From: <wim...@ch...> - 2001-08-08 04:32:25
|
[06:02] Nick change: evil_monkey -> e_monkzzz [06:17] ram (ra...@w-...) left irc: Ping timeout for ram[w-1= 4n251.sequent.com] [06:22] ram (ra...@w-...) joined #gstreamer. [06:37] chillywilly (bau...@d1...) left irc: Re= ad error to chillywilly[d111.as1.nwbl0.wi.voyager.net]: EOF from client [06:45] chillywilly (bau...@d1...) joined #gst= reamer. [06:58] Kuroyi (ri...@ub...) left irc: Ping = timeout for Kuroyi[ubr-35.82.223.wmelbourne.cfl.rr.com] [06:58] jerwin (je...@66...) left= irc: [x]chat [07:00] Kuroyi (ri...@ub...) joined #gstream= er. [07:01] jerwin (je...@66...) join= ed #gstreamer. [07:01] jerwin (je...@66...) left= irc: [x]chat [07:05] chillywilly (bau...@d1...) left irc:=20 [07:17] ram (ra...@w-...) left irc: Ping timeout for ram[w-1= 4n251.sequent.com] [07:54] omega_ (om...@om...) got netsplit. [07:54] taaz (dlehn@66.37.66.32) got netsplit. [07:54] wtay-zZz (wi...@ca...) got netsplit. [07:54] vektor (cooldude@HSE-Kitchener-ppp231173.sympatico.ca) got netspl= it. [07:54] Kuroyi (ri...@ub...) got netsplit. [07:54] terrys_zzz (te...@hs...) got netsplit. [07:55] wtay-zZz (wi...@ca...) returned to #g= streamer. [07:55] vektor (cooldude@HSE-Kitchener-ppp231173.sympatico.ca) returned t= o #gstreamer. [07:55] omega_ (om...@om...) returned to #gstreamer. [07:59] vektor (cooldude@HSE-Kitchener-ppp231173.sympatico.ca) got netspl= it. [07:59] omega_ (om...@om...) got netsplit. [07:59] omega_omicron (om...@om...) joined #gstreamer. [07:59] taaz (dlehn@66.37.66.32) returned to #gstreamer. [07:59] Kuroyi_ (ri...@ub...) joined #gstrea= mer. [08:05] terrys_zzz (te...@hs...) got lost in the net-split. [08:05] Kuroyi (ri...@ub...) got lost in the= net-split. [08:09] Topic changed on #gstreamer by ChanServ!s@ChanServ: 0.2.1 release= : http://gstreamer.net/releases/0.2.1/ [08:09] Nick change: omega_omicron -> omega_ [08:09] Possible future nick collision: omega_ [08:10] vektor (cooldude@HSE-Kitchener-ppp231173.sympatico.ca) got lost i= n the net-split. [09:05] Jacmet (ja...@98...) joined #gstreamer. [09:06] <Jacmet> hi [09:08] arik (ar...@sd...) joined #gstreamer. [09:08] <arik> lo all [09:13] <Jacmet> hi arik=20 [09:13] <arik> howdy [10:08] Nick change: ajbusy -> ajmitch [10:09] <arik> wb [10:09] <ajmitch> hi ;) [10:09] thomasvs (th...@21...) joined #gstream= er. [10:10] <thomasvs> morning/evening [10:10] <ajmitch> hi [10:10] <arik> school is finally done ;-) [10:11] <arik> which means [10:11] <arik> gstplay hacking [10:11] <ajmitch> lucky you ;) [10:11] <thomasvs> arik: good for you [10:11] <arik> ;-) [10:11] <arik> finally [10:11] <ajmitch> more hacking time? [10:11] <ajmitch> yay, so we'll get a great gstplay in a couple of days? = ;) [10:11] <arik> hehe [10:11] <arik> well maybe not a couple days [10:12] <arik> bit i do finally want to try to fix stop [10:12] <arik> get playlist added [10:12] <arik> and a few other things [10:12] <ajmitch> omega_ has started to commit some of the event system [10:12] <ajmitch> i think stop might work soon [10:12] <thomasvs> ajmitch: it does already ;) [10:12] <arik> does it? [10:12] <ajmitch> it does? cool ;) [10:12] <arik> finally [10:12] <arik> couldn't add stop till that was done [10:12] <arik> er [10:12] Action: ajmitch hasn't used it for awhile [10:13] <arik> couldn't add playlist [10:13] <thomasvs> wtay fixed it a few days ago [10:13] Action: arik couldn't compile the lastest cvs today [10:13] <thomasvs> it's probably a hack [10:13] <arik> sweet [10:13] <arik> hehe [10:13] <arik> well it still lets me add playlist [10:13] <arik> and a few other things [10:13] <arik> like repeat [10:14] Action: ajmitch should try & compile cvs=20 [10:14] <thomasvs> arik: of course, no one is stopping you ;) [10:14] <Jacmet> hi thomasvs=20 [10:14] <arik> hehe [10:14] <thomasvs> hi Jacmet [10:14] <arik> i've been waiting to have time for weeksn ow [10:15] Action: ajmitch starts cvs build [10:15] <arik> if it works [10:15] <arik> let me know [10:15] <arik> cause i failed on errors [10:15] <ajmitch> ok [10:16] <thomasvs> hey guys, can I ask your opinion on something ? [10:16] <arik> sure [10:16] <ajmitch> arik: i'm gonna leave for a couple of hours tho [10:16] <arik> heh [10:16] <thomasvs> Ok, I have an idea for a simple user interface I'm goi= ng to code [10:16] <ajmitch> thomasvs: go ahead... [10:16] <arik> well, tell me tonmorrow then [10:16] <ajmitch> :) [10:16] <thomasvs> it uses transparent icons overlayed on top of the desk= top [10:16] <ajmitch> interesting... [10:16] <arik> cool [10:16] <thomasvs> the icon system is tree-based, in that going to the ri= ght for an icon opens up a new menu [10:16] <arik> hmm [10:17] <thomasvs> so actually, you have a tree, and each further item wo= uld be a GList of new menu icons [10:17] <arik> not positive i follow you, but ok [10:17] <thomasvs> so here's the deal : I have the library for displaying= xpm's made in C ... [10:17] <thomasvs> ... I know I could get this working a lot faster in pe= rl than in C if I wrote perl bindings ... [10:17] <thomasvs> ... but is a perl interface not cool enough ? [10:17] <thomasvs> meaning, too slow or whatever ? [10:17] <arik> perl is fine i would think [10:18] <thomasvs> arik: if you want a more detailed idea of how it works= , see the top article at http://davedina.apestaart.org/ [10:18] <arik> alrigh [10:18] <thomasvs> it has some graphics to explain the concept [10:18] <ajmitch> anything slower than hand-optimised asm would crawl on = this computer ;) [10:18] <Jacmet> thomasvs: the perlbindings are for the highlevel stuff, = right? then it should be fine [10:19] <ajmitch> actually, asm would crawl too, time to get a new comput= er ;) [10:19] <Jacmet> ajmitch ;) [10:19] Action: ajmitch waits for gstreamer to just get to 'configure' [10:19] <thomasvs> Jacmet: yeah, for stuff like "take this icon and show = that at that place on the desktop" [10:19] Action: Jacmet coded alphablending routines in x86 (mmx) asm the = other day ;) [10:19] <Jacmet> thomasvs: yes [10:19] <omega_> Jacmet: oooh, me want [10:20] arik (ar...@sd...) left irc: Read er= ror to arik[sdn-ar-001waseatP290.dialsprint.net]: EOF from client [10:20] <Jacmet> omega_: ok - I might clean things up a bit and release i= t [10:21] <omega_> ok, it's something that should go in a libcodec-style li= brary soon [10:21] <Jacmet> omega_: I use it for an OOP system - but you can just th= row away the C++ stuff, if you only wants the blending functions ;) [10:21] <Jacmet> omega_: yes [10:21] <omega_> right [10:22] <omega_> now, vektor was telling me that there's a lot more to al= pha than meets the eye..... [10:22] <omega_> any thoughts on that? [10:22] <Jacmet> omega_: I've made varios 'add/subtract images with satur= ation', overlay functions and so on [10:22] <Jacmet> omega_: well, there are various blending modes (as gimp = has) [10:23] <omega_> he indicated that there were fundamental color issues wi= th alpha, probably nonlinearity and such [10:23] Action: ajmitch leaves now [10:23] <Jacmet> like multiply, maxiumum, minimum and so on [10:23] Nick change: ajmitch -> ajbusy [10:23] <omega_> I couldn't get him to write it up in an email to me ;-( [10:23] <Jacmet> omega_ ;) [10:24] <Jacmet> omega_: I only made blend with a fixed ration between sr= c and dst for all of the image, and per pixel alphablendning (using alpha= value of the src to decide the ratio between src and dst) [10:24] <omega_> right [10:25] <Jacmet> but ofcause the rest can be made ;) [10:25] <omega_> yeah [10:26] <Jacmet> oh, and I have done my routines in 32bit RGBA - I notice= d that the sdlvideosink uses YUV [10:26] <omega_> yeah, I think that may have been what vektor was talking= about more than linear rgba [10:26] <Jacmet> ok [10:26] <omega_> I can see how that could be really hard, probably LUT-ba= sed [10:27] <Jacmet> ofcause we have the colorspace filter, but that probably= wouldn't be good performancewise [10:27] <omega_> definitely want to alpha-blend on the native YUV... [10:27] <omega_> (thinking DVD here) [10:28] <Jacmet> omega_: yes - but where are you going to get the alpha v= alues from ? (no alpha in YUV [10:28] <omega_> subtitles in DVD have a 4-bit alpha in addition to a 4-b= it palettized color [10:28] <omega_> subtitles and menus [10:28] <Jacmet> ok - I don't know anything about dvd ;) [10:31] <Jacmet> Ok, I wouldn't mind looking into coding routines for suc= h.. [10:31] <omega_> problem is researching the issues with alpha and YUV... [10:31] <Jacmet> but we better make sure we are doing it probably the fir= st time, before doing too much asm coding ;) [10:31] <omega_> heh [11:19] Action: omega_ must sleep [11:19] <omega_> l8r [11:19] omega_ (om...@om...) left irc: [x]chat [11:40] Nick change: ajbusy -> ajmitch [13:19] asmod (stevec@64.5.222.2) joined #gstreamer. [14:12] Jacmet (ja...@98...) got netsplit. [14:12] Jacmet (ja...@98...) returned to #gstreame= r. [14:13] <asmod> Hey Jacmet [14:16] terrys (te...@hs...) joined #gstreamer. [15:28] Nick change: e_monkzzz -> evil_monkey [15:54] <Jacmet> /quit [15:54] <Jacmet> argh ;) [15:54] Jacmet (ja...@98...) left irc: =02BitchX=02= : a new fragrance for men, by Calvin Klein [16:07] Nick change: ajmitch -> ajzzzz [16:09] vektor (cooldude@HSE-Kitchener-ppp231173.sympatico.ca) joined #gs= treamer. [17:04] omega_ (om...@om...) joined #gstreamer. [17:05] <asmod> Hey omega [17:05] <omega_> yo [17:05] <vektor> Hey omega. [17:05] <vektor> Question. [17:05] <vektor> Do you set realtime priority on a per-thread basis or pe= r-process basis? [17:05] <vektor> It's per-thread, right? [17:06] <omega_> per-thread afaik [17:08] <asmod> omega_: When I play videos (mpeg specifically) the audio = and video streams are out of sink. [17:08] <vektor> asmod: s/sink/sync/ [17:09] <omega_> yeah, the timestamp handling of the mpeg demuxer isn't i= deal afaict [17:09] <asmod> Good call vektor :) [17:10] <asmod> omega_: So it's just a mpeg problem. Not an overall gstre= amer problem. [17:11] <vektor> omega_: How can you get it so wrong? [17:11] <vektor> :) [17:11] <vektor> I added some neat experimental code to movietime which t= imes how long it takes to blit to the screen, then uses the average offse= t to guess how early I should start drawing future frames. [17:12] <omega_> should be added to videosink [17:12] <vektor> I've also added my code to throttle based on /dev/rtc. [17:13] <vektor> With the linux scheduler (not as root), you get about a = resolution of +- 1500 PTS values for your frame blits. [17:13] <vektor> That's like, fucking ridiculous. [17:13] <vektor> For proper video playback, I've decided you MUST run as = root. [17:13] <omega_> is that with a lowlat kernel or not? [17:13] <vektor> Then, you can sync off the refresh signal on port 3da an= d throttle using /dev/rtc, or just go by the system clock and /dev/rtc. [17:14] <vektor> I'm not sure how much benefit there is to using the ll k= ernel. [17:14] <vektor> Since like, well, without it you should only miss an int= errupt when there's like a load of disk activity or something. [17:14] <vektor> Which shouldn't happen too much. [17:14] <vektor> But I'll do the timings without ll tonight and see how g= ood I can get it. [17:14] <vektor> After my German exam. [17:14] Action: vektor has german tonight. [17:14] <thomasvs> anyone still have autoconf 2.13 ? [17:14] Action: vektor sucks ass at german. [17:15] Action: omega_ does [17:15] <thomasvs> vektor: don't mention the war, you'll be fine [17:15] <vektor> :) [17:15] <thomasvs> omega_: you never had problems with AC_TRUE and AC_FAL= SE undefined during autogen ? [17:15] Action: vektor idle [17:15] <omega_> I did, yes [17:16] <thomasvs> omega_: ok, I think I found the root of the problem. = Did you upgrade to 2.50 ? [17:16] <omega_> nope [17:16] <omega_> testing CVS HEAD now with 2.13 [17:17] <omega_> hrm, still spews [17:17] <thomasvs> oh, so how did you solve it then ? or not solved yet ?= I think I have a simple solution [17:17] <omega_> still has HAVE_FLAC [17:17] <thomasvs> omega_: look for the lines in configure.base that ment= ions thomas, I just put that in [17:17] <thomasvs> uncomment the last of three, comment the first [17:17] <thomasvs> re-run, and no errors [17:17] <thomasvs> Anyway, I solved it for me by renaming HAVE_FLAC to HA= VE_FLACLIB everywhere ... [17:18] <thomasvs> ... and not check for an actual flac function but chec= k with function main [17:18] <thomasvs> seems to work [17:18] <omega_> ok, go ahead and commit the complete set of changes as n= eeded [17:18] Action: omega_ goes to prepare for camping [17:18] <thomasvs> omega_: ok, will do [17:19] <omega_> make sure you update the relevant Makefile.am's <g> [17:19] <thomasvs> omega_: I'll check out fresh after committing and re-t= ry ;) [17:19] <omega_> ok [17:19] <thomasvs> so, like in an hour or so ;) [17:19] terrys (te...@hs...) left irc: Ping timeout for terry= s[hs2-236.magma.ca] [17:23] <asmod> omega_: Is there any way for me to start looking at the n= ew event system stuff? Is it in CVS? [17:28] <thomasvs> asmod: basic code is [17:28] terrys (te...@hs...) joined #gstreamer. [17:28] <taaz> yo, if autoconf is broken fix it, dont just add crappy hac= ks to our code [17:29] <taaz> i'm using 2.52 and haven't seen that error you all are com= plaining about [17:38] <thomasvs> taaz: I'm pretty sure that newest autoconf will fix it= , but that doesn't mean we can't work around it if it doesn't break anyth= ing else [17:40] <omega_> taaz: keep in mind that we do have to maintain backwards= compatibility still ;-( [17:41] Nick change: wtay-zZz -> wtay [17:41] <wtay> yo [17:41] <wtay> omega_: you mind if I hack some more on events? [17:42] <asmod> hey wtay [17:42] <wtay> hi asmod [17:42] <vektor> I'm now convinced you need to be root for smooth video p= layback. [17:42] <wtay> vektor: or set to realtime prio and drop privileges after = that.. [17:43] <omega_> wtay: do so really quickly <g> [17:43] <omega_> or make a list of what you're gonna change, because I'm = out of here in probably 30min [17:44] <wtay> I dunno what I'm about to change really.. probably making = EOS event passing as it used to be.. [17:44] <vektor> wtay: Ah, yes, that would maybe work. [17:44] <wtay> omega_: I'll start with removing the qos/eos crap we have = now [17:45] <wtay> omega_: is the idea still to create an EOS struct that sub= classes GstEvent? [17:47] jerwin (je...@66...) joined= #gstreamer. [17:47] <wtay> hi [17:53] <thomasvs> wtay: are you going to check out cvs ? [17:55] <omega_> wtay: um, not sure [17:55] <omega_> but if I'm gonna work on eos while I'm gone, we're gonna= have to not collide for 2+ days [17:55] <wtay> thomasvs: sure, soon [17:57] <wtay> omega_: I'll wait then [17:57] <omega_> ok [17:57] <omega_> work on something else, like testing, then <g> [17:57] <omega_> quick brainstorm though: what major projects need to be = done still? [17:57] Action: omega_ will brb [17:58] <asmod> Does seeking forward and back work now? [17:58] <wtay> asmod: somewhat, nothing has changed [17:59] steveb (st...@no...) joined #gstreamer. [18:00] <wtay> omega_: can I try a tresting framework with xsl (looks lik= e a fun thing...)? [18:00] <wtay> s/trest/test [18:01] <asmod> Ok. I'm just trying to think of what stuff is left to be = done from my perspective (application development). [18:02] <omega_> wtay: ok, sounds good [18:02] <wtay> omega_: thinking about linking plugins into an app (no dlo= pen) looks interesting too [18:02] <omega_> yup [18:02] <wtay> and it probably requires an API change too.. [18:03] <omega_> need to restructure the plugin init struct so it's a mac= ro, then switch the macro to a __constructor when building in [18:03] <asmod> I've finally got a project started on Mozdev.org for the = mozilla plugin. [18:03] <omega_> (gcc-specific) [18:03] <wtay> omega_: yup [18:04] <omega_> the player lib probably needs major work, esp once we fi= gure out how to get the autoplug to work better, in an element form [18:05] <wtay> probably. [18:05] <omega_> the autoplug code needs a major restructuring, including= initial and incremental graph generation and path calculation separated = from the autoplug stuff itself [18:05] <omega_> gstautoplug would probably go away, or be totally restru= ctured, to serve the autoplug element(s) [18:05] <wtay> if you have some time left you could also think about caps= nego.. [18:05] <omega_> um, no <g> [18:06] <wtay> um, yes :) [18:06] <omega_> I don't know how the current code works, let alone how n= ew stuff might <g> [18:06] <wtay> that's why you could think about it <g> [18:06] <omega_> remember, I don't want to have to think too hard..... [18:06] <wtay> oh right.. :) [18:07] <thomasvs> wtay: ok, if you have the time check out cvs and see i= f autogen.sh still works, but without errors [18:07] <omega_> I'm checking it on 2.13 [18:07] <wtay> thomasvs: I never had errors.. [18:07] <wtay> thomasvs: but I'm testing it right now [18:08] <omega_> works [18:10] <wtay> config.status: error: cannot find input file: libs/colorsp= ace/Makefile.in [18:10] <thomasvs> wtay: isn't colorspace as a lib removed ? [18:10] <thomasvs> omega_: do you have flac installed ? [18:10] <wtay> yes, that's why it fails [18:11] <thomasvs> wtay: ok, then I probably haven't removed that. checki= ng [18:12] <thomasvs> wtay: ok, that's gone. [18:12] <thomasvs> other than that, no problems ? [18:12] <wtay> nope, seems to work [18:15] <omega_> ok, I'm gonna shut down now..... l8r all [18:16] <thomasvs> bye omega [18:16] <wtay> cya [18:16] omega_ (om...@om...) left irc: [x]chat [18:17] <asmod> wtay: Is the video4linux plugin capable of handling input= from a TV tuner card? [18:18] <wtay> asmod: yes [18:18] <wtay> I have to eat now.. brb [18:18] Nick change: wtay -> wtay-food [18:27] thomasvs (th...@21...) left irc: [x]ch= at [18:37] jerwin (je...@66...) left i= rc: Read error to jerwin[66-44-60-9.s9.tnt4.lnhva.md.dialup.rcn.com]: Con= nection reset by peer [18:54] Nick change: taaz -> taaz-food [19:02] Zeenix (gg...@ho...) joined #gstreamer. [19:11] <Zeenix> hi [19:21] ChiefHighwater (pa...@te...) joined #gstreamer. [19:28] Zeenix (gg...@ho...) left irc: Ping timeout for = Zeenix[host-151.netzone.net.pk] [19:29] <vektor> So, what's the deal with MPEG2 video PTS values. Are th= ey given for the current coded frame ? [19:30] <vektor> So like, do they go back in time appropriately re: coded= order vs display order ? [19:30] Zeenix (gg...@ho...) joined #gstreamer. [19:30] Nick change: wtay-food -> wtay [19:30] <wtay> vektor: yup [19:31] <Zeenix> wtay: hi [19:31] <vektor> Does each frame have to have a PTS ? [19:31] <wtay> hi [19:31] <vektor> Do you know? [19:31] <wtay> vektor: the PTS is in the PES packets [19:31] <wtay> they don't need a PTS [19:31] <vektor> Sure. [19:31] <vektor> Well, doesn't the first one ? [19:31] <wtay> I suppose [19:31] <vektor> Like, what I mean is, am I guarenteed that the stream wi= ll provide a PTS for every frame ? [19:32] <vektor> Or will I have to guess sometimes ? [19:32] <wtay> I think you'll have to guess [19:32] <vektor> k. [19:32] <vektor> Thanks! :) [19:32] <Zeenix> wtay: did omega or you reached to any conclusion on that= prob. ? [19:32] <wtay> Zeenix: no [19:32] <vektor> wtay: Did I tell you my pulldown remover kicks fucking a= ss? [19:33] <wtay> vektor: heh. no [19:33] <vektor> wtay: It works awesome now, and it's basically stateless. [19:33] <vektor> wtay: NTSC movies are sooooo much more watchable now. [19:33] <wtay> vektor: cool.. is it a gst plugin or a lib? [19:33] <vektor> It's a lib. [19:33] <wtay> can I download it? [19:33] <vektor> Well, need to move some code back into the lib. [19:33] <vektor> Not yet. [19:33] <vektor> Sorry. :( [19:33] <Zeenix> wtay: you mean you & omega couldnt find which plugin was= the prob. in the pipeline, right ? [19:34] <vektor> Well, you can see the logic at movietime.sf.net in the '= videobuffer.c' file. [19:34] <wtay> Zeenix: it seems like udp was the problem.. [19:34] <vektor> There's some comments there describing how I do it, and = some bugs too. [19:34] <wtay> ok [19:34] <vektor> wtay: I'll try and clean it up after my exam tonight. I= jsut got it working in realtime reliably on saturday. [19:34] <vektor> Transformers: The Movie never looked better. [19:35] <vektor> Clean, crisp, progressive-scan goodness. [19:35] <wtay> neat [19:35] <Zeenix> wtay: what would i do now, what would i tell my Dept. Ch= airman about my project, i only have one month more [19:36] <ChiefHighwater> nice mail asmod 8-] [19:36] <wtay> Zeenix: bug omega some more or fix it yourself I suppose [19:37] <asmod> ChiefHighwater: Thanks :) [19:37] <Zeenix> wtay: you always force me indirecty through some way to = sit in the plugin writer's cockpit <g> [19:38] <wtay> Zeenix: yes :) [19:39] <asmod> wtay: Did you have a chance to look at that test code I s= ent you yesterday? [19:40] <Zeenix> wtay: you made it, you can do it better & i think you sh= ould take this responsibility [19:41] <wtay> asmod: not yet, sorry.. [19:41] <wtay> Zeenix: It works for me.. [19:43] <Zeenix> wtay: i am sure, you dont intend to make gst plugins for= you only [19:45] <wtay> Zeenix: no, I expect people to fix my mistakes [19:46] <Zeenix> wtay: sure if they have enough knowlege & experience you= have [19:47] <Zeenix> wtay: & you sure want gst to be completelly transparent = to the application programmers [19:49] <wtay> Zeenix: we really need to debug it on a PC that has the pr= oblem [19:49] <Zeenix> wtay: you are welcome to come to my house [19:50] <wtay> Zeenix: you're paying the plane ticket? <g> [19:50] <wtay> Zeenix: can I send you a new gstupdsrc.c file with some de= bugging info in it? [19:51] <Zeenix> wtay: just send it to me, i'll try ? [19:51] <wtay> ok [19:52] <wtay> email? [19:53] <Zeenix> pro...@li... [19:53] <wtay> pro...@li...? [19:53] <wtay> ok [19:55] <wtay> ok, sent [19:57] <Zeenix> wtay: replace the old one with this one, right ? [19:57] <wtay> yes [19:59] <Zeenix> wtay: didnt recieved it yet=20 [20:03] <Zeenix> wtay: plz re-send it=20 [20:03] <wtay> ok [20:04] <wtay> done [20:04] <Zeenix> wtay: got it [20:05] <Zeenix> wtay: badblocks searches for badblocks, now i that i hav= e found them, how to fix them ? [20:06] <wtay> Zeenix: I usually press "yes" on all questions [20:07] <Zeenix> wtay: there are no question when i run badblocks [20:08] <wtay> Zeenix: badblocks cannot be fixed AFAIK [20:09] <wtay> brb [20:09] <Zeenix> wtay: i know that, but they are no real badblocks i dont= know why i am getting these at the end of each linux partition [20:10] <Zeenix> wtay: but the the only cure of badblocks is to mark them= off for no future use [20:11] chillywilly (bau...@d6...) joined #gstre= amer. [20:11] <Zeenix> wtay: win98 asks to do this when it finds them [20:13] <wtay> yes.. [20:21] ajzzzz (aj...@p3...) left irc: Ping timeout = for ajzzzz[p34-max11.dun.ihug.co.nz] [20:22] Zeenix (gg...@ho...) left irc: Ping timeout for = Zeenix[host-151.netzone.net.pk] [20:25] Zeenix (gg...@ho...) joined #gstreamer. [20:28] Zeenix (gg...@ho...) left irc:=20 [20:29] Nick change: wtay -> wtay-TV [20:33] ram (ra...@w-...) joined #gstreamer. [20:45] ajzzzz (aj...@p3...) joined #gstreamer. [21:28] Jacmet (ja...@98...) joined #gstreamer. [21:28] <Jacmet> hi [21:29] <asmod> Hi Jacmet [21:30] <Jacmet> hi asmod=20 [21:30] Uraeus (csc...@c2...) joined #gstreamer. [21:30] <Uraeus> hi [21:31] <asmod> Hi Uraeus [21:31] Nick change: wtay-TV -> wtay [21:31] <wtay> hi [21:32] <Jacmet> hi Uraeus and wtay=20 [21:33] <Uraeus> today I had a 'Linux is not ready for the commons' exper= ience, I was trying to get my USB scanner to work at it turns out I need = to both patch Sane and the Kernel to do so, ugh, unfortunatly the kernel = patch was for another kernel than the one I use [21:34] <Jacmet> Uraeus: damn [21:34] <Uraeus> changing kernel is not that easy since I use the one fro= m SGI with XFS support [21:34] <Uraeus> so, no scanner for me under linux at this time :) [21:36] <Jacmet> Uraeus: then you have more time for hacking on gstreamer= instead ;) [21:37] <Uraeus> ehem, well only of you define the term hacking very wide= :) [21:37] <Jacmet> Uraeus ;) [21:38] <Jacmet> Uraeus: I am btw going to Norway on thursday, but I'm no= t passing Oslo.. [21:38] <Uraeus> to bad :( [21:39] <Jacmet> yes - but oh well, then I get to see Kristiansand instea= d ;) [21:39] <wtay> Uraeus: maYam is still learning norwegian :) [21:39] <Uraeus> Jacmet: should be done in a day :), actually I served th= e first three months of my army time near Kristiansand [21:40] <Uraeus> wtay: du f=E5r hilse henne fra meg og si lykke til [21:40] <Jacmet> Uraeus: Ok ;) We are goint to sail to Kristiansand and t= hen follow the coastline north [21:40] Action: wtay calls maYam to translate that... [21:40] <Jacmet> wtay ;) [21:40] <vektor> I wanted to buy a scanner a few months back. [21:41] <vektor> I didn't because I found out that linux sucks ass. [21:41] <Jacmet> wtay: any specific reason she is learning norwegian? [21:42] <Uraeus> vektor: actually it seems that if you have a scanner San= e supports by default it should be relativly easy [21:43] <vektor> Uraeus: And finding a scanner Sane supports is not easy. [21:43] <wtay> Uraeus: uhm... she'll have a conversation with you in a mo= nth or so :-) [21:43] <vektor> Uraeus: All of the scanners out there are USB scanners n= ow, for one. Sane doesn't seem to support much of them. [21:43] <wtay> Uraeus: you said you could help her if she wants? [21:44] <Jacmet> wtay: If she is learning norwegian, she might as well le= arn danish as well, it's almost the same ;) [21:44] <Uraeus> wtay: I can chat with her in norwegian if she wants :) [21:44] <Uraeus> vektor: true [21:45] <wtay> Uraeus: heh, yeah.. when she's able to.. <g> [21:45] jerwin (je...@66...) joi= ned #gstreamer. [21:46] <Uraeus> wtay: we have this thing in Norway we call 'Svorsk' whic= h is how we speak norwegian when speaking with swedes (mixing norwegian a= nd swedish words), I can try doing that with maYam by merging german and = norwegian, the result should be somewhat like flanders :) [21:47] <wtay> Uraeus: ugh [21:48] <Uraeus> wtay: talking with danes like Jacmet is even easier, the= n we just put a potato in our mouth and try talking, sounds exactly like = danish :) [21:49] <Jacmet> Uraeus: hmm... I always thought that was how dutch sound= ed ;) [21:49] <wtay> dutch? not really.. I think.. [21:50] evil_monkey (trevo@208.141.162.68) left irc: Client Exiting [21:50] <Jacmet> wtay: Hollandish dutch, not flamish - It sounds like the= y are trying to swallow something while they speak ;) [21:51] <Uraeus> swallow something while they speak =3D=3D danish [21:51] <Uraeus> <g> [21:51] <wtay> Jacmet: hmm, possibly.. that's not really how I hear it.. [21:51] <Jacmet> wtay: ok ;) [21:51] <Jacmet> Uraeus ;) [21:51] <vektor> http://www4.netbsd.org/Letters/20010803-dolby.html [21:51] <vektor> Holy fuck dudes! [21:52] <Jacmet> Uraeus: but as I'm going to bring my flamish girlfriend = to Norway, we'll see what she thinks about norwegian ;) [21:53] <Uraeus> Jacmet: Norwegian =3D a language so beutiful even angels= cry of joy hearing it (at least that is what I heard) [21:53] <Uraeus> vektor: hmm, that could be bad [21:53] <Jacmet> Uraeus: and you are modest too ;) [21:54] thomasvs (th...@ad...) joined #gstreame= r. [21:55] <thomasvs> hi [21:55] <Uraeus> Jacmet: yes, modesty has always been the trademark of no= rwegians :) [21:55] <Jacmet> Uraeus ;) [21:55] <Jacmet> hi thomasvs=20 [21:56] evil_monkey (trevo@208.141.162.68) joined #gstreamer. [21:56] <Uraeus> hmm, I thought AC-3 was just a storage format like Quick= time, not something which contained encryption or algotrithms that can be= protected under US law [21:58] <vektor> It's an audio codec similar to MP3. [21:58] <Uraeus> aha, not even close then :) [21:58] <vektor> It uses a psychoacoustic model, afaik. [21:58] evil_monkey (trevo@208.141.162.68) left irc: Client Exiting [21:58] <vektor> It's a fundamental piece of a DVD player. [21:58] <vektor> For example. [21:58] <Uraeus> I know [21:58] <Uraeus> any other sites gotten such letters? [21:58] <vektor> So, now we have CSS _and_ AC3 to contend with. [21:59] <vektor> Not that I know of. [21:59] <vektor> But that was just sent a few days ago. [21:59] Action: steveb installs ac3dec while he still can [21:59] <thomasvs> AC3 is dolby property [21:59] <wtay> ah fuck there goes ac3 [21:59] <Uraeus> well, I would think mailing the ac3dec developers would = be a natural step for the dolby people=20 [21:59] <thomasvs> vektor: how are subtitles stored in DVD's ? as text wi= th timestamps ? [22:00] <Uraeus> thomasvs: as pictures [22:00] <thomasvs> Uraeus: really ? that's stupid. [22:00] <vektor> Pictures or text. [22:00] <vektor> Usually pictures. [22:00] <vektor> How is that stupid? [22:00] <thomasvs> I mean it makes sense, but it makes a lot less sense a= s well [22:00] <wtay> thomasvs: not stupid, unless you have all the funny chines= e chars in your font [22:00] <thomasvs> wtay: yeah, but then you'd go for pictures [22:01] <vektor> You can tdo all kinds of stuff with pictures. [22:01] <vektor> They're RLE encoded. [22:01] <vektor> You can do MST3k style overlays, for example (re: Ghostb= usters). [22:01] <thomasvs> It would be neat if you could enlarge fonts and stuff [22:01] <vektor> thomasvs: Also, all of the menu buttons are done using t= he subtitle overlay functionality. [22:01] <vektor> thomasvs: Makes for easier implementation. [22:01] <thomasvs> Anyway, I'd think it wouldn't have been much of an iss= ue to include it as text as well [22:01] <thomasvs> but I can see how it's easier for hardware players of = course [22:01] <vektor> Line21 CC data is stored as ascii text. [22:01] <vektor> So, sometimes you get it. [22:02] <vektor> But I don't know why they would have put it as text. [22:02] <thomasvs> but suppose you're watching with dutch subtitles and y= ou have norwegian friends over ... [22:02] <vektor> No DVD player can support unicode, unless it's a softwar= e player. [22:02] <thomasvs> .. you could display both subtitles [22:02] <vektor> And the DVD standard definitely does NOT have software p= layers in mind :) [22:02] <thomasvs> vektor: you mean, hardware players don't at the moment= , or never could do unicode ? [22:02] <thomasvs> vektor: heh, true [22:02] <vektor> thomasvs: Unicode is very heavy duty. [22:02] <thomasvs> too bad, xine subtitling makes my screen flash a lot [22:02] <Uraeus> thomasvs: your norwegian friends would probably prefer n= o subtitles if the movies was in english :) [22:03] <thomasvs> I was hoping to use my own project to do the title ove= rlaying [22:03] <vektor> That's because they don't do it properly. [22:03] <wtay> not to mention storing all possible fonts in the DVD playe= r... [22:03] <thomasvs> Uraeus: me too, btw, so it's not that much an issue [22:03] <vektor> I'm going to tackle subtitles soon. [22:03] <vektor> You hve to downsample the chrominance nicely and alpha b= lend correctly. [22:03] <vektor> The xine stuff seems a bit hacked. [22:03] <thomasvs> so, they're stored like kind of transparent pixmaps ? [22:03] <vektor> Kinda. [22:03] <vektor> They have a (shitty) alpha channel. [22:03] <vektor> It's a palletted image. [22:04] <thomasvs> While I'm at it, is there a dvd player that does full-= screen display without rescaling ... [22:04] <vektor> (4 bits of alpha is good enough for nobody) :( [22:04] <thomasvs> ... provided you run x at 720x576 ? [22:04] <vektor> Hah! [22:04] <vektor> Even then you need to scale if it's anamorphic :) [22:04] <thomasvs> vektor: I hope you'll say movietime ? [22:04] <vektor> I dunno, I could easily hack that into movietime. [22:05] <thomasvs> vektor: uh, you've lost me now [22:05] <vektor> I intend to anyways. [22:05] <vektor> DVDs can have two aspect ratios. [22:05] <vektor> xine gets this wrong, btw. [22:05] <thomasvs> vektor: it's no problem to run X at that resolution fo= r me, especially if it'll take some of the glitches out [22:05] <vektor> For TV-output ? [22:05] <thomasvs> vektor: I've noticed, I can't get a fullscreen mode wh= ere it doesn't cut off left and right parts [22:05] <thomasvs> vektor: the glitches besides the ones from field order= errors that is [22:05] <thomasvs> yeah for tv out [22:09] <Jacmet> vektor: so the texts are with 4bit alpha - what colorspa= ce and how many colors? [22:11] Nick change: taaz-food -> taaz [22:11] <vektor> Jacmet: The pallette is a series of Y'CbCr component val= ues. [22:13] <Jacmet> vektor: ok, and a 4bit palette for colors as well? [22:14] <Jacmet> I mean, does it use a palette of 16 colors? [22:16] <wtay> Jacmet: the pallete is 16 colors, the colors are stored in= the IFO file [22:16] <Jacmet> wtay: ok, s=E5 16 colors and 16 'shades' of transparency [22:17] <wtay> Jacmet: I don't think there are shades of transparency, no [22:17] <Jacmet> wtay: didn't it have 4bit of alpha? [22:18] <wtay> not in my code AFAICS, it could be wrong though... [22:18] <Jacmet> [22:06] <vektor> (4 bits of alpha is good enough for nob= ody) :( [22:19] <Jacmet> ok, I just understood that as the texts had 4bits of alp= ha [22:19] <wtay> maybe... [22:19] <wtay> actually, while looking at the code there might be 16 shad= es of alpha after all.. [22:20] <Jacmet> oki [22:21] <wtay> hmm, the gstreamer subtitle merger doesn't use the alpha v= alues... :( [22:21] <wtay> it looks like the alpha values are tied to the colors too.. [22:22] steveb (st...@no...) left irc: Read error to steveb= [node1ee03.a2000.nl]: EOF from client [22:23] <Jacmet> I btw talked with omega today about some mmx routines I = have made for add/sub with saturation, various forms of (alpha)blending a= nd so on - they are all for 32bit BGRA, but they might be of use in gstre= amer.. [22:23] <Jacmet> ofcause with fallback routines in C=20 [22:23] <wtay> Jacmet: of course [22:25] <Jacmet> but it seams most (if not all) video in gstreamer is in = YUV (YCbCr for you, vektor ;) [22:25] <wtay> Jacmet: yeah.. [22:30] <jerwin> Y'CbCr you mean [22:31] <Jacmet> yes ;) [22:32] <wtay> you mean I420 or YV12 or YUV2? [22:33] <Jacmet> wtay: me? the plugins I have seen uses I420 (mpeg_play a= nd sdlvideosink) [22:33] <Uraeus> wtay: GStreamer compiled for me on Solaris yesterday [22:33] <wtay> Uraeus: cool, did it run? [22:33] <Uraeus> wtay: well, I am making some more prereq libs now since = I didn't have any plugins to test with [22:34] <wtay> Uraeus: -launch fakesrc ! fakesink? [22:35] <vektor> thomaHi. [22:35] <vektor> Hi. [22:36] <vektor> wtay: those are just structural formats (and 4:2:0 vs 4:= 2:2 sampling). [22:36] <wtay> vektor: I know.. [22:36] <vektor> k. :) [22:37] <asmod> wtay: Is there a chance you'll be able to look at that te= st code tonight? [22:37] <wtay> asmod: oh, ok... forgot about it again... [22:38] <vektor> Jacmet: Do you use Jim Blinn's algorithm for fast alpha = blending? [22:38] <vektor> Jacmet: I mean fast alpha multiplying. [22:38] <Jacmet> does GST_MAKE_FOURCC handle fourcc aliases like I420 and= IYUV ? [22:38] <Uraeus> wtay: I try that, but I need to compile again since I de= leted my first compile [22:38] <asmod> wtay: Thanks :) [22:38] <wtay> Jacmet: yes.. [22:39] <Jacmet> vektor: hmm, no - I just use src*aplha + dst*(1-alpha) =3D= =3D dst + (src-dst)*alpha [22:40] <Jacmet> wtay: great [22:40] <Jacmet> vektor: sorry, I don't remember Jim's algorithm.. [22:40] <vektor> Ack! [22:40] <vektor> Jacmet: That's not really, erm, correct. :) [22:41] <Jacmet> vektor: It's sommerholidays now =3D=3D brain is not runn= ing at full speed ;) [22:41] <vektor> So, how is your alpha stored? As a double or something? [22:41] <Jacmet> vektor: well, that depends on what you mean by correct. = With this I can decide the ratio between src and dst [22:42] <vektor> And it's constant alpha per image? [22:42] <Jacmet> vektor: no, just the top 8 bits of my 32 bits RGBA [22:42] <Jacmet> vektor: no, its perpixel alphablending [22:42] <vektor> Well, usually if you're doing compositing, you'll be com= positing like 4 or 5 things. So, I don't think it's fair to assume that = you only have two maps, and that their alphas add up to one. [22:42] <Jacmet> vektor: I have made constant blending functions too [22:42] <vektor> So, you have to take the alpha value, do a float divisio= n by 255, and then multiply ? [22:42] <Jacmet> vektor: true [22:42] <vektor> Brutal :) [22:43] <Jacmet> vektor: no - mmx muls (16bit) [22:43] <Jacmet> vektor ;) [22:43] <vektor> But you have to divide by 255=20 [22:43] <Jacmet> vektor: I cheat and divide by 256 ;) [22:43] <vektor> That's nasty. [22:43] <vektor> And that's where Jim comes in. [22:43] <Jacmet> vektor: and yes, I know it's wrong [22:43] Action: Uraeus just made the gstreamer webpage viewable with M$ e= xplorer :) [22:43] <Jacmet> wtay: what is the fourcc code of BI_RGB supposed to be? [22:44] Nick change: ajzzzz -> ajbusy [22:44] <wtay> Jacmet: "RGB " [22:44] <Jacmet> vektor: It's quite long ago since I saw Jim's algorithm = - do you have a link to it? [22:44] <Jacmet> wtay: ok [22:44] <vektor> Jacmet: Also, you should DEFINITELY special case when al= pha =3D=3D 0 and alpha =3D=3D 0xff. This speeds stuff up ALOT. [22:44] <vektor> Jacmet: I'm looking for it now. [22:45] <wtay> asmod: ok, confirmed the problem.. [22:45] <Jacmet> wtay: and "RGBA" for alpha as well? [22:45] <wtay> Jacmet: uhm.. sec.. [22:45] <Jacmet> vektor: Yes I do that as well, the problem is ofcause th= at the framerate isn't constant then [22:46] <asmod> wtay: Do you know what caused it? [22:46] <wtay> Jacmet: http://www.webartz.com/fourcc/ it looks like RGBA = is ok [22:46] <wtay> asmod: not yet.. [22:46] <Jacmet> vektor: even though you time motion after the clock it s= till looks a bit stupid [22:46] <vektor> Jacmet: How does that matter? [22:47] <vektor> Jacmet: I don't understand. You want to take a constant= amount of time per frame? Why? :) [22:47] <vektor> That doesn't make sense :) [22:47] <vektor> If you can save CPU cycles, _do_it_! [22:47] <wtay> asmod: do you actually remove the audiosink from the pipel= ine? [22:47] <asmod> wtay: In that version of the code I attempt to remove the= audio and video sinks from the element returned by the autoplug. I'm not= sure if that's correct. [22:47] <vektor> I have to go to my German exam too. [22:47] <wtay> ok [22:47] <asmod> wtay: I also tried to remove them from the pipeline itsel= f.=20 [22:48] <Jacmet> vektor: I made it for demoeffects - you can see when it = skips from say 70 fps to 35 fps [22:48] <wtay> vektor: viel spass :) [22:48] <ChiefHighwater> wtay:hehe [22:48] <Jacmet> wtay: the last letter of "RGB " - is that a space or a n= ullbyte? [22:48] <wtay> Jacmet: space [22:48] <Jacmet> wtay: ok, thanks [22:49] <vektor> I think this is 'ok': [22:49] <vektor> http://mail.lokigames.com/ml/sdl/9370.html [22:49] <wtay> viel spa=DF even.. [22:49] <vektor> Jacmet: Ah, odd. [22:49] <vektor> wtay: I have no idea what that means :) [22:49] <vektor> Jacmet: I'll find you a better description of it after m= y exam ;-) [22:50] <wtay> vektor: have fun.. :) [22:50] <vektor> wtay: Thanks! [22:50] <Jacmet> vektor: odd? you can see the difference between 70 fps a= nd 35 fps right? - an effects which changes between 70 fps and 35 fps loo= ks fairly shaky [22:50] <Jacmet> vektor: thanks - and good luck! [22:51] <vektor> Jacmet: Sure, but, I would have assumed you'd be outputt= ing at constant framerate, handled at the output. [22:51] <wtay> vektor: tell you teather he's a "schwein" :) [22:51] <vektor> Jacmet: Like, render your frames for a constant output f= ramerate. [22:51] <vektor> wtay: Will do! [22:51] <wtay> vektor: rather, don't :) [22:51] <vektor> ;-) [22:51] Action: vektor idle [exam] [22:52] <Jacmet> vektor: then I waste time at the output instead of in th= e alphablending routine - it more or less gives the same ;) [22:52] <Jacmet> wtay ;) [22:57] <vektor> You shouldn't ever be wasting time. [22:58] <Jacmet> vektor: then how would you guarantee a fixed framerate (= ofcause you can use threads and start processing the next frame, but you = are still going to wait) [23:00] asmod (stevec@64.5.222.2) left irc: [x]chat [23:10] Nick change: wtay -> wtay-q3 [23:11] Nick change: wtay-q3 -> wtay [23:12] <Jacmet> wtay: short game ;) [23:12] <Uraeus> wtay: hmm, not sure I like this new computer gaming habi= t of yours :) [23:12] <wtay> no response from the master server :( [23:12] <Jacmet> heh [23:12] jerwin (je...@66...) lef= t irc: [x]chat [23:12] <Jacmet> wtay: It's all part of our master plan to get you to hac= k on gstreamer instead of gaming ;) [23:13] <wtay> damn.. [23:13] <Uraeus> sh redworm.sh |q3.masterserver.org [23:13] <Jacmet> Uraeus ;) [23:13] <wtay> apt-get update time then :) [23:14] <Uraeus> wtay: we need a Gconf backend option for gstreamer-regis= ter :) [23:15] <wtay> Uraeus: hmm. no [23:16] <Uraeus> wtay: hmm, ok we need a flash plugin?=20 [23:16] <wtay> Uraeus: yeah.. [23:17] <Jacmet> wtay: it should be possible to do something like buzz th= ough gstreamer fairly easy, right? [23:17] <Jacmet> ..not that I have really used buzz ever.. ;) [23:17] <wtay> me neither.. [23:18] <wtay> heh, I just got a sircam virus :) [23:18] <Uraeus> wtay: having a gconf backend to gstreamer-register doesn= 't make any sense? (I am writing my second edition gnome-gstreamer mail) [23:18] <Jacmet> but it seams clever with the filters and stuff [23:18] <wtay> Uraeus: not really, it's more suited for an application or= a gnome-capplet [23:19] <Uraeus> wtay: ok, thanks then I remove that item [23:19] <wtay> Uraeus: the core lib should be as small as possible [23:22] <Uraeus> wtay: yup [23:23] <Jacmet> wtay: I'm not quite sure I understand these mime/fourcc = types - where do you tell bit depth and so on? as gst_props? and whats th= e "official" names of the properties? [23:23] Nick change: taaz -> taaz-away [23:23] Action: Jacmet is confused [23:24] <wtay> Jacmet: look at videosink for an example of valid properti= es [23:24] <Jacmet> .. or just stupid ;) [23:24] <Jacmet> wtay: ok, thanks [23:24] <wtay> Jacmet: they are not official, but choosen by convention [23:24] Action: wtay tries the master server again [23:25] <Jacmet> wtay: ok [23:25] <wtay> hmm [23:25] <Jacmet> videosink? do you mean the sdl or the X one - or any of = them? [23:25] <wtay> xvideosink [23:25] <wtay> gstreamer-inspect xvideosink [23:26] <wtay> for GRB images we sorta use the XVisual stuff [23:26] <wtay> s/GRB/RGB [23:27] <Jacmet> It might be me being stupid, but I can't see anything ab= out the format [23:27] <Jacmet> Pad Templates: [23:27] <Jacmet> SINK template: 'sink' [23:27] <Jacmet> Availability: Always [23:27] <wtay> uhm [23:28] <Jacmet> or should I just grab it from cvs? [23:28] <wtay> well.. my copy is currently broken [23:28] <wtay> Jacmet: not really [23:28] <wtay> "format", GST_PROPS_FOURC= C (GST_MAKE_FOURCC ('R', 'G', 'B', ' ')), [23:28] <wtay> "bpp", GST_PROPS_INT (= GST_XIMAGE_BPP (ximage)), [23:28] <wtay> "depth", GST_PROPS_INT (= GST_XIMAGE_DEPTH (ximage)), [23:28] <wtay> "endianness", GST_PROPS_INT (= GST_XIMAGE_ENDIANNESS (ximage)), [23:28] <wtay> "red_mask", GST_PROPS_INT (= GST_XIMAGE_RED_MASK (ximage)), [23:28] <wtay> "green_mask", GST_PROPS_INT (= GST_XIMAGE_GREEN_MASK (ximage)), [23:29] <wtay> "blue_mask", GST_PROPS_INT (= GST_XIMAGE_BLUE_MASK (ximage)), [23:29] <wtay> "width", GST_PROPS_INT_R= ANGE (0, G_MAXINT), [23:29] <wtay> "height", GST_PROPS_INT_R= ANGE (0, G_MAXINT), [23:29] <Jacmet> wtay: ok, I'll look at the source - thanks again [23:30] <Jacmet> but wouldn't it be a good idea to specify such things? h= ow does the colorspace plugin works if there isn't really a standard for = it? [23:31] <wtay> Jacmet: yes, we should have a written standard for that [23:32] <wtay> Jacmet: you don't see properties in -inspect? [23:32] <Jacmet> wtay: ok, lots of work to do ;) [23:34] <Jacmet> wtay: I have 'Capabilities', but not properties [23:34] <wtay> Capabilities: [23:34] <wtay> 'xvideosink_caps': [23:34] <wtay> MIME type: 'video/raw image/raw': [23:34] <wtay> format: FourCC: 'YUY2' [23:34] <wtay> width: Integer range: 0 - 2147483647 [23:34] <wtay> height: Integer range: 0 - 2147483647 [23:34] <wtay> ? [23:34] <Jacmet> and there are none for xvideosink, but they are there fo= r sdlvideosink [23:35] <Jacmet> gstreamer-inspect xvideosink|grep Capab = ~ [23:35] <Jacmet> INFO ( 927:-1) Initializing GStreamer Core Library [23:35] <Jacmet> INFO ( 927:-1) CPU features: (00000000) MMX=20 [23:35] <Jacmet> none [23:35] <Jacmet> gstreamer-inspect sdlvideosink|grep Capab = ~ [23:35] <Jacmet> INFO ( 929:-1) Initializing GStreamer Core Library [23:35] <Jacmet> INFO ( 929:-1) CPU features: (00000000) MMX=20 [23:35] <Jacmet> Capabilities: [23:36] <wtay> you ran -register with a connection to the X server? [23:36] <Jacmet> ohh, that might actually be why - I su'ed root before.. [23:36] <Jacmet> much better! [23:36] <wtay> yeah, xvideosink probes the X server.. [23:37] <Jacmet> MIME type: 'video/raw image/raw': [23:37] <Jacmet> format: FourCC: 'RGB ' [23:37] <Jacmet> width: Integer range: 0 - 2147483647 [23:37] <Jacmet> height: Integer range: 0 - 2147483647 [23:37] <Jacmet> endianness: Integer: 1234 [23:37] <Jacmet> why is it btw both video/raw and image/raw? [23:39] <wtay> some plugin probably decladed the video/raw mime type equi= valent to image/raw [23:39] <wtay> s/decladed/declared [23:39] <Jacmet> ok - I'm not really into MIME types.. [23:39] <Jacmet> but thanks again [23:55] maYam (ma...@ca...) joined #gstream= er. [23:55] <maYam> hellooo [23:55] <wtay> yo maYam [23:55] <wtay> maYam: how's life? [23:56] <maYam> pff.. [23:56] <maYam> i'm a terrible negotiator and i lack commercial talent [23:57] <Uraeus> hei maYam, hvordan g=E5r norsken? [23:57] <maYam> bwaa.. ;) [23:57] <maYam> god kvelden?? [23:57] <maYam> uhm.. [23:58] <maYam> it's going ok [23:58] <Uraeus> maYam: god natt er vel mer treffende p=E5 denne tiden av= d=F8gnet [23:58] <Uraeus> maYam: correct answer, wrong language :) [23:58] <Jacmet> Uraeus: det kan du have ret i ;) [23:58] <maYam> hmm.. i thought 'god natt' was too personal..? [23:58] <Uraeus> maYam: why is night more personal than evening? [23:58] <maYam> or maybe i'm learning some oldfashioned list of words.. [23:59] <Jacmet> maYam: why are you learing norwegian? [23:59] <maYam> Jacmet: cause it sounds so..sweet :) [23:59] <Jacmet> maYam: naah, danish is better ;) [23:59] <maYam> maybe, but they talk too fast! [23:59] <Jacmet> maYam: but it's quite easy to learn once you know norweg= ian [00:00] --- Wed Aug 8 2001 [00:00] <Jacmet> maYam: Yes, my flamish girlfriend complains about that t= oo ;) [00:00] <Uraeus> maYam: actually you might be right, 'god kveld' is a gre= eting, but 'god natt' is more of a goodbye statement :) [00:00] <maYam> yes, i already saw the resemblance.. [00:00] <maYam> aha! so i'm learning it the right way.. =20 [00:01] <Uraeus> maYam: yeah, even if 'kveld' do mean 'evening' and it is= the middle of the night :) [00:01] <maYam> it's fun, but pronounciation is a bit of a prob [00:01] Action: Jacmet is going to learn flamish in a few months [00:02] <Jacmet> now that is going to be "fun" ;) [00:02] <maYam> Uraeus: maybe you could send me a tape 'best bedtime stor= ies, told by Uraeus' [00:02] <Uraeus> maYam: I would think saying the words with a flamish pro= noncuiation would give you something useable [00:02] <maYam> Jacmet, it's easy.. those scandinavian languages look a = lot like flemish.. [00:02] <maYam> ..i think.. [00:03] <Uraeus> maYam: heh, well I don't know if you want a tape that I = read in, but if you want some tapes with norwegian tales or something I c= an send it too you [00:03] <Jacmet> maYam: yes, written flamish is not so hard, but I'm not = used to some of the sounds - like the 'rolling R' [00:03] <Uraeus> maYam: yeah, I can almost read flemish since the words a= re often so similar to ours [00:03] <maYam> waw, that'd be great actually.. the norwegian balck meta= l i'm listening to, is not always very clear ;) [00:04] <maYam> black [00:04] <Jacmet> heh [00:04] <Uraeus> Jacmet: 'rolling R'? here we call it the spitt sound :) [00:04] <Jacmet> Uraeus: yes, if you mix a little german and english into= it.. [00:04] <Jacmet> Uraeus: heh [00:04] <maYam> haha the spitt sound [00:04] <wtay> do we have a rolling R? [00:05] <Uraeus> yeah [00:05] <maYam> de kat kRabt de kRollen van de tRap [00:05] <Jacmet> wtay: yes - except for the french people who doesn't pro= nounce it [00:05] Nick change: Kuroyi_ -> Kuroyi [00:05] <Jacmet> or atleast, so I'm told=20 [00:06] <maYam> oh wait, Jacmet, you're talking about the french 'r' [00:06] <maYam> french r, that's the spitt sound, yes.. you won't need t= hat.. [00:07] <maYam> confusing.. [00:07] <Jacmet> well, 'goede nacht' from here [00:08] <maYam> woohoo.. ook goeienacht [00:08] <wtay> slaapwel [00:08] <Uraeus> maYam: french R? it is in dutch movies I hear it (not th= at I see that many dutch movies, think the last one was called 'Little Si= ster' or something) [00:09] <Jacmet> jij ook ;) [00:09] Jacmet (ja...@98...) left irc: sleep [00:09] <maYam> Uraeus: some do use the french r, but the majority just s= ays it as in norwegian=20 [00:11] <Uraeus> opps, past midnight, need sleep, I have been choosen to = act as guide tommorow to a new female employee at Oracle, might be a grea= t change to aquire a gf :) [00:11] <Uraeus> see ya [00:11] <maYam> heh.. you get to see dutch films.. strange.. i don't eve= n know Little Sister [00:11] Uraeus (csc...@c2...) left irc: Client Exiting [00:12] maYam (ma...@ca...) left irc: Clien= t Exiting [00:34] thomasvs (th...@ad...) left irc: Client= Exiting [00:49] <ChiefHighwater> gotta love it...someone is using online transfer= s to their credit card to steal money out of the account the church uses = to help those in need [00:50] <ChiefHighwater> not the brightest, since the credit card leads r= ight back to an individual [00:59] <wtay> I gotta sleep, cya [00:59] Nick change: wtay -> wtay-zZz [01:02] ajbusy (aj...@p3...) left irc: Ping timeout fo= r ajbusy[p3-max6.dun.ihug.co.nz] [01:03] terrys (te...@hs...) left irc: Ping timeout for terry= s[hs2-236.magma.ca] [01:04] terrys (te...@hs...) joined #gstreamer. [01:55] gtb (Guillaume@Mix-Lyon-203-2-99.abo.wanadoo.fr) joined #gstreame= r. [01:55] <gtb> hi [01:59] dobey (do...@di...) joine= d #gstreamer. [02:00] Nick change: terrys -> terrys_zzz [02:01] evil_monkey (trevo@208.141.162.68) joined #gstreamer. [02:03] Zeenix (pro...@13...) joined #gstreamer. [02:03] <Zeenix> hi [02:08] dobey (do...@di...) left = irc: Ping timeout for dobey[dialup-63.214.72.235.Dial1.Boston1.Level3.net= ] [02:17] gtb (Guillaume@Mix-Lyon-203-2-99.abo.wanadoo.fr) left irc: Ping t= imeout for gtb[Mix-Lyon-203-2-99.abo.wanadoo.fr] [02:21] <Zeenix> what has happend to the gst site, i can't see the pages [02:21] <Zeenix> ? [02:23] <evil_monkey> Zeenix: works fine for me [02:26] <Zeenix> evil_monkey: strange [02:26] <Zeenix> evil_monkey: i am using IE 5.0 [02:26] <evil_monkey> Zeenix: what browser are you using [02:26] <evil_monkey> Zeenix: I am using galeon=20 [02:27] <Zeenix> evil_monkey: unforutatelly i have a win modem, i couldnt= afford an external one [02:27] <evil_monkey> Zeenix: I feel your pain, I had one too [02:28] <Zeenix> evil_monkey: if you really do, why dont you give me some= loan to buy one <g> ? [02:29] <evil_monkey> Zeenix: cause I am dirt poor and doing so would put= me out of a home [02:30] <Zeenix> evil_monkey: did you see omega today ? [02:30] <evil_monkey> Zeenix: haven't been on much, but no, he was going = on vacation sometime soon though [02:32] <Zeenix> evil_monkey: do you know the url where i can get that ta= r ball containing windows avi codecs for gst ? [02:32] <evil_monkey> for gst, do you mean avifile? [02:32] <evil_monkey> or are they the same [02:33] <Zeenix> evil_monkey: no, the windows dlls & others windows files= ( avi codecs ) needed [02:33] <evil_monkey> Zeenix: I think they are the same dlls, and they ar= e at http://avifile.sourceforge.net/binaries-010122.zip [02:33] <evil_monkey> Zeenix, but other then that I don't know [02:43] <Zeenix> evil_monkey: yes i can see the pages, but i had to move = the slider to see it [02:44] <evil_monkey> Zeenix: weird, works perfect here [02:46] Zeenix (pro...@13...) left irc: Ping timeout for= Zeenix[13-28.brain.net.pk] [02:53] gtb (Guillaume@80.9.175.251) joined #gstreamer. [02:54] <gtb> do i need to have glib2 in order to compile gstreamer cvs ? [02:58] jerwin (je...@66...) join= ed #gstreamer. [03:03] <evil_monkey> gtb: no, it cross compiles to both [03:03] <evil_monkey> gtb: I compiled cvs yesterday, I don't have glib2 [03:08] <gtb> ./configure fail with the folowing error : [03:08] <gtb> ./configure: line 2682: syntax error near unexpected token = `PKG_CHECK_MODULES(GLIB2,' [03:08] <gtb> ./configure: line 2682: ` PKG_CHECK_MODULES(GLIB2, glib-2.= 0 gobject-2.0 gthread-2.0 gmodule-2.0,' [03:09] <gtb> i don't understand wath the problem is [03:09] <gtb> s/wath/what [03:09] <evil_monkey> gtb: hmm, don't know about that, I didn't have a pr= oblem [03:10] <evil_monkey> gtb: let me grab cvs and see if I get the same [03:10] <gtb> evil_monkey: ok, thx! [03:11] <evil_monkey> gtb: what version of pkgconfig do you have? [03:11] <gtb> evil_monkey: 8.0 [03:11] <evil_monkey> gtb: well, it isn't that then [03:13] ram (ra...@w-...) left irc: Read error to ram[w-14n2= 51.sequent.com]: EOF from client [03:13] <evil_monkey> gtb: is this when configure gets run by autogen.sh? [03:14] <gtb> it happen when i launch ./autogen.sh and when i launch ./co= nfigure [03:15] <evil_monkey> gtb: k, just curious [03:17] <gtb> evil_monkey: are you sure you hadn't glib2 when you compile= d the cvs ? [03:17] <evil_monkey> gtb: the only glib I have glib-1.2.10 [03:18] <gtb> evil_monkey: what version of autoconf do you use ? [03:18] <evil_monkey> gtb: 2.13 [03:19] <gtb> evil_monkey: me too [03:19] <evil_monkey> gtb: rh71 by chance? [03:20] <gtb> evil_monkey: mandrake 8.0 [03:20] <evil_monkey> gtb: mine ran just fine so it is something on you m= achine, although I am not sure what [03:21] <gtb> evil_monkey: damn, thank you for having tried ! [03:21] <evil_monkey> gtb: I think that to build wiht glib2 you have to e= nable it yourself [03:21] <evil_monkey> gtb: but it cross compilse just fine [03:29] <gtb> evil_monkey: what version of libtool do you have ? [03:29] <evil_monkey> 1.3.5 [03:29] <evil_monkey> gtb:1.3.5 [03:30] <gtb> evil_monkey: same as me... [03:34] <evil_monkey> gtb: sorry I can't help more [03:36] <evil_monkey> gtb: what module from cvs did you get? [03:36] <gtb> evil_monkey: no prb ! [03:37] <gtb> evil_monkey: all modules [03:37] <evil_monkey> gtb: you didn't get gst-glib2-test did you? [03:38] <gtb> evil_monkey: i did [03:38] <evil_monkey> gtb: only get gstreamer [03:38] <evil_monkey> gtb: so cvs -d:pserver:ano...@cv...ur= ceforge.net:/cvsroot/gstreamer login [03:38] <evil_monkey> gtb: so cvs -d:pserver:ano...@cv...ur= ceforge.net:/cvsroot/gstreamer co gstreamer [03:39] <gtb> evil_monkey: i didn't tried to compile gst-glib2-test [03:39] <gtb> evil_monkey: do you think it can cause a problem ? [03:39] <evil_monkey> gtb: well, it isn't needed [03:39] <evil_monkey> gtb: I would try it with out and see if that helps [03:40] jerwin (je...@66...) left= irc: [x]chat [03:40] <evil_monkey> gtb: I never get the glib2 stuff [03:40] <evil_monkey> gtb: I was just wondering because my line 2682 in c= onfigure has nothing about that error [03:41] <gtb> evil_monkey: what version of autogen do you have ? [03:42] <gtb> evil_monkey: sorry, i have no autogen [03:42] <evil_monkey> gtb: yeah, autogen is just a script that uses autoc= onf and that I think [03:44] <gtb> evil_monkey: i found an autogen package at ftp.gnu.org/pub/= gnu/autogen, don't know what it is... [03:48] <evil_monkey> gtb: does it ever get to the creating of make files= or it die while checking for things [03:51] <gtb> evil_monkey: ??? [03:52] <evil_monkey> gtb: how far does the running autogen.sh get? what = is the last thing it prints out before it quits [03:55] <gtb> evil_monkey: since i have updated libtool and automake, ./a= utogen.sh don't work anymore... ;-( [03:55] <evil_monkey> gtb: that sucks [03:55] <gtb> evil_monkey: yes! ! [03:55] <evil_monkey> gtb: revert back then [03:57] <gtb> evil_monkey: i updated using .tar.gz, so now i don't know h= ow to get back [03:58] <gtb> evil_monkey: i think i will try to keep the updated version= s because i know that omega use them [03:59] <gtb> evil_monkey: i think i'll try to get an updated version of = autoconf... perhaps will it work... [03:59] <evil_monkey> gtb: if you want to revert back and you still hvae = the=20 [03:59] <evil_monkey> gtb: folder you compiled from you can do a make uni= nstall and remove it all [04:00] <evil_monkey> then install the original rpms from your cd :) [04:00] <gtb> evil_monkey: i didn't know that, thx [04:01] <evil_monkey> gtb: neither did I till about a month ago and I dec= ided to try it and it worked ;) [04:02] <gtb> are you doing something on/with gstreamer ? [04:03] <gtb> evil_monkey: are you doing something on/with gstreamer ? [04:04] <evil_monkey> gtb: just messing around with it now and then [04:04] <evil_monkey> gtb: I've thought about writing some plugins but I = don't have time right now [04:04] <gtb> evil_monkey: what kind of plugins ? [04:05] <evil_monkey> gtb: like an asf plugin so gstreamer can do asf vid= eo [04:05] <evil_monkey> gtb: that is the one I have been wanting to do [04:05] cbk (cb...@cx...) joined #gstreamer. [04:06] <evil_monkey> gtb: why do you ask/ [04:07] <gtb> evil_money: just to know. i'm mostly interested by audio pl= ugins [04:08] <evil_monkey> gtb: that is nice thing about gstreamer, it is it a= ll, and it is a nice way of doing things to [04:09] <gtb> evil_monkey: yes, i think so [04:09] Parapraxis (sp...@us...) joined #gstrea= mer. [04:10] <Parapraxis> howdee [04:10] Action: Parapraxis just got back from vacation [04:11] Action: evil_monkey wishes he was on vacation [04:11] Action: evil_monkey tries to think of a way to magically get more= vacation days....... [04:12] chillywilly_ (bau...@d1...) joined #gst= reamer. [04:12] chillywilly (bau...@d6...) left irc: Pin= g timeout for chillywilly[d62.as8.nwbl0.wi.voyager.net] [04:18] Nick change: chillywilly_ -> chillywilly [04:24] <gtb> cya all ! thx evil_monkey ! [04:24] gtb (Guillaume@80.9.175.251) left irc: KVirc 2.1.0 'D... [truncated message content] |