Back to Home / #mythtv / 2003 / 07 / Prev Day | Next Day
#mythtv IRC Logs for 2003-07-09

00:00<Captain_Murdoch_>oh, since I just called dohandledeleterecording from the event thread?
00:01<Captain_Murdoch_>I'll just wait and read the diff. :)
00:11<Captain_Murdoch_>gotta drop offline for a few minutes, will be back.
00:11<-- Captain_Murdoch_has quit ()
00:11-!-bline [] has joined #mythtv
00:18-!-FreddieD [] has joined #mythtv
00:28<mythtv_>Chutt: have you looked at adding other codecs into myth's player? It appears that many mpeg formats arn't supported correctly, and such as the liba52 for atsc/hdtv
00:28<Chutt>change your nick.
00:29<Chutt>and, what do you mean by 'supported correctly'?
00:30<mythtv_>Chutt: mythtv's player exits with a "avformat error: -1" error when attempting to read an ATSC broadcast stream
00:30-!-mythtv_ is now known as brandon
00:30<Chutt>figure out why
00:31<brandon>should it be supported?
00:31<Chutt>why would it be?
00:31<Chutt>most likely, the liba52 stuff wasn't compiled in
00:31<Chutt>what's that have to do with anything?
00:32<brandon>it appears it should be compiled in, but mythtv isn't handliong it correctly
00:33<Chutt>just because it's present in the source means absolutely nothing about it being compiled properly
00:34<brandon>so has anyone attempted to get it working?
00:34<Chutt>i doubt it
00:35-!-yourface [[pA]] has joined #mythtv
00:35<Chutt>provide me with a short clip, and if i'm bored, i'll see what i can do
00:35<brandon>5 second=20 megs, want it?
00:36<Chutt>put it on a website somewhere
00:36<yourface>When I go to play an in progress recording it randomly stop playing it. It's done this since 0.9. Any suggestions?
00:36-!-Captain_Murdoch_ [] has joined #mythtv
00:36<Chutt>yourface, figure out why and send in a patch
00:38<Chutt>couple people have reported the same thing, but it doesn't happen to me
00:39<yourface>I don't get an error, and I don't know if I should try sending in debug info?
00:39<Chutt>debug info wouldn't help at all
00:39<Chutt>you need to figure out why it thinks it's at the end of the fiel
00:40<brandon>working on an upload..
00:40<mikegrb>Chutt: It doesn't happen to me but from descriptions it seems like the frontend is playing faster or skips some frames or some such and ends up reaching the end
00:40<Chutt>mikegrb, wow, really? :p
00:40<Chutt>it's not playing faster, it's prebuffering
00:40<mikegrb>that makes more sense
00:40<Chutt>but, for some reason, it thinks it's at the eof
00:40<Chutt>if i can't reproduce it, i can't fix it
00:41<Chutt>so someone that it happens to will have to fix it
00:41<mikegrb>of course
00:41<ahbritto>yourface: Do you have automatic commerical skipping on?
00:41<brandon> (20% uploaded)
00:42<Chutt>brandon, just lemme know when it's done
00:42<yourface>Which lib source file contains frontend playback for watching an in progress recording?
00:42<mikegrb>oh and just so it doesn't seem otherwise I was agreening with you, not being sarcastic or some such ;-)
00:43<Chutt>yourface, libs/libmythtv/ -- RingBuffer.cpp, NuppelVideoPlayer.cpp, and either nuppeldecoder.cpp or avformatdecoder.cpp depending on if you're using software or hardware encoding
00:43<Captain_Murdoch_>is this the problem where you're watching a recording that's currently in progress and it keeps thinking it's at the end every so often so you have to restart?
00:43<Chutt>captain_murdoch, ayup
00:43<Chutt>kicks you out
00:43<Captain_Murdoch_>I've experienced that but haven't tried to track it down.
00:43<Chutt>it's probably your code, too =)
00:43<Captain_Murdoch_>nah, I don't turn on auto-commercial-skip. :)
00:44-!-Ripp [] has quit [Read error: 110 (Connection timed out)]
00:44<Chutt>ok, so it's probably the prebuffer
00:44<Captain_Murdoch_>I think it's not getting the updated file length properly from the encoder.
00:44<Captain_Murdoch_>doesn't it check the length or does it just read until it can't read any more?
00:45<-- yourface([pA] has left #mythtv
00:46<Captain_Murdoch_>and it only re-queries the length if it needs a keyframe position from the encoder right?
00:47<Chutt>not sure
00:49<Captain_Murdoch_>looking at the code now to see.
00:53<-- tmkhas quit ()
00:54<Captain_Murdoch_>looks like it only calls parent->SetEOF when it can't read from the ringbuffer. only calls fillpositionmap when it needs to fastforward. looks pre/ring-buffer related so he can start checking there. :)
00:54<Captain_Murdoch_>I rarely use LiveTV. :)
00:55<Chutt>this is watching an in-progress recording
00:55<Captain_Murdoch_>same to me. :) I always wait until it's done and commercial detection is complete.
00:56<Captain_Murdoch_>but yeah, I said that wrong.
00:57<Chutt>ah, good
00:57<Chutt>colo box is back
01:03<brandon>Chutt: done
01:16<-- Captain_Murdoch_has quit ()
01:19<brandon>Chutt: I have to go for the evening. If you need anything regarding the clip, I'll check back in tomorrow. The clip though is taken from a PBS ATSC stream.
01:19<brandon>Chutt: have a good evening.
01:20<brandon>.. The clip works xine too. fyi
01:20<brandon>works in xine that is
01:21-!-brandon [] has quit ["night"]
01:35-!-bbeattie [] has quit [Remote closed the connection]
01:43-!-dub20 is now known as dub20|afk
01:43-!-techniq [] has quit [Read error: 104 (Connection reset by peer)]
01:53<bline>I'm going to have to learn the mpeg format to fix these broken mpegs output from my pvr
01:53-!-FreddieD [] has quit [Read error: 110 (Connection timed out)]
02:56<Chutt>that wasn't too bad
02:57<Chutt>Input #0, mpegts, from '/mnt/store/clip.ts':
02:58<Chutt> Stream #0.0[0x21]: Video: mpegvideo, 704x480, 29.97 fps, 15000 kb/s
02:58<Chutt> Stream #0.1[0x24]: Audio: ac3, 48000 Hz, stereo, 192 kb/s
02:59<bline>Chutt: do you have any shows recorded with your pvr?
02:59<bline>Could you do me a favor and run mplayer -identify on one?
02:59-!-techniq [] has joined #mythtv
03:00<techniq>i am getting an error "QSqlDatabase warning: QMYSQL3 driver not loaded", do i need to just install the mysql-devel rpm and recompile qt with -qt-sql-mysql on configure?
03:00<Chutt>techniq, you just need to install the qt-mysql rpm
03:01<Chutt>bline, what info are you looking for?
03:01<bline>the length it thinks it is
03:01<techniq>Chutt: I thought i had it on (i looked earlier, let me look again)
03:01<bline>yeah, it shouldn't be printed for VDR streams
03:01<bline>err VBR
03:02<Chutt>hour long show
03:02<techniq>Chutt: I think i know why, i'm using my kdecvs qt libs to compile against
03:02<Chutt>then install the proper qt-mysql bit for those.
03:02<bline>and it thinks it's 28 minutes
03:02<bline>I bet the bitrate listed is the peak
03:02<Chutt>bline, because it's treating the max bitrate as the average bitrate
03:03<techniq>Chutt: If i still want to use them (that way i don't have to change the source for the suse problems)
03:03<bline>Chutt: yeah, I dug in the code a bit and it only outputs the length if it thinks it's not VBR
03:04<Chutt>i should probably generate a patch and send it to the ffmpeg guys
03:05<bline>what are you fixing?
03:05<Chutt>made the mpeg-ts demuxer more robust to crap in the stream and added ac3 support to it
03:06<Chutt>fixes that guy's hdtv clip
03:06<Chutt>well, it's sdtv
03:06<Chutt>not hdtv, but, whatever
03:06<Chutt>480p w/ ac3
03:06<bline>ffmpeg does strange things with the output mpeg from the pvr heh
03:07<Chutt>yeah, the mpeg files have problems :(
03:07<Chutt>ie, that audio skipping issue
03:07<bline>from what I can see, things think they are CBR
03:08<bline>but there are other issues too, changing to cbr recording doesn't solve the problems
03:09<bline>I wonder if this is a driver issue
03:10<bline>initialization perhaps, I wish I had windows to test on with the windows driver
03:10<bline>see if it's mpeg is as busted
03:15<ahbritto>bline: careful what you wish for: Windows 98 - $93 at
03:16<Chutt> Stream #0.0[0xe81]: Video: mpegvideo, 1920x1088, 29.97 fps, 18030 kb/s
03:16<Chutt> Stream #0.1[0xe84]: Audio: ac3, 48000 Hz, stereo, 384 kb/s
03:17<Chutt>(from mythtv, of course)
03:21-!-FreddieD [] has joined #mythtv
03:50-!-choenig [] has joined #mythtv
03:52-!-dub20|afk is now known as dub20
03:58-!-froggy19 [] has joined #mythtv
03:59-!-FryGuy [] has quit [Read error: 104 (Connection reset by peer)]
03:59-!-FreddieD [] has quit [Read error: 60 (Operation timed out)]
03:59-!-froggy19 is now known as FryGuy
04:38-!-choenig [] has quit ["wuuuusch..."]
04:43-!-vektor [] has quit [Read error: 110 (Connection timed out)]
05:11-!-vektor [~bbiggs@nat-pc.cs.Dal.Ca] has joined #mythtv
05:21-!-FreddieD [] has joined #mythtv
05:58-!-FreddieD [] has quit [Read error: 110 (Connection timed out)]
06:11-!-zaheer [] has joined #mythtv
06:39-!-bline [] has quit ["Client exiting"]
07:09-!-jkolb [] has joined #mythtv
07:11-!-FryGuy [] has quit [Read error: 54 (Connection reset by peer)]
07:12-!-bline [] has joined #mythtv
07:27-!-FreddieD [] has joined #mythtv
07:40<-- jkolbhas quit ()
07:49-!-FreddieD [] has quit [Read error: 60 (Operation timed out)]
07:49-!-jkolb [] has joined #mythtv
08:50-!-rizzi [] has quit [Read error: 60 (Operation timed out)]
08:50-!-jkolb [] has quit [Read error: 104 (Connection reset by peer)]
09:09-!-jkolb [] has joined #mythtv
09:15-!-just1nux [] has joined #mythtv
09:22-!-stenis [] has joined #mythtv
09:27-!-creepy [~lightning@] has joined #mythtv
09:27-!-FreddieD [] has joined #mythtv
09:31<stenis>Hi, Is Mythtv is gentoo now ?
09:31<stenis>Hi, Is Mythtv in gentoo now ?
09:32<stenis>I did a emerge sync but can't find it
09:34<thor>I know nothing about Gentoo, just saw that on the mailing list
09:34<stenis>Yes I have read that but if i do emerge xmltv it can't find it
09:34<-- creepy(~lightning@ has left #mythtv ("gone..")
09:35<thor>sorry ... have never used Gentoo (and not really clear on what an emerge is). Maybe try the mailing list?
09:36<stenis>haa I found it now in media-tv
09:46<bline>thor: that code is discusting
09:47<thor>the Perl stuff ?
09:47<bline>(the perl code you referenced from the list)
09:47<thor>discusting as in unreadable, or disgusting as in really crappy way to do things
09:49<bline>found it with google heh
09:52<bline>it doesn't get the size, only the id3 tags
09:54<bline>actually, hold on that.
09:54<bline>It does get the size, it uses MP3::Info for that
09:56<thor>is MP3::Info a standard part of Perl?
09:56-!-choenig [] has joined #mythtv
09:56<bline>no, you have to install it
09:57<thor>if it's not too much to ask, can you tell me how MP3::Info calculates the playing time ?
09:57<bline>$i->{SIZE} = $vbr && $vbr->{bytes} ? $vbr->{bytes} : $h->{size}
09:58<thor>I have no idea what that means
09:58<thor>what's the && and ? operators do in Perl ?
09:58<mdz>thor: same as C, short-circuited
09:58<mdz>that means if $vbr and $vbr->{bytes} are nonzero, the result is $vbr->{bytes}, otherwise $h->{size}
09:59<thor>bline, is "size" just file size, or do they pull off header/etc ?
09:59<mdz>thor: can't you tell from the descriptive variable name?
10:00<bline>one sec
10:00<bline>it's header info
10:01<bline>$h is a hash reference to info from the header
10:01<thor>they ask the header what the size of the music data is?
10:01<bline>bytes it from the header as well
10:02<bline>heh, nice and confusing
10:02<thor>and no comments, I'll bet
10:02<bline>no, it's the size. They pass in $bytes to _get_header() and return it as part of a hash
10:03<bline>no, no comments
10:03<thor>so if it's not vbr, sounds like they're taking a short cut
10:04<thor>damn ... I'm going to have to actually figure out what the mp3 file format contains to do this properly
10:05<bline>$byte comes from read $fh, $byte, 4; after they seek to the start of the file
10:06<thor>say again
10:06<thor>file header .... byte .... 4?
10:07<bline>seek $fh, 0, 0; read $fh, $byte, 4;
10:07<thor>what's that 4 mean ?
10:07<bline>just like C
10:07<bline>read 4 bytes into $byte from $fh
10:07<bline>$fh is like a FILE *fh
10:09<bline>I'm sorry, this shit is confusing
10:09<thor>I do appreciate it
10:09<bline>after they return it, from _get_header() they reassign it to $eof - $off;
10:09<bline>so it is the file size - the header
10:09<thor>plus 4 bytes
10:09<thor>- 4 bytes
10:10<thor>one last question ... where/how do they set vbr ?
10:11<bline>from the header info
10:11<thor>ah, mp3 header tells them if it's vbr
10:11<thor>that's probably the speed up
10:12<bline>no problem
10:13<thor>now to go dig around in the mad library
10:13<Chutt>hey guys =)
10:13<bline>a string Xing in the header says if it's vbr, don't ask me what the means
10:13<bline>hey Chutt
10:13<thor>Chutt, nice HDTV update
10:13<Chutt>xing was the company that did the first vbr stuff
10:14<bline>that makes sence
10:14<thor>is it a logical assumption that if it doesn't say xing, it ain't vbr
10:15<thor>s/logical/usefully practical
10:15<Chutt>not always, unfortunately
10:15<Chutt>that'll work most of the time
10:16<thor>user option: scan quickly *OR* scan accurately
10:16<mdz>mp3 wasn't vbr initially
10:16<Chutt>how slow is the current stuff?
10:16<thor>I have no mp3's
10:16<thor>almost none
10:16<mdz>user option: mp3 or vorbis
10:16<Chutt>pretty much
10:17<thor>but post says 45 seconds with this Perl thing vs. 2 hours with Myth
10:17<Chutt>thor, they can fix it if they want it faster
10:17<Chutt>really =)
10:17<mdz>or they can buy some CDs
10:17<thor>yeah, I know
10:18<Chutt>i guarantee that the perl thing gets the wrong length at least a percentage of the time
10:18<mdz>instead of downloading all their music from peer to peer networks
10:18<thor>but, I need to go in there to fix the hang ... and while I'm in there, I thought, should fix scanning time if it's doable (same code)
10:19<Chutt>are there qt compile options for xft?
10:19<bline>I would have done this module in C
10:19<Chutt>and xrender?
10:19<Chutt>what module?
10:19<thor>default on
10:19<bline>this mp3 module
10:19<thor>I think
10:20<Chutt>bline, it's using libmad for everything, it's practically in c :p
10:20<bline>MP3::Info, it would have been about the same length in C
10:20<bline>it's not using any C
10:21<thor>bline, it doesn't link (or whatever the Perl term is) to libmad?
10:22<Chutt>what's the difference between a c function and a c++ function?
10:22<thor>is this a test?
10:22<bline>thor: nope, there is no C in it, it's pure perl
10:22<mdz>Chutt: calling convention
10:22<Chutt>that's it
10:23<bline>initialization in the declairation?
10:23<mdz>name mangling
10:23<mdz>and hence overloading
10:23<Chutt>decoding 1080i takes 90% cpu on here
10:23<bline>(i've only seen that done once)
10:23<Chutt>xine does it with slightly lower, i wonder if that's libmpeg2 vs ffmpeg
10:24<thor>Chutt, someone send you an HDTV toy?
10:24<Chutt>just playing with a demo track
10:36<poptix>MS sent me some discovery channel movie that's in hdtv format
10:36<poptix>haven't even looked at it yet
10:38<Chutt>probably their wmv9 stuff
10:41<vektor>I have a copy of T2 on DVD which came with a 1080i wmv9 version.
10:41<vektor>sucks, mpeg2-ts would have been so much easier for me.
10:42<jkolb>Chutt: The guy I had testing the mythweb stuff reported that the latest patch worked for him, too. (I don't think you ever downloaded the one that wouldn't put an img tag in if the file didn't exist)
10:42<Chutt>naw, i didn't
10:42<Chutt>sorry =)
10:43<mdz>vektor: wmv? really?
10:44<Chutt>mdz, it's their 'we can fit hdtv in the same space as a regular dvd, honest!' stuff
10:44<vektor>yeah, that's the reason they use wmv.
10:44-!-stenis [] has quit ["Client exiting"]
10:45-!-FreddieD [] has quit [Read error: 60 (Operation timed out)]
10:45<vektor>ah well, it was cheaper and better quality than the T2: Ultimate DVD set.
10:45<vektor>at least for the SDTV MPEG2 version.
10:45<mdz>I wish mp3 would just die
10:46<mdz>"There's also a newer incompatible format from Fraunhofer..."
10:46<vektor>You're talking about VBR problems or something?
10:46<thor>good old Fraunhofer
10:46<vektor>What's wrong with MP3?
10:47<thor>patents? badly thought out header information? varying tagging systems?
10:48<vektor>it's no worse than MPEG2 or MPEG4, the header information is well thought out, and tagging systems don't belong in a media stream.
10:48<vektor>use a proper external metadata format.
10:48<thor>tell me the playing time of an mp3 file without actually playing it and I'll be your friend ...
10:49<vektor>it's a streaming format
10:49<vektor>you have metadata formats to tell you that
10:49<vektor>just like on DVD all that information is in the IFO files
10:49<vektor>you can't tell the length of an MPEG2 video stream by looking at headers.
10:49<vektor>it doesn't make sense for a streaming format.
10:49<vektor>say you're streaming an mp3 stream from the net, how can it tell you the length? there is none.
10:49<vektor>if it was in the headers, how could you concatenate two streams together?
10:49<vektor>you couldn't
10:50<vektor>indexing information doesn't belong in a streaming format, it wasn't designed for that.
10:51<mdz>vektor: I see no justification for putting up with all of the random incompatible crap just for the benefit of using a proprietary system which offers me nothing over an open system
10:52<vektor>What random incompatible crap? MP3 has proven extremely stable, look how many hardware MP3 players there are, nobody would dare encode something that doesn't work with them.
10:53<vektor>The benefit is that there is a spec, it was tested during its writing, you can implement your own encoder and decoder straight from the spec and know that it will interoperate.
10:53<mdz>vektor: I don't own one, so I don't know whether they play any MP3 you might find, and hardware vendors have done stupider things in the past
10:54-!-Drikus_ [] has joined #mythtv
10:54<vektor>Personally, I see no reason to use a format like ogg, which took forever to come up with a spec, and afaik there are no two cleanroom implementations of it.
10:54<vektor>like, implementations straight from the spec.
10:54<mdz>vektor: the implication being that ogg vorbis doesn't have a spec?
10:54-!-radsaq [] has quit [Read error: 104 (Connection reset by peer)]
10:54-!-radsaq [] has joined #mythtv
10:55<vektor>mdz: They didn't until recently, apparently they do now.
10:55<mdz>I think "I can write an encoder and not have to pay royalties" is a pretty good reason
10:55<thor>vektor, ogg is open, which is pretty much a slam dunk from my perspective
10:56<vektor>Hey, don't get me wrong, I like things that are open too. I am just not happy with how they went about it.
10:56<vektor>I much prefer open standards processes like the IETF.
10:56<mdz>vektor: er...ogg is an i-d
10:56<vektor>There were no drafts of the ogg _spec_, simply the code itself. I don't believe that's enough of a spec or design.
10:56-!-dopez [] has joined #mythtv
10:56<mdz>standards track
10:56<vektor>It is now.
10:57<mdz>but honestly, I'd rather have an open source reference implementation than a spec if I had to choose
10:57<vektor>hey wait
10:57<vektor>do they have an I-D for the media part or just like the mime type etc.
10:57<Chutt>the mad guy wanted to do a vorbis implementation
10:57<Chutt>but couldn't without a spec =)
10:57<vektor>Chutt: exactly.
10:57<vektor>i wanted to do a vorbis implementation.
10:57<vektor>they told me 'just use our source code'
10:58<vektor>and yeah, ogg is their stream format, there is an i-d. i don't think there is for vorbis.
10:58<vektor>there's a docbook spec that is still incomplete.
10:59-!-hfb [] has joined #mythtv
11:00<mdz>like I said, I'd rather have a known working implementation that I could freely use in any project I wanted, rather than a spec
11:01<vektor>i care much more about interoperability.
11:01<thor>I just want to know playing time
11:01<vektor>thor: scan the file, store in some metadata format.
11:01<mdz>no point in solidifying something into a spec when it hasn't demonstrated its correctness in the real world
11:01<vektor>same as with mpeg2 video
11:01<mdz>vektor: that's what he does, and it takes 2 hours
11:01<thor>vektor, that's what we do now, just that it's slow
11:02<vektor>mdz: that's the point of a spec writing process, multiple implementations test it and their knowledge goes back into the spec
11:02<vektor>that way the spec isn't biased towards any one implementation
11:02<vektor>shouldn't take two hours.
11:02<Chutt>it does :p
11:02<bline>write it
11:02<mdz>vektor: or, you implement it, then write a spec, and others implement from the spec and thereby hammer out any problems
11:02<mdz>which sounds like what they are doing
11:02<thor>there are crazy people out there with 10 Gigs of mp3's
11:03<vektor>thor: i have 80G.
11:03<thor>well, then there's the truly insane
11:03<bline>only 2.6G here
11:04<vektor>i bet i can scan mp3s faster.
11:04<thor>I'll take that bet
11:04<thor>he he
11:04<vektor>mdz: ok, so basically, you're saying (what i've been saying) that ogg/vorbis is still way immature.
11:04<vektor>since there is basically like 1 implementation still.
11:05<vektor>and so the problems haven't been hammered out yet.
11:05<vektor>so i, personally, don't trust it compared to mp3.
11:05<vektor>anyway, i'm off to lunch. i think i'll tackle the mp3 problem when i get back. :)
11:06<thor>any and all help appreciated
11:06<Chutt>there are 2 implementations
11:06<Chutt>the integer one for embedded apps
11:09<bline>what's a WIP?
11:10-!-figgy [] has joined #mythtv
11:11<Drikus_>Work in Progress
11:16<mdz>vektor: just because you can't buy them in the store doesn't mean they don't exist
11:16<mdz>vektor: there is the reference implementation
11:16<mdz>vektor: there is a pure java implementation
11:16<mdz>vektor: someone did an ASIP thing according to google
11:17<mdz>and the integer one, though that's by the same people as the reference implementation I think, so maybe it counts for half
11:18-!-schultmc [] has joined #mythtv
11:18<mdz>make that two pure java implementations, jOrbis and freedomaudio
11:19<mdz>I use one of them at work
11:41<bline>there is a perl implementation too heh
11:43<mdz>really? not just an xs module which uses libvorbis?
11:43<mdz>Ogg::Vorbis at least is just a wrapper for the library
11:44-!-poptix [] has quit []
11:44-!-hfb [] has quit []
11:44-!-dopez [] has quit []
11:44<-- Snow-Manhas quit (
11:44-!-dub20 [] has quit []
11:46--> Snow-Man( has joined #mythtv
11:46<bline>ahh, well there is a header parser in pure perl
11:49-!-dub20 [] has joined #mythtv
11:56-!-mecraw [~mecraw@] has joined #mythtv
11:58-!-sfr [] has joined #mythtv
11:59-!-hfb [] has joined #mythtv
12:13* blinegoes to bed
12:13-!-bline [] has quit ["Client exiting"]
12:21<rkulagow_>chutt: are you here?
12:24-!-FreddieD [] has joined #mythtv
12:29-!-dopez [] has joined #mythtv
12:39-!-zaheer [] has quit [Ping timeout: 14400 seconds]
13:02<Chutt>rkulagow, yeah
13:02<Chutt>what's up?
13:02<rkulagow_>just checking: the preroll didn't make it for 0.10, correct?
13:03<Chutt>i think it did
13:03<rkulagow_>(lost access to some of my mail when i turned in my work laptop)
13:14-!-choenig [] has quit [Remote closed the connection]
13:20-!-choenig [] has joined #mythtv
13:21-!-FreddieD [] has quit [Read error: 110 (Connection timed out)]
13:27-!-Chang [] has joined #mythtv
13:33-!-bline [] has joined #mythtv
13:34-!-poptix_ [] has joined #mythtv
13:34-!-poptix_ is now known as poptix
13:35<bline>someone got a sec for a quick qt question?
13:35<bline>how do I get a literal %2 in a string?
13:36<bline>ok, I'll try that
13:36<bline>it's breaking the query in mythvideo
13:36<bline>if the filename has a %20 in it for example
13:36<poptix>man sprintf
13:36-!-jkolb_lurk [] has quit [Read error: 113 (No route to host)]
13:36<bline>poptix: it's a Qt/C++ thingy
13:36<bline>arguments to QString
13:37* blinedoesn't pretend to fully understand it
13:37<poptix>doesn't matter
13:37<poptix>man sprintf has most of the documentation for format strings
13:37<Chutt>he's not using sprintf
13:38<Chutt>it's not a format string
13:38<poptix>what is it being interpreted as?
13:38<jkolb>Why are you using the filename as a format string and not passing it in as an arg?
13:38<bline>the second argument poptix
13:38<Chutt>poptix, replacement
13:38<Chutt>jkolb, if there's replacments further down, then it's seen as the next format string to replace
13:38<bline>jkolb: it is being passed as .arg
13:39<jkolb>That's hideous.
13:39<Chutt>no, it's not
13:39<bline>oh, heh. I can just fix this by making it one replacement
13:40<Chutt>you can do fairly neat things with it
13:40<jkolb>Maybe I'm misunderstanding what you're saying, then. Is there a way to not have it interpreted as a format string?
13:40<just1nux>Chutt: noticed a problem with the delete dialog, it does not wrap or truncate the "subtitle". If a show has a very long title (Amazing Race) then the dialog can end up being wider than the screen is.
13:40<Chutt>it's not being interpreted as a format string
13:41<Chutt>it's being interpreted as a place to insert replacement args
13:41<Chutt>just1nux, hrm, it should be truncating it
13:41<Chutt>i'll look into it
13:42<jkolb>It's not a printf-style format string, but it's still a format string.
13:42<sfr>Chutt also happens when select a show for recording in the epg, which nxtvepg delivers sometimes.
13:43<sfr>aehm, the descriptions which nxtvepg delivers of course
13:45* mikegrblikes hawaiian pizza
13:45<mikegrb>sorry guys, wrong window
13:46<choenig>sfr: I hope, you're not in a hurry with the testing stuff, I just killed my MythBox and now have to reinstall ;)
13:48<sfr>no, how did you kill it; h/w probs?
13:49<choenig>a simple 'rm * -rf' in the wrong window
13:49<jkolb>Yay root!
13:49<sfr>:-) or better :-(
13:50<jkolb>I did something similar once. 'rm -rf /*' instead of 'rm -rf *'
13:50<choenig>hehe :-)
13:50<sfr>well, or as nmap puts it: "warning: you are NOT root"
13:51<bline>ack, % or \% doesn't fix
13:51<choenig>unfortunately I realize things like that exactly in the momant when its to late
13:51<jkolb>When the prompt didn't come back after two or three seconds, I re-read what I had typed and hit ^C. Too late.
13:51<bline>err %%
13:51<bline>this thing expands strings it just expanded
13:52<poptix>that's almost as bad as my 'chown -R apache:apache / www'
13:52<bline>I guess that can be interesting for some things, but not this
13:52<poptix>someone interrupted me, and i put an extra space in there
13:52<sfr>happened to me once with customer production data, luckily they just made backup the day before.
13:53* blinegoes to readup on QString
14:03<Chutt>mdz, yo
14:11<bline>Chutt: do you see anything wrong is replacing %20 in the title for a move to a space? It is a common enough thing
14:11<Chutt>that's fine with me
14:12<bline>ok, I'll submit it along with the query fix
14:18-!-dopez [] has quit [Read error: 104 (Connection reset by peer)]
14:21<mdz>Chutt: yes?
14:21<Chutt>would it be possible for you to pull out a change from cvs and put it in the mythtv packages?
14:23<Chutt>only change since 0.10
14:23<Chutt>i still haven't tested it
14:23<Chutt>but it should work
14:35-!-dopez [] has joined #mythtv
14:36<mdz>Chutt: is this for the new xmltv?
14:39-!-FreddieD [] has joined #mythtv
14:41<mdz>slow cvs
14:43<Chutt>any better?
14:48<mdz>ok, it's in my CVS for the next release
14:50<mdz>I had a couple of other things pending, so maybe I'll do one tonight or tomorrow
14:50<mdz>are a lot of people complaining, or are you anticipating it?
14:53<sfr>Chutt: i see a strange message from mythfrontend when deleting a recording without the actual .nuv file. this is the latest CVS. The message is: 'Received a: 49529 message from the backend. But I don't know what to do with it.' the delete-recordings screen is empty afterwards, other recordings show up only when i exit this screen, and enter it again.
14:53<Chutt>mdz, someone asked
14:53<Chutt>on the users list
14:54<sfr>the db is updated, though.
14:55-!-bline [] has quit ["Client exiting"]
15:01-!-tmk [] has joined #mythtv
15:10-!-Justin__ [] has quit [Read error: 54 (Connection reset by peer)]
15:10-!-Justin__ [] has joined #mythtv
15:23-!-Chang [] has quit ["ChatZilla 0.8.34 [Mozilla rv:1.4b/0]"]
15:35<mdz>it's that time of afternoon again
15:35<mdz>when someone volunteers to change the database around
15:35<thor>and explains how it will make *everything* faster
15:41-!-FreddieD [] has quit [Read error: 110 (Connection timed out)]
15:42<Chutt>i wonder if i should check this code in
15:46<Chutt>70-80% cpu useage for playback of 1080i =)
15:46<Chutt>down from 85-100%
15:47<Chutt>however, i think i broke, oh, transcoding, commercial detection
15:47<Chutt>probably seeking
15:47<Chutt>ah well
15:47<jkolb>Ehn. Nobody uses those anyway.
15:48<thor>yeah, but 1080!
15:48<thor>1080i with 10% less CPU
15:48<thor>woooo hooooo
15:48<Chutt>it helps out mpeg playback
15:48<Chutt>i fixed things so it's 100% direct rendered now, instead of 66%
15:49<-- jkolbhas quit ()
15:49<Chutt>playback of pvr-250 files is like 1-2% now, instead of around 5
15:49<thor>trying to think of something supercillious to say ...
15:51-!-ahbritto [] has quit [Read error: 60 (Operation timed out)]
15:52-!-stenis [] has joined #mythtv
16:01<just1nux>Chutt: if you use the setting to start recordings early, when the recording starts it doesnt appear in yellow until the actual start time of the program. Is this intentional?
16:02<Chutt>no, but it's a side effect of how it decides if it's recording
16:02-!-Justin__ [] has quit [Read error: 54 (Connection reset by peer)]
16:03<just1nux>ok. just seemed odd.
16:03-!-Justin__ [~justin@H239-211.STATE.RESNET.ALBANY.EDU] has joined #mythtv
16:03<Chutt>all it does for that is see if the current time's in between the recording time to decide to color it or not
16:03<just1nux>oh... that aint good.
16:08-!-dopez [] has quit [Remote closed the connection]
16:10-!-ahbritto [] has joined #mythtv
16:16<Chutt>recompiling globalsettings.cpp, whee
16:22-!-Justin__ [] has quit [Read error: 60 (Operation timed out)]
16:32-!-Justin__ [~justin@H239-214.STATE.RESNET.ALBANY.EDU] has joined #mythtv
16:35-!-Justin__ is now known as Justin_
16:59-!-FryGuy [] has joined #mythtv
17:02--> deflux-( has joined #mythtv
17:02<deflux->Hiya all. :)
17:03<deflux->Has anyone else had an issue with the TV-In channels being off by one and the picture being very light in color and grainy?
17:04<Chutt>search the archives
17:04<mikegrb>deflux-: that is then faq
17:04<Chutt>look in the docs
17:04<Chutt>it's covered in both places
17:04<mikegrb>deflux-: er yha docs... under troubleshooting
17:04<deflux->Looking in the docs. :P
17:05<deflux->And I see no entry concerning my issue.
17:06<deflux->And what FAQ? doesn't have a faq.
17:06<Peit|Home>deflux-: it does, one sec
17:06<Chutt>it doesn't
17:06<deflux->The FAQ module isn't active.
17:07<Chutt>it's in the docs
17:07<deflux->Thanks for the help, btw.
17:07<Chutt>and you can search the archives for 'off by one' and i think that'll get it as well
17:07<mikegrb>I mis-spoke when I said FAQ
17:07<Peit|Home>I thought it did
17:07<Peit|Home>i could have sworn
17:07* Peit|Homeis loosing his marbles
17:07<mikegrb>Peit|Home: it used to, it's been rolled into the docs
17:08<deflux->lol, but I look in the docs and no entry under troubleshooting is relevant.
17:08<deflux->And do I search through the ml archives using a mail command, or is there a web interface to the archive for searching?
17:09<Peit|Home> <-- deflu-
17:09<Chutt>there's a link to the searchable archives on the webpage
17:09<deflux->The canada-cable topic mislead me. ;)
17:10-!-FreddieD [] has joined #mythtv
17:10-!-extremis [] has joined #mythtv
17:11<deflux->hrm...I'm going to have to manually look at the tuner on the card to find the proper tuner identification?
17:11<Chutt>it's generally 2 or 19
17:12<deflux-> detected 19
17:12<deflux->I'll try 2.
17:12<Chutt>make sure the tuner module is loading with the proper type, as well as the bttv module
17:13-!-dopez [] has joined #mythtv
17:15<deflux->in /etc/modules.conf I have now options bttv card=63 tuner=2 | options tuner type=2
17:15<deflux->That look solid?
17:15<Chutt>looks reasonable
17:16* deflux-starts up xawtv.
17:17<deflux->Forgot to mute line-in.
17:17<Conaz>deflux-, good luck :)
17:18<Conaz>deflux-: I think we've all done that
17:19<deflux->It looks good.
17:19<deflux->Getting random lines sometimes in the signal tho.
17:19<Chutt>black lines, all the way across the picture?
17:20<deflux->No, it's extremely random, but looks like signal line distortion.
17:20<Chutt>interlacing artifacts?
17:20* deflux-tries different channels.
17:20<deflux->Hm..I'm not sure if they are interlacing artifacts or not.
17:21<extremis>Does anyone use vdr?
17:21<Chutt>there's a whole big vdr mailing list
17:21<deflux->I'm turning off interlacing playback right now.
17:22<deflux->Appears to have been interlacing artifacts.
17:22<extremis>Chutt: searching for a noneu site for that :)
17:22<Chutt>i doubt there are any
17:23<deflux->Is there anyway to run mythtv without a window manager and have it properly run mplayer and gain focus, etc?
17:23<deflux->I want to reduce memory usage as much as possible.
17:23<Chutt>run something lightweight
17:23<Chutt>focus is only an issue with mythgame and mythvideo
17:23<Chutt>as they both execute external apps
17:24-!-figgy [] has quit ["Client Exiting"]
17:24<deflux->MythVideo doesn't regain focus.
17:24<Chutt>it tries to
17:24<deflux->Is this an ongoing issue or something I can resolve with a change of window manager or wm config?
17:24<Chutt>it's a window manager issue
17:25<deflux->I wonder if fluxbox will do it properly.
17:25<deflux->aroo, mythtv just froze up.
17:27<deflux->So far, I love it tho. *,*
17:27* deflux-hugs his streamzap usb remote control.
17:42<deflux->Here's an issue.
17:43<deflux->I start trying to watch TV, and it's a black screen, then after awhile, a solid blue.
17:43<Chutt>wrong input on the tuner?
17:43<deflux->I haven't changed anything on the tuner config side.
17:43<deflux->Good news, seems focusing works in fluxbox.
17:43<Chutt>try playing something else that uses xv
17:45<ahbritto>deflux-: try xine or mplayer
17:45<deflux->Well, MythVideo works.
17:45<Chutt>no, try it now
17:45<deflux->with mplayer
17:46<ahbritto>mplayer -vo xv _your_file_
17:46-!-Drikus_ [] has quit ["toedeledoki"]
17:46<deflux->one moment
17:46-!-ahbritto [] has quit ["Client exiting"]
17:47<deflux->mplayer -vo xv worked fine.
17:48<Chutt>then it's most likely on the wrong input
17:49<deflux->But I didn't change the input.
17:49<deflux->If I switch back to gnome, it works.
17:50<deflux->Now all of a sudden, it works.
17:51-!-moegreen [] has quit [Read error: 113 (No route to host)]
17:51<deflux->but fluxbox definately has focusing working properly.
17:52-!-stenis [] has quit ["Client exiting"]
18:01<extremis>how much work is being done to run mythtv without X?
18:03<Chutt>i'm not doing any work on it
18:04<Chutt>last i checked, everthing but video output worked in qt/embedded
18:10<deflux->Looks like I'm completely finished setting it up.
18:10<deflux->Now I gotta customize the redhat startup to properly set things up.
18:11-!-sfr [] has quit ["Client exiting"]
18:21-!-jkolb_lurk [] has joined #mythtv
18:25-!-choenig [] has quit [Remote closed the connection]
18:26<-- deflux-( has left #mythtv
18:28-!-hadees [] has joined #mythtv
18:32--> deflux-( has joined #mythtv
18:32<deflux->What FFT library is MythMusic talking about during a visualization?
18:32<deflux->I tried fftw, but that doesn't satisfy it I guess.
18:45<Chutt>i believe it's fftw
18:47<hadees>does mythmusic make it's own visualizations or does it use some other visualization libary
18:47<Chutt>what do you mean make its own?
18:47-!-mecraw [] has quit ["Trillian ("]
18:48<hadees>I mean was the visualization writen just for mythtv
18:48<Chutt>a couple of them were
18:48<Chutt>others i incorporated from other programs
18:49<hadees>did you do any from xmms?
18:49<Chutt>i don't think so
18:49<Chutt>goom was standalone, as was synaesthesia
18:49<Chutt>oh, the bumpscope plugin was originally an xmms plugin
18:50<hadees>the only reason i ask is there is xmms-winamp plugin, its broken now but i am trying to fix it
18:50<hadees>if i can figure out how to fix it i want to add it into mythmusic
18:51<hadees>xmms-winamp visualization plugin
18:55-!-ahbritto [] has joined #mythtv
18:57<-- hadeeshas quit ()
19:03<-- deflux-has quit (Read error: 110 (Connection timed out))
19:17-!-schultmc [] has quit ["Client exiting"]
19:23-!-moegreen [] has joined #mythtv
19:26-!-hadees [] has joined #mythtv
19:35<Chutt>junk committed to cvs
19:38<tmk>woo! junk
19:38<Chutt>lowers cpu usage for mpeg2 playback
19:38* tmkblinks
19:39-!-radsaq [] has quit []
19:39-!-techniq [] has quit []
19:39-!-tdb30_ [] has quit []
19:39-!-rkulagow_ [] has quit []
19:39-!-rkulagow [] has quit []
19:39-!-term [] has quit []
19:39<tmk>i think
19:39<Chutt>it was low already
19:39<tmk>from what to what
19:39<Chutt>now it's even lower
19:39-!-rkulagow_ [] has joined #mythtv
19:39<Chutt>eh, average before was 5% on full-sized decodes, now it'w 1-2%
19:39<tmk>what was the optimization?
19:39<Chutt>better direct rendering
19:40<Chutt>doing it so hdtv playback was easier
19:40-!-radsaq [] has joined #mythtv
19:40<Chutt>_that_ was 85-100% before, now it's 70-80%
19:40<Chutt>(on this xp 1800+)
19:40<tmk>is that just resolution based
19:40<tmk>or is there more to it
19:40<Chutt>that's most of it
19:40<tmk>hdtv i mean
19:41<Chutt>this clip i've been testing with is 1920x1088
19:41<Chutt>just copying that much data around takes a lot of time
19:41<tmk>yeah i know
19:41<Chutt>direct rendering gets rid of some extra copies
19:41<Chutt>so, helps out
19:41<tmk>i think the people who are worried about piracy
19:41<tmk>should just increase the resolution of their video
19:42-!-jkolb [] has joined #mythtv
19:42<Chutt>jkolb, hey, that stuff doesn't always seem to request new pixmaps
19:42<Chutt>i haven't looked into it much
19:43<tmk>just processing that much data will kill most cpu's
19:43<Chutt>but it only asked for the first half dozen or so
19:43<jkolb>It shouldn't if the pixmap already exists.
19:43<tmk>ie dvb
19:43<Chutt>they don't
19:43<Chutt>i'll look into it once i get back from dinner
19:43<jkolb>Eh? I've had it ask for 45 or so during a single pass.
19:43<Chutt>i was testing what i just checked in
19:43<Chutt>so i deleted all the pixmaps
19:43<Chutt>and mythweb only requested 6 or so
19:44<jkolb>Huh. Lemme test it again here.
19:44<Chutt>all of em generated fine in mythfrontend
19:44<Chutt>tmk, so, going to work on decoding now that the evil whine's gone?
19:44<tmk>that's the idea
19:44<tmk>there are a few other issues i may hit first
19:44<tmk>ioremap thing
19:45<tmk>and the audio's not getting set right for some people
19:45<Chutt>lowering how much memory it snarfs?
19:45<Chutt>gets set right for me, screw them =)
19:45<tmk>yeah.. some people are running out
19:45<Chutt>ah well
19:45<tmk>right now we map the whole pci space
19:45<Chutt>bbiab, dinner
19:45<tmk>64mb that is
19:46<tmk>chutt's a real team plaer
19:50<extremis>mikegrb: are you alive?
19:51-!-term [] has joined #mythtv
19:52-!-rkulagow [] has joined #mythtv
19:52-!-tdb30_ [] has joined #mythtv
19:53<jkolb>Just generated 55 pixmaps in one pass.
19:53-!-hfb [] has quit ["Client exiting"]
19:55-!-hadees [] has quit [Read error: 104 (Connection reset by peer)]
20:02-!-Captain_Murdoch_ [] has joined #mythtv
20:05<tmk>via looks to be releasing specs/drivers for their micro-atx stuff
20:05-!-ahbritto [] has quit ["Client exiting"]
20:08<extremis>jesus, this other homebrew doesn't work either
20:08<extremis>I must be an idiot
20:10-!-ahbritto [] has joined #mythtv
20:21* extremiscrys
20:22<jkolb>What are you trying to do?
20:22-!-Sembiance [] has joined #mythtv
20:23<jkolb>Aside from drive yourself to the brink of insanity, apparently.
20:23<Sembiance>Do I need to re-set up MythTV if I get a new motherboard and the TV Tuner cards have changed PCI slots? (I have 3 PVR 250's) Or does myth only care that /dev/video0 video1 and video2 are there and they are what they are configured as
20:23<Sembiance>Basically, Myth doesn't keep track of PCI slot id's at all does it?
20:27<Captain_Murdoch_>myth doesn't care as long as video 0, 1, and 2 are still the same kind of cards as before.
20:28<extremis>trying to get lirc working with my dishnet 6000 receiver
20:28<extremis>unfortunately the actisys 200l I bought 3 weeks ago didn't work
20:28<extremis>so I was told to make a homebrew
20:28<extremis>I made the blars infra device
20:28<extremis>it didn't work
20:28<extremis>so I made the 'improved_transmitter'
20:29<extremis>it doesn't work either
20:29<extremis>I can see the voltage increase across it, so I know lirc is sending pulses
20:29<extremis>I have roughly -10v potential during idle
20:29<extremis>and it jumps about 500mV across the led's during pulse
20:29<extremis>the non infra-red led doesn't illuminate
20:30<jkolb>Ahh, so driving yourself insane *is* your only goal.
20:30<tmk>is your led installed correctly?
20:30<tmk>if you get em backwards they don't work
20:30<extremis>I've asked the mailing list repeatidly for alternatives but with no response
20:31<Sembiance>captain_murdoch: Thank you for your answer :)
20:31-!-tmk [] has quit ["Client exiting"]
20:31<extremis>tmk: yes and if one is backward then potential doesn't flow
20:31<extremis>I have everything correct
20:31<extremis>the only thing that deviates is that my .5k ohm resister in front of the led is 470ohms
20:35<Captain_Murdoch_>extremis: have you tried jvc_send or do you know if it works for your receiver?
20:35<extremis>I'm using jvc_raw now
20:35<extremis>I hate this
20:35<extremis>I don't know enough about what I'm doing
20:36<Captain_Murdoch_>I haven't tried it. I did get my home-made transmitter to controll my dishnetwork box a little but since I also have cable I just have my myth systems using that.
20:36-!-root [] has joined #mythtv
20:36<extremis>can I buy your homebrew?
20:36<extremis>I've already spent $90 on this so far
20:36<extremis>I need it done
20:37<extremis>I'll send you my actisys
20:37<extremis>it should work fine for cable
20:37<Captain_Murdoch_>:) it's not in the transmitter. dish is very picky that's why there's a jvc_send instead of just a normal lirc config file.
20:37<extremis>well the LED should illuminate right?
20:37<extremis>or are you speaking of the actisys?
20:37<Captain_Murdoch_>mine only worked part of the time and I never spent the time to tweak it since I have cable as well. it would work fine for controlling my VCR but not the dish.
20:37<Captain_Murdoch_>homemade from the diagram
20:38<extremis>but I should see the LED illuminate reguardless right?
20:38<extremis>thats why it is there
20:38<extremis>so you know its tx'ing
20:39<extremis>how much of a voltage increase did you have across the infra-LEDs?
20:39<Captain_Murdoch_>never checked. mine was the simple transmitter, didn't have a led.
20:41<extremis>ok, I just used jvc_send
20:41<extremis>no flash of the led and nothing on the receiver
20:43<Captain_Murdoch_>I never even tried that, I was just curious if you had. I just played with lirc with my dish
20:46<extremis>has anyone made this one:
20:49<Chutt>rkulagow, want me to sync the docs?
20:52-!-phar0e [] has joined #MythTV
20:52<Chutt>captain_murdoch, if you get a chance, would you be able to test if commercial skipping produces the same data with current cvs that it did before my most recent checkin?
20:54<Captain_Murdoch_>I should be able to test in a little while. I'm still remote but can get a current comm list and then re-run mythcommflag and compare the new list.
20:55<jkolb>Chutt: I just generated previews for all of my recordings from mythweb. 54 pixmaps, all there in one pass.
20:55<Chutt>jkolb, ya know what, it's because you're joining against the channel table
20:55<-- Sembiance( has left #mythtv ("free(sembiance);")
20:56<Chutt>when you're making that genpixmap query
20:56<jkolb>If I don't need the info from the channel table, I'll drop that join.
20:56<Chutt>well, you don't
20:56<Captain_Murdoch_>be back in a few.
20:56<-- Captain_Murdoch_has quit ()
20:56<Chutt>you _should_ just be sending back what it sends you :p
20:57<Chutt>but, not joining should be fine
20:57<Chutt>see, i've changed sourceids since then
20:57<Chutt>so all the old chanids are left hanging
20:57<Chutt>no big deal, until it tries to do that join
20:59<jkolb>I see.
20:59<jkolb>I'll rewrite it so I avoid hitting the database at all.
21:02<jkolb>I also have the patch for allowing a timestamp to be sent along with the QUERY_GENPIXMAP call. If there's no timestamp sent, it just uses 64 like it used to. Of course, since nothing sends a timestamp yet, it will always just use 64, so it's kinda pointless for me to give it to you yet.
21:03<Chutt>time in seconds into the program?
21:03<jkolb>I think so.
21:03<Chutt>makes the most sense
21:04<jkolb>Yeah, I send a variable to GetScreenGrab now instead of a constant 64.
21:08<Chutt>mini-itx machine takes soooo long to recompile
21:10-!-term [] has quit [Read error: 110 (Connection timed out)]
21:11<ahbritto>Chutt: Try using distcc?
21:12<jkolb>Which kind of epia is it?
21:13<jkolb>Probably still goes faster than my Celeron 850.
21:14<Chutt>they're slow
21:21-!-Captain_Murdoch_ [] has joined #mythtv
21:21-!-term [] has joined #mythtv
21:22-!-tdb30_ [] has quit [Remote closed the connection]
21:22<jkolb>Would you much mind if I had static strings for the unnecessary parts?
21:23<jkolb>For instance, I received a null string for all the rankings. So I can either just send a space, or I can test for null.
21:24<Chutt>doesn't really matter, i think most of those are ignored
21:24<extremis>Does anyone have a homebrew to sell?
21:25<jkolb>So I can just send nothing for those
21:25<Chutt>need to send at least a space
21:26<Chutt>but, yeah, i think all it really needs is the path
21:26<Chutt>but i'm not sure on that =)
21:27<jkolb>I'll just keep taking things out until it stops generating previews.
21:27<Chutt>that works
21:36-!-Brandon [] has joined #mythtv
21:36<Chutt>brandon, your clip plays in current cvs.
21:36<Chutt>took all of a half hour, mainly because it's got junk in the stream that the mpeg-ts demuxer had to be made to handle
21:37<Captain_Murdoch_>Just ran Myth local on my laptop connected through a ssh tunnel over a 46K modem session to my backend and database at home
21:38<Brandon>Chutt: amazing. :-) I'm downloading it now.
21:38<Captain_Murdoch_>kinda slow updating. :)
21:39<Brandon>Chutt: Thanks for fixing that part.
21:40<jkolb>Chutt: It seems to require chanid, filename, starttime, endtime, and hostname.
21:42<jkolb>Chutt: Look in your backend log. Did you get a lot of 'Unable to connect to database' errors?
21:43<jkolb>Maybe the genpixmap hanlder isn't closing a connection right away?
21:45<jkolb>I have 102 mysqld processes running. I think it decided not to spawn more or something.
21:45<Captain_Murdoch_>Chutt: commercial detection still appears to detect the same exact frame numbers on the one show I tested.
21:57-!-extremis [] has quit ["changing servers"]
21:57<Captain_Murdoch_>the MaxEpisodes limit seems to be working fine. auto-deleted some of my wife's shows today it looks like according to the logs. :)
21:57<Captain_Murdoch_>need to install current cvs though. this is code from 2 days ago.
21:58<-- Captain_Murdoch_has quit ()
22:00<jkolb>Chutt: should work out.
22:02<Chutt>jkolb, i'll try it
22:02<Chutt>but, not right now
22:04<term>whee, downloading channel info
22:04<Chutt>jkolb, i didn't see any such errors before, btw
22:05<jkolb>Could you check to see how many mysql processes you have after generating a lot of pixmaps?
22:05<Chutt>when i test the next patch, sure
22:06<jkolb>It still has 60 connections open.
22:11-!-tdb30_ [] has joined #mythtv
22:11<tdb30_>can someone help me out? I'm getting choppy playback and I'm wondering what optimizations I can do?
22:13<ahbritto>tdb30_: Do you have the basics? Xvideo, dma on hard drives?
22:14<tdb30_>dma yes
22:14<tdb30_>but not sure of Xvideo
22:14<ahbritto>So you checked your hdparam -d /dev/hdX after boot?
22:14<tdb30_>yup and just now to be sure
22:14<tdb30_>it shows dma enabled
22:14<ahbritto>you have mplayer?
22:14<ahbritto>does that work smoothly?
22:15<tdb30_>err hold on.
22:15<ahbritto>mplayer -vo xv FILE
22:15<tdb30_>I had mplayer I dont' think its currently installed
22:16<tdb30_>I am also only getting about 31M transfer on the drive according to hdparm is that too slow? Also I'm running an athalon 1.1 incase that maters
22:18<jkolb>Chutt: MythContext::OpenDatabase doesn't check to see if the database that's passed in is already open. Might that be causing this?
22:18<-- Snow-Manhas quit (Read error: 113 (No route to host))
22:19<ahbritto>tdb30_: Your drive sounds fine. Try play back with something else before blaming myth. Install mplayer or try xine.
22:19<jkolb>Wait, nevermind. That db connection was just created. I don't see a close() call made on it before the removeDatabase() call.
22:25<Chutt>doesn't need one
22:25<Chutt>i'm pretty sure removedatabase implies close
22:25-!-phar0e [] has quit ["Client exiting"]
22:25<jkolb>"Warning: There should be no open queries on the database connection when this function is called, otherwise a resource leak will occur."
22:26<jkolb>Not sure if that means the connection has to be closed, or what, but that's from the documentation for QSqlDatabase::removeDatabase
22:33<jkolb>Back to lurking. Night.
22:33<-- jkolbhas quit ()
22:34-!-dopez [] has quit ["gn"]
22:36-!-tdb30_ [] has quit ["using sirc version 2.211+KSIRC/1.2.4"]
23:28-!-FreddieD [] has quit [Read error: 110 (Connection timed out)]
23:45-!-tmk [] has joined #mythtv
23:59-!-tdb30_ [] has joined #mythtv
23:59<tdb30_>what does the error Can't open file "recordedmarkup.MYI mean?