Back to Home / #mythtv / 2003 / 04 / Prev Day | Next Day
#mythtv IRC Logs for 2003-04-13

00:15<orangey>_shad: the first two were OK, yeah..
00:15<orangey>not like the set from your myth directory
00:15<_shad>those were with new camera
00:16<_shad>those were old camera
00:16<_shad>they were asstastic
00:16<orangey>hahahaha. nice : )
00:16<orangey>Alright, time for bed.. nightie : )
00:16<_shad>night :)
00:16-!-orangey [] has quit ["Client Exiting"]
00:21-!-bhills [] has joined #mythtv
00:48<Chutt>mdz, around?
00:52<_shad>Chutt: You see the pics of my newest setup?
00:54<Chutt>got a couple machines there
00:55<_shad>yep :)
00:55<_shad>Hopefully all will run mythtv
01:05-!-NonToxic is now known as zZzToxic
01:17<mdz>Chutt: yep
01:24-!-r1ckt3r [] has joined #mythtv
01:24<Chutt>mdz, was wonder how hard you thought it'd be to make the TunerCardInput setting in videosource.h fill itself automatically when the videodevice setting was changed
01:24<Chutt>right now it's a static list, and that doesn't really work =)
01:31<Chutt>the ivtv driver uses stuff like 'Tuner 0' instead of 'Television', and 'Composite 1' instead of 'Composite1'
01:40<rkulagow>moegreen: are you here?
01:53-!-hfb [] has joined #mythtv
02:11-!-_shad [] has quit []
02:11-!-Tack [] has quit []
02:11-!-mdz [] has quit []
02:12<-- r1ckt3r( has left #mythtv
02:20-!-Tack [] has joined #mythtv
02:20-!-_shad [] has joined #mythtv
02:20-!-mdz [] has joined #mythtv
02:28<mdz>Chutt: sorry, got distracted
02:28<mdz>Chutt: I don't imagine it'd be too hard
02:28<mdz>always meant to do it
02:30-!-TheAsp [] has quit [Remote closed the connection]
02:35-!-Chang [] has joined #mythtv
02:39-!-PeteCool [] has quit ["I like core dumps"]
02:45-!-_shad [] has quit []
02:45-!-mdz [] has quit []
02:45-!-Tack [] has quit []
02:45-!-Tack [] has joined #mythtv
02:45-!-_shad [] has joined #mythtv
02:45-!-mdz [] has joined #mythtv
02:48-!-Chang [] has quit [Read error: 104 (Connection reset by peer)]
02:52<mdz>Chutt: the code is already in videosource.cpp; just need to factor it out into a common function
03:08-!-Morph [] has joined #mythtv
03:17<mdz>Chutt: do I have to worry about any of this v4l2 stuff?
03:22<mdz>Chutt: do you plan to make mythtv able build without v4l2 kernel headers again?
03:23<mdz>ok, I've got it written, don't have time to test it tonight. I'll check it out tomorrow and commit it
03:31<mdz>people need to stop answering merle reine's questions on mythtv-dev
03:31<mdz>and he needs to go to mythtv-users
03:51-!-choenig [] has joined #mythtv
04:15<poptix>mdz: unsub him, heh.
04:22-!-choenig [] has quit [Remote closed the connection]
04:24-!-_shad [] has quit []
04:24-!-mdz [] has quit []
04:24-!-Tack [] has quit []
04:26-!-choenig [] has joined #mythtv
04:31-!-Tack [] has joined #mythtv
04:31-!-_shad [] has joined #mythtv
04:31-!-mdz [] has joined #mythtv
05:05-!-hfb [] has quit ["Huh! Wasn't me!"]
05:08-!-bhills [] has quit ["Client Exiting"]
05:25-!-FryGuy [] has quit [Read error: 54 (Connection reset by peer)]
05:26-!-froggy19 [] has joined #mythtv
05:36-!-_shad [] has quit []
05:36-!-mdz [] has quit []
05:36-!-Tack [] has quit []
05:51-!-mdz [] has joined #mythtv
05:51-!-Tack [] has joined #mythtv
06:45-!-_bDove [] has joined #mythtv
06:45<_bDove>anyone alive?
06:57-!-_shad [] has joined #mythtv
08:46-!-_bDove is now known as bDove
08:57-!-Brainstorm [~Brainstor@] has joined #mythtv
08:58<-- Brainstorm(~Brainstor@ has left #mythtv ("Client Exiting")
09:36-!-sc00p [~oldendic@] has joined #mythtv
09:58-!-davehunn [] has joined #mythtv
09:58<davehunn>hi i am having a lot of bttv problems
09:59<davehunn>i have lost all audio from the line out as well as from mythtv
10:02<davehunn>i have no idea why the audio has sudenly stopped working the chip is found during insmod and xawtv reports the sound as being stereo but i can not hear anything
10:03<davehunn>also trying to pass a signal from the line in the the card does not work
10:29-!-TheAsp [] has joined #mythtv
10:29<TheAsp>This app was compiled against libmyth version: 0.9.04112003-1
10:29<TheAsp>but the library is version: 0.9.04122003-1
10:29* TheAspscratches head
10:29<Tack>Recompile the module giving you that message.
10:30<Tack>They're the same version. :)
10:30<Tack>No they're nottttt.
10:30<Tack>0411 vs 0412
10:30<TheAsp>stupid dates
10:30<TheAsp>does it refuse to run?
10:30<TheAsp>or just whine?
10:31<Tack>I think it refuses to run. At least it has for me. Just recompile the module.
10:33<Chutt>sorry, i got tired of stuff that could be fixed by a make distclean :p
10:33<Tack>It's only a big deal if compiling takes hours. :)
10:33<Chutt>and that was a lot easier than figuring out why qmake didn't always generate full dependency information
10:34<Tack>What does myth use Imlib2 for?
10:34<TheAsp>hehe, it needs a "i've recompiled myth 10 times in the last day, shut up" switch
10:36<Chutt>tack, nothing?
10:37<Tack>Chutt: Just wondering, because ./configure reports if Imlib2 is present.
10:37<Chutt>ah, notice the top line of configure
10:37<Chutt>of the output
10:38<Tack>Oh. :)
10:38<TheAsp>hmm... xf fork..
10:40<Chutt>i believe libavcodec can use imlib2 for some simple text rendering
10:41<Chutt>but i don't even have those bits of code in the mythtv source at all
10:43<TheAsp>*sigh* why do i read /.?
10:43<Chutt>there's a parts list on the lirc page
10:44<Chutt>along with really simple schematics
10:44<Tack>Chutt: If I bring Conrad part numbers to Radio Shack will they be able to look them up?
10:44<TheAsp>do I need anything more then ducttape, wires, and the obvious parts? :)
10:45<Chutt>there's a radioshack parts list on the lirc page.
10:45<Tack>There is?
10:45<Chutt>yes, there is
10:45<TheAsp>Did the thread that used to flag comercials after recording get changed to run while recording?
10:46<Chutt>previously recorded shows can be running while something else is recording
10:53<TheAsp>damn, that receiver is simple...
10:55<Tack>Yeah, I was thinking that too.
11:16<Captain_Murdoch>Hey tack, you said something yesterday (I'm reading the scrollback buffer) about wishing cvs had a changelog. Do a google search for "". It generates a normal changelog file from cvs logs.
11:17-!-gora [] has quit ["Client exiting"]
11:17<Tack>Captain_Murdoch: Yeah, I didn't realize (or rather the obvious didn't occur to me) that commit logs were being kept using cvs. :)
11:18<Captain_Murdoch>I find it easier to look through a ChangeLog file sometimes so I use occasionally.
11:18<Captain_Murdoch>course now that I said that someone will probably tell me there's a way to do it through the web and see all logs. :)
11:18<Tack>Looks useful; thanks.
11:19<Captain_Murdoch>theasp, the flag commercials thread runs after a show is finished, but if you have back to back recordings, the first flag commercials thread is running during the 2nd recording.
11:20<TheAsp>Captain_Murdoch: Hmm... it looked like it ran during the recording for me last night
11:20<Captain_Murdoch>latest code is doing much better at flagging them also using blank frame detection. I also have a little code to try to detect places where they fade almost to black then start another commercial.
11:20<Captain_Murdoch>TheAsp: it can't. it's started during by teardownrecorder which called when the recording finishes.
11:21<Captain_Murdoch>my typing is terrible today, need to wake up fully I guess.
11:21<TheAsp>well it ate more cpu then normal, didnt see the thread running after and i could skip comercials in it quickly :P
11:22<TheAsp>not that i dont beleive you :P
11:22<TheAsp>(btw, star trek v sucks.)
11:25-!-Ndiin [] has quit [Read error: 60 (Operation timed out)]
11:25<Captain_Murdoch>if you see a "niced" mythbackend running that probably is the flag commercials thread.
11:26<Captain_Murdoch>it shouldn't take up much cpu though with current cvs.
11:27<TheAsp>yesyes, i know
11:29<Captain_Murdoch>TheAsp: well, the blank frame stuff works in 2 ways. during recording (in the recording thread), blank frames are detected and kept track of. After recording that blank frame list gets saved to the database. Then the flag commercials thread fires off. It actually re-reads the file and redetects blank frames and does some more processing (eventually other detection methods as well) to flag the commercials. It then saves the new
11:29<Captain_Murdoch>d blank frame lists to the database. The player can read either the blank frame list or preferably the commercial break list. If only the blank frame list is available, the player will generate a quick commercial break list from the blank list and will use that to allow you to skip around. This list is never saved, just used by the player since the break list didn't exist.
11:29<Captain_Murdoch>wow, xchat automatically split my post. :)
11:32<Captain_Murdoch>so if you play a recording right after it finishes, you are able to skip because the player generated the break list using the blank list that the recorder saved. This functionality is available whether you have auto flagging turned on or not. the auto-flagging does do some more/better processing though and is more accurate usually.
11:33<TheAsp>well, is it doing new stuff during recording (last week or so)?
11:33<TheAsp>it used to take 30% last night it was taking 50-60%
11:33<Captain_Murdoch>the main mythbackend thread is doing that?
11:35<Captain_Murdoch>I don't believe so. that would have taken my box through the roof on cpu. I normally run at about 85% total usage on my athlon 700 mythtv box with 15% or so free. That is helping keep me conscious of what cpu the flagging stuff is using.
11:37<TheAsp>well it was bringing me through the roof :P
11:37<TheAsp>had other stuff going on
11:37<TheAsp>thats why i noticed
11:43<Captain_Murdoch>my "i don't believe so" was replying to your question, not the 50-60% cpu you mentioned. :) I just reread that and figured some might take it the wrong way.
11:43<Captain_Murdoch>when do you plan on resyncing with cvs?
11:44<TheAsp>whenever something gets committed i want?
11:44<TheAsp>i usually do it once a day
11:48-!-rkulagow_ [] has joined #mythtv
11:48-!-rkulagow [] has quit [Read error: 54 (Connection reset by peer)]
11:48<Captain_Murdoch>so you're running pretty current then? only thing that changed was on the 10th (so I'm running this code on my box) when I changed it to check max 500 pixels instead of 200 if it didn't find a bright pixel. that shouldn't cause that high of a cpu spike though since it short-circuits as soon as it finds a couple bright pixels.
11:51<Captain_Murdoch>I've rewritten the blank frame detection routine to be more accurate though I believe and will be testing that some more before committing to cvs.
11:53<TheAsp> my autodeletion code is going to get nasty in the next few days
11:53<TheAsp>im going to rip off tivo's method
11:53<TheAsp>and add locking to playing/recording so i dont delete shows being watched or recorded
11:55<TheAsp>man, i love when you randomly copy some albums to a dir and it magically makes 700mb
11:55<Captain_Murdoch>I'd like in the future to have the ability to say "only record X instances of this show" so for example I only have to keep around 5 instances of some of my wife's TLC shows. :) also the ability to flag a show for non-deletion (even through playback/delete-box)
11:57-!-wigginjs_ [] has quit ["Client exiting"]
11:57<TheAsp>well, tivo does this (aparently): you can set not to delete, you can set to expire after a certin time, then it goes through and finds the oldest expired show
11:57<TheAsp>currently i have a field with ints 0-100, 100 is never delete, and it picks off the lowest one with the oldest date
11:58<TheAsp>so, you'll be able to say to keep a few days of the shitty wife shows, not by count but by time
11:59<TheAsp>it doesnt work to bad currently, but i dont have any ui to set the stuff nicely
12:01<TheAsp>I'm waiting on mdz to design some new screens for managing schedule'd recordings... I hate doing UI stuff
12:01<Captain_Murdoch>yeah, but that means that when I record my 2 nascar races over the weekend that it could bump out her whole week's worth of "trading spaces" since they're low priority.
12:01<TheAsp>yeah, thats why i want the expire stuff
12:02<TheAsp>though, is that any real loss if trading spaces gets deleted? :)
12:02<Captain_Murdoch>mine doesn't normally fill up anyway, but once auto-delete/expire is working I might start recording more stuff, especially when if we get any sort of intelligent recording like saying "record anything with Clint Eastwood"
12:03<Captain_Murdoch>no real loss, but could cause an argument. :)
12:03<TheAsp>hehe :)
12:03<TheAsp>she can blame me :)
12:03<Captain_Murdoch>nah, I'd just use it to convince her I needed to add another hard drive to the file server to store more stuff.
12:06<Captain_Murdoch>anyway, gotta head out. I'm recompiling my latest test code so I can put it on my production myth box to do some final testing before committing to cvs. It's getting pretty accurate on some of my worst case shows like Law & Order and a couple sci-fi movies I recorded.
12:40-!-PhracturedBlue [] has joined #mythtv
12:40-!-PBlue_away [] has quit [Read error: 111 (Connection refused)]
12:53-!-Ndiin [] has joined #mythtv
13:13-!-bhills [] has joined #mythtv
13:34-!-zZzToxic is now known as NonToxic
14:20-!-paperclip [] has joined #mythtv
14:22-!-TheAsp [] has quit [Remote closed the connection]
14:47-!-PeteCool [] has joined #mythtv
15:07-!-Tack3 [] has quit [Remote closed the connection]
15:18-!-hfb [] has joined #mythtv
15:34-!-poptix [] has quit [Read error: 110 (Connection timed out)]
15:37-!-hfb [] has quit ["Huh! Wasn't me!"]
15:57-!-poptix_ [] has joined #mythtv
16:06<-- bDove( has left #mythtv ("In a world without fences, who needs GATES?")
16:08<PeteCool>moegreen: about the playbackbox: here with SasQuatch, all shows with != 9 episodes show the down arrow
16:08-!-nyquiljer [] has joined #mythtv
16:18<nyquiljer>i just grabbed cvs and make dies with /bin/sh: cd: libavformat: No such file or directory
16:18<nyquiljer>is this to be expected?
16:32-!-mdz [] has quit ["Client Exiting"]
16:46<Captain_Murdoch>did you "make distclean"? it might require that since makefiles need to be regenerated.
16:47<Captain_Murdoch>or if you don't have libs/libavformat, then you need to do a "cvs update -d" I think to get new directories
16:47<nyquiljer>oh, thats the trick
17:07-!-mdz [] has joined #mythtv
17:22<Tack>Chutt: You around?
17:45-!-choenig [] has quit [Remote closed the connection]
17:56<nyquiljer>does rtjpeg or mpeg4 give smaller filesize?
17:56<nyquiljer>with defautlt quality settings
18:00<poptix_>nyquiljer: neither.
18:00-!-poptix_ is now known as poptix
18:07<nyquiljer>poptix: so they are exactly the same?
18:07<poptix>nyquiljer: nope.
18:07<nyquiljer>poptix: um.
18:07<nyquiljer>i guess i'm not following you
18:08<poptix>i don't like people following me
18:09<nyquiljer>how helpful
18:19<Morph>nyquiljer: Id suggest google. its full of answers.
18:28-!-PeteCool [] has quit ["Client exiting"]
18:48-!-nyquiljer [] has quit ["Download Gaim []"]
20:30-!-Timon [] has joined #mythtv
20:32-!-nyquiljer [] has joined #mythtv
20:45<Morph>hmm anyone using cvs mythtv with a pvr250 or 350?
20:47-!-Kuwanger [] has joined #mythtv
20:47<Timon>Isaac is
20:48<Timon>Isaac -- Chutt
20:48<Kuwanger>Hmm..mythbackend just stopped working.
20:49<Kuwanger>Running mythbackend again returns "couldn't open db"
20:50<Morph>Yeah..he's been posting to the list thats working on the driver for that card. He mentioned that he had it working but had to do some DB stuff manually. was just wondering what it was. :)
20:50<Timon>Kuwanger: Is mysql running :-)
20:51<Timon>Kuwanger: Is mysql.txt pointed to the right mysql server?
20:51<Kuwanger>Hmm..restarting mysql seemed to fix the problem.
20:51<Kuwanger>What's mysql.txt?
20:52<Timon>thats the file which stores the settings myth uses to connnect to the db
20:53<Timon>restarting was going to be my next suggestion :-)
20:53<Kuwanger>No idea what broke..
20:55<Timon>I think it was the flux capasiter
20:56<nyquiljer>one point twneytone jigawatts???
20:56-!-nyquiljer [] has quit ["Download Gaim []"]
20:57-!-nyquiljer [] has joined #mythtv
20:58<Kuwanger>Why didn't he just round to 2?
21:21<Chutt>morph, with current cvs, just run setup as normal, but select 'Hardware MPEG encoder' as the cardtype
21:21<Chutt>and 'Tuner 0' as the default input
21:21<Chutt>everything else is like a normal tuner card, i believe
21:22<Tack>Chutt: Any ideas what my sound problem could be? Seems very obscure if nobody else is having that problem.
21:22<Morph>Chutt: ah. excellant.
21:23<Chutt>tack, no idea, sorry
21:24<Chutt>morph, was that you that said it was hanging on live-tv?
21:24<Morph>Chutt: on the ivtv list?
21:24<-- Kuwanger( has left #mythtv
21:24<Morph>yeah. it was. this was having it as a V4L2 type though. lemme see about Hardware MPEG Encoder
21:25<Chutt>if it does that, without any error messages to the console, it's usually something sitting on /dev/dsp
21:25<Chutt>artsd, esd, whatnot
21:27<Tack>Is it normal to take several seconds (5-10 seconds) to enter the Playback section? I only have 5 shows recorded.
21:27-!-wigginjs [] has joined #mythtv
21:28<Morph>hmm anything needed in Input Connections?
21:30<Chutt>tack, if you're running at a larger resolution, yeah
21:30<Chutt>tack, it's scaling all those graphics and stuff
21:31<Chutt>morph, yeah, that all needs to be setup just like a normal card
21:31<Chutt>i'm assuming you've used mythtv before, though =)
21:31<Morph>Okay, im guessing that needs to be done manually since nothing shows up..
21:31<Chutt>are you running _current_ cvs?
21:31<Chutt>it should be probing the inputs
21:32<Chutt>i just committed that late last night
21:32<Morph>Yup. just rechecked it out and rebuilt.
21:32-!-lydickaw [] has joined #mythtv
21:32<Tack>Chutt: Only running at 800x600.
21:32<Chutt>try going into the setup subdir and doing a make clean there
21:32<lydickaw>can anyone else get to ?
21:32<Chutt>tack, that shows up pretty much instantaneously, here
21:32<Tack>Chutt: What's your CPU?
21:33<Chutt>try turning off the transparency effects in setup, though
21:33<Tack>Hmm, I'm not _that_ far behind you at 1400.
21:33<Tack>Chutt: Okay.
21:36<Tack>Seems to help.
21:36<Tack>Starting live tv still takes several seconds. If that's not normal, maybe it's related to my sound problem.
21:36<Timon>Chutt: Am I correct in presuming that if I edit the cardname.xmltv and comment out/remove a chan, that it will eventually fall off the epg in about a week as mythfilldatabase runs and does its mojo?
21:40<Morph>Chutt: no dice. when I go into the Input Connections screen, I get an ioctl error...0x803c7601
21:44<Chutt>morph, sounds like it's not trying to use the v4l2 stuff
21:44<Chutt>timon, you should just remove it from the db, it'll still show up with 'unknown' programs
21:45<Timon>But, when I run mythfilldatabase, won't it get re-added?
21:45<Chutt>tack, it should take a couple seconds to start live tv.. need to give the recorder time to encode some stuff
21:45<Chutt>don't want to be _right_ behind the recorder during playback
21:45<Chutt>timon, not if it's not getting downloaded anymore, no
21:46-!-poptix [] has quit [Read error: 110 (Connection timed out)]
21:50<Morph>Chutt: interesting. the machine should be set to use V4L2
21:52<Tack>Chutt: Yeah, a couple seconds would be okay. 10 seems a lot though?
22:07-!-mdz [] has quit ["Client exiting"]
22:13<nyquiljer>does the commercial mark skipping get done on the nuv level (ie, will mencoder respect them), or at the myth playback level?
22:14<nyquiljer>like, if you've gone into edit and loaded the cutpoints from the commercial skipper
22:14<nyquiljer>at that point is it done in the nuv file, or does myth read those points from somewhere external?
22:15<Morph>ah ha.
22:18<Captain_Murdoch>commerciak breaks and blank frames are marked in the database in the recordedmarkup table.
22:18<nyquiljer>Captain_Murdoch: thats kinda what i figured, but i was hoping :)
22:18<Captain_Murdoch>the cutlist is also stored in the database in the recorded table but I plan on moving it to the recordedmarkup table at some point.
22:20<Captain_Murdoch>need to get a "strip" function working that will take a .nuv file and apply the cutlist and make a new file. :)
22:20<nyquiljer>that would be really handy
22:21<nyquiljer>i dont know anything about this stuff, but just out of curiosity, are the cut points referenced by time, or frames, or what?
22:22-!-poptix_ [] has joined #mythtv
22:22-!-poptix_ is now known as poptix
22:25<Chutt>morph, figure it out?
22:26<Morph>Chutt: yeah. I think so. I think you were right. I dont think it was using V4L2
22:27<Captain_Murdoch>nyquiljer: it's frame based but doing a straight copy of the frame numbers from the recordedmarkup table and trying to use them in mencoder might not work. if a frame is skipped/missing, it is still accounted for in the frame counts used for blank frames, comm break list, and cutlist.
22:27<Morph>rebuilding now.
22:28<Chutt>i had some issues with the videodev2 header
22:28<Chutt>i dunno really what to do to fix things, had to take some includes out of it
22:29-!-lydickaw [] has quit ["Client Exiting"]
22:29<Chutt>i'm just pointing at <linux/videodev.h> in mythtv, which will usually point to the vendor's /usr/include/linux/videodev.h, not the kernel sources
22:29<Timon>Chutt: Where does / what sets the channel when livetv first starts?
22:29<Chutt>and so not be the v4l2 stuff
22:29<Chutt>timon, the db
22:30<Chutt>happens in the constructor of TVRec
22:30<Chutt>tv_rec.cpp, lib/libmythtv/
22:30<Timon>Ok, thanks
22:31<Morph>Chutt: the videodev.h I have has an ifndef in there for including videodev2..
22:35<Chutt>compiling mpegrecorder.cpp in mythtv'll emit a warning if it thinks it doesn't have v4l2
22:35<Chutt>if that helps =)
22:36<Chutt>i'm going to have to futz with things more, though
22:36<Chutt>because apparently mandrake includes an assold version of v4l2
22:36<Chutt>in their kernel
22:36<Chutt>that, obviously, breaks things
23:02<Morph>Chutt: that did it. I now have a list of inputs.
23:09-!-mdz [] has joined #mythtv
23:16-!-froggy19 [] has quit [Read error: 104 (Connection reset by peer)]
23:20-!-froggy19 [] has joined #mythtv
23:32<Morph>rock recording working.
23:33<Morph>whats being used for playback?
23:38<Ndiin>Chutt: hmm, Just managed to get myself that backtrace very similar to the one posted a few days ago (in mythmusic) by next()ing right when starting playback a few times
23:38<Ndiin>Still points to the buffer destruction, though
23:39<Chutt>morph, same stuff it uses for playback of the stuff it records
23:39<Chutt>libavcodec for decoding, custom stuff for playback
23:39<Morph>Chutt: Nod im getting an Illegal Instruction in both cases. recording works great though.
23:39<Chutt>what kind of cpu?
23:39<Morph>VIA C3
23:40<Chutt>turn off deinterlacing in the settings/playback stuff
23:40<Chutt>should fix it up
23:41* Morphchecks
23:41<Chutt>'course, you'll get the standard interlacing artifacts, but =)
23:41<Chutt>i just never really bothered to fix that to fallback to C properly
23:42<Morph>hmm Now its a Floating Point Exception.
23:42-!-froggy19 [] has quit [Read error: 104 (Connection reset by peer)]
23:43<Chutt>i'd like a backtrace, if you have the time
23:44<Morph>how would I grab that?
23:44<Chutt>familiar with gdb?
23:44<Morph>a bit.
23:45<Morph>ah cool.
23:45<Chutt>step-by-step directions on how to recompile for debugging
23:45<Chutt>and what to do in gdb
23:45<Morph>heh. rebuild time again.
23:45<Chutt>yeah, sorry
23:45-!-froggy19 [] has joined #mythtv
23:45<Morph>no prob.
23:46<Morph>btw, mythtv and friends are very impressive.
23:47<mdz>Chutt: processors which don't have the extensions would probably just be too slow with a C version anyway :-)
23:47<Chutt>mdz, that's mainly why i never bothered =)
23:47<Chutt>morph, thanks
23:47<nziarek>moegreen: your new mythweather isn't in CVS is it?
23:48<Chutt>but it'd be nice to automatically turn itself off or something
23:48<moegreen>nziarek: no, i've been dragging my feet. I've just got to finish up the default weather-ui.xml file and a few other misc. items
23:49<nziarek>oh - ok...i thought it was, but everytime I downloaded it, I didn't get the new stuff, and I was wondering what I was doing wrong :)
23:49<Ndiin>Does this make any sense:
23:49<Ndiin>34 delete buffers[i];
23:49<Ndiin>(gdb) p i
23:49<Ndiin>No symbol "i" in current context.
23:49<Chutt>where's that from?
23:49<Ndiin>the mythmusic bt on next()
23:50<Ndiin>line 33
23:50<Ndiin>frame 5, but i'm going through each..
23:50<moegreen>Chutt: approx. what is your cpu usage when watching livetv? I've got 30% free, but it's studdering some
23:50<Chutt>with the pv4?
23:50<Chutt>err, pvr
23:51<Chutt>'bout 85% free
23:51<Chutt>try pausing it a bit
23:51<Chutt>then unpause and see if that helps the studder
23:51<mdz>heh, the mythtv packages are showing up in popularity-contest
23:51<Chutt>i've seen that occasionally
23:51<Chutt>mdz, where?
23:52<Chutt>moegreen, i think my network code isn't _quite_ up to the higher-bitrates
23:52<Ndiin>This is supposedly the thing crashing it, though.. the top three frames are delete[], delete, and free():
23:52<Ndiin>23 delete [] data;
23:52<Ndiin>(gdb) p data
23:52<Ndiin>$2 = (unsigned char *) 0x422bfc58 "\214\xFF\r"
23:52<Chutt>p *data
23:52<Ndiin>from buffer.h:23
23:53<Chutt>so, nothing wrong with that
23:53<Chutt>i dunno
23:53<Chutt>unless it's been deleted twice..
23:53<thor>this in buffer?
23:53<Chutt>mdz, wow, 2
23:53<Ndiin>I thought it might have been the recycler, so I added an if in it, didnt help any
23:54<mdz>Chutt: 2 current, 2 old
23:54<-- mdz( has left #mythtv ("Client exiting")
23:54-!-mdz [] has joined #mythtv
23:54<mdz>damn new xchat broke ^W
23:54<Chutt>you didn't commit your videosource change, did you?
23:54<mdz>anyway it's actually 2 votes and 2 recently upgraded
23:54<mdz>no, I have it sitting in my working directory
23:55<mdz>need to build it on a box with a v4l device to test it
23:55<mdz>it's building now, but it's slow
23:55<Chutt>thanks for doing that =)
23:55<mdz>no problem, one less XXX
23:55<Chutt>i wasn't sure how to make it auto-update when the devicename changed
23:55<mdz>all the settings emit a valueChanged signal
23:55<Chutt>right, but how to tie the two together
23:55<Chutt>just some signal/slot stuff?
23:56<Ndiin>thor or Chutt, any ideas on this bt? its quite odd, basically the same one we saw last week..
23:56<Chutt>you saw my dialog->wizard stuff?
23:56<mdz>I should move that stuff into videosource.cpp one of these days
23:56<Chutt>ndiin, ever heard of valgrind?
23:56<mdz>Chutt: saw the commit, didn't look at the changes
23:56<Ndiin>hmm, ill check it out, see if it gives any more info
23:56<Chutt>ndiin, good memory debugger
23:56<mdz>but it looked fine when I ran setup
23:56<Chutt>might help
23:57<Chutt>since i can't reproduce it at all
23:57<mdz>seemed like a good idea; setup was a little unintuitive with the navigation
23:57<Chutt>mdz, yeah, and this'll allow for help text
23:57<Ndiin>grabbing it now
23:57<thor>Ndiin, the seg fault is so rare that it's almost certainly something very subtle
23:57<Chutt>would it be possible for the big list boxes to be help text-ified?
23:57<Ndiin>thor: I was able to reproduce it three or four times in a row now.. by nexting really quickly when first starting playback
23:57<mdz>been meaning to look into putting in initial help text for the wizard
23:57-!-froggy19 [] has quit [Read error: 104 (Connection reset by peer)]
23:57<Chutt>not sure if the 'wizard' is best for that, but at least a help box would be nice for those
23:57<mdz>so that when any wizard starts, it gives a one-sentence explanation of the navigation
23:58<mdz>big list boxes like the profile selector?
23:58<Chutt>heh, use for the mythpopupbox, there
23:58<Chutt>mdz, yeah
23:58<mdz>or something else?
23:58<Chutt>the profile selector, the tuner card selector
23:58<Chutt>should be fairly obvious, but just trying to be consistant
23:58<thor>Ndiin, sometimes yes ... but other times I keep going till Qt complains it can't give me anymore threads (and still no seg fault)
23:58<mdz>those are kind of a bad hack
23:59<mdz>in a pseudo-elegant sort of way
23:59<mdz>the list box is actually a setting
23:59<mdz>which happened because I wrote it to use as a selector
23:59<Ndiin>thor: well, im going to try it with valgrind, just reading the manual first ;)
23:59<Chutt>thor, hey, just remembered i had intended to add a hostname to the playlist db entry =)
23:59<mdz>but then it turned into an editor
23:59<Chutt>ah, right