Back to Home / #mythtv / 2003 / 06 / Prev Day | Next Day
#mythtv IRC Logs for 2003-06-19

00:00<billytwowilly>rkulagow__: there's a whole bunch of stuff. here's the only audio one I think: alias sound-slot-0 i810_audio
00:00<billytwowilly>Shado07: I recommend mandrake for a mythtv box. there are really good directions for the mandrake distro.
00:00<rkulagow__>you don't have:
00:00<rkulagow__>above snd-intel8x0 snd-pcm-oss
00:00<rkulagow__>alias sound-slot-0 snd-intel8x0
00:00<rkulagow__>in there?
00:00<billytwowilly>don't think so. let me check
00:01<Shado07>Where are the directions? on the site you are saying/
00:01<billytwowilly>Did I skim again and make a mistake?;)
00:01<billytwowilly>should I put that stuff.
00:01<billytwowilly>Shado07: yah. great installation instructions. I think they talk about redhat too. I'm talking about mythtv.
00:02<rkulagow__>billy: no, not necessarily. that's what i have for my audio; i'm just trying to figure out what you've got. i don't have an nforce board, and from the looks of it, you've got multiple apps complaining that the mixer isn't there.
00:02<Shado07>huh. I'll give it a shot instead of RH9.. whynot ;)
00:02<billytwowilly>rkulagow__: I know;) it's freaking me out;)
00:02<billytwowilly>Shado07: If you have to reinstall you might as well;)
00:03<rkulagow__>billy: do a lsmod | grep mixer
00:03-!-arj [] has joined #mythtv
00:03<billytwowilly>rkulagow__: I got nothing.
00:03<rkulagow__>well, there you go then. that's a clue.
00:04<rkulagow__>[root@livingroom mythtv]# lsmod | grep mixer
00:04<rkulagow__>snd-mixer-oss 14488 0 [snd-pcm-oss]
00:04<rkulagow__>snd 40868 0 [snd-seq-oss snd-seq-midi-event snd-seq snd-pcm-oss snd-mixer-oss snd-intel8x0 snd-pcm snd-timer snd-mpu401-uart snd-rawmidi snd-seq-device snd-ac97-codec]
00:04<billytwowilly>heh, unfortunately, I have no idea how to act upon this clue;)
00:04<rkulagow__>are you running alsa?
00:04<rkulagow__>(you should be, if it's mandrake)
00:04<billytwowilly>I don't know.. Is that default for mandrake 9.1?
00:04<billytwowilly>then I am.
00:04<billytwowilly>I'll follow the alsa instructions then.
00:04<rkulagow__>cd /etc/rc.d/init.d
00:04<rkulagow__>./alsa status
00:05<billytwowilly>sound loaded
00:05<rkulagow__>(as root probably)
00:05<billytwowilly>I'm guessing that is the prob then. Thanks man.
00:06<rkulagow__>mandrake 9.1 may not have the correct drivers setup for the nforce board. what kernel?
00:06<billytwowilly>2.4.21-18 or something.
00:06<billytwowilly>it isn't the stock one. I updated.
00:06<billytwowilly>via rpm.
00:07<billytwowilly>rkulagow__ the sound drivers come from nvidia.
00:07<rkulagow__>did you load the nvidia nforce driver from nvidia?
00:07<billytwowilly>it also has the net driver in it.
00:07<billytwowilly>I couldn't be here without it;)
00:09<rkulagow__>did you install the nvidia driver from source or from an RPM?
00:09-!-Kuwanger [] has quit [Read error: 60 (Operation timed out)]
00:09<billytwowilly>from source
00:11<rkulagow__>paste in your modules.conf
00:11<billytwowilly>probeall usb-interface usb-ohci ehci-hcd
00:11<billytwowilly>alias ieee1394-controller ohci1394
00:11<billytwowilly>alias sound-slot-0 i810_audio
00:11<billytwowilly>alias eth0 nvnet
00:11<billytwowilly>alias /dev/nvidia* nvidia
00:11<billytwowilly>alias char-major-81 bttv
00:11<billytwowilly>options tuner type=2
00:12<rkulagow__>as root, do a modprobe i810_audio
00:12<billytwowilly>it's there.
00:13<rkulagow__>do you see a mixer?
00:13<billytwowilly>there is this right below it: ac97_codec 12488 0 [i810_audio
00:13<billytwowilly>one sec.
00:13<billytwowilly>lsmod | grep mixer turns up nothing.
00:15<rkulagow__>take a look at
00:16<billytwowilly>I'm already there;)
00:16<rkulagow__>modprobe snd-intel8x0;modprobe snd-pcm-oss;modprobe snd-mixer-oss;modprobe snd-seq-oss
00:17<billytwowilly>that first one brings up an error of no such device.
00:17<rkulagow__>at this point, i think you're on your own. i don't have the hardware, so try the nvidia linux forums or the alsa stuff
00:17<rkulagow__>right; because you're using i810_audio
00:18<billytwowilly>I think I can figure it out. there is a bunch of stuff in the installation guide about this. I think I just have to modprobe those modules and then put some stuff in /etc/modules.conf
00:18<rkulagow__>but you probably don't want to mix alsa and OSS stuff at the same time; not sure what the nvidia driver is.
00:18<billytwowilly>thanks for the help.
00:18<rkulagow__>sure. good luck.
00:18<billytwowilly>I'll keep you posted;)
00:22<rkulagow__>cutt: you still here?
00:22<rkulagow__>duh, chutt: you still here?
00:23<-- hfb( has left #mythtv ("Client exiting")
00:31-!-billytwowilly [] has quit [Remote closed the connection]
00:34-!-chris [] has joined #mythtv
00:37-!-chris [] has quit [Client Quit]
00:39-!-billytwowilly [] has joined #mythtv
00:40<billytwowilly>it doesn't appear to be using the i810_audio module..
00:41<billytwowilly>wait, I played sound and now it's being used..
00:42-!-FryGuy [] has quit [Read error: 104 (Connection reset by peer)]
00:45<Chutt>rkulagow, kinda
00:46<rkulagow__>the website FAQ is still the old one; the CVS faq has been cut down to one line that says, "look at the HOWTO". did you have plans for the website FAQ link?
00:47<Chutt>yeah, i'm going to get rid of it
00:52<rkulagow__>huh. i was just going to mention something about how rewinding seems to stutter, so to get my ducks in a row and remind myself of the issue i fast forwarded to a program, then hit "p". that paused it, but now it's locked up.
00:52<rkulagow__>i'm going to control-C it...
00:54<rkulagow__>ok, i was able to stop it.
00:54<rkulagow__>back to gdb
00:55-!-rkulagow_ [] has joined #mythtv
00:55-!-billytwowilly [] has quit ["Client exiting"]
00:56<rkulagow_>i guess this is the interesting bit:
00:56<rkulagow_>Thread 1 (Thread 16384 (LWP 4238)):
00:56<rkulagow_>#0 0x40b33306 in nanosleep () from /lib/i686/
00:56<rkulagow_>No symbol table info available.
00:56<rkulagow_>#1 0x00000001 in ?? ()
00:56<rkulagow_>No symbol table info available.
00:56<rkulagow_>#2 0x0807d29f in PlaybackBox::play(ProgramInfo*) (
00:56<rkulagow_> this=0xbfffe890, rec=0x865b998) at playbackbox.cpp:1231
00:56<rkulagow_> tvrec = (ProgramInfo *) 0x83c2e18
00:56<rkulagow_> db = (struct QSqlDatabase *) 0x845bee8
00:56<rkulagow_>---Type <return> to continue, or q <return> to quit---
00:56<rkulagow_> tv = (class TV *) 0x857fab0
00:56<rkulagow_> doremove = 255
00:56<rkulagow_> doprompt = 255
00:56<rkulagow_>#3 0x0807d040 in PlaybackBox::selected() (this=0xbfffe890)
00:56<rkulagow_> at playbackbox.cpp:1198
00:56<rkulagow_>No locals.
00:56<rkulagow_>#4 0x0807fb64 in PlaybackBox::keyPressEvent(QKeyEvent*) (
00:56<rkulagow_> this=0xbfffe890, e=0xbfffe480) at playbackbox.cpp:1622
00:56<rkulagow_> handled = true
00:56<rkulagow_>#5 0x405755f5 in QWidget::event(QEvent*) ()
00:56<rkulagow_> from /usr/lib/qt3/lib/
00:57<rkulagow_>No symbol table info available.
00:57<rkulagow_>#0 0x40b33306 in nanosleep () from /lib/i686/
00:57<rkulagow_>all the other threads are in nanosleep
00:58<Chutt>that's one's where it's supposed to be
00:58<Chutt>can you just email me the bt full?
00:58<rkulagow_>sure can
01:00<bline>what's up?
01:01<Chutt>bline, hey, you still having problems with playback?
01:01<bline>I've been playing my recordings with xine
01:01<Chutt>would you mind trying a patch out?
01:02<bline>Sure, but I really think my problem is hardware related
01:02<Chutt>could be, but could also be a bug =)
01:02-!-hfb [] has joined #mythtv
01:02<rkulagow_>chutt: mailed. still at gdb.
01:02<bline>I guess I can watch enterprise later
01:02<Chutt>oh, don't interrupt anything
01:03<bline>It's no big deal, it's a recording :)
01:03<bline>sure, or you can dcc it
01:03<Chutt>and a repeat
01:03<bline>they are all repeats now
01:04<bline>but I didn't watch it for most of the season
01:04<rkulagow_>and actually, the original question i had was that it looked like when rewinding, i'd get 10 seconds of rewind, 2 seconds forward, 10 seconds rewind, two seconds forward, etc. unless i made rewind a high number, i could actually get stuck in a pseudo-pause...
01:04<Chutt>oh, wait
01:04<Chutt>pvr-250 recordings, or live-tv?
01:04<rkulagow_>pvr-250 recordings
01:04<Chutt>remote or local?
01:04<rkulagow_>whoops, sorry, me or bline?
01:04<Chutt>that was to bline
01:04<bline>are you talking to me or rh?
01:05<rkulagow_>ok, i'l wait.
01:05<Chutt>well, this may help live-tv
01:05<Chutt>i dunno
01:06<bline>live-tv does the same thing recordings do
01:06<bline>the video speeds up shortly after the sound cuts out
01:08<Chutt>rkulagow, just got the email
01:08<rkulagow__>ok, let me know if there's anything i can do.
01:08<Chutt>you've got the mandrake threaded backtrace problem
01:08<Chutt>so, nothing i can do with it :(
01:09<Chutt>i'd want to know what the other threads were doing
01:09<Chutt>thread 1 (the one with data) is right where it's supposed to be -- waiting for the playback to stop
01:11<rkulagow__>hrmm. i've done other bt's that you were able to use; wonder what's gone wrong with this one...
01:11<Chutt>have you done just hangs like this?
01:11<rkulagow__>well, i guess there's no point in keeping it in gdb if you can't use it.
01:11<Chutt>i don't remember..
01:11<rkulagow__>actually, i think everything else has been a segfault...
01:12<Chutt>segfaults would work fine
01:12<rkulagow__>history: there were two niced mythbackends running; one is doing "walking with cavemen", so it's sitll running. i believe that the other one was doing the program i tried to bring up.
01:13<Chutt>maybe the commercial skip stuff affected it somehow
01:13<rkulagow__>both are pvr-250's, courtesy of the dual-pvr-250 pundit in the living room
01:13<rkulagow__>both are mpeg-2, sorry.
01:14* blinegoes back to enterprise
01:14<Chutt>i don't think it'll help local playback at all
01:14<rkulagow__>so is this particular session a lost cause?
01:14<bline>I'm going to replace the motherboard tomarrow
01:15<rkulagow__>let me see if it works now that the mythbackend has finished.
01:15<rkulagow__>(to try to get back to the original issue that prompted this in the first place...)
01:15<bline>arggg, I've seen this one
01:15<thor_>what was the original issue?
01:16<thor_>oh ... never mind ... complicated video stuff
01:17<thor_>(looked at scrollback)
01:23<rkulagow__>ok, i've got what i think is the original issue happening again. mpeg-4, ~13 minutes into "Farscape". hit left arrow to start rewinding. Rewind 1X is looping between 13:07 and 13:08. rewind 1.5x loops between 13:07, 08, 09. rewind 2X, looping between 13:03, 04, 05. unless i go to super-high rewinds, i'm not really rewinding at all.
01:24<Chutt>is there a commercial skip table defined?
01:24<rkulagow__>there should be. this was a successful capture (59:54)
01:24<Chutt>might be a bug with that
01:24<rkulagow__>let me check.
01:26<rkulagow__>well, i have 140000 rows in recorded markup, so i guess a select * might take a while. let me check the table definition, unless someone has it handy...
01:27<Captain_Murdoch>select chanid, starttime, type, count(*) from recordedmarkup group by chanid, starttime, type order by starttime, chanid, type;
01:27<thor_>I'm not sure this is a new problem
01:27<Chutt>so, i had my source hacked up to produce huffyuv (ie, lossless compression) video
01:27<Chutt>to test with huge bitrates
01:27<Captain_Murdoch>that was cut-and-pasted from my mysql scrollback buffer.
01:27<Chutt>left the backend running with that
01:27<rkulagow__>146 rows.
01:28<Chutt>recorded a show
01:28<Chutt>filled up my drive =)
01:28<rkulagow__>let me correlate the show / starttime with the results. hold a sec...
01:29<thor_>I'm sure Tyler would know what to do
01:30<Chutt>thor, he was in here
01:30<rkulagow__>ok, here's the culprit:
01:30<rkulagow__>| 2044 | 20030615230000 | 3 | 5 |
01:30<rkulagow__>| 2044 | 20030615230000 | 4 | 1 |
01:30<rkulagow__>| 2044 | 20030615230000 | 5 | 1 |
01:30<rkulagow__>| 2044 | 20030615230000 | 7 | 2741 |
01:30<Chutt>when i was telling snowman about his emails
01:30<rkulagow__>looks like there's data in the recordedmarkup table for my program that doesn't like to actually rewind, even though it says that it is.
01:31<Captain_Murdoch>do you have auto-skip on?
01:31<rkulagow__>no; auto flag, blank+scene is enabled.
01:31<Captain_Murdoch>ok, so if auto-skip is off the only time it should use the commercial break table is if you hit 'z' or 'q'.
01:32<Captain_Murdoch>how long is that show?
01:32<rkulagow__>one hour
01:32<Captain_Murdoch>you should have about 7000 '7' records for a 1-hour show with 2 GOPs per second.
01:33<rkulagow__>baby's crying; i may be done for tonight.
01:33-!-Morph [] has quit [Read error: 60 (Operation timed out)]
01:34<thor_>holy crap, it's 2:30 am
01:34<thor_>gotta sleep
01:34<Chutt>can't sleep, it's patch night!
01:34<Chutt>or something
01:34<thor_>anyone here in the San Jose area or interested in relocating?
01:34<thor_>I've got a head hunter with a stick up his butt giving me trouble
01:35<bline>you sound like an ad :)
01:35<thor_>Sequioa funded, ex-rightworks people
01:36<thor_>I just don't want to talk to this guy anymore
01:36<thor_>CTO level position ... anyone
01:36-!-extremis [] has quit ["My damn controlling terminal disappeared!"]
01:37<thor_>Chutt, "patch night" ? is there something I should be doing?
01:37<Chutt>i'm just applying that manual schedule thing
01:38<Chutt>and the ff/rew key stuff
01:38<Chutt>and going to look into ben's movie patch
01:38<thor_>FWIW, I've seen rewind get caught up many times
01:38<Chutt>i never have, though
01:39<thor_>never paid attention as it's on the main system in the living room
01:40<thor_>Can we put up mdz as for that via arena thing?
01:40<rkulagow__>guys, got to go. get some sleep!
01:40<rkulagow__>thanks for the help.
01:40<-- rkulagow__( has left #mythtv
01:40<Chutt>sorry =)
01:47<Captain_Murdoch>Chutt: any objections to adding the ability to make a LineEditSetting and MythLineEdit read-only? On the recording profile editor, sometimes you want to allow editing the profile name and sometimes not. I changed the name field in the editor from a combobox to a lineedit since it makes more sense also.
01:48<Chutt>err, no objections
01:48<Captain_Murdoch>ok. good. :) already coded it.
01:48<Captain_Murdoch>works good. just created a new profile and was able to change the name, but when editing an existing profile the name can't be changed.
01:49<Captain_Murdoch>could make it so that you can only change profile ID's > a certain number to protect Default and LiveTV and let people change the names of their own profiles.
01:49<Captain_Murdoch>err, change profile names with ID's greater than a certain number.
01:50<Captain_Murdoch>that could cause problems though since ID is an auto increment field. have to look at keying off the name and just not let people edit Default, LiveTV and Transcode.
01:51<Chutt>that works
01:53-!-arj [] has quit ["leaving"]
01:58-!-bline [] has quit ["Client exiting"]
02:09<thor_>Chutt, you have any interest in talking to Dell about a Myth-based commercial product?
02:10<thor_>(if you're still up?)
02:11<bigguy>dell wants a myth based product?
02:11<thor_>That would be over-stating things
02:11<thor_>Just dealing with e-mails, nothing serious ....
02:12<Chutt>it'd still have to be GPLd
02:12<Chutt>what'd be involved?
02:12<thor_>I really don't know
02:12<thor_>Old friend of mine ...
02:13<thor_>started reach networks ...
02:13<thor_>got bought out for a gazillion something or other
02:13<thor_>I gave him a Myth based box a few weeks ago ....
02:13-!-poptix [] has joined #mythtv
02:14<bigguy>the only thing that they could really change and keep internal to them would be making a legal commercial zap2it client
02:14<thor_>and he's been shopping it to Dell
02:14<bigguy>or whoever they made a pact with
02:14-!-bline [] has joined #mythtv
02:15<thor_>listings are a big problem
02:15<thor_>harumph ... the whole thing frustrates me
02:15-!-mechou [] has joined #mythtv
02:16<Chutt>i imagine listings won't be that big of a problem
02:16<thor_>myth is supposed to be something to do in my spare time for fun
02:17<Chutt>what's that?
02:17<thor_>hmmm ....
02:17<bigguy>you can have fun?
02:17<thor_>to no obvious monetary gain ?
02:17<bigguy>I thought that was banned in all 50 states
02:17<mechou>hey isaac, do you ever sleep??
02:18<Chutt>i'll be going to sleep in a bit
02:18<Chutt>thor, anyway, i dunno. perhaps
02:19<thor_>Let me make a couple of calls ...
02:19<mechou>just wondering, cause I see you post on threads virtually 24hrs....
02:19<thor_>if there's something beyond idle chat
02:19<thor_>I'll send you an e-mail
02:20<thor_>probably nothing
02:20<Chutt>mechou, bah, usually only up until 3 am or so
02:20<Chutt>then i go through the list once i wake up
02:20<Chutt>gotta do email first before anything else =)
02:20<mechou>you're an addict ;-)
02:21<Chutt>thor, ok
02:21<Captain_Murdoch>Chutt: just committed the recording profile edit fix.
02:22<Chutt>has anyone used ramon's manual scheduling stuff?
02:27<Chutt>captain_murdoch, in the future, if you change the size of one of the classes in linmyth (adding/deleting a member variable), could you update the libmyth binary number?
02:27<Chutt>i'll do it this time, since i'm making a commit in a few seconds, but..
02:27<thor_>Oh, good moment for a question ...
02:28<thor_>should that number be bumped only when you change the interface (.h) ?
02:28<Chutt>only when the size of the object is changed
02:28<Captain_Murdoch>ok, sorry about that.
02:29<Chutt>not a problem
02:29<thor_>here's another question ... to display my C++ ignorance ...
02:30<Captain_Murdoch>so that's MYTH_BINARY_VERSION in mythcontext.h right
02:30<Captain_Murdoch>ok. hadn't messed with that before. :) I'll remember next time.
02:30<thor_>should I worry about how many members something like a GenericTree class has? I have no idea if that increases the memory requirements of a given executable on a per instance basis ?
02:31<Chutt>thor, not unless you're allocating tons and tons of stuff
02:31<thor_>and thanks
02:31-!-hfb [] has quit ["Client exiting"]
02:33<Chutt>the myth_binary_version stuff is because
02:34<Chutt>otherwise, a user of the library was built against a different header file
02:34<Chutt>so it thinks the size of the class is different from what it is in the library
02:34<Chutt>allocates the wrong amount of space, bad things happen
02:35<thor_>... but it seems that all of those dependencies got a lot simpler in the "quest for one window" drive
02:35<Chutt>all the widgets and stuff are still going to cause problems if they change size
02:35<Chutt>actually, the uitypes probably won't
02:36<Chutt>since most of those are allocated by xmlparse, right?
02:36<Chutt>so, yeah, stuff internal to the lib shouldn't matter
02:36<thor_>ah good
02:36<Chutt>but, better to be safe than cause totally random segfaults
02:37<thor_>what can you tell the guy who set everyone's right channel to 0
02:37<Chutt>it didn't die or anything
02:37<Chutt>just made everyone deaf in their right ear
02:38<thor_>easy to say now
02:38<thor_>it was a big screw up
02:38<thor_>I will never ...
02:38<Chutt>i was just annoyed at having to do a bugfix release
02:38<thor_>... ever ...
02:38<thor_>be eager to add just one more feature
02:38<Chutt>didn't really care about the bug =)
02:39<thor_>still, that dynamic volume control in MMusic is kind of nice
02:39<Chutt>it is
02:40<thor_>btw, I can't think of what else to tweak
02:40<Chutt>in mythmusic?
02:40<thor_>so I guess I need to seperate out the backend
02:41<Chutt>want to write a menu for the tv osd?
02:41<Chutt>i'm trying to pawn that off on someone
02:41<Chutt>be similar to one of the uitypes widgets
02:41<thor_>it's just a widget, no?
02:42<Chutt>pretty much
02:42<Chutt>just displayed differently
02:42<Chutt>and uses different low-level stuff to build off of
02:42<thor_>does it have to write itself to a pixmap ...
02:42<thor_>or to some weird ass NuppelVideo stream?
02:42<Chutt>draws itself directly to the screen
02:43<Chutt>but it should be able to build this off of existing widgets
02:43<Chutt>so it'd be more of a wrapper
02:43<thor_>I will have a look
02:43<thor_>may take me a while though ....
02:43<thor_>it's in code I haven't looked at before
02:43<Chutt>no worries if you don't want to
02:44<thor_>no, not that
02:44<Chutt>i don't want to do it =)
02:44<Chutt>which is why i'm trying to get someone else to =)
02:44<thor_>just that I tend to be very pendantic about new code ...
02:44<thor_>I eschewed moegreen's lists
02:44<thor_>on that principal
02:45<thor_>not superiority ...
02:45<thor_>probably just arogance
02:47<thor_>so, what's the short version of the best *-ui.xml --> OSD interface?
02:47<thor_>or *.xml -->
02:48<Chutt>how the osd themes work?
02:48<thor_>I have no idea ... I guess I'll check
02:48<Chutt>it's pretty similar to the uitypes stuff
02:48<Chutt>only less widgets
02:48<Chutt>moegreen patterned the uitypes set after the osd code
02:48<Chutt>osd.cpp parses the osd.xml file
02:49<Chutt>creates OSDType* stuff
02:49<Chutt>which are in osdtypes.cpp
02:49<thor_>is there some reason we really want these to be seperate .xml files?
02:49<Chutt>containers are turned off/on by calls to osd.cpp
02:49<Chutt>well, i like to have the osd be separate from the rest of the ui theme
02:49<thor_>pour quoi?
02:50<Chutt>separate themes, i dunno
02:50<Chutt>more choice for the user
02:50<Chutt>i use defaultosd with the blue theme, not blueosd =)
02:50<Captain_Murdoch>I like the ability to use 2 different themes. I use blue currently for theme and Corners for osd.
02:51<thor_>confusing to the user?
02:51<thor_>or, better yet ....
02:51<thor_>if theme not in osd theme
02:51<thor_>do default
02:51<Chutt>i think the current stuff's fine for now
02:52<thor_>I will have a look
02:52<Chutt>before each video frame gets shown, it calls a function in the OSD class
02:52<Chutt>which goes through all the active osdtype containers
02:52<Chutt>and tells them to draw themselves
02:53<Chutt>passing along a pointer to the video data
02:53<thor_>on every frame?
02:53<Chutt>it's not as inefficient as it sounds =)
02:54<thor_>25 to 30 times a second?
02:54<Chutt>well, yeah
02:54<Chutt>have to if you want it to be alpha-blended
02:54<thor_>can't it just check a bool?
02:54<mechou>darn, that's just amazing
02:54<thor_>signal--> I have something to show!!!!
02:55<Chutt>thor, oh, you mean why bother even checking the osd containers?
02:55<thor_>well, I dunno
02:55<thor_>where do keypresses go?
02:55<Chutt>to the TV object
02:56<thor_>no idea ...
02:56<Chutt>they get sent along to the OSD stuff if necessary
02:56<thor_>haven't looked at the code
02:56<thor_>video codecs frazzle my limited brain
02:57<Chutt>i need to go to bed
02:57<Chutt>g'nite =)
03:09<-- mechou( has left #mythtv
03:30-!-Morph [] has joined #mythtv
03:41-!-bline [] has quit ["Client exiting"]
03:46-!-Morph [] has quit [Read error: 60 (Operation timed out)]
04:06-!-ahbritto [] has joined #mythtv
04:45<-- SamNmaXhas quit ()
05:54-!-Morph [] has joined #mythtv
05:59-!-thor_ [] has quit ["using sirc version 2.211+KSIRC/1.2.1"]
07:24-!-jkolb [] has joined #mythtv
07:29-!-Morph [] has quit [Read error: 60 (Operation timed out)]
08:18-!-bline [] has joined #mythtv
08:18-!-jkolb [] has quit [Read error: 104 (Connection reset by peer)]
08:45-!-Morph [] has joined #mythtv
08:47-!-rkulagow__ [] has joined #mythtv
09:05-!-rkulagow___ [] has joined #mythtv
09:05-!-rkulagow__ [] has quit [Read error: 104 (Connection reset by peer)]
09:15-!-bline [] has quit ["Client exiting"]
09:29-!-jkolb [] has joined #mythtv
09:45-!-JAC [] has joined #mythtv
09:46-!-bline [] has joined #mythtv
09:46-!-bline [] has quit [Client Quit]
09:47-!-bline [] has joined #mythtv
09:49-!-Captain_Murdoch_ [] has joined #mythtv
10:23-!-JAC is now known as joecap
10:23-!-jkolb [] has quit [Read error: 104 (Connection reset by peer)]
10:40-!-schultmc [] has joined #mythtv
11:15-!-hfb [] has joined #mythtv
11:35-!-Drikus__ [] has joined #mythtv
11:43-!-mecraw [~mecraw@] has joined #mythtv
11:49-!-jkolb [] has joined #mythtv
11:57-!-mechou [] has joined #mythtv
12:04<Conaz>slow news day?
12:07-!-poptix [] has quit ["leaving"]
12:08-!-poptix_ [] has joined #mythtv
12:08-!-poptix_ is now known as poptix
12:10-!-ForsGump [] has joined #mythtv
12:10<bline>Wow, a lot of commits yesterday
12:11-!-Chief00 [] has joined #mythtv
12:11<Chief00>OK there, chiefs?
12:38-!-rkulagow___ [] has quit [Read error: 54 (Connection reset by peer)]
12:43<-- Chief00has quit ()
12:43<mechou>Anybody still having trouble with we interface being extremely slow?
12:44<mechou>mythtv-dev and mythtv-commits seem fine
12:44<Chutt>use the gossamer-threads archive linked to from the web page instead
12:44<Conaz>The one post I replied to this morning only took 1 minute to get back to me
12:45<Conaz>Oh, web interface, sorry
12:53<Snow-Man>You all can buy me a bigger pipe, mkay?
12:54<jkolb>I have a nice lead pipe right here.
12:54<Chutt>we just bought a bigger pipe, so quitcher bitchin
12:55<jkolb>I found mine in the Ballroom.
12:55<jkolb>With Ms. Scarlet.
12:56-!-Morph [] has quit [Remote closed the connection]
12:56-!-Morph [] has joined #mythtv
12:58<Snow-Man>Chutt: yea, yea, that one just doesn't go to my house. :)
13:04<mechou>Issac, your penultimate cvs commit "connect to master backend only when first needed...",
13:05<radsaq>yeah except he didn't give props to srl, bma, or me
13:05<mechou>does that fix the FE crash BE over the networrk we talked about some time ago?
13:05<Chutt>no, it doesn't
13:08<mechou>OK, just wondering if it had anything to do with it...
13:15-!-Morph [] has quit [Read error: 110 (Connection timed out)]
13:33-!-ForsGump [] has quit [Read error: 60 (Operation timed out)]
13:35-!-rickter [] has joined #mythtv
13:44-!-ForsGump [] has joined #mythtv
13:46<mechou>Chutt: How do you determine when a cvs snapshot is ready for full fledged release (e.g. 9.0)?
14:02-!-dopez [] has joined #mythtv
14:16-!-vektor [~bbiggs@nat-pc.cs.Dal.Ca] has joined #mythtv
14:16<vektor>Chutt: hey you there/
14:16<Chutt>what's up
14:16<vektor>So, I'm about to add V4L2 support in my shittacular app.
14:16<vektor>And if I use videodev2.h then I get complaints that struct timeval is redefined.
14:16<vektor>Cause it includes linux/time.h
14:16<Chutt>copy videodev2 and videodev.h to your app
14:16<vektor>and that seems to not like /usr/include/time.h
14:16<Chutt>and use your own copies
14:17<vektor>yeah but do I have to remove the #include ?
14:17<vektor>this seems really ass busted.
14:17<Chutt>it is
14:17<vektor>fucking lame.
14:17<vektor>so you just removed the #include in your copy?
14:17<Chutt>pretty much
14:17<Chutt>i've just got myth_videodev.h and myth_videodev2.h
14:17<Chutt>in my source tree
14:18<vektor>AHAHAHAHA fucking gerd knorr did too in his copy
14:18<Chutt>i don't think i made any other changes than that include
14:18<vektor>in xawtv/libng
14:18<Chutt>it's really dumb, but, whatever
14:18<vektor>i'm with you on the dumb part.
14:18<Chutt>there's barely any v4l2 shit anyway
14:19<Chutt>and for most stuff, the v4l1 compatibility layer works
14:19<vektor>Well, I have this user who's coming up my ass on this one.
14:19<vektor>Since the DVB driver doesn't work.
14:19<dopez>does this also mean if i compiled mythtv and patch my kernel with v4l2 lateron that v4l2 support is allready in mythtv?
14:19<vektor>dopez Well yeah, the only sane thing is for v4l2 apps to have their own copy of the API header file.
14:19<Chutt>dopez, yup.
14:20<Chutt>you don't need to patch your kernel with v4l2
14:20<vektor>Chutt: I have my PVR-250 and I still haven't put it in my bawx.
14:20<vektor>it's sitting on my table.
14:20<vektor>school has been sucking.
14:21<vektor>Chutt: So, Miguel Freitas of xine did a tvtime plugin for them.
14:21<Chutt>dvb cards produce yuv data?
14:21<vektor>Chutt: Yes they do.
14:21<Chutt>didn't know that
14:21<vektor>Has an MPEG2 decoder.
14:21<vektor>At least the popular card that the linux-dvb driver supports.
14:21<Chutt>just an output plugin?
14:22<vektor>It's not an output plugin, it's a processing plugin.
14:22<vektor>sends it to the xine output driver.
14:22<vektor>it's inefficient in many ways.
14:22<vektor>but a nice proof of concept.
14:22<vektor>i was very happy that someone else did that work :)
14:22<vektor>i got some ok patches but it needs work
14:22<vektor>basically it converts 4:2:0 to 4:2:2 and sends it into tvtime.
14:22<vektor>we can do alot better
14:22<vektor>since we should be converting as we read the image to maintain cache coherency
14:23<vektor>but it's nice to have someone else doing some of the work of doing a 'tvtime plugin'
14:23<vektor>so it'd be alot easier to stick into your renderer.
14:23<vektor>i'll of course keep you posted.
14:23<Chutt>thanks =)
14:23<vektor>anytime. :)
14:23<vektor>still, i'm going to add mpeg support to tvtime
14:23<vektor>there are other things that xine is doing stupid
14:23<vektor>like buffering like mad
14:24<vektor>fuck that shit.
14:24<vektor>plus it's just fun to do your own stuff.
14:24<vektor>how's mythtv going?
14:24<Chutt>quite well
14:24<vektor>does the ivtv driver suck as much as it seems it does?
14:24<vektor>i've been following the list and stuff and it's fucking hilarious :)
14:24<Chutt>doesn't work at all for some people
14:25<poptix>works great for me, with my no msp init stuff =p
14:25<poptix>vektor: basicly, it works
14:25<Chutt>for the most part
14:25<poptix>but it doesn't have all the workarounds for shitty motherboards/chipsets that a more robust driver does
14:25<Chutt>kevin needs to add actual v4l2 ioctls for driver parameters
14:25<vektor>this james pattie guy is a bit silly too :)
14:25<Chutt>is that the script guy?
14:26<Chutt>every fucking cvs version of his crap gets announced like it's the most important thing ever
14:26<vektor>yeah :)
14:26-!-Morph_ [] has joined #mythtv
14:26<poptix>he wants to feel special
14:26<vektor>i had an extensive email conversation with him about frequency tables
14:26<vektor>it didn't go too well :)
14:30<bigguy>cause of the suckage
14:30<bigguy>caused by the hole in his head
14:33<bigguy>anyway had to do some minor automotive egine work now it's time to head out on the road
14:39<poptix>kind of sucks that plasmajohn is gone
14:39<-- rickter( has left #mythtv
14:40<Chutt>he wasn't doing any coding
14:40<poptix>shrug =p
14:41<vektor>Why is he 'gone'?
14:41<vektor>oh, external reasons :)
14:41<bline>Looks job related, or lack of job related.
14:42<vektor>Ah, the mythical job.
14:44<GreyFoxx>ick, work... where is the fun in that ?
14:46<vektor>GreyFoxx: $ ?
14:48-!-Morph_ is now known as Morph
14:48<GreyFoxx>True, true
14:49<vektor>Sigh, they changed the tuner API too in V4L2 ??
14:49<GreyFoxx>The act of spending the $ does have it's rewards :)
14:49<vektor>why did they do this?
14:49<Chutt>vektor, recently?
14:49<Chutt>or the whole thing?
14:49<vektor>no I mean from V4L1.
14:49<Chutt>to cause problems
14:50<Chutt>it's pretty similar, though
14:50<Chutt>except that you have to set the standard _after_ you set the input
14:58<vektor>in my code btw, i'm doing both at once here :)
14:58<vektor>like if( v4l2 ) blah else blah;
14:58<vektor>so some of those might hurt me
14:58<Chutt>that's what i do in the mythtv code
14:58<vektor>i just want to make this v4l1/v4l2 distinction really, really transparent.
14:59<Chutt>works ok
14:59<vektor>coool maybe i'll go through yours then
14:59<Chutt>i didn't do the brightness controls yet
14:59<Chutt>those are still v4l only
14:59<Chutt>but, channel changing / standard setting works fine
14:59<vektor>what about audio
15:00<vektor>do you do any audio stuff, or just btaudio ?
15:00<Chutt>ya know, i ignore that =)
15:00<Chutt>i'm always recording the audio
15:00<Chutt>so i just have to make sure it's not disabled
15:00<Chutt>otherwise, i don't touch it
15:06<-- jkolbhas quit ()
15:28-!-olivier [] has joined #mythtv
15:36-!-Chang [] has joined #mythtv
15:39-!-radsaq [] has quit [Read error: 60 (Operation timed out)]
15:41-!-olivier [] has quit ["Client exiting"]
15:54-!-radsaq [] has joined #mythtv
16:05-!-arj [] has joined #mythtv
16:06-!-Jon [] has joined #mythtv
16:06-!-Jon is now known as jkolb
16:30-!-Viddy [] has quit [Remote closed the connection]
16:30-!-BubbaBlue [] has joined #mythtv
16:31-!-ForsGump [] has quit [Read error: 60 (Operation timed out)]
16:32-!-BubbaBlue is now known as ForsGump
16:32-!-Morph [] has quit [Connection timed out]
16:38-!-Viddy [] has joined #mythtv
16:43-!-Morph [] has joined #mythtv
16:45-!-Viddy [] has quit [Remote closed the connection]
16:50-!-Morph [] has quit ["Reconnecting"]
16:50-!-Morph [] has joined #mythtv
16:51-!-zaheer [] has joined #mythtv
16:51<jkolb>Is the allrecord table used anymore? I can't find any references to it in the code.
16:51<jkolb>If not, are there any other tables in the database that are obsolete?
16:52-!-Morph [] has quit [Client Quit]
16:53-!-Chang [] has quit [Remote closed the connection]
16:54-!-bdavis [] has quit [Read error: 60 (Operation timed out)]
17:04-!-ForsGump [] has quit [Connection reset by peer]
17:04-!-ForsGump [] has joined #mythtv
17:09-!-claybo [] has joined #mythtv
17:09-!-rkulagow__ [] has joined #mythtv
17:10<claybo>hey, i came in here about a week ago to see if anyone's had any luck with a via epia m10000 and wintv pvr tuner card in mythtv. i want to set up a home theater pc and give it tivo-like functionality, and wanted to introduce as little power consumption and heat as possible.
17:11<claybo>is an epia m10000 alongside a wintv pvr enough for mythtv to be able to record scheduled shows, and play back everything, and pause live tv, etc?
17:14<-- joecap( has left #mythtv
17:14<-- claybo( has left #mythtv
17:16-!-mechou [] has quit [Read error: 113 (No route to host)]
17:16-!-radsaq [] has quit [Read error: 104 (Connection reset by peer)]
17:17<zaheer>claybo: i dont think theres an mpeg2 hardware driver for the epia m10000 yet for linux
17:17<zaheer>oh he disappeared
17:20-!-radsaq [] has joined #mythtv
17:28-!-radsaq [] has quit [Read error: 113 (No route to host)]
17:28-!-Drikus__ [] has quit ["toedeledoki"]
17:31-!-mechou [] has joined #mythtv
17:37<-- Captain_Murdoch_has quit ()
17:58-!-FryGuy [] has joined #mythtv
17:59-!-dopez [] has quit ["reboot"]
18:00-!-zaheer [] has quit ["reboot"]
18:04-!-dopez [] has joined #mythtv
18:09-!-zaheer [] has joined #mythtv
18:13-!-Viddy [] has joined #mythtv
18:40-!-arj [] has quit ["leaving"]
18:42<-- ForsGumphas quit ()
18:43-!-bdavis [] has joined #mythtv
19:04-!-schultmc [] has quit ["Client exiting"]
19:07-!-zaheer [] has quit ["Client exiting"]
19:24<-- mechou( has left #mythtv
19:26<mdz_>Chutt: I have an idea to improve myth's stability
19:26<mdz_>Chutt: we link in a copy of gdb and run it with a debugger attached all the time
19:26<mdz_>Chutt: because I'll be damned if I can ever get it to crash under gdb :-)
19:27<Chutt>i can't get it to die at all, really
19:27<mdz_>I get the occasional inexplicable crash
19:27<mdz_>but I just ran it for 5 days straight under gdb
19:27<mdz_>and nothing
19:27<mdz_>(the backend)
19:28<mdz_>I had to detach it before it filled up the process table with zombies
19:28<mdz_>that's an exaggeration. but 351 threads is a bit much
19:28<mdz_>starting to consume huge amounts of address space
19:33<bline>is it normal, that if I restart the backend with the frontend running the frontend doesn't work anymore? ie is there some code to reconnect automatically if the frontend looses it's connection
19:37<Chutt>the frontend doesn't reconnect
19:51<Chutt>it might be fairly easy to add reconnecting now
19:51<Chutt>i'll look into that sometime
19:52<Chutt>as long as it doesn't die in a 'bad' place, at least
19:52-!-hfb [] has quit ["Client exiting"]
19:55<FryGuy>does cvs version of mythtv still stall when you play a recording when the .nuv doesn't exist?
19:55<Chutt>i dunno
19:55<Chutt>does it?
19:56<FryGuy>i don't know. the cvs version I have from right before .9 came out did
19:57<FryGuy>i don't have net connectivity for my myth box here
19:57<FryGuy>so I can't test it, and I want to write a patch if it hasn't been fixed
20:01-!-mechou [] has joined #mythtv
20:12-!-poptix [] has quit [Read error: 110 (Connection timed out)]
20:17-!-kslater [] has joined #mythtv
20:32-!-poptix_ [] has joined #mythtv
20:32-!-poptix_ is now known as poptix
20:39<-- jkolbhas quit ()
20:44-!-rkulagow_ [] has quit [Read error: 113 (No route to host)]
21:23<Captain_Murdoch>mdz, I can change the recording profile name box back to a combobox if it makes sense, I didn't think of what you mentioned on the list. I can picture lots of "How come I can't name my own profile' question on the lists though if it's not editable in some way.
21:26-!-extremis [] has joined #mythtv
21:26<extremis>woohooo this new geforce4 card works great
21:26<extremis>now its starting to come together
21:29<kslater>man, you guys have been pounding out the code of late.
21:37<-- rkulagow__( has left #mythtv
21:38<extremis>ca1vin: I'm playing with overscan
21:39<extremis>I got the height right
21:39<extremis>now to get the width right
21:40<vektor>odd, really those two should be linked.
21:40<vektor>in tvtime we have one overscan value and it affects both width and height.
21:40<vektor>otherwise you get invalid aspect ratios.
21:41<vektor>or are you talking about the tv output on the gf4...
21:41<extremis>I'm talking about tvoutput
21:41<extremis>but I think you are right
21:41<extremis>maybe I shouldn't use 1024x768
21:41<Chutt>seeing as it's pointless for tv out, yeah :p
21:42<extremis>Chutt: looks smoother
21:42<extremis>the fonts are smoother
21:42<vektor>Why would you have your TV output set to 1024x768 ??
21:42<vektor>That makes no sense.
21:42<Chutt>see? :p
21:43<extremis>1024x768 is not 16:9
21:43<vektor>what the hell are you doing
21:43<vektor>oh, and btw, a resolution is not an aspect rati
21:43<extremis>trying to get teh aspect right
21:44<extremis>well there are pixes off my screen
21:44<mechou>hey vector, you know anything 'bout nvtv and geforce3 tv out?
21:44<extremis>if I reduce the number of pixes in the width then it won't be off my screen
21:44<vektor>extremis: regardless, aspect ratio and resolution are different variables.
21:44<vektor>extremis: you can change your resolution without changing your aspect ratio.
21:44<extremis>vektor: my aspect ratio is 16:9 on my tv
21:45<vektor>the two important quantities are resolution in pixels, and pixel aspect ratio
21:45<extremis>but at 1024x768 and overscan .7 the width is too large
21:45<-- scanline( has left #mythtv ("Client exiting")
21:45<vektor>ok, but your pixels don't have to be square
21:45<extremis>I want my pixels square
21:45<vektor>mechou: I don't know alot about it specifically, no. I do know that most of the TV output chips in the geforce cards aren't built for video output.
21:45<vektor>extremis: Why do you care about square pixels?
21:45<vektor>And what is the input to your television?
21:46<extremis>vektor: svideo right now
21:46<vektor>Is it VGA?
21:46<vektor>S-Video ?
21:46<extremis>I will use dvi next week
21:46<vektor>It has DVI input?
21:46<vektor>Ok, but regardless, over S-Video you're limited to 480 scanlines.
21:46<vektor>So I doin't know why you're doing like '1024x768'
21:46<vektor>(sorry, 486 scanlines)
21:46<vektor>Your TV encoder chip is resizing.
21:46<Chutt>you don't want your tv encoder chip to be resizing
21:46<vektor>No, you don't.
21:47<mechou>vektor: youu wouldn't happen to know if "Display" settings in XF86config affects offset and centering on tv?
21:47<vektor>mechou: No idea, but that sounds scary and stupid.
21:47<vektor>The TV encoder chip knows how to position itself correctly.
21:48<extremis>the "mythTV" logo looks like crap at 640x480 but looks smooth at 1024x768
21:48<vektor>extremis: Well, 640 is some poor horizontal resolution for TV.
21:48<vektor>Use at least 720.
21:48<vektor>And who resized it to 640x480 ?
21:48<vektor>And how did they do it?
21:48<Chutt>720x480 =)
21:48<vektor>Yeah, I thought the graphics were rendered to 720x480.
21:48<Chutt>but it's still going to be 4:3 over svideo
21:49<vektor>Yep, of course.
21:49<Chutt>they're rendered at 800x600 in mythtv
21:49<Chutt>and scaled on demand
21:49<vektor>Unless you tell your television that it should expect anamorphic content.
21:49<vektor>Chutt: What sort of scaling algorithm do you use?
21:49<Chutt>qt's smooth scaling stuff
21:49<Chutt>looks nice
21:49<vektor>Is it a pre-process?
21:49<Chutt>interpolated and stuff
21:50<vektor>I should give you a better scaler.
21:50<Chutt>naw, this is fine
21:50<vektor>I want to pick up a gf4, maybe tomorrow if I can swing it.
21:53<extremis> resolution: 100x100 dots per inch
21:53<extremis>ok, got that fixed
21:53<extremis>and I'm using 800x600
21:53<extremis>but 1024x768 still looks smoother
21:54<vektor>So you're letting the TV encoder scale from 800x600 down to 720x480 ...
21:54<extremis>sure, why not
21:54<vektor>Because it will suck for video content.
21:54<vektor>Since you'll be scaling it up to 800x600 just to have the TV encoder scale back down.
21:55<vektor>Maybe it's doing a better scale on your generate content, but .. ugh.
21:55<vektor>Also, how are you going to deal with interlaced content?
21:56<extremis>it won't display at 720x480
21:56<vektor>What won't?
21:56<extremis>my geforce4
21:56<vektor>Which TV encoder chip do you have on your card?
21:56<extremis>just a sec...
21:58<vektor>Ideally, the bt869..
21:58<vektor>Conexant rocks.
21:58<vektor>Or the 870/871
21:58<vektor>uh ...
21:58<vektor>You mean the NVIDIA proprietary TV encoder?
21:59<vektor>Not an OEM'ed chip?
21:59-!-extremis [] has quit ["[BX] Leggo my Eggo!"]
21:59-!-bline [] has quit [Remote closed the connection]
21:59-!-extremis [] has joined #mythtv
21:59<vektor>need I repeat my question?
21:59<extremis>(--) NVIDIA(0): Detected TV Encoder: NVIDIA
21:59-!-bline [] has joined #mythtv
21:59<vektor>Do you have a digital camera photo of the card?
21:59-!-FryGuy [] has quit [Read error: 113 (No route to host)]
22:00-!-FryGuy [] has joined #mythtv
22:01-!-Scrye [] has joined #mythtv
22:02<vektor>That's your card exactly?
22:03<vektor>Well, the docs are extremely vague, there is like no way of knowing wtf it is doing or what it expects.
22:04<vektor>the valid modes for your TV encoder
22:04<vektor> will be reported in a verbose XFree86.0.log file (generated with
22:04<extremis>what are you looking for?
22:04<vektor> `startx -- -logverbose 5`) when X is run on a TV.
22:04<bline>"like no way" or "no way"?
22:04<vektor>What 'valid modes' are reported for your tv encoder?
22:05<vektor>bline: Well at least for the conexant chips the specs are online and they discuss stuff like how a DVD player is supposed to work with it, or expected to work with it.
22:05<vektor>But the sad reality of consumer video cards is that until Microsoft does a proper DirectVA or a stronger DirectX architecture for interlaced video and TV outputs, this is all going to suck ass.
22:05<vektor>The primary purpose of most TV encoder chips hasn't been video playback.
22:05<extremis>(II) NVIDIA(0): TV Standard: NTSC-M: 720x480
22:06<extremis>it said that is supported
22:06<bline>vektor: I was giving you are hard time for your overuse of 'like' =)
22:06<vektor>bline: bah :)
22:06<bline>kids these days
22:06<vektor>extremis: ok, so that's like what you want, yo
22:06<extremis>right, well it dies when I specify it
22:06<vektor>bline: dude what's funny is that at OLS people make fun of me since i talk irl like i talk on irc :)
22:07<vektor>extremis: well, maybe you're not specifying it right :)
22:08<extremis>(II) NVIDIA(0): Not using mode "720x480" (no mode of this name)
22:08<extremis>of course I'm not doing it right
22:08<extremis>I'm a lamer ;)
22:08<vektor>I'm scared to suggest it (since I think it's the wrong thing to do) but have you tried defining a 720x480 modeline?
22:09<extremis>hell I don't even know how I record the proper values for that modeline for both my tv and video card
22:09<vektor>well, I don't think it's what you should do.
22:09<vektor>Chutt might know what's up here though
22:09<vektor>otherwise, i do recommend emailing NVIDIA's support email
22:19-!-Ripp [~chris@] has joined #mythtv
22:29-!-joecap [] has joined #mythtv
22:35-!-joecap [] has quit ["Client exiting"]
22:36-!-joecap [] has joined #mythtv
22:37-!-mechou [] has quit [Read error: 60 (Operation timed out)]
23:11<-- joecap( has left #mythtv
23:13<extremis>the nvidia docs don't say alot
23:13<extremis>they tell me what commands are valid
23:13<extremis>and they say I should be able to use 720x480
23:13<extremis>but thats not true it seems
23:15<dopez>extremis: not sure tho, but shouldnt it be 640x480 for NTSC and 768x576 for PAL (720 sounds like a mpeg2/dvd resolution so i might be wrong)
23:16<vektor>dopez: Standard digital sampling for NTSC is 720x486.
23:16<vektor>Standard digital sampling for PAL is 720x576.
23:16<vektor>See CCIR-601 or 709 or any other broadcast video standard.
23:20<dopez>off to bed....nn
23:20<extremis>vektor: are you doing 720x486?
23:21-!-dopez [] has quit ["zzz"]
23:21<vektor>I don't own a GF4.
23:21<vektor>However, I have a G400 and it can do 720x486 interlaced.
23:21<vektor>And it gives me an interrupt every field.
23:21<vektor>So I can do broadcast quality TV output. :)
23:23<extremis>and you specify this in your Screen section as a mode ?
23:23<extremis>can you send me your XF86Config?
23:24<vektor>No, the X driver can't do this.
23:24<vektor>Someone hacked a driver for DirectFB based on some leaked source code. :)
23:25<vektor>but you didn't hear that from me
23:25<vektor>in windows it's a huge pain too
23:26<vektor>TV output that is
23:26<vektor>I worked for a company that did broadcast TV graphics
23:26<vektor>and we tried very hard to convince consumer cards to do high quality TV output
23:26<extremis>does the g400 have dvi out?
23:26<vektor>ATI basically told us they have no idea when their TV encoder chips are sending a top or bottom field
23:26<vektor>they just have nobody who really knows anything about video and stuff
23:27<vektor>microsoft knows nothing either, but they're desperately trying to learn, at least for the next media player and stuff
23:27<vektor>since right now windows dvd players suck compared to any hardware dvd player in most situations, afaict
23:27<vektor>the G400 does not have DVI output
23:27<vektor>it doesn't make sense for TV though
23:27<vektor>DVI is a computer standard i dunno what you do for video
23:27<vektor>the G550 has DVI but it's only for the VGA head of course
23:28<vektor>if your TV can do VGA or DVI input, then you can probably have it sync to a computer refresh rate or something, but then you're fucked because you have to do temporal interpolation on your video
23:29<extremis>you are speaking greek
23:29<vektor>i am sorry.
23:30<bline>I thought it was french
23:30<vektor>my point was just that the APIs aren't there in either linux or windows to properly deal with sending video to a TV.
23:31<vektor>so no matter what there will be problems.
23:31<bline>make it work vektor, I know you can.
23:31<extremis>You seem to have quite a bit of experience with this, how do you think you could contribute to this project?
23:31<vektor>Well, about a year and a half ago we had a big discussion on the xpert list about an XVIDEO API for interlaced video.
23:32<vektor>We even decided on how it should work, at least to start.
23:32<vektor>Nobody has implemented it yet, I was going to for the MGA driver, but haven't had time.
23:32<vektor>But that really wasn't for TV output either.
23:32<vektor>There really aren't many TV output drivers at all, to be honest, and alot of the problems are hardware problems.
23:32<vektor>I'd like to get a GF4 and then try and experiment some more myself.
23:32<vektor>The DirectFB driver was a big step forward, I helped out with that a bit.
23:33<vektor>I mean, I'm trying :)
23:33<bline>No excuses.
23:33<vektor>I did a webpage describing some of the problems (
23:33<extremis>the directfb driver is for what hardware?
23:33<vektor>It's just for the G400/G450/G550 and no software apps really use it properly.
23:33<vektor>We support it in our app (tvtime) but I haven't had enough time to spend on fixing it up.
23:33<vektor>There's an mplayer patch for it, but there are issues.
23:33<vektor>Since mplayer syncs only to audio not to the video (at least mplayer-g1)
23:34<vektor>Really, I need help before I can do anything.
23:43<extremis>ok, now to get my pvr250 ready
23:44<ahbritto>Uh oh.
23:46<extremis>heh, v4l2?
23:46<extremis>must be a seperate driver
23:57-!-PeteCool [] has joined #mythtv