve the following:
[22:47] <benow> DEBUG GST_PLUGIN_LOADING(16717) gstplugin.c(395):gst_plugin_load_file: Error loading plugin /usr/lib/gstreamer-0.8/libgstoptscheduler.so, reason: /usr/lib/gstreamer-0.8/libgstoptscheduler.so: undefined symbol: _gst_bin_type
[22:47] BBB ([email protected]) joined #gstreamer.
[22:47] <BBB> bah
[22:47] <BBB> gaim started acting weird
[22:47] <BBB> anyway
[22:47] <BBB> you don't have decoders
[22:47] <BBB> make sure you've got libmad, libmad-devel, those sort of things
[22:48] <pel> bbb, well - i don't see mp3-suport there. so you're right I guess =)
[22:48] <benow> link is: gcc   -Wl,--export-dynamic -pthread -lgstreamer-0.8 -lgobject-2.0 -lgmodule-2.0 -ldl -lgthread-2.0 -lxml2 -lpthread -lz -lm -lglib-2.0   -g -O2 -shared *.o -o ../libGSTJava.so
[22:48] <BBB> you don't even have vorbis support ;)
[22:48] <benow> and /usr/lib, /usr/lib/gstreamer-0.8 is in ld_library_path
[22:49] <pel> hehe.. 
[22:50] <pel> damn code that compiles even when the functionality I want in it isn't available... ;-)
[22:50] <pel> typical example of 'if it compiles, it works'
[22:51] <benow> where's the neural configuration interface when you need it...
[22:51] <pel> my point exactly ;-)
[22:52] <pel> bbb, thanks for the help anyways. 
[22:52] Action: pel will try to not to die an utter clueless moron
[22:53] <BBB> no prob :)
[22:53] <BBB> hope it helps a bit
[22:54] <BBB> id you can't find it; there will be 0.8.0- RPms on freshrpms/fedora in a few days
[22:54] <pel> hehe... it probably will, at least a bit =)
[22:54] <BBB> ok:)
[22:56] <benow> so, anyone have any hints for undefined symbol?
[22:56] <BBB> not really...
[22:56] <BBB> are you linkingstatically?
[22:58] <benow> -shared
[22:58] <benow> should it be static?
[22:58] <pel> hmm.. installing libmad-dev and libvorbis-dev didn't help much
[22:59] <benow> pel: you'll have to re ./configure and make && sudo make install
[22:59] <benow> and re gst-register-*
[23:00] <BBB> benow: no, it should be shared
[23:00] <BBB> did you disable any subsystem?
[23:00] <BBB> this is java?
[23:00] <benow> yeah, it's dieing during jni run, but problem may be with linking in native compile 
[23:01] <benow> during jni run, /usr/lib and /usr/lib/gstreamer-0.8 are visible, so it should be able to resolve libs (and does)
[23:01] <benow> so I think its something in the native gcc'ing
[23:02] <BBB> hm
[23:02] <BBB> I really don't know :(
[23:02] Action: BBB knows nothing of java wrapping et al
[23:03] <benow> well, it has little to do with the java wrapping and more with the compilation... does the above gcc line look correct, afa gst requirements/libraries go?
[23:03] <benow> I'm using `pkg-config gstreamer-0.8 --[libs|cflags]` during compilation.
[23:04] <BBB> yes
[23:04] <BBB> it looks allright
[23:04] <spyder482> mmm, gstreamer and java
[23:04] <BBB> do you need to "extern C" parts of the C thing that you wrap in java?
[23:04] <BBB> hm
[23:04] <BBB> it's c
[23:04] <BBB> so no
[23:04] Action: BBB thinks
[23:04] <spyder482> i think si
[23:04] <spyder482> so
[23:05] <spyder482> oh...
[23:05] <benow> spyder482: yeah, gst+java'd kick some ass
[23:05] <spyder482> yeah, not if it's C...
[23:05] mxpxpod ([email protected]) left irc: "To quit or not to quit.. wait... was that a question?"
[23:05] Action: spyder482 likes java but it's too slow
[23:05] <BBB> I know those symbols are not being exported
[23:05] <BBB> do you use those type symbols in your own object?
[23:05] <BBB> we only export GST_* symbols
[23:05] <benow> meh, little of an issue on a good box (dual mp2k+ here)
[23:06] <Company> we should link the plugins to libgstreamer
[23:06] <BBB> I didn't say that
[23:06] <BBB> I said we don't export __gst_* symbols
[23:06] <BBB> ;)
[23:06] <Company> yes we do export those
[23:06] <BBB> oh, we do?
[23:06] <BBB> hm
[23:06] <Company> we don't export those from plugins
[23:06] <BBB> I thought only gst_* symbols
[23:06] <BBB> ah, ok
[23:06] <Company> but we do export them from libgstreamer
[23:07] <Company> the problem is the linker can't find the symbols from libgstreamer, because libgstreamer symbols aren't loaded globally
[23:07] <Company> for whatever reason
[23:07] Nick change: Ura_out -> Uraeus
[23:08] <benow> the code in question can be found at:
[23:08] <benow> http://benow.ca/misc/gst-java/
[23:08] <benow> native code, anyway.
[23:08] <Company> and if they're not, the linker fails to link those because plugins aren't explicitly linked to libgstreamer
[23:09] <Company> if we'd link every plugin to libgstreamer, there wouldn'tr be a problem
[23:09] <benow> the Makefile is pretty simple, and called after setting env from ./compile
[23:09] <Company> but omega didn't want that for whatever reason
[23:10] <Company> every other plugin using framework does that, so they don't have any problems
[23:10] <BBB> this fails during coimpiling of the *lib* wrapper
[23:10] <BBB> not plugin wrapper
[23:10] <BBB> so it's not appliccable
[23:10] <Company> no it fails during running the wrapper
[23:11] <Company> when running the wrapper, the loaded plugins fail to find libgstreamer
[23:11] <Company> probably because the wrapper isn't loaded with RTLD_GLOBAL
[23:12] <Company> so if we'd do what everybody else does we'd link our plugins to libgstreamer and all the binding developers wouldn't have no more problems
[23:12] <Company> ask ds-work why we don't do this, he understood that back then ;)
[23:13] <benow> Company: so I need to link against /usr/lib/gstreamer-0.8/*.so?
[23:14] mxpxpod ([email protected]) joined #gstreamer.
[23:14] <Company> benow: no
[23:16] <benow> can I modify the gcc to have the linker to find the symbols from libgstreamer?
[23:16] <benow> (please excuse my C inexperience, I am a stranger to the ways of the un-runtimed)
[23:17] <Company> as a quick fix:
[23:17] Nick change: mxpxpod -> mx|gone
[23:18] <Company> try adding this line before calling gst_init and tell me if it works:
[23:18] <Company> g_module_open ("libgstreamer-0.8.so", 0);
[23:19] <Company> (if that works, it's a hack, not a clean solution)
[23:19] Action: BBB goes home
[23:19] <BBB> bye
[23:19] <BBB> :)
[23:19] BBB ([email protected]) left irc: "Download Gaim: http://gaim.sourceforge.net/"
[23:20] <benow> Company: yup, that resolved the could not find 'opt'
[23:20] <benow> but now it complains about not finding filesrc
[23:21] <benow> INFO  GST_PLUGIN_LOADING(17311) gstplugin.c(369):gst_plugin_load_file: plugin "/usr/lib/gstreamer-0.8/libgstoptscheduler.so" loaded
[23:21] <benow> Error during execution: Could not find plugin named: filesrc
[23:21] <benow> org.gstreamer.NoSuchPluginException: Could not find plugin named: filesrc
[23:21] <Company> what's the debug output with GST_DEBUG=GST_PLUGIN*:5 ?
[23:22] <benow> says nothing of filesrc, here's the whole line (excuse the dump!)
[23:22] <benow> $ GST_DEBUG=GST_PLUGIN*:5 bin/MP3Player ../test.mp3
[23:22] <benow> Initializing with args: --gst-debug-level=5
[23:22] <benow> LOG   GST_PLUGIN_LOADING(17311) gstplugin.c(114):_gst_plugin_register_static: queueing static plugin "gstcoreelements" for loading later on
[23:22] <benow> LOG   GST_PLUGIN_LOADING(17311) gstplugin.c(120):_gst_plugin_register_static: attempting to load static plugin "gstcoreelements" now...
[23:22] <benow> LOG   GST_PLUGIN_LOADING(17311) gstplugin.c(208):gst_plugin_register_func: plugin "(NULL)" initialised
[23:22] <benow> INFO  GST_PLUGIN_LOADING(17311) gstplugin.c(124):_gst_plugin_register_static: loaded static plugin "gstcoreelements"
[23:22] <benow> LOG   GST_PLUGIN_LOADING(17311) gstxmlregistry.c(462):plugin_times_older_than: comparing plugin times from /usr/lib/gstreamer-0.8 with 1079814792
[23:22] <benow> DEBUG GST_PLUGIN_LOADING(17311) gstpluginfeature.c(103):gst_plugin_feature_ensure_loaded: loading plugin gstoptscheduler for feature
[23:22] <benow> DEBUG GST_PLUGIN_LOADING(17311) gstplugin.c(299):gst_plugin_load_file: attempt to load plugin "/usr/lib/gstreamer-0.8/libgstoptscheduler.so"
[23:22] <benow> LOG   GST_PLUGIN_LOADING(17311) gstplugin.c(351):gst_plugin_load_file: Plugin 0x826b0a8 for file "/usr/lib/gstreamer-0.8/libgstoptscheduler.so" prepared, calling entry function...
[23:22] <benow> LOG   GST_PLUGIN_LOADING(17311) gstplugin.c(208):gst_plugin_register_func: plugin "/usr/lib/gstreamer-0.8/libgstoptscheduler.so" initialised
[23:22] <benow> INFO  GST_PLUGIN_LOADING(17311) gstplugin.c(369):gst_plugin_load_file: plugin "/usr/lib/gstreamer-0.8/libgstoptscheduler.so" loaded
[23:22] <benow> Error during execution: Could not find plugin named: filesrc
[23:22] <benow> org.gstreamer.NoSuchPluginException: Could not find plugin named: filesrc
[23:23] <benow>         at org.gstreamer.GstElementFactory.makeElement(GstElementFactory.java:34)
[23:23] <benow>         at org.gstreamer.GstBin.makeContainedElement(GstBin.java:45)
[23:23] <benow>         at test.org.gstreamer.MP3Player.main(MP3Player.java:35)
[23:24] <Company> that looks like some error in the bindings
[23:25] <benow> yeah, it may be... just checking that now.  Thanks for that hack!  I'll document it ;)
[23:25] <Company> you could try running with GST_DEBUG=GST_ELEMENT_FACTORY:5
[23:26] <Company> learn the debugging stuff, it helps :)
[23:28] <Company> ds-work: ping?
[23:28] <benow> Company: yeah, seems handy
[23:32] <benow> yup, is a binding problem... string conversion.
[23:39] benow ([email protected]) left irc: "thanks for your help BBB, Company."
[23:39] mathrick ([email protected]) joined #gstreamer.
[23:41] <mathrick> yo
[23:41] <apoc_> hi mathrick
[23:45] <Uraeus> mathrick: how goes things with libmms? ;)
[23:47] jdahlin ([email protected]) joined #gstreamer.
[23:47] <Uraeus> hi jdahlin
[23:48] <Uraeus> jdahlin: I filed a g-s-r bug today :)
[23:48] <jdahlin> Uraeus: horray!
[23:48] <jdahlin> Uraeus: bug #? :-)
[23:48] <Uraeus> I knew you would say that :)
[23:49] Action: jdahlin is just on his way out to "retrieve" a couple of beers
[23:50] <Uraeus> jdahlin: http://www.w3.org/Graphics/SVG/Test/20030813/htmlframe/full-pservers-grad-04-b.html
[23:50] <Uraeus> jdahlin: err, sorry
[23:50] <jdahlin> Uraeus: not really?
[23:50] <Uraeus> jdahlin: http://bugzilla.gnome.org/show_bug.cgi?id=137751
[23:50] <jdahlin> hihi
[23:50] Action: jdahlin goes to Uraeus porn site first
[23:50] <Uraeus> that is the URL I was meaning to give you :)
[23:50] <Uraeus> heh, that is really geek porn :)
[23:51] <jdahlin> porn as porn
[23:51] ChrisHJW ([email protected]) left irc: Read error: 110 (Connection timed out)
[23:51] <jdahlin> Uraeus: will check it out on monday
[23:51] <Uraeus> jdahlin: ok good :)
[23:51] <jdahlin> Uraeus: and that's the only bug you managed to find? :)
[23:52] <Uraeus> jdahlin: well it is a rather major one :) I find more when that one is fixed :)
[23:52] <jdahlin> Uraeus: don't wait for me to fix that one ;-)
[23:53] <Uraeus> I will hold my breath until you fix it
[23:53] <jdahlin> anyway time to leave
[23:53] <jdahlin> damn norrmän
[23:54] <jdahlin> see you around
[23:54] jdahlin ([email protected]) left irc: "Leaving"
[23:58] <mathrick> Uraeus: poorly. I still can't get myself together, which annoys the hell outta me
[23:59] <Uraeus> mathrick: think to yourself this: if I don't have 300 loc done by tommorow I have to hit my hand with a hammer; if you believe your own treat it will help you pull yourself together
[23:59] <mathrick> Uraeus: hmm, gonna try this :>
[00:00] --- Sun Mar 21 2004
[00:03] <pel> seems my problems stem from libid3tag
[00:04] <pel> the configure test fails, but if I just run it myself it works like a charm.
[00:04] <pel> odd.
[00:07] spyder482 ([email protected]) left irc: "Leaving"
[00:10] <mathrick> Uraeus: btw, why announce gst win32 support now? did something significant happen on that front? I thought it was in compilable-and-even-runnable state for a while now?
[00:11] <Uraeus> mathrick: well I just noticed that it was at a working level yesterday in the Matroska channel, and I am not sure an advogato entry counts as 'annoucing it' :)
[00:12] <mathrick> Uraeus: well, you are aware it's linked to PGO, aren't ya ;)
[00:13] <Uraeus> yes, well some good mindshare building announcements has never hurt any project :)
[00:20] jcsston ([email protected]) joined #gstreamer.
[00:21] <Uraeus> hi jcsston
[00:21] <jcsston> lo Uraeus
[00:21] <Uraeus> jcsston: how goes the mingw porting effort?
[00:22] <mathrick> lo jcsston
[00:23] markey ([email protected]) left irc: "nite"
[00:23] <jcsston> Uraeus: I don't know, Mosu was going to try it today or tomorrow (I'm not good with configure/automake stuff)
[00:24] <Uraeus> jcsston: nobody is good with that we just juryrig it somehow ;)
[00:24] <jcsston> hehe
[00:25] <Uraeus> ok, time for me to sleep, night
[00:26] Uraeus ([email protected]) left irc: "Client exiting"
[00:26] <mathrick> Uraeus: you know you just disrespected thomasvs severely? ;)
[00:26] <mathrick> hmm, he left
[00:27] <mathrick> jcsston: so, how's it going wrt varargs MSVC doesn't like? were you able to sort it out completely?
[00:28] <jcsston> mathrick: We had agree to use MinGW instead of MSVC.
[00:31] <mathrick> jcsston: ic
[00:31] <mathrick> jcsston: was it only with variadic macros, or functions too?
[00:31] <jcsston> mathrick: just marcos
[00:33] <mathrick> jcsston: and is there ISO standard for these? or free for all in that regard?
[00:34] <jcsston> mathrick: ISO, variadic marcos are included in the C99 standard, but MSVC does support them :/
[00:34] <jcsston> s/does/doesn't
[00:35] <mathrick> heh, what do you expect from (supposedly) one-stop IDE that can't even do simple diffs?
[00:51] <mathrick> hehe: http://www.cse.unsw.edu.au/~malcolmr/intercal/threaded.html
[00:51] <mathrick> http://www.catb.org/~esr/intercal/overloading.html
[00:51] <mathrick> now that's _evil_ :)
[01:00] Action: Company wants some happily ever after potion
[01:05] <mathrick> Company: ?
[01:06] <Company> mathrick: watch the shrek 2 trailer
[01:07] wheels ([email protected]) left irc: Read error: 110 (Connection timed out)
[01:07] wheels ([email protected]) joined #gstreamer.
[01:08] <mathrick> Company: wow, shrek 2? trailer? haven't heard of it
[01:09] <Company> www.shrek2.com
[01:16] Rotty ([email protected]) left irc: Remote closed the connection
[01:17] <pel> if gstplay plays a file w/o problem, but totem fails (problem with padding), is that necessarily totems fault?
[01:22] bilboed_ ([email protected]) left irc: Read error: 110 (Connection timed out)
[01:22] bilboed_ ([email protected]) joined #gstreamer.
[01:28] Shoragan ([email protected]) left irc: "Leaving"
[01:28] Shoragan ([email protected]) joined #gstreamer.
[01:30] pel ([email protected]) left irc: "ChatZilla 0.9.52B [Mozilla rv:1.6/1]"
[01:31] <mathrick> Company: looks like fun's coming into town again :)
[01:32] Action: mathrick wonders how long he'll have to wait for polish dubbing to come out
[01:33] <mathrick> Shrek was one of these extremely rare cases, where dubbing was actually _better_ than original play
[01:34] <wheels> mathrick: are most things subtitled or dubbed there?
[01:34] <mathrick> wheels: depends
[01:35] <mathrick> wheels: kids stuff is mostly dubbed (obvious), rest is subbed, but in cinema only. In TV, we have that strange and unique concept of "reader"
[01:36] <mathrick> wheels: it's the only thing I consider worse than dubbing, btw
[01:37] <wheels> mathrick: Hmm, not sure really what it is, but sounds bad.  :-)
[01:37] <mathrick> wheels: it's really awful. It's poor-man's dubbing preformed off-screen by one guy
[01:38] <mathrick> wheels: it's kinda legacy thing, back from communists regime days, but it doesn't make it any better :>
[01:38] <wheels> ouch -- it's hard for me to imagine anyone prefering such to subtitles...
[01:39] Action: wheels just has a hate for all things dubbed, but being a native English speaker I'm somewhat biased...
[01:39] <mathrick> wheels: but it doesn't really matter to me, as I don't watch TV more often than 15 minutes in 3 months :)
[01:39] <wheels> (But I never watch forein films dubbed when I can avoid it...)
[01:39] <mathrick> wheels: no, you're perfectly normal :)
[01:39] <wheels> mathrick: Well, I don't watch TV (don't even have one), but just films...
[01:39] <wheels> mathrick: Well, I live in .de and eveyone here is used to having everything dubbed...
[01:40] <mathrick> wheels: subbing is the only acceptable way of translation
[01:40] <mathrick> wheels: yeah, I know, it's way too popular
[01:41] Action: mathrick notes that .de dubbing is particularly bad. Worst in the world probably
[01:41] <wheels> No way -- I've seen much worse.  :-)
[01:41] <mathrick> wheels: you're kidding
[01:41] Nick change: kh_zZz -> khalek
[01:42] <wheels> mathrick: I've seen some stuff dubbed into Spanish that's just horrible...
[01:42] <wheels> mathrick: Like at least in .de there's some attempt to get pretty close to what the lips are doing.  ;-)
[01:42] <mathrick> wheels: humm, never tried that. But .de is gross enough :)
[01:43] <wheels> mathrick: Yeah, I can't handle it on principal.  Even though I can usually understand the German at this point I refuse to see something dubbed out of my native language into another one.  :-)
[01:44] <mathrick> wheels: heh, the absoluttely worst thing I've ever happened to watch was probably uber gross hentai anime with absolutely disgusting .de sub... boy, that was horrible :\
[01:44] <mathrick> s/sub/dub/ of course
[01:48] <mathrick> http://www.mines.edu/students/b/bolmstea/malbolge/
[01:50] <wheels> mathrick: Must be GObject based.  :-P  
[01:51] Action: wheels can't help the jabs every once in a while.  I've been conditioned for far too long...
[01:51] <mathrick> wheels: no, but it's so evil that most complex piece of code written in it is NOP ;)
[01:51] <mathrick> wheels: gah. GObject is extremely clean design, IMO better than C++ for example :)
[01:52] <wheels> mathrick: Actually I think it's not bad for C.  But I just find subclassing in C always a bit ad-hoc...
[01:53] <wheels> And it requires much more work than "class Foo : public Bar {}"  :-)
[01:53] <mathrick> wheels: you mean whole that macro and boilerplate stuff?
[01:53] <wheels> mathrick: yeah
[01:53] <mathrick> wheels: there's GOB for that, if you find it really cumbersome
[01:55] <wheels> mathrick: Well -- but here's what I always come back to -- why not just use a language that's meant for OOP...
[01:55] <mathrick> and I think some meta macro was added to Glib 2.4, so it looks like inheritance and class creation is something like one line anyway
[01:55] <mathrick> wheels: C++ is OO hack on C anyway
[01:55] <wheels> mathrick: Like I'm not usually a C++ zealot, but for OOP I still really prefer it.  (I'm mostly a C programmer at work.)
[01:55] <mathrick> wheels: and I find multiple inheritance way inferior to interfaces
[01:56] <wheels> mathrick: C++ supports MI
[01:56] <wheels> mathrick: It's Java that does interfaces...
[01:56] <mathrick> wheels: I know, that's why I don't like it
[01:56] <wheels> mathrick: Oh, I misread...
[01:57] <wheels> mathrick: I think that's arguable.  I'd actually appreciate a language that supported both.  But really to support interfaces instead of MI you need a top-to-bottem OOP structure to the language with a base object (i.e. Objective-C or Java).
[01:57] <wheels> mathrick: But you can obviously simulate interfaces trivially with MI...
[01:58] <mathrick> wheels: I must admit, templates are nice, and STL is really something, but I don't think C++ is some killer lang
[01:58] <mathrick> wheels: not always true, unfortunately
[01:59] <wheels> mathrick: No, me either.  I think in 5-10 years the C++ programmers will be the stubborn ones that people say the same things about that we say about C programmers now.  :-)
[01:59] Kurosu__ ([email protected]) joined #gstreamer.
[02:00] <mathrick> wheels: I was badly bitten by MI, and that become my primary gripe over C++
[02:00] <mathrick> s/become/became/
[02:01] <Company> C++ is the GStreamer of programming languages
[02:01] <mathrick> Company: care to elaborate?
[02:01] <Company> it can do everything and the output is real cool, but it's hard to do right
[02:01] <wheels> Yeah -- really that's a problem with a lot of languages -- supporting too much syntax.
[02:02] <mathrick> wheels: and why do you say you need top-to-bottom iheritance for ifaces?
[02:02] <Company> the subset of syntax i like is named C ;)
[02:02] <mathrick> true
[02:03] <mathrick> C++ is no more that cute small language fitting in its entirety into thin book, together with programming tutorial :)
[02:03] <wheels> mathrick: It doesn't really make sense to restrict yourself to single-inheritance if you can't function with the assumption that you should always be able to stick a certain set of functionality into the right point of the inheritance hierarchy.
[02:04] <mathrick> wheels: hmm, you lost me a bit
[02:04] <wheels> Company: Well, but C is way too limited.  Whereas C++ went too far, C was about right for 1970.  ;-)  There are just some things that aren't natural with it -- like OOP.  And there are some things that are really nice like type safety with generics...
[02:05] <Company> true
[02:05] Shoragan ([email protected]) left irc: "Leaving"
[02:05] <mathrick> wheels: templates == generics? or something else?
[02:06] <wheels> mathrick: Well, MI is basically just so that you can stuff things together that weren't always designed to work together.  If everything in the language were an object you can say, "Well, this should be stuck into the hierarchy between these level..."
[02:06] <Company> MI does too much wrong
[02:06] <wheels> mathrick: Yeah.  Java calls them "generics" and the general C++ term for it is "generic programming".
[02:06] <mathrick> wheels: as are interfaces meant for that
[02:07] <Company> interfaces suck for that purpose
[02:07] <wheels> Company: Well, it's a convenience -- but of course one can ask at what cost.  It saves a lot of forwarding methods.  :-)
[02:07] <Company> they don't do enough
[02:07] <mathrick> wheels: primary difference is that interfaces expose you to a bit more of internal stuff, but in exchange don't get in your way when you need that extra bit of functionality
[02:08] <wheels> mathrick: hmm?
[02:08] <wheels> mathrick: Well, interfaces are really a subset of MI -- like you can simulate interfaces perfectly in C++...
[02:08] Marsupilami23 ([email protected]) joined #gstreamer.
[02:09] <wheels> mathrick: class FooInterface { public: virtual void doSomething() = 0; }
[02:10] <Company> i'd be pretty much happy if an interface where a class
[02:10] <wheels> Company: Oh, another thing I want in C -- overloading.  :-)
[02:10] <mathrick> wheels: well, true, as long as you din't find suddenly that API you're going to use wasn't using virtual inheritance for standard interfaces
[02:10] <Company> so you could only do multiple inheritance with classes specifically meant for that purpose
[02:10] <mathrick> wheels: then it becomes painful
[02:11] <Company> wheels: i'll never touch a language allowing the user to add overloading
[02:11] <Company> :p
[02:11] <mathrick> hmm, overloading is nice, it's just hard to get 100% right :)
[02:12] <Company> overloading is allowing abbreviating for laziness' sake
[02:13] <wheels> Company: Well, or readability's sake.  ;-)
[02:13] <Company> i don't call that readability
[02:13] <Company> x << y;
[02:13] <Company> very readable
[02:13] <wheels> Company: Well, operator overloading is a different debate.  I meant function overloading.
[02:14] <Company> function overloading requires OOP
[02:14] <wheels> Not really.
[02:14] <wheels> printSomething(int i);
[02:14] <wheels> printSomething(const char *s);
[02:14] <mathrick> add(int, int); add(float, float)
[02:14] <Company> yeah
[02:15] <Company> ok
[02:15] <wheels> printSomething(const struct Foo *f);
[02:15] <Company> it's kinda the same then OOP
[02:15] <mathrick> not really
[02:15] <Company> x.printSomething();
[02:16] <mathrick> myfloat = add(10.0, 20.0); myint = add(10, 20);
[02:16] <mathrick> there's nothing OO about it
[02:16] <mathrick> just syntactic sugar
[02:16] <Company> well, the difference between OOP and !OOP is that OOP prints the first argument in front of the object
[02:16] <Company> s/object/function/
[02:17] <Company> so it's a.func (b) instead of func (a, b)
[02:17] <mathrick> hmm, that means GObject isn't real OOP then?
[02:17] Action: wheels won't answer  :-)
[02:17] <Company> no, GObject isn't real OOP
[02:18] <mathrick> I must disagree here
[02:18] <Company> GObject is emulating OOP with C
[02:18] <mathrick> it just lacks sugar in C, but is first class OOP
[02:18] <Company> though you can do a->func (b) even in C  ;)
[02:18] BBB ([email protected]) joined #gstreamer.
[02:18] <mathrick> OOP is paradigm, not syntax :)
[02:18] <Company> actually you need a->func (a, b) but who cares
[02:18] <mathrick> yo BBB
[02:19] <Company> that's the nice thing about C - it warns you when it overloads functions
[02:19] <BBB> hi
[02:19] <mathrick> Company: what do you mean by "overloads" here?
[02:20] Kurosu ([email protected]) left irc: Read error: 110 (Connection timed out)
[02:20] <Company> mathrick: calling a->func (a, b) is overloading func
[02:20] Nick change: Kurosu__ -> Kurosu
[02:20] <mathrick> Company: howso?
[02:21] <mathrick> in my dictionary, overloading is registering more than 1 function with the same name in common scope
[02:22] <Company> yeah, yeah, it's not totally automatic
[02:22] <Company> but it's the same principle
[02:23] <mathrick> Company: I'd rather say it's virtual function invocation, rather than overloading
[02:24] <Company> whatever, you can pretty much achieve the same results with it
[02:24] <wheels> mathrick: And there's one that I'd actually not mind seeing go in C++ -- I prefer the Java way of things being virtual by default.
[02:25] <wheels> (Of course in C++ it's a performance concern since having virtual methods implies new BC and size requirements...)
[02:25] <mathrick> BC?
[02:25] <wheels> "binary compatibility"
[02:26] <wheels> Sorry, common usage in the KDE world -- "BC" and "BIC"...
[02:26] <mathrick> ie, incompatibility with C?
[02:27] <BBB> double declarations of the same function
[02:27] <wheels> mathrick: No, you can't reorder the virtual functions in a C++ class that part of a public API -- or add virtual functions -- they mess up the vtable.
[02:28] foser ([email protected]) left irc: "[ I want to believe ]"
[02:28] <mathrick> wheels: ahh, I didn't know you can't add vfunc w/o breaking BC
[02:28] BBB ([email protected]) left irc: "Client exiting"
[02:28] BBB ([email protected]) joined #gstreamer.
[02:28] <mathrick> reordering always breaks it, I think
[02:29] <wheels> mathrick: Since behind the scenes there's just a static per-class table with function pointers to virtual functions in the same order as they occur in the header.  That table size can't change (since it's part of the object size) and if you move things around in the header the table entries will point to the wrong functions in shared objects.  :-)
[02:29] <BBB> did I miss anything?
[02:29] <mathrick> BBB: no
[02:29] <BBB> my chat clients hates me
[02:29] <BBB> ok
[02:29] <wheels> BBB: mathrick insulted your mother, but other than that...  :-)
[02:29] <BBB> oh, that's normal
[02:29] <BBB> I always do the same thing to his
[02:30] Action: BBB updates gst-rec to gtk-2.3
[02:30] Action: BBB sort of likes the new GtkComboBox
[02:30] Action: BBB still needs to find images
[02:30] pb_ (~pb@2002:5169:e72f:1:240:95ff:fe30:aa84) left irc: "Client exiting"
[02:30] <mathrick> images of what?
[02:31] <BBB> think about codecs
[02:32] <BBB> I'd like a small image representing the file format, the video source/audio source
[02:32] <BBB> such things
[02:32] <BBB> like a soundcard for 'osssrc', and a wave-image (sinus) for sinesrc
[02:32] <BBB> and white for silence ;)
[02:32] <BBB> such things
[02:32] <mathrick> ahh, that things
[02:32] <BBB> dunno how exactly, but that's possible with GtkComboBox
[02:33] <BBB> with GtkOptionMenu, that was pure evil
[02:33] <mathrick> I still think about that "hint" property for elements
[02:33] <BBB> hint property?
[02:33] <BBB> email the list to explain what you mean ;)
[02:34] <mathrick> BBB: yep, something that would give you hint about how to interpret other props, what the element does, how to organize UI maybe
[02:35] <BBB> ah, that
[02:35] <BBB> maybe :)
[02:35] <BBB> we'll see... I'd say such things shouldn't be properties but interfaces, but anyway ;)
[02:35] <BBB> it'd be really cool to be able to specify ('hint') a property descriptive image in gobject
[02:35] <mathrick> hehe, how do you call interface for XviD encoder then?
[02:36] <BBB> but that's probably too far
[02:36] <BBB> dunno yet
[02:36] <BBB> I'm still looking at gst-rec as a testbed for cool encoding features in gst
[02:36] <BBB> it really rocks
[02:36] <mathrick> BBB: that's the kind of things I'd like to see in it. Dunno if images are still reasonable, but who knows :)
[02:36] <BBB> it just crashes easily
[02:36] <BBB> I get what you mean... dunno
[02:36] <BBB> maybe we should ask owen
[02:37] <mathrick> BBB: maybe, I'll think that over and come up with some writeup... someday ;)
[02:38] <BBB> heh :)
[02:38] Action: BBB has a sweet profile editor and interface locally
[02:38] <BBB> it's waiting for broader testing (read: commit to CVS) once 0.8 is branched of off HEAD
[02:38] <mathrick> ok, it's 2:39AM, a bit late :)
[02:39] <mathrick> nite
[02:39] <BBB> g'ntie
[02:39] Nick change: mathrick -> mathrick|sleep
[02:48] khalek ([email protected]) left irc: Connection timed out
[02:56] smoke ([email protected]) got netsplit.
[02:59] sublett ([email protected]) left irc: "Why are all the really good songs short?"


-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click