Back to Home / #mythtv / 2003 / 09 / Prev Day | Next Day
#mythtv IRC Logs for 2003-09-11

---Logopened Thu Sep 11 00:00:31 2003
---Daychanged Thu Sep 11 2003
00:00-!-dopez [] has quit ["nn"]
00:10-!-monkeyBox [] has quit ["Client exiting"]
00:20-!-mechou [] has joined #mythtv
00:21<mechou>howdy bbeattie
00:23-!-grogan_ [aaron@eviltwin.STUDENT.CWRU.Edu] has quit [Read error: 110 (Connection timed out)]
00:31<bbeattie>mechou: hey
00:33<mechou>bbeattie, how's the pcHDTV and myth coming along?
00:34<bbeattie>.. it works partially, about the same as it always has. I've been very busy, so busy that pcHDTV is starting work on myth for it's futher support now. I'm a little frustrated with my htpc, two dvi cables both bad and both mail ordered. And I can't get abc hdtv. :-/
00:35<bbeattie>.. just not a good time irl for me too. many complications.
00:35<abhishek>what do you need for hdtv? plain old antenna work or need a special one?
00:36<mechou>man, sorry to hear that....but really, isn't that why we got HTPC in the first place?
00:37<bbeattie>:) I'm over 300 hours invested in this, if I had done extra contract work, which was possible, I could have bought some very nice AV equipment. All I have to show is basic hdtv for myth and a how-to to show what a pain it is,. ;)
00:38<mechou>abhishek, depends how far you live from the tansmitter towers. Regular antenna can work ok if you got strong signal
00:38<bbeattie>.. there's too much that needs to be addressed for hdtv in linux, program listing especially, that just isn't ready for common use yet.
00:39<abhishek>that linux hdtv tuner has gotta be the cheapest digital tuner in existance
00:39<mechou>bbeattie, I hear ya and feel that pain, man.
00:39<bbeattie>I'm resticted to using an antenna in a basement, 8 miles from two hdtv towers withen 10 degrees, I get 8/11 well at any time
00:39<abhishek>mechou: how many people broadcast hdtv stuff now?
00:39<bbeattie>depends on where you live,
00:39<abhishek>la, ca, us
00:39<mechou>well, PBS does almost everywhere...
00:39<hadees>YES, i am finaly getting my crazy expensive htpc case next week... costs so much but it is just so cool check it out
00:40<mechou>hadees, I just dont know about spending $$$ on a case, no matter how nice it is....
00:41<hadees>bbeattie: does htpc come through standard cable or just digital usally?
00:41<mechou>hadees, that case probably set you back $400, right?
00:42<hadees>mechou: i usally get the cheapest cases how ever when it comes to my entertainment system I don't want somting that looks crappy sitting next to high end stuff
00:42<bbeattie>I'm using rg6 cable, but rg58, rg59, and even thicknet network cable can work. all depends on what you have to run for cable length.
00:43<mechou>sheesh, you could have scavenged an old amp case off fleabay for <$100
00:43<hadees>well one big thing about the case is how it handles heat
00:43<hadees>the case will be almost slient, not a whole lot of need for fans
00:44<mechou>old amps produces more heat than athlong XP...:)
00:44<hadees>yeah but if you just throw it in there the vents aren't in the right place and the air doesn't flow over the stuff right
00:45<mechou>yeah, u were just looking for an excuse to part with $400 :)
00:46<hadees>hehe, nah it was 3 reaons, heat, looks, and actully what i think might be the biggest i got it exaclty the way i wanted it, it was totaly custom
00:47<hadees>he pretty much would have done anything if it was possible
00:48<hadees>it is also i never really got into home theater stuff till htpcs
00:49<hadees>felt like since my friends spend big $$$ on recivers and speakers and what ever i would shell some out for a cool computer
00:49-!-Captain_Murdoch [] has quit [Read error: 110 (Connection timed out)]
00:49<mechou>hadees, for $400 I'd want the outer case to be a heat sink, if you know what I'm talking 'bout
00:49<mechou>like those amps they put in cars....
00:51<hadees>eh... probaly never goning to talk you into this being a good purchase for me but i am happy with it and i don't plan on ever buying another case for any htpc again
00:52<hadees>now all i need to do is add powermate support to mythtv
00:52-!-Timon [] has quit [Remote closed the connection]
00:52<mechou>I predict in a year or so all you'll see are small form factor pc for HTPC use. You're starting to see that already with shuttles, epia, etc....
00:53<hadees>i don't think htpcs will ever get that big really, not custom ones anyway
00:53<mechou>eventually tuner will come integrated on the mobo...
00:53<hadees>things like tivos will be what most people use
00:54<mechou>and when that time comes you're going to look for another case....
00:54<hadees>the playstation 2.5 is supposed to have media pc qualities
00:55-!-bbeattie [] has quit ["Client Exiting"]
00:55<mechou>whats a powermate?
00:55<hadees>mechou: for that stuff to happen build your own htpcs would have to go main stream, and nothing says they won't still have the same tuners and what not for normal sized pcs
00:55-!-phar0e [] has joined #mythtv
00:55<hadees>a powermate is device that looks like a volume controller
00:55<phar0e>=] MythTV rocks!!
00:56<mechou>and the purpose of the powermate is...?
00:56<hadees>you can make it do what ever you want
00:57<mechou>so it an input switcher of sorts, right?
00:57<mechou>for your htpc.
00:57<hadees>handle the volume, scroll, show somthign from its led, or you can push it and make a click
00:57<hadees>yeah but if you put it on right it resenables a volume knob on a reciever
00:58<hadees>the case i bought has a mount for it
00:58<hadees>as well as a mount for a irman and a lcd screen
01:05<mechou>playstation 2.5, haven't heard 'bout that one....
01:07<hadees>it is what sony plans on releasing, it will play playstation 2 games but use some of the tech that they are making for the playstation 3
01:08<hadees>from what i heard it will kind of tivo like
01:08<hadees>just all going towards what sony wants the playstation to be
01:08<hadees>which is the center of the entertaiment center
01:09<mechou>I assume by that time it will have a hdtv receiver?
01:09<mechou>playstation 2.5, that is?
01:10<mechou>massive hard drive too to store all those transport streams....
01:10<hadees>i am not sure on the specifics
01:11<mechou>It would be cool if it had that. But I was really disappoined with ps2 dvd picture quality.....
01:11<hadees>well they hurried it out
01:11<mechou>enven with component cable....
01:11<hadees>what they need is for nintendo to get back in the game so there is more competion
01:12<hadees>although i don't think that will happen, i think nintendo is going the way of sony becuase they don't seem to understand they need high capacity discs and fast processors
01:12<hadees>not the way of sony, i meant sega
01:13<mechou>well, sega and nintendo have differnt business model than sony and xbox....
01:14<mechou>hard to say if that business model will remain successful....
01:15<hadees>sony and xbox?
01:15<mechou>ps2 and xbox....
01:15<hadees>unfortanly xbox will because bill gates will keep throwing money at it untill it works
01:16<hadees>and i think sony will win out if they just make thier next game system better than the xbox
01:16<hadees>xbox also (unfortantly because i hate micro$oft) has the right idea by not reinventing the wheel
01:17<hadees>they use the good quality computer hardware that is still way ahead of game systems
01:18<mechou>seriously, though, the only game I really played (like) on ps2 is GT and that "mad clown" driving shooting game (name escapes me atm)
01:18<mechou>and I don't like any xbox titles....
01:18<hadees>you a gamecube man?
01:18<mechou>no, never even touched a gc
01:19<mechou>not really a game console person at all, really....
01:19<hadees>i have all 3 systems, i find that xbox has no real titles to themselves how ever the ones on all 3 systems are better on the xbox
01:20<mechou>what ar e your faves on the xbox?
01:20<hadees>well Halo is a great game, xbox got lucky with that one
01:21<hadees>Splinter cell is fun
01:21<hadees>so is unreal
01:21<mechou>FPS are a dime a dozen, don't you think? FPS I prefer playing on PC myself.
01:21<mechou>better control...
01:21<hadees>FPS are a dime a dozen how ever good ones are rare
01:22-!-paulproteus_ is now known as paulproteus
01:22<hadees>i actully think i get better controls with the xbox controler i feel i can move my head and body more together
01:22<mechou>I'm still playing RTCW on my pc since its initial release....
01:22<hadees>i am going to get an xbox controller converter for my computer
01:23<mechou>ratshack sells them...
01:23<hadees>the one thing that playstation 2 has over all the systems though is FF and GTA
01:24<hadees>final fantasy
01:25<mechou>ok..I found GTA "unresponsive." Physics modeling was a bit "funky"
01:26<hadees>well if you were looking for real physics you should get a flight sim, GTA is just a fun game
01:26<hadees>and GTA Vice City is amazing
01:26<hadees>the music and the story are great
01:27<mechou>GTA I played on the PC. I think you had more detail....
01:27<hadees>and i think the controlls could have been better but really this is the first time GTA was in that mode the first two games were over head
01:27<hadees>well like in Vice City they fixed stuff
01:28<hadees>from GTA3
01:28<hadees>i don't know GTA will be the next mario
01:28<hadees>for the new gaming gen
01:29<hadees>the funny thing is i read on slashdot more 20 somthing women play video games then young boys
01:29<mechou>I got tired of GTA, but havent tired out on RTCW
01:29<mechou>castle wolfenstein.
01:29<hadees>oh, i just got that for the xbox
01:29<hadees>haven't played it much, been playing that new star wars game
01:30<hadees>for the xbox not galaxies
01:30<mechou>RTCW is all about network opponents...
01:30<hadees>i don't need another game like Ulitima online sucking away all my free time
01:30<hadees>on the xbox it is on live
01:31-!-tmk [] has joined #mythtv
01:31<hadees>i like the live model better than the playstation one, live is cheap but you get more out of it
01:31<hadees>not but
01:31<hadees>anyway i need to go to bed i got class tomorrow
01:32-!-hadees is now known as hadees_sleep
01:33<tmk>chutt, you about?
01:38<bline>about what
01:40<tmk>sup bline
01:40<bline>hey tmk, hows the decoder coming?
01:40-!-activelow [] has joined #mythtv
01:40<bline>err encoder
01:41<bline>I'll get one as soon as Isaac finishes adding support to myth
01:42<activelow>anyone ever use a digital cable box from time warner?
01:47<tmk>bline: it's in myth now :)
01:47<tmk>active: is it a motorola DCT2000
01:52<bline>I saw a commit: Starting to get the pvr-350 decoder working.
01:53<tmk>yeah it's not that great
01:54<tmk>but video sometimes comes out the tv-out
01:57<bline>sometimes heh
01:57-!-mechou [] has left #mythtv []
01:57<Chutt>tmk, yo
01:57<Chutt>get it compiled/working?
01:57<tmk>i think it's sending data too infrequently
01:58<tmk>when i play, it works great
01:58<tmk>but if i ff
01:58<tmk>it gets all jumpy
01:58<Chutt>same if it pauses
01:58<tmk>could be
01:58<tmk>also there was a bug in your gop offset code
01:58<tmk>i think
01:58<Chutt>try pausing, it goes all wonky afterwards
01:58<tmk>i patched it
01:58<Chutt>what'd you change?
01:58<tmk>lemme diff it
01:58<tmk>like 3 lines
01:59<tmk>you want it in email?
01:59<Chutt>or just paste it here
02:00<tmk>--- ./videoout_ivtv.cpp 2003-09-10 22:49:55.000000000 -0700
02:00<tmk>+++ ../../../m2/mythtv/libs/libmythtv/videoout_ivtv.cpp 2003-09-10 15:10:07.000000000 -0700
02:00<tmk>@@ -88,14 +88,13 @@
02:00<tmk> perror("Cannot open ivtv video out device");
02:00<tmk> else
02:00<tmk> {
02:00<tmk>- ivtv_cfg_start_decode sd;
02:00<tmk>- memset(&sd, 0, sizeof(sd));
02:00<tmk> if (skipframes > 0)
02:00<tmk> {
02:00<tmk>- sd.gop_offset = skipframes;
02:00<tmk>- }
02:00<tmk>+ ivtv_cfg_start_decode sd;
02:00<tmk>+ memset(&sd, 0, sizeof(sd));
02:00<tmk>- ioctl(videofd, IVTV_IOC_S_START_DECODE, &sd);
02:00<tmk>+ ioctl(videofd, IVTV_IOC_S_START_DECODE, &sd);
02:01<tmk>+ }
02:01<tmk> }
02:01<tmk> }
02:01<Chutt>where do you set the offset, then?
02:01<tmk>err i think i did that diff backwards
02:01<tmk>swap the +'s and -'s
02:02<Chutt>oh, wait
02:02<tmk>must have been dinnertime eh
02:02<Chutt>yeah, not thinking
02:02<activelow>tmk: I'm not sure about the model #, I want to upgrade for this season of HBO, and want to make sure I can setup mythtv to change the channels
02:02<Chutt>that's pretty damn obvious, heh
02:03<tmk>anyhow, where do you set the buffer size?
02:03<tmk>128k or w/e
02:03<tmk>yeah, if i found it, you shoudl be ashamed :P
02:03<tmk>also, looking in my logs, i don't see any closes actually happening
02:03<Chutt>ivtvdecoder, GetFrame()
02:03<tmk>are you sure it's closing all the way
02:04<Chutt>far as i know
02:04<Chutt>might need to add a couple mutexes in there
02:04<tmk>shame on you, that's not 128k
02:04<tmk>thats 128000 bytes
02:04<Chutt>bah, whatever
02:04<tmk>it matters!
02:04<Chutt>i'm lazy
02:05<tmk>hehe me too
02:05<Chutt>it does?
02:05<tmk>yeah cause it fills 32k buffers
02:05<tmk>if there's leftovers
02:05<Chutt>the decoder should be buffering internally, no?
02:05<tmk>yeah, but it's less efficient
02:05<Chutt>so, what, 131072?
02:05<tmk>yeah that looks right
02:05<tmk>i changed it already
02:05<tmk>i wanna try 65536 first
02:05<tmk>thats what i've been using
02:06<Chutt>lemme know what works best
02:06<tmk>can you add those mutex's then?
02:06<tmk>could be printk's overflowing one another
02:06<Chutt>i'm adding a mutex lock 'round the decoder calls right now
02:06<tmk>k cool
02:06<tmk>i'll wait on your commit
02:07<tmk>it'd be nice to have some errors.. i had an old version of the driver loaded (ie non video16) and it started displaying to my monitor
02:07<tmk>which i found strange ;)
02:08<tmk>good job failing to software mode though
02:08<tmk>i was impressed at least
02:08<Chutt>has to, else the previews don't work
02:09<tmk>perhaps make it return false on candecode() if the video size is < 100x100 or something
02:09<Chutt>naw, no need
02:09<Chutt>it already knows
02:09<tmk>hows that commit comin'?
02:10<tmk>what editor do you use if i may ask?
02:10<Chutt>apparently, i broke something
02:10<tmk>me too
02:11<tmk>crazy germans taking over my mailing-list heh
02:11-!-abhishek_ [] has joined #mythtv
02:11<tmk>i think i inadvertently referred them to yours also
02:11<tmk>sorry bout that
02:12<tmk>i was just trying to get rid of them
02:12<Chutt>it's gone all black&white on me again
02:13<Chutt>earlier tonight
02:13<Chutt>and, yeah, just reloading saa7127 fixed it
02:13<Chutt>something's broken..
02:15<phar0e>did mythweb2 get moved out of mythtv's CVS repository?
02:15<Chutt>it's mythweb now.
02:15<Chutt>i sent an email to both lists about that.
02:15<bline>there was a post about that to the list
02:15<Chutt>'Small CVS changes.'
02:17<Chutt>still seems to stop a lot.
02:17<Chutt>especially after seeking/pausing
02:17<tmk>or stop
02:17<tmk>128k still
02:17<Chutt>outright stop, or break up
02:17<Chutt>i changed it to an actual 128k
02:18<tmk>lemme test something realquick
02:18<Chutt>removed the locking, too =)
02:18<Chutt>since it was blocking on that
02:18<Chutt>i may open the videodev twice
02:18<Chutt>one to do ioctls on, one to write to
02:18<Chutt>will that work?
02:19<tmk>which ioctls
02:19<Chutt>the framesync one
02:19<tmk>the start_Dec and stop_dec ones it doesn't matter
02:20<tmk>what''s framesync do?
02:20<tmk>i don't think we need that
02:20<tmk>must be one of matt's
02:20<tmk>don't need it
02:20<Chutt>doesn't it get the frame #/pts of what's playing?
02:20<Chutt>i need to know where in the stream we are
02:20<tmk>didn't i write one for that?
02:20<Chutt>and the play/pause ioctls
02:20<Chutt>i dunno
02:20<Chutt>i copied his code
02:21<Chutt>writing to the device is happening in one thread
02:21<Chutt>play/pause and that framesync are in another
02:21<Chutt>writing to the device blocks, so i _know_ it's calling those other ioctls on the same fd
02:23<Chutt>ah, should i use GET_TIMING, then?
02:23<tmk>hold on a sec
02:23<tmk>i'm seeing how his code works
02:24-!-FryGuy [] has quit [" HydraIRC -> <- Nine out of ten l33t h4x0rz prefer it"]
02:24<tmk>yeah use get_timing
02:24<tmk>i think the vsync might be causing teh freezes actually
02:27<Chutt>now i'm not getting video at all
02:27<tmk>with get_timing?
02:28<tmk>what's it returning?
02:31<tmk>what's 'frame''
02:31<tmk>is it a frame #
02:31<tmk>ie 1,2,3 etc
02:31<tmk>or some timing thing
02:31<Chutt>i'm using the pts
02:32<Chutt>to calculate the frame number, i guess
02:32<tmk>that''s silly
02:32<tmk>it gives you frame #
02:32<Chutt>just doing what was in the original code
02:32<Chutt>but, relative to what?
02:32<tmk>even get_timing gives you frame #
02:32<Chutt>if i seek, close + open it
02:32<tmk>to start of decode
02:32<Chutt>what's the frame number there?
02:32<tmk>relaltive to start of decode
02:32<Chutt>yeah, i'd have to calculate it
02:32<tmk>pts is in teh stream already i suppose
02:32<Chutt>i was going to do that, but this was how the stuff was in that patch
02:33<tmk>well i spose it doesn't matter
02:33<Chutt>the pts doesn't give the frame number
02:33<Chutt>so it'd be better to calculate it
02:33<Chutt>pts doesn't always start at 0
02:33<Chutt>but, anyway
02:34<Chutt>i'm going to have to look at this later
02:34<tmk>i'll play iwth it some too
02:35<tmk>did you commit anything yet
02:35<Chutt>using GET_TIMING seems to make it all jerky
02:35<Chutt>naw, gimme a sec and i'll put what i've got in there
02:35<tmk>well framesync may not be needed at all
02:35<tmk>or it may be called too frequently
02:36<Chutt>i need to know when it's displaying frames
02:36<Chutt>i did add a usleep in there, so it _should_ only be getting called once per frame
02:37<tmk>but does it need to be updated every frame?
02:37<tmk>that seems excessive
02:38<Chutt>'specially for edit mode.
02:38<Chutt>need to know exactly where we are
02:38<Chutt>i just committed stuff
02:38<tmk>yah i'll play with it
02:38<Chutt>left it using FRAMESYNC
02:38<Chutt>but i fixed that bug you pointed out
02:38<Chutt>and changed it to 128k
02:38<tmk>i think the problem will end up being that framesync blocks
02:39<tmk>and the decoder may not have enough free buffers to handle a long block
02:39<Chutt>switch it to the GET_TIMING real quick, you'll see it go wrong =)
02:40-!-brainless [] has joined #mythtv
02:40<Chutt>i expect a patch in my inbox by morning, now, so chop-chop
02:40<Chutt>i did add another fd variable to that class
02:41<Chutt>let me know if i should make it use another fd to write to
02:41-!-phar0e [] has quit []
02:42<tmk>same fd
02:42<tmk>is preferable
02:42<tmk>but i'll try a diff one for framesync
02:42<Chutt>but, writing the actual data
02:42<Chutt>and stuff
02:42<Chutt>blocks, right?
02:42<tmk>so does framesync
02:42<Chutt>so what happens if i come along in another thread and do an ioctl
02:43<tmk>depends on the ioctl
02:43<Chutt>while it's blocked for writing?
02:43<tmk>there's very little internal checking
02:43<tmk>all the decoder ioctls just set internal variables
02:43<Chutt>the play/pause/framesync calls are all currently in a different thread
02:43<tmk>that don't' affect decoding
02:43<tmk>so framesync and write are concurrent?
02:43<Chutt>from the write/close/open for seeking
02:43<Chutt>and so's play/pause, i imagine
02:44<tmk>we don't pause while ff'ding tho
02:44<tmk>i haven't tried the actual pause feature yet
02:44<Chutt>hit p =)
02:44<tmk>go to bed
02:44<Chutt>have fun
02:49-!-steelep [signwatchr@] has joined #mythtv
02:55-!-abhishek_ [] has quit [Client Quit]
02:57-!-activelow [] has quit ["Client exiting"]
02:58-!-abhishek [] has quit [Read error: 104 (Connection reset by peer)]
03:18-!-choenig [] has joined #mythtv
03:22-!-tmk [] has quit []
03:30-!-KikoV [] has quit [Read error: 113 (No route to host)]
03:39-!-brainless [] has quit []
04:13-!-o_cee\zZz [] has quit [Read error: 60 (Operation timed out)]
04:19-!-fr32 [] has joined #mythtv
04:28-!-Viddy [] has quit [Remote closed the connection]
04:29-!-Viddy [] has joined #mythtv
04:36-!-choenig [] has quit [Remote closed the connection]
04:37-!-steelep [signwatchr@] has quit ["me is bugging out"]
04:44-!-fr32 [] has left #mythtv []
04:44-!-fr32 [] has joined #mythtv
04:47-!-fr32 [] has quit []
04:53-!-sfr [] has joined #mythtv
05:56-!-o_cee [] has joined #mythtv
06:25-!-fr32 [] has joined #mythtv
06:41-!-fr32 [] has quit []
07:44-!-jkolb [] has joined #mythtv
08:17-!-o_cee [] has quit [Read error: 60 (Operation timed out)]
08:17-!-jkolb [] has quit [Read error: 104 (Connection reset by peer)]
09:09-!-jkolb [] has joined #mythtv
09:36-!-Captain_Murdoch_ [] has joined #mythtv
09:43-!-pahli_bar [] has joined #mythtv
09:44-!-pahli_bar [] has quit [Client Quit]
09:47-!-rkulagow [] has joined #mythtv
09:48-!-fr33 [] has joined #mythtv
09:50<fr33>isaac sometimes here?
09:51<sfr>fr33, sure just ask for Chutt
09:53<fr33>someone saw this technique :
09:53<fr33>looks very good! download the sample from there page
09:54-!-fr33 [] has left #mythtv []
10:13-!-FryGuy [] has joined #mythtv
10:23-!-m0j0 [~m0j0@] has joined #mythtv
10:51-!-schultmc [] has joined #mythtv
11:03-!-dopez [] has joined #mythtv
11:04-!-kja [] has joined #mythtv
11:05<Chutt>rkulagow, thanks for updating the docs
11:06-!-sc00p_ [] has joined #mythtv
11:06<rkulagow>no problem. "maybe it should be in the FAQ"
11:10<rkulagow>i can't believe it's taken this long to get the official recommended modules.conf for the pvr's. if the "new" one works, i'll update the HOWTO.
11:16-!-sc00p [] has quit [Read error: 104 (Connection reset by peer)]
11:17-!-sc00p__ [] has joined #mythtv
11:19-!-mecraw [~mecraw@] has joined #mythtv
11:19<rkulagow>hey, chutt. broken GDB issue again. if i were to open a bug for the Mandrake GDB because it keeps stopping for SIG32, what specific technical thing would I actually need to put in the bug so that the packager could say, "oh, yeah, ok, that makes sense"?
11:19-!-sc00p_ [] has quit [Read error: 104 (Connection reset by peer)]
11:20<Chutt>i dunno
11:20<Chutt>like, a log of it running
11:20<mdz_>rkulagow: how about "makes it impossible to debug programs which use pthreads"?
11:20<Chutt>that, too
11:20<mdz_>well, not impossible, I guess
11:20<rkulagow>ah, ok, i figured it was some sort of compile option where they've used the wrong one or something.
11:21<mdz_>probably it's just old
11:21<mdz_>gdb used to suck like that everywhere
11:21<rkulagow>[mythtv@masterbackend mythtv]$ rpm -qa|grep gdb
11:21<rkulagow>[mythtv@masterbackend mythtv]$ rpm -qa|grep gcc
11:22<rkulagow>let me check gnu for latest version and see if that's it.
11:23<rkulagow>Version 5.3 (stable) released on 2002-12-16
11:24<rkulagow>hrmm. seems like it's current.
11:25<Chutt>if your libc was released much more recently than the gdb, that's a problem
11:26<Chutt>mythbackend started segfaulting again when i do a make install
11:26<Chutt>it hadn't for the longest time
11:30-!-sc00p_ [] has joined #mythtv
11:43-!-StarHeart [] has joined #mythtv
11:43-!-_rkulagow [] has joined #mythtv
11:45-!-hfb [] has joined #mythtv
11:45-!-sc00p__ [] has quit [Read error: 110 (Connection timed out)]
11:53-!-choenig [] has joined #mythtv
12:01-!-rkulagow [] has quit [Read error: 110 (Connection timed out)]
12:20-!-Drikus_ [] has joined #mythtv
12:20<_rkulagow>damn. what was the URL of that real-time #mythtv IRC log? my session reset, and i lost scrollback about that GDB issue we were talking about.
12:22-!-steelep [~signwatch@] has joined #mythtv
12:24<_rkulagow>thanks for the URL.
12:33-!-kja [] has quit ["..."]
12:34-!-tmk [] has joined #mythtv
12:45<Chutt>tmk, hi
13:06<Chutt>hey, ever get that decoder working?
13:08<tmk>did you get my email
13:08<Chutt>tmk, yeah
13:08<tmk>myth takes like an hour to compile
13:08<tmk>so i went to bed
13:09<Chutt>yea yea
13:09<Chutt>you probably didn't need to do a full recompile
13:09<tmk>i didn't
13:09<tmk>i just typed 'make'
13:09<tmk>i guess something got flipped that said it needed a full recompile
13:09<Morph>tmk: your "prize" has been shipped.
13:09<tmk>Morph: excellent!
13:10<tmk>i never got clarification: does it come with a case
13:10<tmk>or is it just the board
13:10<Chutt>tmk, so it should be very slightly faster than your existing machine, but it's still pretty damn slow
13:10<Morph>hmm. I think it might have been just the board.
13:10<tmk>hmm. perhaps i'll mount it on my wall somehow, or on the underside of one of my A/V shelves
13:11<tmk>Chutt: yeah, but does the mpeg decoder work?
13:11<tmk>well enough i mean
13:11<Chutt>but it takes an age to compile
13:11<tmk>no problem there
13:11<tmk>that's what daytime during work is for
13:11<tmk>or nighttime during sleep
13:11<tmk>how is the tv-out on it? i have a feeling i'll prefer the pvr350
13:12<Chutt>you'll prefer the 350
13:12<tmk>i thought as much
13:12<Chutt>the tv-out's ok, though
13:20<tmk>grr why won't x-chat recognize ^U
13:38<bline>use irssi
13:39-!-kja [] has joined #mythtv
13:39-!-Justin_ [] has quit [Read error: 110 (Connection timed out)]
13:43-!-Justin_ [] has joined #mythtv
13:48<sfr>Chutt: don't tell me it's actually THAT easy? Only 1h ago did i remember streaminput.* which i had completely removed from my tree. Before i spent hours starring at the code trying to figure out how to do it.
13:48<Chutt>pretty much
13:49* sfr listens to - Best of the 80 's
13:49<Chutt>got it working, then?
13:49<sfr>took about 15 minutes now after remebering the streaminput.
13:51<Chutt>got a quick patch?
13:51<sfr>BUT, i have no idea how the socket connection is handled in the decoder.
13:51<Chutt>it's really just inheritance
13:51* sfr believes that.
13:52<Chutt>the decoder uses a QIODevice for all it's reading/seeking and stuff
13:52<Chutt>a QFile is a QIODevice, and so is a QSocket
13:52<Chutt>so they both look the same to the decoders
13:52<sfr>and QIODevice knows that seeking isn't working for a socket?
13:53<sfr>think i need a good book/resource on qt programming.
13:53<sfr> isn't much more than a refernce.
13:53<Chutt>check out the top of vorbisdecoder.cpp
13:53<Chutt>the oggseek() function, in particular
13:54<sfr>Chutt: ok, will do later on.
13:54<Chutt>well, that's just how it knows it can't seek on it
13:54<jkolb>Eh? Can't seek oggs?
13:54<Chutt>http streaming.
13:55<jkolb>Oh, that.
13:55<sfr>Chutt: So what you actually meant with new input method, was to create a new class based on streaminput?
13:55<sfr>one per stream type? e.g. shoutcast, realaudio ...
13:55<Chutt>well, one per url type
13:55<sfr>url/audio protocol, no?
13:56<Chutt>http would be one
13:56<sfr>i mean. are there shoutcast streams sending ogg vorbis?
13:56<Chutt>it's a similar format
13:56<Chutt>and the http input should be able to do static files on a webserver, too
13:57<sfr>yes, but there must/should be a matching decoder, too.
13:57<Chutt>well, the decoder should be separate from how it's getting the data
13:58<sfr>like a realaudio feed via rtsp
13:58<Chutt>so you'd need a realaudio decoder, and a rtsp input thingie
13:58<sfr>oh, sure. like it is now.
13:59<sfr>But one thing i haven't looked at yet: how to intercept the mp3 stream to filter the tags. is readyread called during streaming?
13:59<Chutt>should be, yeah
14:00<Chutt>i'm not exactly sure, myself
14:00<tmk>Chutt: i got another email from matt: he said he's going to look into the decoding-stops
14:00<sfr>i mean, whenever the decoder requests more data.
14:00<Chutt>tmk, i think i've gotten most of it
14:00<Chutt>bugs in his code =)
14:00<tmk>i meant when the ivtv-driver stops decoding
14:01<tmk>does that sometimes to me
14:01<Chutt>i've fixed things around a little, and it hasn't done it yet
14:01<sfr>so Chutt, you once worked for trolltech? or this Brad Hughes just contributed this code.
14:01<tmk>what was the issue? other than mattbugs? :)
14:01<Chutt>sfr, well, he didn't contribute it, i took it out of his mq3 program :p
14:01<sfr>Chutt: :))
14:02<Chutt>tmk, after an unpause, it was calling IOC_PLAY very, very often
14:02<tmk>i see
14:02<Chutt>and that seemed to lock it occasionally
14:02<Chutt>and the FRAMESYNC stuff would cause it to lock as well
14:02<sfr>Chutt: before sending a patch, i'll create a proper shoutstream class, clean it up, and add support for the pls playlists.
14:02<tmk>well this does it for me on normal dd if=blah.mpg of=/dev/video16
14:03<tmk>well i'm glad you're on top of it
14:03<Chutt>sfr, ok, that's up to you
14:03<Chutt>i don't mind applying smaller patches, though
14:03<tmk>i guess our balance of favors is now even, or perhaps in your favor ;)
14:04<sfr>Chutt: but you sure wouldn't like to listen to 80's rock all the time, cause it's hardcoded now.
14:04<tmk>infact, i'm quite sure i owe you one now :/ (thanks morph ;)
14:05<tmk>looking at via's site: - VIA FliteDeck? Luxurious Utility
14:05<tmk>wth does that mean
14:07<Morph>isnt that their tool to boot up quickly and play a DVD?
14:09<tmk>seems strange to call it 'luxurious'
14:10<Morph>heh. no one truly understands everything that VIA does.
14:16<Chutt>tmk, hey
14:16<Chutt>uhm, why am i getting 22763904 back from the GET_TIMING call's .frame member?
14:17<Chutt>for the very first frame decoded
14:17<Chutt>you're not initializing something =)
14:22<tmk>yeah seems so :<
14:22<Chutt>i'll work around it for now
14:22<tmk>i think the framesync api call migth be the right one
14:23<tmk>seems the same as get_timing
14:23<tmk>but better synced
14:23<Chutt>i'm calling it in the same thread as the write() now
14:23<Chutt>seems to work ok
14:23<Chutt>i don't get updated every frame, but...
14:23<Chutt>in WriteBuffer()
14:23<Chutt>the GET_TIMING
14:24<tmk>hmm i'm just passing along what the card sends in get_timing
14:25<tmk>are you checking for negative retval? the api call might be failign
14:25<Chutt>i am
14:26<tmk>but it works on subsequent calls right
14:26<Chutt>yeah, once it's started decoding
14:27<tmk>perhaps i need to put a wait in until it's sent some data
14:28<tmk>oh also, not sure what this means, but it seems like the stream isn't closing down properly with mythtv
14:28<tmk>even when watching video
14:28<tmk>err wait, i may have ^C'd mythfrontend
14:29<Chutt>i just made it reset to blank
14:29<Chutt>after it's opened the device
14:29<Chutt>i'll be committing soon
14:29<Chutt>this seems to work better, at least
14:29<tmk>i'll probably start compiling it soon then ;)
14:29<tmk>if i want it done by dinnertime
14:30<Chutt>make sure you remove the 'return false' at the top of CanDecode()
14:31<tmk>i know 'bout that :)
14:32<Chutt>it did just stop decoding
14:34<Chutt>there, committed
14:34<tmk>i think it's due to sending dma's too frequently
14:34<tmk>since it didn't do it before we switched to matt's algorithm
14:34<tmk>which sends 32k at a time
14:34<tmk>before we were doing 64k
14:35<tmk>your docs sure do get updated a lot
14:37<tmk>Chutt: would it make sense to wrap that 'return false;' in an #ifdef
14:37<tmk>so i can just edit
14:37<Chutt>it's only temporary
14:37<Chutt>i just need to get around to adding some setup options for it
14:37<tmk>and it's 'CanHandle' :
14:38<Chutt>yeah, whatever
14:38<Chutt>so, for the osd
14:38<Chutt>can i bypass the ivtv-fb driver, or do i have to go through it?
14:38<tmk>i'd say bypass is fine
14:38<tmk>in fact, preferred
14:39<Chutt>ok, so, are there ioctls for that?
14:39<tmk>no i don't thinkso
14:39<tmk>also, i want to do direct memory writes for that
14:39<tmk>instead of dma
14:39<tmk>since i don't want to deal with the fact that only one dma at a time is allowed
14:40<tmk>but that's transparent to you
14:40<Chutt>well, i need a means to get at that memory
14:40<tmk>just be aware that it'll eat cpu
14:40<tmk>yeah there's an ioctl for that
14:40<tmk>but i think it's in the FB code
14:40<tmk>i may move it
14:40<Chutt>that's what i was just looking at :p
14:40<tmk>for now, just point the call at the fbdev
14:41<tmk>i need to rework that a bit given that it's in global alpha mode
14:41<Chutt>i'm just looking through source
14:41<Chutt>for now
14:41<Chutt>seeing how that all works
14:42<tmk>anyone know of any cases that have ir headers
14:42<tmk>for mini-it
14:42<tmk>and a pci riser if they're flat
14:42<Chutt>the -350 is about the same length as the motherboard
14:42<Chutt>which causes problems in the tiny cases
14:43<tmk>ideally, i want a wide one
14:43<Chutt>i'm using a full-sized desktop case
14:43<tmk>well i want to put it on my a/v rack
14:45-!-extremis [] has left #mythtv []
14:46<Chutt>my case is only slightly taller than the dvd player i have
14:46<Chutt>but i have a huge old dvd player
14:48-!-bline [] has quit [Read error: 104 (Connection reset by peer)]
14:49-!-bline [] has joined #mythtv
14:52<jkolb>That's always a good clue to check -users.
15:02<hadees_sleep>i can run mythtv fine from just a window manager right?i don't need gnome or kde right?
15:02<Chutt>of course
15:02<Chutt>and if you don't use mythvideo or mythdvd, you don't even need a window manager
15:02-!-hadees_sleep is now known as hadees
15:03<Morph>Dont need a window manager?! Thats madness! how can it be done?!
15:03* Morph gets out of infomercial mode
15:04<hadees>Chutt: well i want mythvideo and mythdvd, any suggestions on a window manager? what do most peole who set up myth like that use?
15:04<Chutt>i run blackbox on my living room box
15:04<Chutt>pretty lightweight
15:05<hadees>cool, thx for the info
15:06<Morph>openbox works well too..
15:06<jkolb>Yeah, I think I'm using openbox.
15:07<Morph>using it with epist is handy..
15:09<jkolb>I used to use epist, but I only use the box for myth, so I just launch it at startup now.
15:10<Morph>I found it handy when myth doesnt work for whatever reason and I need a terminal on the box :)
15:10<jkolb>Ah. That's sshd's job.
15:11<Morph>ssh doesnt work too well when you need to start myth front or backend by hand..
15:11<Morph>well. frontend anyways.
15:11<jkolb>Why not?
15:12<jkolb>nohup mythfrontend > mythfrontend.log &
15:13<Morph>sure sure. run to next room and see if it started, run back and check the log.
15:14<jkolb>Oh. You don't have a laptop and 802.11. That's your own fault.
15:14<jkolb>Or just set up a series of mirrors.
15:15<Morph>are you just arguing this for a sake of arguing? :)
15:15* tmk watches myth compile
15:16<tmk>is the correct configure/make process this:
15:16<jkolb>I suppose. I really do have a laptop and 802.11 in my house, though. It's nice.
15:16<tmk>make doesn't seem to work till i run qmake
15:16<tmk>but qmake doesn't do anything
15:16<Morph>as do I.
15:16<Morph>tmk: you have QTDIR defined?
15:17<Chutt>qmake makes the Makefile
15:17<tmk>yeah i figured as much
15:17<jkolb>Morph: Then why are you running around between rooms?
15:18<Morph>jkolb: I usually dont. thats why I have a keyboard handy, if there are problems I hook the kyboard up, opena term. and test.
15:19<Morph>hrm. lag.
15:21<tmk>i run wires
15:21<tmk>fast ethernet is good ethernet
15:22<jkolb>I have both. Jacks in every room, and 802.11 for the laptops.
15:23<jkolb>I even ran 5e in case I ever buy a gig-e switch.
15:28<_rkulagow> chutt / mdz: i'm going to enter a bug for gdb for Mandrake cooker because of the SIG32 events that we mentioned earlier. IIRC, there's also instances where gdb doesn't have the full stack so that the bt isn't complete. am i remembering correctly?
15:36-!-jkolb [] has quit []
15:37<Chutt>rkulagow, i think so
15:38<_rkulagow>chutt: can you please sync docs?
15:38<Chutt>i was just about to =)
15:41<_rkulagow>i wish i had a testcase for the Mandrake people so they can see what it is exactly that i'm talking about.
15:46<mecraw>should the ChannelInfo know what videosource it comes from?
15:48<mecraw>i have 2 videosources, when i go to the epg and select to view a channel that isn't on the currently playing source it doesn't do anything
15:48<mecraw>it just returns to the currently playing channel
15:48<Chutt>yeah, you can't switch cards/sources like that
15:49<mecraw>i'd like to, any suggestions on what to change?
15:49<mecraw>i could try to look up the source by channel string
15:49<Chutt>well, switching capture cards in live-tv mode would be difficult
15:50<mecraw>or i could also put the sourceid in the channel info
15:50<Chutt>if it were only inputs on a single card, that shouldn't be too difficult
15:50<mecraw>actually i'm just using 2 sources on my pvr250
15:50<Chutt>just have to look up which input to change to
15:50<Chutt>and, then, well, change to it =)
15:52<mecraw>i'll give it a shot, i think the best solution is to add to the ChannelInfo class
15:52<Chutt>probably, yeah
15:52<mecraw>why is it hard to change capture cards in live tv?
15:52<Chutt>you've got the whole connection to the backend thing to change
15:52<Chutt>the desired card has to be available, etc
15:53<mecraw>ah, ok, thanks
15:56-!-chimaera [] has joined #mythtv
15:58-!-deflux- [] has joined #mythtv
15:58<deflux->I keep getting IO BOUND- blocking in ThreadedFileWriter::Write(), then eventually the frontend freezes.
15:58<deflux->Anyone run into this issue?
15:58<chimaera>when i try to fill my database or run the cronjob, i get a message that qpixmap can't be created and nothing else happens..
16:00<Chutt>deflux, it means that it can't write to your disk fast enough.
16:00<Chutt>chimaera, that's not a fatal error message.
16:01<deflux->Anyway to have a larger buffer or something?
16:01<deflux->It's on a DMA100 controller. o,o
16:01<Chutt>it's got a fairly large buffer already
16:01<Chutt>make sure dma's turned on.
16:01<deflux->It is. o,o
16:01<Chutt>don't know what to tell you, then.
16:03<chimaera>Chutt: so nothing to worry about..
16:04<Chutt>there's something else wrong, i'd say.
16:12-!-davipt [] has joined #mythtv
16:13<thor_>Chutt, well put
16:13<_rkulagow>chutt / mdz: if you have a second, can you take a quick look at the response i got to the gdb issue?
16:18<tmk>oh my! can it be? myth finished compiling
16:18<Chutt>rkulagow, if it's not dealing with the SIG32s, that means it's not dealing with threads properly
16:18<Chutt>it should be handling those internally
16:18<Chutt>since it does need to do stuff when it gets them
16:18-!-tmk is now known as tmk|out
16:21<_rkulagow>so the nostop commands are a workaround to something that shouldn't be broken in the first place? (if i'm getting the gist of it)
16:22<Chutt>the nostop commands are _probably_ causing it to mess up even more
16:22<_rkulagow>ok. i'm going to look for the GDB readme and see if there's some sort of compile / config flag that should be flipped.
16:22-!-chimaera [] has quit ["leaving"]
16:31-!-davipt [] has quit ["brb"]
16:31-!-davipt [] has joined #mythtv
16:50-!-ahbritto [] has quit [Client Quit]
17:00-!-_rkulagow [] has left #mythtv []
17:09-!-dopez [] has quit []
17:11<sfr>suppose i have a pvr350. Now if i add a dvb (t, c or s type) card, could the dvb-stream (mpeg2 afaik) be directly handled by the pvr350 mpeg-decoder logic?
17:12<Chutt>it really depends on how picky the decoder is
17:12<Chutt>i haven't tried to play anything but the files it produces through it
17:12-!-Captain_Murdoch_ [] has quit []
17:13<sfr>Aha, so producer #1 mpeg2 ~= producer #2 mpeg2?
17:13<sfr>:( sucks
17:14<sfr>what is the difference in cpu-load with/without using the pvr- mpeg decoder?
17:20-!-dopez [] has joined #mythtv
17:21-!-Drikus_ [] has quit ["toedeledoki"]
17:21<Chutt>with the decoder, it basically uses just a tiny amount of cpu
17:21<Chutt>for playback
17:22<sfr>so it would be a noticable difference, say on a 2x850MHz machine.
17:25<sfr>And just a clarification: the decoder output is NOT usable as a regular display, only programs like mplayer, myth and so on can make use of it?
17:25<Chutt>that's not true
17:26<sfr>i can use this output as a X display?
17:28<sfr>and that applies to both pvr types? they become more and more interesting..
17:29<sfr>ok, not to bother you any longer. Should i search the ivtv site/ml for more?
17:30<Chutt>or ask tmk when he comes back
17:30<choenig>sfr: aren't you talking about the output only? the pvr 350 doesn't have a decoder or did I miss smth?
17:31<Morph>the pvr350 has a decoder.
17:31<sfr>choenig: yes. i think both cards have a h/w decoder. the pvr350 only adds the radio iirc.
17:32<choenig>oh, I ment the 250 has no decoder
17:32<Morph>the 250 doesnt have a decoder.
17:32<thor_>some 250's do have a decoder
17:32<choenig>just a typo :)
17:32<thor_>just no video out
17:32<Morph>really? Hmm..hadnt heard that.
17:33<thor_>ivtv faq, 2 revs of 250
17:33<sfr>thor_: you got me confused now. So i connect the tv to the pvr350' decoder output. And i can use this output also as a X display?
17:34<choenig>hmm, sfr, I thought that's what you meant above ;)
17:34<thor_>according to Chutt (who understands these things), the 350 can present itself as a framebuffer ... onto wich you could run X .... but I only have a slim grasp of this
17:35<sfr>thor_: Yeah, thats we he also just told me.
17:35<thor_>ah, sorry, I just walked in
17:36<sfr>but i guess, only X cvs has support for the pvr350 fb. Chutt?
17:36<Chutt>doubt it
17:36<Chutt>it should just use the standard X fb driver
17:37<sfr>in this case, i think i'll get one. :)
17:40<m0j0>is there a chance that the 250v1 decoder will be useful in the future?
17:41<Chutt>a very slight chance
17:42<sfr>Hm, is PAL support being actively worked on? Or even working?
17:43<Chutt>i don't know, sorry
17:44-!-rkulagow [] has joined #mythtv
17:45<sfr>time to sleep, Good Night.
17:45-!-sfr [] has quit ["Client exiting"]
17:52<m0j0>does it look like there will be advantages to using 350 decoder beyond lower cpu?
17:52<Chutt>better quality tv out
17:52<thor_>Chutt, you doing that mytDVD patch or you want me to ?
17:52<Chutt>the serialize one?
17:52<Chutt>figured you could =)
17:53<thor_>serialize I'm on top of, I meant the translation one (just posted)
17:53<Chutt>ah, allright
17:53<Chutt>if you want to apply it, you can, but i don't mind doing it
17:54<thor_>you go ahead, save me messing up my half finished serialize code
17:55<thor_>not sure you should really be translating "ac3" and "dts" and stuff ...
17:55<Chutt>probably not
17:57-!-choenig [] has quit [Remote closed the connection]
18:00-!-tmk|out is now known as tmk
18:01<tmk>sfr: only the pvr-350 can decode currently
18:02<tmk>the 250 may have the chip, but it doesn't seem like it'll work as a general purpose decoder (ie decode to mem)
18:05<tmk>oh, of course he's gone now
18:09<m0j0>sfr is gone, but that bit of info was interesting to me. ;) i may have to invest in a 350 at some point since it doesn't look like my 250v1 decoder will be of much use.
18:09<kja>I see that on init of mpeg2 stuff in avformatdecoder keyframedist is initialized to either 12 or 15, is these the only possible values? (Sorry for bugging you all, but I don't have the specs...)
18:11<tmk>m0j0, i wish it were possible, but without some better research
18:11<tmk>i won't be able to
18:11<tmk>kja: nah it could be anything i guess
18:11<tmk>but 12 and 15 are standard
18:12<tmk>2 gop's per second
18:12<tmk>pal and ntsc respectively
18:12<Chutt>that's why it calculates it as well
18:12<tmk>i didn't realize that it calc'd. n33t
18:13<Chutt>it tries to, at least =
18:13<Chutt>er, =)
18:13<Chutt>ah well, later
18:13<kja>Chutt, Yea I know that, the reason I ask is that most streams from dvb is either 12/15, but occationally theres somthing wrong in the stream, so it comes out lower
18:14<kja>wondering if a calculation every time is better...
18:16<kja>could be that I'm missing a pic_start on the boundary of the buffer though, I'll investigate further, thanks for the info
18:24-!-m0j0 [~m0j0@] has quit ["Client exiting"]
18:52-!-hfb [] has quit ["Client exiting"]
18:52-!-steelep [~signwatch@] has quit [Read error: 54 (Connection reset by peer)]
18:57-!-steelep [~signwatch@] has joined #mythtv
19:04-!-schultmc [] has quit ["Client exiting"]
19:39-!-mechou [] has joined #mythtv
19:57<tmk>Chutt: not sure if you used it, but select/poll works properly now for the encoder
19:58<Chutt>heh, cool
19:59<Chutt>i just read() big chunks from it
19:59<tmk>k, i'm adding it to the decoder too
19:59<tmk>not sure if you care there either :)
20:00<Chutt>the decoder might be more interesting
20:00<Chutt>wouldn't block as long
20:00<tmk>well it sleeps on the same waitq
20:00<tmk>i think
20:01<tmk>but you would only be guarenteed writing 32k without blocking in any event
20:03-!-steelep [~signwatch@] has quit ["me is bugging out"]
20:06<tmk>not sure the implications of blocking in mythtv
20:06<Chutt>doesn't matter in that thread
20:06-!-StarHeart [] has quit ["Client exiting"]
20:06<Chutt>as long as it's not for _too_ long
20:08-!-Captain_Murdoch [] has joined #mythtv
20:09-!-mechou [] has quit [Read error: 60 (Operation timed out)]
20:14-!-mecraw [~mecraw@] has quit ["Trillian ("]
20:15<tmk>well, it has to xfer 1Mbps or so for full-size video
20:15<tmk>so it won't be waiting too long in any case
20:18* tmk goes home
20:18-!-tmk [] has quit ["Client exiting"]
20:22<kja>Chutt, would you accept me making the 'setup' able to configure any backend and making it a bit more flexible with different cardtypes?
20:22-!-Captain_Murdoch_ [] has joined #mythtv
20:23<Chutt>how are you going to probe cards on different backends?
20:23* kja is tired of waiting for someone to step up and do dvb-setup, so he's halting dvb backend coding, hehe :)
20:23<kja>hmmm, good question
20:23<Chutt>if you can do that cleanly, then, sure
20:24<kja>only needs a small extension to the back-frontend protocol../me thinks
20:27<Chutt>backend doesn't run without the cards defined
20:27<Chutt>and doesn't allow you to change things without a restart
20:29<kja>what if I added a pre-configured state to it, i.e. if it's not configured it starts waiting for configuration probe, if it's configurated already and receives a reconfigure event it will restart itself by forking...
20:29<Chutt>i dunno
20:31<kja>sounds just nasty?
20:32<Captain_Murdoch_>I thought it was changed recently to allow the backend to start even if cards weren't defined. instead of aborting it just doesn't start the scheduler, or am I thinking of something else.
20:33<kja>yea capt. /me thinks that too
20:34<kja>Chutt, anyways if you think that the restart if already configured potentially destroy things, it could wait until it's not doing anything
20:35<Chutt>i dunno
20:36<kja>and we get the advantage of not having to have X on backend (or am I missing sumthin, never configured a separate backend)
20:36<Chutt>you still need X on the backend
20:36<Chutt>since qt needs X in most instances
20:37<Chutt>and it's not difficult to export a display
20:37<kja>aha :)
20:37-!-bbeattie [] has joined #mythtv
20:39<kja>so, really, remote config could be implemented as a call to backend to start setup on a remote display?
20:40-!-poptix_ [] has joined #mythtv
20:40-!-poptix_ [] has quit [Client Quit]
20:41<Chutt>since that'd require logging in and stuff
20:42-!-poptix [] has joined #mythtv
20:42<poptix>mythtv via 802.11g
20:42<Chutt>works ok?
20:43<poptix>works great
20:43<poptix>even with my hardware mpeg-2
20:43<poptix>i get around 22mbit _real_ throughput
20:43<poptix>mythtv is using an average of 9mbit
20:44<poptix>i'm in my car outside my 3rd floor apartment at the moment
20:44<poptix>still looking good =p
20:44-!-bbeattie [] has quit ["off to find dinner."]
20:47<kja>Chutt, anyway, i'm going to expand the card-config stuff, so I can figure out the remote stuff while i'm doing that...
20:52<Chutt>is there any real reason that the external channel changer script waits until the script's done?
20:55<Chutt>ABC is all fucked up
20:55<Chutt>it looks like they're getting a station that's being transmitted at a _really_ low bitrate
20:55<Chutt>thought it was myth there for a second =)
20:56<Chutt>mdz, hey, are you there?
20:58<mdz>Chutt: yep
20:58<Chutt>so, is there any reason i can't get rid of the waitpid() in the external channel change bit?
20:59<mdz>I don't think so
20:59<mdz>I'll have a peek
20:59<mdz>I guess the only reason is to check whether it failed
21:00<mdz>(and to prevent it from being a zombie)
21:01<Chutt>ah, true for that
21:01<Chutt>but, what if i don't want to wait? =)
21:01<mdz>then you need to handle SIGCHLD
21:02<mdz>or, hell, just have the shell put it in the backgroun
21:02<Chutt>slap a & on the end?
21:02<mdz>it is invoking the shell anyway
21:02<mdz>%1 %2 &
21:02<mdz>that way it'll be the shell's child, the shell will exit, and it'll be orphaned to init
21:03<Chutt>which'll reap it
21:03<mdz>the only failure case which really matters is when they've botched the configuration
21:03<mdz>in which case, the shell should complain to stderr
21:03<mdz>I think
21:04<Chutt>ah well
21:04<Chutt>i'll commit that, and if people bitch, i can revert it
21:05<mdz>sounds good
21:05<mdz>I agree that it should be more usable if it doesn't wait
21:06<Chutt>channel changing's a tad faster with my current stuff
21:06<Chutt>though waiting on the prebuffering is still the longest thing
21:16<davipt>"<poptix:#mythtv> mythtv via 802.11g"
21:16<davipt>I've been using standard 802.11 (11MB) at 352x288
21:17-!-markr [] has joined #mythtv
21:17<davipt>mdz: I've been updating mythtv for 0.12/cvs, do you want diffs for debian dirs ?
21:17<mdz>davipt: sure, ok
21:22-!-markr [] has quit []
21:23<Chutt>stuff's changed that much?
21:26<mdz>I have no idea
21:28-!-schwin97 [] has joined #mythtv
21:37<vektor>Chutt: Do you ever set the V4L 'volume' control?
21:38<vektor>I have this fun problem. On my card, any value other than 100% means mute.
21:39<vektor>But on some drivers, it seems it's an actual volume control, and 100% means really loud and clipping.
21:39<vektor>Not sure what to do about this.
21:39<Chutt>yeah, people have been asking about that
21:39<Chutt>for me as well
21:39<vektor>mplayer seems to set the volume to 0 on exit.
21:39<vektor>This is why the problem is annoying.
21:40<vektor>I can't have tvtime start and leave the volume at 0.
21:51<mdz>solution: don't run mplayer :-P
21:52<Chutt>people insist on using it for some reason
21:52<Chutt>well, i was averaging 800ms to do all the background work to change channels
21:52<Chutt>now it's 300ms
21:53<Chutt>guess that's not too bad
21:53<Chutt>still takes awhile to buffer up enough data, though
21:53<vektor>Chutt: Is this V4L work or like back-end DB work and stuff?
21:54<thor_>assuming that's a time issue (data to arrive) rather than a cpu issue (assembly of data)?
21:54<Chutt>thor, yeah
21:54<Chutt>though, now i've broken things
21:55<Chutt>and all i get is a black screen =)
21:55<thor_>it's your audio ....
21:55<thor_>ha ha
21:55<Chutt>converted some stuff to use QWaitConditions
21:55<Chutt>i probably missed something in one of em
21:57<mdz>oh, nice
21:58<hadees>what player do you guys use if you don't use mplayer? xine? ogle?
21:59<thor_>player .... that would entail sitting in front of a tv and watching something .... don't have time for that
21:59<hadees>to busy programing mythtv to use it eh?
22:29<kja>hmm, seeking backwards in livetv is a bit acward, it seem to hit the same seektable index multiple times...
22:29* kja is trying to figure out how the myth ui is workin..
22:32<kja>and there is the occational 'freze', but usually seeking like crazy 'fixes' it :)
22:33<mikegrb>2.6.0 is uber-responsive
22:38* kja ..while watching mtv :)
22:40-!-tmk [] has joined #mythtv
22:42<mikegrb>ran 'updatedb' without nice'inh it, wouldn't try myth with that but I could browse and switch tabs etc as if updatedb wasn't running
22:42<mikegrb>that and all the hardware on my laptop is finally working
22:44<Chutt>yeah, i want the docs updated by someone that spells like 'nessacary'
22:45* Chutt rips apart more code
22:47-!-mecraw [] has joined #mythtv
22:48-!-mecraw [] has quit [Client Quit]
22:51-!-mecraw [] has joined #mythtv
22:52<kja>sure, that 26' is very nice on hard tasks
22:56<tmk>evening boys
22:56<tmk>and girls if there are any
22:57<Chutt>so, the decoder code work for you?
22:58<tmk>trying now
22:58<tmk>did you update anymore?
22:58<Chutt>not yet
22:58<tmk>hey did you see my note about the make install
22:58<tmk>and strip
22:58<Chutt>it doesn't strip here when compiled in debug mode
23:00<tmk>perhaps my cvs'ing reset my
23:00<tmk>err yeah pretty sure it did
23:00* tmk withdraws comment
23:00<tmk>hmm. mythbackend exited with 'bus error'
23:01<tmk>(before now)
23:01<tmk>so what does the wife think of all this computing? my g/f is starting to get jealous :)
23:02<Chutt>she's used to it
23:02<Captain_Murdoch_>pretty sure files are stripped here when in debug mode also. maybe it's because debug and release are both defined.
23:02<Chutt>captain_murdoch, i comment out release mode when enabling debug mode
23:03<Chutt>ah well, back later
23:03<Captain_Murdoch_>I'll test that. my binaries are stripped but I have the debug line uncommented in
23:05<tmk>chutt: seems real good
23:10<tmk>it stops decoding sometimes
23:10<tmk>but that's the driver's fault
23:10<Chutt>tmk, did you wrap decoder call 0x05?
23:10<Chutt>if i wanted to only decode one frame..
23:12<mdz>I hate qmake; I wish it would leave me alone
23:12<tmk>no not yet
23:12<tmk>but i plan to
23:13<tmk>this is quite nice
23:13<tmk>btw, good work :)
23:13<tmk>now if only emeril wasn't on tv...
23:14<tmk>oh chutt: one thing i did notice
23:14<tmk>channel changing doesn't seem to work
23:14<tmk>by #
23:14<tmk>up/down work
23:14<tmk>but i typed in '02' and nothing happened
23:14<Chutt>oh, it's probably waiting on the osd
23:14<Chutt>and the osd isn't doing anything, so, doesn't work
23:15<tmk>well i have about 1.5 hrs to code
23:15<tmk>anything you really want done?
23:15<Chutt>i dunno
23:15<Chutt>whatever you want, really
23:16<Chutt>a way to get the osd address and size and stuff would be cool, then i could finish that off
23:16<Chutt>if you're looking for something to do, of course :p
23:18<tmk>and you're using the GET_TIMING ioctl right
23:19<tmk>how do you want to specify the fb?
23:19<tmk>i think it'd be best for myth to query the card somehow for which fb it's registered as
23:19<tmk>and then use the fb for the osd
23:19<tmk>don't you
23:20<Chutt>i dunno
23:20<Chutt>however you want me to do it, i don't really mind
23:20<tmk>ok, i think that makes more sense.. keep functionality separate
23:20<Chutt>all i need is a way to get a buffer to write to
23:20<mikegrb>vektor: I'm walking someone through install of tvtime...... on an xbox :) I don't know what his motivation is but oh well
23:23<vektor>mikegrb: Join #tvtime.
23:24<Chutt>man, i've broken lots of stuff tonight
23:26<tmk>chutt: the 'indent' program is great
23:27<tmk>cat ivtv-fb.c | indent -kr > new-ivtv-fb.c
23:27<tmk>and it's readable!
23:31<moegreen>Chutt: do those confirmation numbers for the -dev list work? I've never actually had it work for me, it always says it is invalid
23:32<Chutt>they should work
23:34<Chutt>i don't think snow-man is around
23:34<Chutt>else i'd ask him
23:34<Chutt>moegreen, what url are you going to to sign up?
23:35-!-mecraw_ [] has joined #mythtv
23:35-!-mecraw_ [] has quit [Client Quit]
23:35<moegreen>Chutt: well, I got the email asking for confirmation that my account is valid. So I try clicking on the url (of course I've deleted the email), so I just go to the login url that it gives me and use my password
23:35<Chutt>what host are you using, though?
23:36<Chutt> or
23:36<moegreen>oh, umm...
23:36<Chutt>ok, since isn't hosting the lists anymore
23:37<Chutt>anyway, i'll see if snowman can look at it when he's around tomorrow
23:37<moegreen>ok, whenever - not a big deal, just kinda strange that I can never get it to work
23:38-!-dopez [] has quit ["..."]
23:38<Chutt>i can't do much on my own, he runs all that stuff =)
23:40<moegreen>heh, I got booted last time because the person running my mail server had subscribed to a spam blocker service that crashed so all my messages were rejected.
23:40<moegreen>having the two backup mx's doesn't matter much then
23:40<Chutt>no, no it doesn't =)
23:46<tmk>ok chutt, so here's how i see this working
23:46<tmk>you request framebuffer_id from ivtv via an ioctl
23:46<tmk>it gives you an int
23:46<tmk>and then you do all ioctls to /dev/fb<id>
23:46<tmk>it'll return -1 if no framebuffer's installed
23:47<kja>Chutt, while we are on the list subject, would it be possible to do subject substitution on something like '([Rr][Ee][:])?([\[mythtv\]])?([Rr][Ee][:])\.*' ?
23:47<kja>I can see that the linux-dvb list does this, and it looks funcky
23:48<hadees>anyone in here use gentoo and mythtv?
23:48<hadees>i meant with
23:49-!-mecraw [] has quit [Read error: 113 (No route to host)]
23:49<mikegrb>hadees: what is your problem? I've used the two though I didn't use the ebuilds
23:54-!-Captain_Murdoch_ [] has quit []
23:56<Chutt>tmk, so how does the framebuffer stuff work? :p
23:56<hadees>well the question is about ebuilds
23:57<hadees>specificly lirc, how do you configure it because ebuilds don't really stop do they?
23:57<Chutt>tmk, just give me example code of how to get a pointer to a buffer to draw to, and i'll be happy =)
23:57<tmk>just an ioctl
23:57<tmk>i'll get to that in a sec
23:57<tmk>it's not written yet ;)
23:58<mikegrb>hadees: generally they install default config files for you that you have to edit afterwords as for compile time options, which I think you are more interested in the USE enviroment variable is used some lemme look at some stuff real quick... see if I can help you out a bit
---Logclosed Fri Sep 12 00:00:01 2003