Back to Home / #mythtv / 2003 / 10 / Prev Day | Next Day
#mythtv IRC Logs for 2003-10-03

---Logopened Fri Oct 03 00:00:36 2003
00:06-!-DogBoy [] has joined #mythtv
00:08<Chutt>captain_murdoch, are you here?
00:13<Captain_Murdoch>just walked in
00:14<Chutt>someone just posted to the users list about an autoexpire issue
00:14<Chutt>i must've missed it before, but in your analysis from a couple weeks ago, you were wondering about the gContext->dispatch()
00:14<Chutt>not waiting or something?
00:14<Captain_Murdoch>the sendevent vs postevent.
00:14<Chutt>it's not supposed to wait, it's suppose to add the message to the message queue and return
00:15<Captain_Murdoch>the original function was supposed to not wait. the function I called in dispatch now was supposed to wait.
00:15<Captain_Murdoch>that's sendevent vs postevent.
00:15<Captain_Murdoch>forget which is which.
00:16<Captain_Murdoch>that's why I was wondering why it wasn't waiting for the guy a couple weeks ago because everything I read said it should wait until the receiver returns.
00:16<Chutt>oh, i don't think that that's safe, anyway
00:17<Chutt>it looks to me like sendEvent assumes it's being called from the event loop thread
00:17<Captain_Murdoch>I was thinking of reworking it to only delete a max of 1 recording each time the expire code fires off (ie, once every 10 minutes or whatever the user sets).
00:17<Chutt>that'd work
00:18<Captain_Murdoch>be safer also, rather than deleting all shows in 5-10 seconds if there's a problem it will take 3-4 hours. :)
00:19<Chutt>i'd change that dispatchNow to just dispatch
00:19<Chutt>and make it delete only that fast
00:20-!-hEximal [] has quit ["Leaving"]
00:20<Captain_Murdoch>taking a look now. I think I was talking with him about that before I lost net connection and power for a few days (though not on the same days strangely) because of the hurricane.
00:22<Captain_Murdoch>really can just change the inner autoexpire while loop to an if statement. so instead of looping through the list of programs it will only delete the first if need be then sleep again.
00:23<Captain_Murdoch>so "while not enough space free delete one program" becomes "if not enough space free delete one program"
00:23<Chutt>you have time to do that anytime?
00:23<Captain_Murdoch>changing now, easy fix. might change wording on the messages printed also to make more sense with the new method.
00:25<Chutt>cool, thanks
00:27<extremis>Chutt: so I should install a minimul -stable and then apt-get upgrade or, should I point it to the -unstable directories on install?
00:27<Chutt>i've done it both ways
00:27<extremis>good, I like the idea of getting it bleeding edge on install
00:28<extremis>its odd it won't use anything other than ext2 for the fs type
00:28<Chutt>old kernel
00:31<warlord>hmm, I think there is a bug in the recording selection algorithm.. If a program is scheduled to record but the system thinks it's a duplicate, and then (after the program starts) you change the setting to enable recording of duplicates, then the system thinks it wants to record, but the recording never starts (and it's never removed from the "to be recorded" list).
00:32<warlord>So right now, even though it's 12:30am on 10/3, my myth thinks it wants to record soccer at 10:25pm on 10/2
00:32<Chutt>it displays it as wanting to record that?
00:32<Chutt>the scheduler doesn't even consider stuff older than now
00:33<Chutt>that's not an accurate list.
00:33<Chutt>check the fix conflicts screen
00:33<Chutt>it won't be there
00:33<warlord>No, it's not there..
00:34<Chutt>the fix conflicts screen is what's actually being recorded
00:35<warlord>Ahh -- so the reason I'm being screwed is that it will never look at a program that started in the past, even if the program is currently airing.
00:35<Chutt>no, if it's currently airing that's fine.
00:35<warlord>Not if it was already marked for recording...
00:36<Chutt>record dups doesn't mean what you think it does, however.
00:36<warlord>it doesn't?
00:36<warlord>what does it mean, then?
00:37<Chutt>check out line 312 or so of scheduledrecording.cpp
00:40<warlord>uhh... that logic seems non-intuitive (if not completely wrong). Care to explain the reasoning behind that logic?
00:40<Chutt>talk to captain murdoch
00:40<Chutt>as that's all his code.
00:40<warlord>I see...
00:40<Chutt>there's two ways to define duplicates in here
00:40<Chutt>you're using one, he's using another
00:40<warlord>So basically what you're saying is there is no way to say "record this program regardless of how it compares to other programs"
00:41* warlord runs around the room screaming a bit..
00:42<warlord>My wife wants a "record this, dammit" flag.
00:43<Captain_Murdoch>the "record duplicates" code tells it to record even if the title, subtitle, and description match an entry in oldrecorded UNLESS they match an entry in the recorded table (since why record something if you've still got it hanging around from the first time you recorded it)
00:43<warlord>Captain_Murdoch: because sometimes I know better than the tv guide.
00:44<warlord>For example, there was a double-header of soccer tonight out of portland oregon. The title, subtitle, and description of both shows was exactly the same, but I WANTED it to record BOTH.
00:44<Captain_Murdoch>mdz was talking about more options like what you want a while back but I think he's been too busy to work on anything. things like adding a "force record this", "force don't record this", etc.
00:45<warlord>IMHO, a "record this once" setting should record it, regardless of what it says in the database. Isn't that what "record this once" is all about?
00:45<Chutt>singlerecord should still bypass all that, no?
00:45<extremis>bah: "Couldn't download libpam0g"
00:45<extremis>well can't install debian
00:45<warlord>Chutt: one would think, but it doesn't/
00:45<warlord>the soccer programs were all defined as "record once"
00:45<Chutt>oh, that's right, it changed when it moved that logic into the select
00:46<warlord>can we change it back?
00:46<warlord>we can't fix the select to change it back?
00:46<Chutt>that's a different question
00:46<Chutt>and the proper pronoun there is 'i'
00:47<warlord>Hey, **i** didn't break the code...
00:47<Chutt>it's not broken.
00:47<warlord>Yes, it is. "record once" doesn't work.
00:47<Chutt>that's your opinion
00:48<warlord>Go ask 100 people on the street... Give them all the details about the situation, and I bet you $100 that 98 of them will agree with me.
00:48<Chutt>and until you contribute some code, i don't really care about your opinion =)
00:48<warlord>Are you this much of an asshole in real life? Or is this just your online persona?
00:48<Chutt>you've got the source, you're just as capable as anyone else of fixing it
00:49<Chutt>and since you care about it so much, that's good motivation for you to fix it and send in a patch
00:49<warlord>Honestly, I bet you could have fixed it in less time than demeaning potential developers.
00:49<warlord>well, honestly, the more I deal with you here, the less I want to help you out.
00:49<Chutt>that's fine with me
00:50<Chutt>i don't owe you any time to fix what you think are problems
00:50<Chutt>unless you want to pay for my time
00:50<FryGuy>can you owe me some time to help me figure out my unix homework? :)
00:51<Captain_Murdoch>would be fairly easy to check the "record duplicates" flag and record anyway if it's a singlerecord item.
00:51<warlord>I'll keep that in mind .... You never know when the roles will be reversed.
00:51<FryGuy>the man for cut is worthless
00:51<Chutt>warlord, oh, i've already spent time fixing stuff for you that i can't reproduce, remember? :p
00:51<warlord>Captain_Murdoch: I did that, but it didn't start recording (I set the flag after the program started)
00:52<Chutt>though, i really don't know how you're still tripping that audio changed stuff
00:52-!-tmk [] has joined #mythtv
00:52<Chutt>tmk, try current cvs
00:52<Captain_Murdoch>in (Myth)TVland, a person's opinion's are about as valuable as their code contributions. (ie, the more important/better the latter, the more important/better the former).
00:52<Chutt>i think i fixed it
00:52<Captain_Murdoch>warlord, I mean in the code, not in the gui.
00:52<warlord>Chutt: Honestly, I dont know either.. I've hit it once since you put in that fix, but I can't reliably reproduce it anymore.
00:52<tmk>which did you fix
00:52<Captain_Murdoch>tmk, I ordered 3 of those cards.
00:52<Chutt>tmk, well
00:53<Chutt>tmk, it looks like the GET_FRAME_
00:53<Chutt>err, GET_FRAME_TIMING stuff was off still
00:53<tmk>still giving bogus values
00:53<Chutt>not bogus
00:53<Chutt>just wrong, until it started displaying something
00:53<Chutt>but lemme know
00:53<tmk>so you queried it too soon
00:53<Chutt>i didn't really test it extensively
00:53<tmk>i'll do so
00:53<tmk>yeah its tough
00:53<Chutt>seemed to work ok, though
00:53<tmk>did you find the framesync err msg bug too
00:53<Chutt>no, i couldn't reproduce that one
00:54<Chutt>even with it recording and stuff
00:54<Chutt>warlord, so in scheduledrecording.cpp, at that line i pointed out to you before
00:54<Chutt>just wrap where it sets the duplicate flag with a check if it's a singlerecording
00:55<tmk>it happens after the recording finishes
00:55<Chutt>i don't want to commit that without testing, and i don't have time to make sure it'll do the right thing
00:55<warlord>Yea, I'm looking at the schema to figure out how to determine that.
00:55<tmk>i'll try cvs
00:55<Chutt>warlord, shouldn't need to touch the db at all for that
00:56<extremis>is there a unstable repository for
00:56<warlord>Well, it needs to grab the record.type, no?
00:56<Chutt>stuff gets uploaded to unstable directly
00:56<Chutt>warlord, it should have it
00:56<Chutt>at that point
00:56<extremis>thats what I thought
00:56<Chutt>but you can generally use the testing repository
00:57<extremis>hrm, should I dist-upgrade first?
00:57<Chutt>i don't use apt to upgrade
00:57<Chutt>so, can't help ya there :p
00:58<extremis>yeah, I think I do, then I upgrade
00:58<Chutt>i really shouldn't inflict dselect on anyone by suggesting that
00:58<extremis>you really should use apt
00:58<extremis>its the only reason I use debian
00:58<extremis>hell its great on redhat too
00:58<Chutt>i use it for little things
00:59<extremis>ported for its l33tness
00:59<Chutt>but i'm used to dselect
01:00<warlord>Chutt: sorry, but record.type is not returned in that Select in schedulerecording.cpp.
01:00-!-ChaosExiguum [] has quit [Read error: 110 (Connection timed out)]
01:00<warlord>take a look at lines 234-241
01:00<Chutt>yeah, that's right
01:00<Chutt>it looks it up later for some reason
01:00<warlord>There is no "record.type" in that string. BUt I suppose I could add one.
01:02<warlord>well, that makes it hard to ignore the "duplicate" checking if 'type' isn't available. ;)
01:02<Chutt>could also do it later on
01:02<Chutt>but adding it there's probably easiest
01:07<warlord>what's proginfo->recording for?
01:07<Chutt>that stores all the recording info
01:07<Chutt>it's what gets filled in later
01:07<warlord>?? looks like a bool?
01:07<Chutt>oh, nevermind
01:07<warlord> if (proginfo->duplicate)
01:07<warlord> proginfo->recording = false;
01:07<Chutt>wrong member
01:08<Chutt>yeah, that just disables it entirely
01:08<warlord>Ah, ok.. Default is true, then, in the constructor?
01:10<extremis>is there a partition size limit with ext2?
01:11<warlord>not as far as I know.
01:11<Chutt>it's fairly large
01:11<warlord>I've got a 180G ext3 partition..
01:11<extremis>I'm trying to make a 200gig partition with ext2
01:11<extremis>I don't understand why unstable debian didn't upgrade this kernel
01:12-!-DogBoy [] has quit [Read error: 110 (Connection timed out)]
01:13<Chutt>it generally doesn't do kernels automatically
01:13<Chutt>since, well, that can be bad =)
01:15-!-DogBoy [] has joined #mythtv
01:16<warlord>is cvs down?
01:16<warlord>cvs [diff aborted]: connect to failed: Connection timed out
01:17<Chutt>still running here
01:17<warlord>Trying again...
01:18<Chutt>yeah, i just verified that pserver access is definitely still working
01:20<warlord>It doesn't like me.
01:21-!-Timon [] has joined #mythtv
01:21<Timon>thor_: you around?
01:29<warlord>Chutt: looking more at this SQL, it looks like the WHERE clause is completely ignoring items "FROM record" where record.type = kSingleRecord
01:30<warlord>Although my SQL is extremely novice, so perhaps I'm not reading this right.
01:31<warlord>Or... maybe it gets covered by the inside of the WHERE where it's matching startdate, starttime, and endtime.. Which should be good enough.
01:34<warlord>Hmm, I can get to cvs from my laptop, but not my myth box...
01:34<warlord>and the problem appears to be at WEIRD.
01:39<warlord>patch sent. g'night
01:39-!-warlord is now known as warlord-afk
01:46-!-dja [] has quit [Remote closed the connection]
01:50<o_cee>rkulagow: here?
01:51<o_cee>rkulagow: section 21.13: "(or files which were are already in the correct format)".
02:05-!-sc00p__ [] has joined #mythtv
02:16-!-sc00p_ [] has quit [Read error: 60 (Operation timed out)]
02:22<tmk>chutt: still buggy
02:22<tmk>seems abaout the same to me
02:40-!-Timon [] has quit [Remote closed the connection]
02:48-!-rkulagow [] has quit [Read error: 60 (Operation timed out)]
03:09-!-tmk [] has quit []
03:11<o_cee>Chutt: you there?
03:12<o_cee>is it safe to assume that there never will be a valid category/subtitle thats only 1 char long?
03:13<o_cee>dammit, you gotta be asleep, forgot about the tz difference.
03:20-!-cbreen [] has quit [Connection reset by peer]
03:49-!-Viddy [] has quit ["Client exiting"]
04:48-!-sfr [] has joined #mythtv
05:03<sfr>o_cee: re: one character long category. what for?
05:12-!-FryGuy [] has quit [Read error: 54 (Connection reset by peer)]
05:59-!-Viddy [] has joined #mythtv
06:32-!-yeyo [] has quit ["Client exiting"]
06:46<o_cee>sfr: it's in mythweb. $show->category that is NULL returns " " (no, i haven't had time to look at the reason yet), so doing a strlen > 0 isn't enough. doing a strlen > 1 would work if we can safely assume that there aren't any categories with only 1 char. otherwise there will have to be a && != " " as well.
06:48<sfr>don't exactly understand that, but to rely on categories to always be > 1 character sounds wrong to me.
06:49<o_cee>yeah, i guess the other way is better (more code though)
07:18-!-atomix [] has joined #mythtv
07:18<atomix>someone out there?
07:18<atomix>sogar in deutsch?
07:19<atomix>habe gestern unter gentoo mythtv installiert
07:19<sfr>ja, but english is preferred.
07:19<atomix>lief auch alles ganz gut
07:20<at07:21<atomix> when i switch to view tv - a black screen appears
07:21<sfr>did you set up your tv channels?
07:21<atomix>i think no - where can i set the channels?
07:23<atomix>mythfilldatabase don't work
07:23<sfr>atomix: exists. you do so by running mythfilldatabase --preset and either --xawchannels or --manual
07:24<sfr>scratch that either or, they can be combined.
07:26<atomix>what it cardinput?
07:27<o_cee>have you run the setup program at all?
07:27<atomix>o_cee: yes
07:27<sfr>atomix how many tuners do you have?
07:28<sfr>then cardinput =1. it's for the case when you have 2 tuners and each one receives different channels.
07:29<atomix>sfr - now are channels in the mysql-table
07:29<atomix>can i start mythfrontend in a single smaller window?
07:30<sfr>atomix: change the settings in mythfrontend.
07:30<o_cee>use two desktops instead
07:31<sfr>atomix: does Murph ring a bell? :)
07:31<atomix>what? i don't understand - sorry
07:31<sfr>atomix: better so.
07:32<sfr>atomix: mythfrontend->Settings->Appearance 2nd screen
07:36<atomix>so now are channels in the database, but the screen was black to
07:37<sfr>did you test your sound setup with aplay?
07:38<atomix>xmms works
07:38<sfr>and look for error messages on the backend/frontend
07:39<atomix>what means 'running as slave backend'
07:39<atomix>ist this correct?
07:39<atomix>i have only one system
07:39<sfr>only if you also have a master backend
07:40<o_cee>my advice is to follow the instructions found in the howto.
07:40<sfr>atomix: did you follow the install/setup instructions?
07:40<atomix>sfr: yes - point for point
07:40<o_cee>what point is _not_ working then?
07:41<o_cee>what output do you get to the console? errors about /dev/dsp?
07:41* sfr thinks mythtv really needs a FAQ
07:41<atomix>no errors
07:41<o_cee>it's all in the howto, should be enough
07:42<atomix>that my problem - so stupid i normaly not
07:42<o_cee>paste here the last few lines
07:42<atomix>from what?
07:42<o_cee>the console, when you get the black screen
07:42-!-GreyFoxx [] has joined #MythTV
07:43<atomix>connecting to backend server
07:43<atomix>the last and only
07:45<o_cee>and that's the correct ip i guess?
07:45<o_cee>what does the backend say?
07:46<atomix>Running as a slave backend
07:46<atomix>Probed: /dev/video0 - Television
07:46<atomix>Probed: /dev/video0 - Composite1
07:46<atomix>Probed: /dev/video0 - S-Video
07:46<o_cee>don't list it all
07:46<o_cee>list what's last
07:46<o_cee>and any errors
07:47<atomix>time stamp - connected.
07:47<sfr>atomix: start setup and tell us the settings of: IP address of $HOST and Master server IP address
07:47<atomix>no errors in /var/log/mythtv/mythbackend.log
07:48<sfr>o_cee: if both address match, then it's a master backend, yes?
07:48<o_cee>i'll see how my db looks like
07:48<atomix>think so
07:49-!-overridex [] has quit ["Because I live in a VAN down by the RIVER"]
07:49<o_cee>put there instead
07:50<o_cee>shouldn't make any difference i guess
07:50<atomix>i try
07:50<o_cee>is the backend running at all?
07:52<sfr>o_cee: i don't see a reference to 'black screen in live-tv' in the docs. do you?
07:53<atomix>now the log saw Could not connect to master server.
07:53<sfr>does the frontend run on the same system
07:53<o_cee>the backend isn't started
07:54<o_cee>prove me wrong
07:55<sfr>it's strange how Chutts attitude spreads in this channel :)
07:57<sfr>o_cee: so what about that black screen thing? do you see it in the docs?
07:58<o_cee>not really, but it's defanately one of the most discussed issues on the mailinglist
07:58-!-atomix_ [] has joined #mythtv
07:58<atomix_>sorry, t-offline disconnet me
07:59<atomix_>why prints mythbackend: could not connect to master sercer?
08:00<sfr>atomix_ b/c it thinks its a slave backend, which is wrong.
08:00<atomix_>i see it - moment
08:00<atomix_>stupid typo
08:00<atomix_>now connection etablished
08:01<atomix_>but the frontend display only the black screen
08:02<sfr>atomix_ audio problem most likely. check
08:02<o_cee>what kind of errors does the console give you?
08:02<atomix_>o_cee: now errors
08:05-!-jkolb [] has joined #mythtv
08:06<atomix_>no errors
08:08<sfr>atomix_ check your soundcard setup. without working audio (in duplex mode) you won't see a picture.
08:08<atomix_>ok - it will test it
08:08<sfr>atomix_ btw, what tuner card do you have?
08:12-!-atomix [] has quit [Read error: 110 (Connection timed out)]
08:28<atomix_>no success - i try it later
08:28<atomix_>now i must go work in kitchen :-(
08:29<sfr>hehe. doing the dishes?
08:29<atomix_>yes - help my wive
08:30-!-atomix_ is now known as atomix|away
08:53-!-toe [] has joined #mythtv
08:54<toe>rm - hi all - I cant get my actisys 220l+ to work with lirc - anyone have this working?
09:20-!-ChaosExiguum [] has joined #mythtv
09:22<o_cee>Murph Crusher <> <-- clever guy.
09:22<o_cee>Chutt: you still asleep?
09:23-!-atomix|away is now known as atomix
09:23<atomix>o_cee i think i have i problem with my soundsetup
09:24<atomix>i try to hear something with line-in mute and capture from xawtv
09:24<atomix>but no sound
09:24<o_cee>noooo, do you think so?
09:24<o_cee><sfr> atomix_ audio problem most likely. 1\xBD hour ago
09:25<atomix>yes - but now i search the bug in my config
09:25<o_cee>just follow the HOWTO, it's covered.
09:26<o_cee>and search the mailinglist, it's been discussed.. _alot_...
09:27<atomix>o_cee: ok
09:28<sfr>atomix: explains testing this well. if you still have problems, well then ask again.
09:29<atomix>sfr: i try anything i read in troubleshooting audio
09:29<atomix>and follow the instruction to install alsa
09:30<atomix>and mute/cature the line/capture
09:30<o_cee>have you emerged alsa-oss?
09:31<o_cee>and followed the alsa installation guide at
09:31<atomix>yes - is in the portage
09:31<atomix>yes - i do the full installation of alsa
09:31<o_cee>have you insmodded snd-pcm-oss and snd-mixer-oss?
09:31<atomix>xmms runs fine
09:31<o_cee>xmms is probably using alsa, not alsa-oss
09:32<atomix>yes - lsmod returns that both in
09:32<atomix>ok i try xmms with also-oss
09:32<o_cee>wich would mean ordinary OSS
09:33<sfr>atomix: do you have sound working in xawtv with line-in NOT muted?
09:33<atomix>sfr - yes
09:33-!-Hanuman [] has joined #mythtv
09:33<Hanuman>Any remote wonder users in here?
09:33<sfr>atomix: checked your mixer settings with alsamixer? spec. the capture setup?
09:34<atomix>xmms works also fine with liboss
09:35<atomix>i set the capture with amixer as it was describe in the howto: amixer set Capture,0 100%,100% mute captur
09:37<sfr>atomix: alsamixer is easier to use imho
09:37<atomix>sfr - i know but i think commandline is more secure
09:37-!-dakeyrus [] has joined #mythtv
09:38-!-omegahamlet [] has quit [Read error: 104 (Connection reset by peer)]
09:39<sfr>atomix: as you wish. but mythtv works fine here with me using alsamixer :-) i'm just mentioning it as i had problems setting up the capture stuff with another mixer prog, aumix.
09:40<atomix>ok - at the moment i try alsamixer
09:43<atomix>but there was no channel that enable the sound that aplay /dev/dsp makes
09:44<sfr>what do you mean?
09:45<atomix>i try any mixer to capture the sound that "aplay /dev/dsp" makes
09:46<sfr>atomix: you have xawtv still running?
09:46<atomix>when i unmute line then i can hear the sound again
09:47<atomix>but my soundcard did't capture anything from the line-in
09:47-!-m0j0 [~m0j0@] has joined #mythtv
09:47<sfr>ok, set master, pcm, line-in, capture and adc channel to 80% volume.
09:48<sfr>far to the right
09:48<o_cee>what's the command to edit /etc/groups?
09:48-!-Hanuman [] has left #mythtv []
09:48<sfr>o_cee: vi
09:49<o_cee>there's another one, at least under gentoo
09:49<o_cee>that takes care of updating and so on
09:49<sfr>addgroup on debian iirc
09:49<jkolb>Sorry, vigr
09:49<o_cee>that might be it :) thx
09:49<atomix>sfr: no adc at my alsamixer
09:49<jkolb>vipw is for /etc/passwd
09:49<o_cee>what's vipw?
09:50<sfr>atomix: ok, maybe you don't need it.
09:50<sfr>atomix: now unmute all channels mentioned before, EXCEPT line-in.
09:50<atomix>ok - i have a soundblaster live card 5.1 sound
09:51<atomix>ok - all unmute, except line-in
09:52<sfr>now move to line-in and hit space to select line-in for capture
09:53<sfr>dito for the capture channel
09:53<atomix>both set to captur
09:53<sfr>now if you're lucky it works
09:53<atomix>no sound :-(
09:54<o_cee>but video?
09:54<atomix>video with xawtv
09:54<atomix>i test the audio-config
09:55<sfr>atomix: leave the mixer settings as they are, stop aply, xawtv and restart the backend/frontend and see what happens
09:55-!-Captain_Murdoch_ [] has joined #mythtv
09:56<atomix>no sound no video
09:56<atomix>only black screen
09:56<sfr>atomix: what does fuser -v /dev/dsp say?
09:57<o_cee>does anyone know how to get a keyboard with funky play/pause/internet buttons to work? haven't looked at all at it yet
09:58<atomix>no error at backend and frontend :-(
09:59<sfr>Captain_Murdoch_ good morning. with auto-skip commercials active, is there an option to disable it temporarily while/before watching a recording?
10:00<Captain_Murdoch_>sfr, no. I don't even use auto-skip. I don't trust the detection that much yet. :)
10:01<Captain_Murdoch_>I've considered an option like that though. or just a way to toggle it on/off without having to go into setup.
10:01<Captain_Murdoch_>might be a good item to put on the popup menu whenever someone gets around to working on one.
10:02<sfr>Captain_Murdoch_: dejavue? Chutt said the same :(. what about a switch in the watch recordings screen for now?
10:04<sfr>Captain_Murdoch_ you wrote that comm.-detect code, or was that someone else?
10:04<Captain_Murdoch_>that was me.
10:04<Captain_Murdoch_>I think it's good, works for me most of the time on shows I watch, but still not near 100% so I don't use auto-skip.
10:05<sfr>atomix: you can try to record audio (with rec for example, comes with sox) to see if that works.
10:06<atomix>sfr: that do not work too
10:06<o_cee>Captain_Murdoch: have you worked any more on the other detection methods?
10:07-!-Edgan [] has quit ["Client exiting"]
10:07-!-ChaosExiguum [] has quit [Read error: 110 (Connection timed out)]
10:07<Captain_Murdoch_>I've got logo detection working pretty well, but it can't work standalone because logos don't start/stop on the exact right times. I was planning on using the logo detection to enhance blank-frame but also give people the option of using just logo detection.
10:08<sfr>Captain_Murdoch_ iirc to detect blank frames you calculate the avg. brightness of a frame. is that level configurable?
10:09<o_cee>Captain_Murdoch: okay, cool. way cool :)
10:09<o_cee>will it work on "transparent" logos?
10:09<Captain_Murdoch_>not average brightness. it does a grid search that looks almost like an interlaced gif/png drawing but as a grid not just lines. it looks for pixels over a minimum value.
10:10-!-Edgan [] has joined #mythtv
10:10<Captain_Murdoch_>o_cee: it detects them reasonably well, but I'm still working on it.
10:10<o_cee>okay.. got a couple of those
10:11<sfr>atomix: you tested it with xawtv running an line-in unmuted? just checking.
10:11<atomix>xawtv with line ummuted sounds fine
10:12<Captain_Murdoch_>sfr: the code uses 2 thresholds. if a pixel is found over the max of the two the function returns false (ie, non-blank). if none are found above the max but some are above the lower then it does an average of the screen to see if it's just dim. if all are below the min of the two thresholds it assumes blank.
10:12<sfr>atomix: but can you record the tv audio using rec?
10:12<o_cee>i promised you to look at the logos. haven't really done it, but stations seem to use the same pattern here. usually it matches up pretty well, or they turn it on after the comercials, when there's announcements for other shows on the same station.. but that doesn't relaly matter imo
10:12<o_cee>i mean, stations does it the same way, but not all stations do it the same way :)
10:13<Captain_Murdoch_>I haven't had a chance to work on the logo stuff in 2-3 weeks because of being out of town on vacation and then the hurricane hit this area but I hope to get back on it soon.
10:13<o_cee>ohkay :)
10:14-!-jbwiv [] has joined #mythtv
10:14<jbwiv>hey guys....anyone know if a 500mhz celeron is powerful enough to run a mythtvbackend? This would be using at least one PVR-250 and perhaps an additional PVR-350?
10:14<Captain_Murdoch_>I have it doing the detecting, just need to add code to allow using only logo detection as well as using logo detection to enhance the "all detection methods" method which currently uses blank frame and scene change info.
10:15<o_cee>sounds promising
10:15<sfr>jbwiv: with only pvr's then i'm pretty sure. but probl. not as an frontend too.
10:15<Captain_Murdoch_>jbwiv, should be ok for at least one pvr-250, not sure about 2. I have a Celeron 466 that I had a pvr-250 in and it worked ok except for the crappy builtin i810 video support.
10:17<Captain_Murdoch_>I use the 466 as my main frontend, it's slow in some things but I liked the size of the box, fits nicely with my other A/V equipment.
10:17<Captain_Murdoch_>plays 480x480 mpeg2 fine
10:18<o_cee>i gotta get some pictures as well.. haven't finished to box yet though
10:18<Captain_Murdoch_>the round thing on top is the creative cmr100 receiver.
10:18<atomix>sfr: so sound from xawtv rec to
10:18<kvandivo>the black thing you can barely see on the left end of the box?
10:19<sfr>Captain_Murdoch_: what's that box on the left? ;)
10:19<atomix>maybe the soundcard are damaged
10:19<Captain_Murdoch_>kvandivo: sorry I haven't got back to you on the recommendations stuff. any more improvements
10:19<Captain_Murdoch_>box on left is vcr, box on right is computer. the thing sitting on the top left of the computer is the IR receiver.
10:19<Captain_Murdoch_>yeah, see the dust on it.
10:20<sfr>Captain_Murdoch_: vcr? ah, i remember now. :)
10:20<Captain_Murdoch_>that's used to transfer myth recordings to ancient spools of "tape" for giving myth recordings to relatives that forget to set their timers.
10:20<o_cee>tape.. hmm, had those on the C64.. never seen them since :)
10:20<kvandivo>i'm just being patient, CP. No real improvements at this time. My next step is to add real 'program rankings' code where the user can rank any program (whether they are recording it or not) positive or negative. oh, and i need options for autorecording recommendations, and that will require a flag in the 'record' table so that we know that something is a recommendation and can note that when it comes time to autodelete, etc.
10:20<Captain_Murdoch_>or so show off the fact that you can cut out commercials using detection or manual edits.
10:21<kvandivo>once we get real rankings, i can do the bayesian i've been wanting all along
10:21<o_cee>kvandivo: how does the recomendation stuff works anyway? haven't tried it
10:21<kvandivo>and, after that... collaborative filtering
10:21<kvandivo>it looks at words in all the shows you've recorded and finds matches in the upcoming shows
10:22<kvandivo>to greatly simplify thingns..
10:22<o_cee>oh, so fully automatic..
10:22<o_cee>might be neat :)
10:22<Captain_Murdoch_>kvandivo: I think it was working ok on my system. I think I may have found a bug in the scheduler's prunelist. the recommendations stuff was listing shows I already had set to record so I checked and "mythbackend --printsched" didn't show the episodes sometimes when they were in the future too far. I made a note to check into it
10:22<kvandivo>well, the determination is... you (currently) have to tell it to record them manually
10:22<o_cee>doesen't tivo do something like that? (me has never used a real pvr, such things doesn't exist here)
10:23<kvandivo>ya.. it shouldn't be doing that.. after I calculate the list, i'm getting the upcoming list of shows to record and explicitly remove them..
10:23-!-PNY_ [[] has quit ["Client Exiting"]
10:23<kvandivo>ahh.. one thing I've noticed, Captain_Murdoch_ .. and I've asked about this, but haven't gotten an answer (and haven't had time to dig yet..)
10:23<sfr>atomix: so you have to fix the recording issue.
10:24<kvandivo>if myth is recording a particular episode of south park (which is notorious for showing the same episode 15 times ina week)
10:24<atomix>yes - i search at this moment at google
10:24<kvandivo>the recommendations code will sometimes suggest one of the alternate showings
10:24<kvandivo>(since it doesn't know that is an alternate showing of a show you've already set it to record)
10:24<kvandivo>understand what I mean?
10:25<kvandivo>i need a function that will return 'all pending recordings, as well as all alternate showings of shows that also don't need to be recorded'
10:25<kvandivo>is it possible that is what you are seeing when you see shows in the recommendations that you have 'already set to record'?
10:26<Captain_Murdoch_>kvandivo: the problem with prunelist is that it's pruning shows it shouldn't so your code recommends recording it when in actualality it is already going to record.
10:27<Captain_Murdoch_>could be. if you have "anytime" recordings the schedule should list all future occurances of an episode until it records the episode the first time then all should disappear.
10:27<Captain_Murdoch_>are you talking timeslot, weekslot or anytime/channel?
10:28<kvandivo>ummm... i probably have south park set to anytime... if episode A is on at 9, 9:30, and 10:00. and is going to be recorded at 9, recommendations knows that, and won't recommend the 9 showing.. however, it doesn't know that the 9:30 and 10 showings are duplicates of one that you are already recording
10:29<kvandivo>so they are eligible to be recommended
10:31<kvandivo>at least that's the behavior it seems like i've noticed.
10:31<kvandivo>perhaps i've been wrong all along and i've actually been seeing the bug you have noticed and have attributed it to the wrong thing.
10:32<kvandivo>recommend.cpp line 138 looks to me like it just grabs every pending recording and removes it..
10:32<kvandivo>seems to me like RemoteGetAllPending returns conflicts and everything, iirc. so it should take all of those out of the possible list
10:34<jbwiv>Captain_Murdoch_: did you say your 486 is frontend and backend? I'd be using the 500mhz celeron for both, if it'll work...
10:35-!-hculver [] has joined #mythtv
10:35<kvandivo>lately, though, Captain_Murdoch_, I've been spending my time working on stuff other than recommendations.. i've about fixed a bug with running mythfilldatabase very close to midnight (for testing, i get one shot per day, since i'm not interested in changing my computer clock just for testing)
10:36<kvandivo>and, i noticed last night that chutt was commenting that 6544 wasn't accurate for the list of upcoming programs, which struck me as strange since it calls the same functions to get the list as the view conflicts screen.... so, i looked into that, and have added some code to tell it to not list upcoming shows for recording if their start time < NOW(). still testing it to verify that it works right, though
10:36<hculver>Are others seeing the mythtv-users mailing list mail stall? I get no mail from the mailing list for a few days, and then several hundred arrive all at once, and its ok for a few days.
10:38<kvandivo>o_cee: tivo does have a 'suggestions' area that will record shows for you.
10:38<hculver>thanx, I'll investigate this end (altho other mail seems to be fine).
10:38<o_cee>ohkay.. something similar?
10:40<sfr>kvandivo: what is that bug?
10:41* KikoV is away: I'm busy
10:42-!-Drikus [] has joined #mythtv
10:52-!-Drikus_ [] has joined #mythtv
10:55-!-hfb [] has joined #mythtv
10:56-!-Drikus_ [] has quit [Client Quit]
10:59-!-Drikus [] has quit [Read error: 60 (Operation timed out)]
11:05-!-tmk [] has joined #mythtv
11:05-!-dopez [] has joined #mythtv
11:16<Chutt>tmk, so, did i fix it or what?
11:19<tmk>forget what i said last night
11:19<tmk>forgot to make install
11:20<tmk>just tested it out now
11:20<tmk>kinda wierd how the osd time-display jumps around now hto
11:20<tmk>always seems to end up in the right spot
11:20<Chutt>yeah, i think the osd is updating too fast
11:20-!-dja [] has joined #mythtv
11:21<tmk>but good work! :)
11:21<tmk>i patched the v4l stuff
11:21<tmk>and added a tuner option to itvt
11:21<tmk>so if people have bttv and ivtv and diff tuner types
11:21<tmk>they can set it via ivtv
11:21<Chutt>ah, cool.
11:21<tmk>i also put in the i2c 2.8.0 patch
11:22<tmk>goin to work now
11:23<sfr>tmk: but it still supports older versions of i2c?
11:23<tmk>yeah, you have to -DNEW_I2C to use the new stuff
11:23<tmk>the version check in the patch didn't work
11:23-!-tmk [] has quit []
11:23-!-mecraw [~mecraw@] has joined #mythtv
11:31-!-FryGuy [] has joined #mythtv
11:34-!-Hanuman [] has joined #mythtv
11:34<Hanuman>Anyone know of a good howto for picture-in-picture on MythTV? Can't find it anywhere in the docs.
11:35<sfr>Hanuman: see keys.txt for how to use it.
11:35<Hanuman>sfr: thanks
11:36<o_cee>Chutt: is it safe to assume that strlen(categories) > 1 ? it's solvable with a bit more code another way as well, an issue with mythweb..
11:36<Chutt>no, the category can be empty
11:37<o_cee>yeah, but i mean a valid category
11:37-!-poo [] has quit ["Download Gaim:"]
11:38<sfr>o_cee: i thought you had already finished that bit more code ;)
11:38<o_cee>mythweb now check for strlen($show->category) > 0, but if it's NULL it returns " ", and passes the > 0 check anyway.
11:38<o_cee>sfr: it's an easy fix, just wanted to know the prefered solution :)
11:39-!-poo [] has joined #mythtv
11:40<o_cee>Chutt: no opinion? then i'll send a patch that does it the proper longer way.
11:42<sfr>that reminds me about something: what's preferred? filling empty fields in the db with default values, or some additional code to set a field to a default value before usage/display?
11:43<sfr>i'm thinking about parts of the program data, like airdate or category.
11:44-!-Hanuman [] has left #mythtv []
11:51<kvandivo>sfr: you said ' kvandivo: what is that bug?' what were you referring to?
11:51<sfr>that bug in mythfilldabase when run at 23:59
11:52<kvandivo>ahh.. it has problems with the current date. the offset passed to xmltv changes meaning based on the current date.
11:52<kvandivo>i've added in code that captures the 'current' date and uses it through the entire 'day capture' process
11:53<kvandivo>this could be a tv_grab_na-only thing.
11:53<warlord-afk>Yea, but xmltv's captures wont deal properly.
11:53<kvandivo>i've handled that, war
11:53-!-warlord-afk is now known as warlord
11:53<warlord>Does it accept offset=-1?
11:53<warlord>(or do you just never attempt offset=1?)
11:53<warlord>er, 0
11:54<kvandivo>basically, the code acts as it did before, but if the day changes, it tweaks the value of 'i' appropriately.
11:54<kvandivo>and, the current date is passed among the function calls, rather than using mysql's CURRENT_DATE, which is what was screwing things up
11:55<warlord>Ahh, so if you ever are trying 'i==0' across the date flip, you're screwed?
11:55<kvandivo>again, though.. i'm still doing some final debugging.. (didn't have any errors last night, and the data looks correct, but 'i' needs tweaked differently.. )_
11:55* kvandivo mutters to himself and goes and loads filldata.cpp to see exactly what warlord is talking about.
11:56<warlord>While you're working on it, can you add a command-line argument for the 'default i value'?
11:56<sfr>got a different question: anyone in here in PAL-country which doesn't need the --preset option in mythfilldatabase?
11:56<sfr>or better: anyone outside the US?
11:57<kvandivo>warlord: i won't add two different things at once. better to submit two seperate patches for that to appease the powers that be
11:57<warlord>kvandivo: Fair enough.. (my request is probably easier ;)
11:57<sfr>warlord: that i value is which day to fetch?
11:58<warlord>sfr: yes. it's the "day offset" value
11:59<kvandivo>warlord: it actually looks like my current i tweaking -might- be using i = -1, which worked fine, btw.. but at 11:59, it did the grab_data on line 1397 for sept 3. (and then the date changed) and by the time it got to the for loop, the first time through it said that it already had data for sept 2. So, I probably still need to tweak the 'i' in that case. I hadn't looked at that code real close before trying it.
11:59<warlord>See line 1443 in filldata.cpp
11:59<sfr>warlord: i might be wrong, but then why don't you run tv_grab_na manually and just feed the xml file to mythtv?
11:59<kvandivo>hmmmm.. 1397 is in _my_ copy of the code.. don't know what it is in cvs.. it's the grab_data befoer the for loop that i'm talking about, though
12:00<warlord>sfr: that's what I'm doing now, but it would be MUCH easier if I could tell it "auto-grab offset=0" instead of offset=1, so when I run at 4am Monday it will grab Monday's data and not Tuesday's
12:00<kvandivo>if you are really concerned about, it, just change the 1 to a zero in your copy of the code, war
12:00<kvandivo>(like what i did with the 1 second vs 5 second SELECT recordChanged FROM thing..
12:01<warlord>Yea, I might just do that..
12:01<Captain_Murdoch_>jbwiv, the 466 is only a frontend. your 500 with a decent video card should be able to do frontend/backend because I was able to do frontend/backend/livetv on my 466 with my pvr-250. tweak the bitrates lower than the old default of 16000/8000 though.
12:02<Chutt>the new defaults are more sane =)
12:03<warlord>Hey, I liked the 16k/8k defaults ;)
12:07<sfr>warlord: instead of adding more command-line parameters to mythfilldatabase what about being just able to say: run /foo/ba/script with parameter '--days 14 --offset 1' and grab the xml file from /whatever/pleases/you ? b/c nobody sent a patch i guess.
12:08<sfr>it'd make mythfilldatabase more flexible
12:08<warlord>sfr: which script? I do want to limit the pounding of the zap2it servers.
12:09<sfr>warlord: any script that works for the user. tv_grab_na in your case.
12:10<warlord>I dont understand -- one can already run the script by hand and then use mythfilldatabase --file
12:11<sfr>yeah, so you don't need an additional command-line option :P
12:11<warlord>My point is that mythfilldatabase is ALWAYS going to download one day of data (and usually 2 when run once a day) -- it grabs offset 1, and then it grabs offset N (because 0 to N-1 are already in the database). I want to allow the user to CHOOSE which day is that 'first day'
12:11<Chutt>change the source.
12:11<warlord>sfr: YES, I do.. because NOW in order to do what I want I have to pull down THREE days.
12:12<warlord>Chutt: Well, I changed my copy of the sources to just 1 -> 0, but that's not really "the right solution"
12:12<Chutt>sure it is
12:12<warlord>It's the right solution for me..
12:12<Chutt>you're the only person to ever want that changed
12:12<warlord>It doesn't work for people who run the fill at 11:50pm
12:13<Chutt>sure it does
12:13<Chutt>it grabs tomorrow
12:13<warlord>So I'm the only one who wants to update my database the day-of?
12:13<sfr>warlord: i don't see your point when you already run tv_grab_na manually anyway.
12:13<Chutt>what would be the point of updating that day when there's 10 minutes left in it?
12:13<warlord>sfr: I dont want to have to run tv_grab_na manually -- it's extra (unnecessary) pounding on the zap2it servers
12:14<warlord>Chutt: right, that's exactly why "changing it to 0" is not the right solution.
12:14<Chutt>keeping it at 1 is
12:14<warlord>Adding a --default-day option lets ME set it to 0, and let's the person running at 11:50 keep it at 1
12:14<Chutt>you're the only one who wants it at 0
12:14<Chutt>no one else has ever brought that up in over a year
12:15<sfr>warlord: why extra? don't tell me you run two cronjobs downloading data from zap2it each night.
12:15<Chutt>so it's just as easy for you to go in and make that one character change
12:15<warlord>sfr: you CLEARLY don't understand what mythfilldatabase does.
12:16<warlord>Chutt: just because I'm the only person being vocal about it doesn't mean nobody else wants the feature.
12:16<Chutt>warlord, and, nice way to completely miss the point on that 'cle266 decoder segment fault' thread
12:17<Chutt>unless i'm completely mistaken and you're using the cle266 hardware decoder.
12:17<sfr>warlord: i know b/c i told it what to do. my point is rather about mythfilldatabase being too limited in which script it can run automatically.
12:18<sfr>warlord: i meant what it should do in my cronjob.
12:18<warlord>sfr: in my cronjob I want to be able to say:
12:18<warlord>mythfilldatabase --update
12:19<warlord>(or perhaps "mythfilldatabase --update --default-day 0")
12:19<warlord>currently, I have to:
12:19<Chutt>the --update option is only meant for people in europe.
12:19<Chutt>since some of those grabbers don't contain all the data, and running without it can destroy user-inputted data.
12:19<warlord>Chutt: ok.. it didn't seem to do any harm (and the docs are less than verbose about what it does)
12:20<warlord>regardless, my point is that right now I need to run:
12:20<Chutt> For running non-destructive updates on the database for
12:20<Chutt> users in xmltv zones that do not provide channel data
12:20<Chutt>that's pretty clear, i think
12:20<warlord>mythfilldatabase; tv_grab_na <channel-set 1>; mythfilldatabase --file; tv_grab_na <channel-set 2>; mythfilldatabase --file
12:21<warlord>It hits the zap2it servers AT LEAST 4 times (offsets 0 and 1) but usually 6 (0, 1, and N)
12:21<sfr>warlord: i agree with that. but adding special command-line arguments for tv_grab_na, another one for tv_grab_XY seems wrong to me.
12:21<warlord>Chutt: only if you know what it means. someone else here suggested I use --update
12:22<Chutt>why don't you change when you run the cron job?
12:22<Chutt>or just change the one single character in the source and be done with this whole pointless thing?
12:22<warlord>you're just the center of the universe, aren't you chutt?
12:22<thor_>or run your clock one hour early
12:22<Chutt>it's open source.
12:23<Chutt>you _have_ the source
12:23<Chutt>you _can_ edit it
12:23<Chutt>it can't be everything to everybody.
12:23<warlord>yes, but making a change just for me isn't helping the community
12:23<Chutt>no one else in the community has ever wanted this
12:23<warlord>how do you know?
12:23<warlord>Did you go ask eveyone?
12:23<Chutt>someone would have asked
12:23<Chutt>if you've seen the stupid things people have asked for on the mailing lists :p
12:23<warlord>In my experience more people are "me, too" people than "I want" people..
12:24-!-steelep [~signwatch@] has joined #mythtv
12:24<thor_>"I want" to be a "me,too" person
12:24<sfr>warlord: the xmltv files must be considerably different in the US compared to Europe/Germany. no offense intended.
12:25<warlord>sfr: yes, they are.
12:25<warlord>thor_: LOL
12:25<thor_>hey, it's Friday
12:25<Chutt>warlord, are you going to post a follow up to your most recent email to -dev, saying that you didn't bother reading the thread before posting, or what?
12:27-!-cai [~cai@] has joined #mythtv
12:28<cai>I need some assistance with debian and mythtv...I am kinda slow...anyone up to the challenge...or provide me the precise RTFM link?
12:28<Chutt>cai, ask on the mailing lists.
12:28<cai>Can't connect to local MySQL server through socket '/var/run/mysqld/mysqld.sock' (2)
12:28<sfr>cai: sorry you asked for it.
12:28<cai>ok Chutt
12:29<sfr>cai: add a ;-)
12:29<cai>sfr going there now
12:29<Chutt>you really want
12:30<thor_>hmmmm .... telnet localhost 2342, reload plugins .... badda-bing .... there's a remote possibility I'm not a complete idiot
12:30-!-lmatter [] has quit ["Client exiting"]
12:31<Chutt>thor, great
12:31<thor_>it's coming
12:31<thor_>a thread for everything, and everything in its thread
12:34-!-hurdel [] has joined #mythtv
12:36<hurdel>anybody have experience with converting .avi files to svcd?
12:37<hurdel>in linux
12:37<sfr>Chutt: when are the <alttext> tags being used?
12:37<cai>dpkg-checkbuilddeps: Unmet build dependencies: libqt3-mt-dev
12:37<cai>getting further but the above is a problem
12:38-!-tmk [] has joined #mythtv
12:38<Chutt>sfr, on load in themedmenu.cpp
12:39<sfr>Chutt: ok, but when is it preferred over the <text> tag?
12:40<Chutt>when there's not enough room for the full text from the <text> tag
12:41<thor_>I think a lot of things use cutDown() though
12:41<Chutt>the menus don't
12:41<cai>Failed to create database (incorrect admin username/password?
12:41<cai>now this is the error =(
12:42<sfr>mdz: where are you? ;)
12:42<sfr>cai: are you using the mythtv debs?
12:45<cai>thought I was sfr
12:45<cai>apt-get install mythtv
12:45<cai>and it seemed to do alot
12:45<cai>fter unpacking 0B of additional disk space will be used.
12:45<cai>Setting up mythtv-database (0.11-6) ...
12:45<cai>Failed to connect to database: Access denied for user: 'root@localhost' (Using password: YES) at -e line 5, <> line 1.
12:46<cai>That is what comes up now with apt-get install mythtv
12:46<thor_>cai, mdz made the debian packages .... something about them creating a random password or something .... I'd search the mailing list archive
12:46<Chutt>what'd you tell it the password is?
12:46<cai>my root password chutt
12:46<sfr>cai: yeah, you answered the question about the Mysql administrator password wrongly probably.
12:46<Chutt>didn't it tell you 'this is not your root password'?
12:47<sfr>thor_: no s/t about people not reading carefully what his packages tell them.
12:48<Chutt>sfr, no, that's a different problem -- the password generation it does isn't always escaped properly
12:49<sfr>Chutt: ok, another reason.
12:49<Chutt>unrelated to this =)
12:51<sfr>built-in lottery, cool.
12:51<sfr>Chutt: but not this time. different error message.
12:55<sfr>cai: _carefully_ edit /var/cache/debconf/passwords.dat and replace the mysql root password for the MYTHTV package/template with the correct one. better make a backup before.
12:55* kvandivo senses some low level voodoo taking place.
12:56<sfr>but white magic in this case. at least that's the intention :)
12:57-!-_kch_ [] has quit ["Terminando cliente"]
12:58-!-lmatter_work [] has quit ["Client Exiting"]
12:59<cai>sfr the error has changed now
13:00<cai>ERROR 1045: Access denied for user: 'mythtv@localhost' (Using password: YES)
13:00-!-lmatter [] has joined #mythtv
13:02<sfr>cai: so you changed that password and rerun apt-get install mythtv-database?
13:03<cai>I am finding some similar errors sby googling
13:03<cai>perhaps I should purge and reinstall mythtv
13:03<cai>what is that command
13:03<cai>dpkg-purge mythtv
13:03<steelep>dpkg -P
13:03<sfr>cai: just wanted to suggest that. you might have to purge the myth* packages one by one.
13:05-!-cai [~cai@] has quit [Remote closed the connection]
13:11<sfr>Chutt: i'll abuse you as a postit :) the section about mythfilldatabase (9.4) doesn't mention the --preset option at all. rkulagow/someone should add it.
13:32-!-hculver [] has quit [Read error: 60 (Operation timed out)]
13:49-!-poo [] has quit [Read error: 54 (Connection reset by peer)]
13:49-!-poo [] has joined #mythtv
13:50-!-hurdel [] has left #mythtv []
14:15<o_cee>my mythbox got _really_ fucked. the capslock/scrlock diods are just blinking.. hmm
14:15<o_cee>can't get in via ssh either
14:16<o_cee>but it's responding to pings.. wtf
14:17<jkolb>That's pretty impressive.
14:17<o_cee>yeah, it is :)
14:17<MinmayIsABimbo>hmm... pings are handled really low level in the kernel
14:17<jkolb>My Mythbox is still dead atm. Haven't had it running for a month now. :(
14:18<MinmayIsABimbo>you could have caused a kernel panic (I have had the same thing happen to me before)
14:18-!-krenin [~krenin@] has joined #mythtv
14:18<jkolb>Have a new ps sitting next to me, though, and I'm going to install it when I get home.
14:18<krenin>I just installed MythTV, man it is AWESOME
14:18-!-paulproteus [] has quit [Read error: 104 (Connection reset by peer)]
14:18<krenin>i am getting a SVGA to TV converter
14:18<sfr>really? :)
14:18<krenin>building an entertainment system out of one of my PCs with it
14:18<o_cee>does the capslock/scrlock blinking mean anything?
14:19<thor_>kernel panic ?
14:19<sfr>iirc its morse code
14:19<MinmayIsABimbo>bad things(tm) happened :)
14:19<o_cee>heheh, c'mon :)
14:19<mikegrb>I have one box that when it loses power to the psu from the wall the scroll/caps/num lock blink when the power comes back... but it doesn't boot back up
14:19<mikegrb>sfr: that was good
14:20<o_cee>mikegrb: got picture on it still.. but it's not responding to anything
14:20<mikegrb>o_cee: have you tried any of the amazing sys-req key combos?
14:20<o_cee>guess i don't know em all
14:20<o_cee>got any good ones?
14:21<sfr>o_cee: only remember alt-sysrq-u for emergency ro mount, and alt-sysrq-b for you guess...
14:21<mikegrb>heh no, not off hand, seen them on webpages but never had a reason to use them so they don't stick in my memory
14:21<mikegrb>jkolb: I use that one :)
14:21<MinmayIsABimbo>o_cee: what kernel you using?
14:21<o_cee>i'll try, going to the living room :)
14:21<o_cee>2.4 gentoo
14:21<jkolb>I've never had it fail to do what I expect.
14:22<MinmayIsABimbo>know which 2.4 kernel? (like 2.4.21 or 2.4.20, etc..)
14:22<mikegrb>jkolb: certainly
14:22-!-paulproteus [] has joined #mythtv
14:22<sfr>i get a lock-up when trying to rmmod bttv when the ivtv driver is loaded.
14:23<o_cee>MinmayIsABimbo: not when i can't get to it, off of my head, no :)
14:23<o_cee>but i can check
14:23<o_cee>what gentoo has
14:23<o_cee>sfr: didn't seem to work
14:24<sfr>guess it's fubared
14:26<krenin>what is regular TV resolution?
14:26<warlord>What TV format?
14:26<o_cee>ntsc or pal
14:27<o_cee>dunno then :) 720x480 i think
14:27<warlord>Actually, I think 640x480
14:27<warlord>but I'm not 100% sure.
14:27<krenin>i've been recording at 320x240 with my 1 ghz duron
14:27<sfr>no s/t like 720x576
14:27<krenin>it works ok
14:28<o_cee>720x576 is pal
14:28<o_cee>MinmayIsABimbo: seems to be 2.4.20.. why?
14:28<DogBoy>or xvcd
14:29<DogBoy>you can do ntsc at 720x576
14:29<sfr>google also suggests: 640x480 for ntsc.
14:29<DogBoy>the google oracle
14:29<o_cee>no more ideas? then i'll have to kill the dragon and hope he wakes up again..
14:29<Chutt>there is no set horizontal resolution for analog video. it's, well, analog :p
14:30<krenin>my Duron 1 ghz can't handle 640x480
14:31<o_cee>then the dragon must die. :/ please wake up
14:31<sfr>o_cee: be confident, the dragon is strong.
14:31<kvandivo>if you live in _that_ much fear that your 'dragon' isn't going to wake up o_cee, maybe it is time to buy a new 'dragon'
14:31<DogBoy>480x480 for recorded shows if you have a 1 gig machine
14:32<Captain_Murdoch_>krenin, try 352x480 or 480x480. I use 352x480 on my Athlon 700 and it leaves me with 20% cpu free.
14:32<sfr>dragon XP maybe ;)
14:32<poo>ne1 using myth with hdtv ?
14:35<MinmayIsABimbo>o_cee: I saw a few posts about 2.4.21 doing that with ide-scsi
14:38<krenin>which should i use mpeg4 or the other
14:38<o_cee>aaaargh, bad superblock
14:38<o_cee>using ext3
14:39<sfr>maybe your dragon caught a flew?
14:39<o_cee>seems like it
14:40<o_cee>MinmayIsABimbo: not using scsi
14:41<MinmayIsABimbo>hmm... prolly just a random kernel panic... cant really say more than that :\
14:41<Captain_Murdoch_>krenin, mpeg4
14:41<sfr>MinmayIsABimbo: random kernel panic? he runs Linux, not W...
14:43<MinmayIsABimbo>sfr: true...
14:45-!-paulproteus [] has quit [Read error: 104 (Connection reset by peer)]
14:46<o_cee>how do i know where my superblocks are?
14:47-!-TheWildgoose [] has joined #mythtv
14:47<sfr>o_cee: you could try 32768 but they are not always at the same position afaik
14:47<sfr>o_cee: depends on the partitions size i think
14:48<mikegrb>well they look like ordinary blocks most of the time until duty calls, then they step into the nearest phone booth and don their super hero attire
14:48<TheWildgoose>thor_ are you around?
14:48-!-paulproteus [] has joined #mythtv
14:48-!-mecraw [~mecraw@] has quit ["Trillian ("]
14:48<sfr>mikegrb: again...
14:48<sfr>mikegrb: oh, maybe not :)
14:49<TheWildgoose>I just updated mythmusic cvs and I notice a problem when ripping flac files. Wondered if you could give me a few pointers...
14:49<thor_>ok ... don't know much about the flac code ... but there were a couple of patches recently
14:49<TheWildgoose>Problem is that file looks like it rips ok, the comments field looks just the same as before. However, the db scan is importing it as "unknown artist", "unknown album", etc
14:49<TheWildgoose>yeah, I saw that there were a few patches
14:49<krenin>i seem to have good luck with 352x288, i think it's what my card does actually
14:49<TheWildgoose>can you think how it might have been broken as described?
14:50<thor_>file plays ok?
14:50<TheWildgoose>yeah fine
14:50<TheWildgoose>and the "tags" look fine
14:50<TheWildgoose>using metaflac
14:51<TheWildgoose>looks like the db scan is importing wrongly
14:51<thor_>hmmm ...
14:52<TheWildgoose>I have an old copy of source hanging round.
14:52<TheWildgoose>going to diff to understand the recent changes, but where should I be looking?
14:52<TheWildgoose>(based on changes made recently...)
14:52<thor_>I'm just looking at some diff's to see if anything is obvious
14:52<TheWildgoose>My previous working version was like 3 weeks old
14:53<thor_>that looks like the problem ....
14:54<TheWildgoose>where are you looking?
14:54<thor_>flacdecoder.cpp, line 577
14:55<TheWildgoose>ok, I'm there
14:56<krenin>can mplayer play rtjpeg files?
14:56<thor_>try dropping in an older flacdecoder.cpp (3 weeks should be fine) and see if that solves it
14:56<thor_>if so, we need to have it return tags properly but not segfault on long ones.
14:58<TheWildgoose>hmm, yeah, I can see it has changed a lot, but I can't see obviously why it now fails...
14:58<thor_>bit garbly
14:59<thor_>unlike most patches, I didn't walk through it by hand
14:59<TheWildgoose>can't see why it segfaults on the old code actually? Seems to only memcopy up to 512 chars? Is it because it needs one extra char for terminator?
14:59<thor_>char []'s are just silly
15:00<thor_>you may have to fiddle with it a bit (some cout <<'s) .... I don't have any flacs to test on .... suppose I could create some if you don't have the time
15:01<krenin>how do you minimize mythtv in the gui
15:01<sfr>krenin: you don't. you could ask it to run in a window though.
15:01<krenin>so can't switch back to my Xfce desktop?
15:02<krenin>...while its recording
15:02<sfr>krenin: the recording is done by mythbackend. you can quit mythfrontend without fear.
15:02<krenin>oh ok
15:02<krenin>how do you quit mythfrontend?
15:03<thor_>how did you tell it to quit?
15:03<thor_>(that was a hint)
15:04<krenin>ok gotcha
15:04<krenin>it's only taking about 20% cpu for mythbackend to record
15:04-!-jkolb [] has quit []
15:05<krenin>this thing is the coolest thing since sliced bread
15:05-!-sc00p__ is now known as sc00p
15:05<TheWildgoose>thor Does it appear to you that this flac code is doing a case sensitive compare?
15:07<TheWildgoose>it appears to call getComment with lowercase, but then the flac files have comments in uppercase
15:08<TheWildgoose>thor ? Are you there?
15:09<thor_>yup, I'm just slow
15:09<TheWildgoose>Yep, definitely appears to be doing a case sensitive search
15:09<TheWildgoose>Does that look right to you?
15:09<TheWildgoose>How would you prefer I fix it?
15:09<TheWildgoose>copy to qstrings?
15:10<TheWildgoose>or is the strcmp fixable?
15:10<thor_>I'd just use QStrings, and judicous .upper() .lower()
15:10<thor_>but it's not my code
15:10<TheWildgoose>oh, sorry. Thought this was you...
15:11<TheWildgoose>or do you mean the update wasn't your code?
15:11<TheWildgoose>Can strncmp be made to do case insensitive match? (I think not)
15:11<TheWildgoose>but I am very rusty on c++ these days
15:11<thor_>original flacdecoder has not changed much since it was first borrowed
15:12<TheWildgoose>Yeah, I understand
15:12<thor_>the method wants a QString back
15:12<TheWildgoose>Tagging is something that got bolted on later - yes?
15:12<TheWildgoose>OK, I will try something and then can I send it to you for commit?
15:13<thor_>I'd just convert the char* label to a QString and work from there (the trolls do strings well)
15:13<krenin>switched X resolution to 640x480 and it runs much smoother...
15:13<thor_>Tagging was always there in the decoders
15:14<thor_>anything you can come up with would be great .... send it to the -dev list so the guy who sent the patch can see if it breaks the thing he was fixing in the first place
15:15<TheWildgoose>Oh really! Hmm. I want to retag some of my files actually. I wonder if the setComment can be used to write something... Seems that I need a "clearComments" first though, otherwise it will just keep appending stuff
15:15<TheWildgoose>perhaps I could adjust the setComment to check for existing comments and overwrite...
15:15<thor_>easyTag is your friend
15:15<TheWildgoose> if they have the same "key="
15:16<TheWildgoose>I did try it and it didn't seem to support tagging the way we are doing it here...
15:16<TheWildgoose>...for flac files that is
15:16<thor_>well .. it works amazingly well on oggs and mp3's, not sure about flacs
15:16<TheWildgoose>yeah, it claims to do flacs
15:16<TheWildgoose>but it certainly doesn't seem to read the comments fields that we have set in myth
15:17<thor_>hey, while you're at it .... absolutely no-one would complain if you came up with a comment editing dialog =)
15:19<TheWildgoose>Yeah, nice to have!
15:19<TheWildgoose>now sorry, you are going to have to help me a bit more...
15:19<TheWildgoose>How do I get this: (const char *)entry->entry;
15:19<TheWildgoose>into a QString
15:19<TheWildgoose>without doing the memcpy that was being used before
15:20<TheWildgoose>does qstring have any fancy methods?
15:21<thor_>QString a_string = (const char *)entry->entry; cout << "a_string is \"" << a_string << "\"" << endl;
15:21<TheWildgoose>Hmm, really. So simple. Rather clever of QString though
15:22<TheWildgoose>So I wonder why the original author used strncpy into a char[] and then assigned to a QString?
15:22<thor_>QString & operator= ( const char * str ) <-- from the Qt docs
15:24<krenin>i bet there are a lot of people making entertainment systems around MythTV
15:24<krenin>better than a tivo
15:24<TheWildgoose>Hmm, sorry probably dumb question, but isn't char[] just the same as char* anyway? (It's been about 7 ish years...)
15:24<thor_>more or less ....
15:25<TheWildgoose>...(I get to write a lot of Visual Basic these days... Perl in my free time...)
15:25<thor_>but both are to be avoided.
15:25<TheWildgoose>(Did you mean VB and perl... ;-)
15:25<thor_>no, char[] and char*
15:26<thor_>although seeing people "code" in VB does make me fairly nauseous
15:27<TheWildgoose>Yep. I'm probably one of them as well...
15:27<TheWildgoose>It's far too easy to "cheat" and write something nasty...
15:27<TheWildgoose>However, the flip side is that I have written more code on my own for a major app than many teams might manage in some other language
15:28<TheWildgoose>There is a health debate about .Net vs Java in our office at the mo....
15:28<thor_>far too easy to think you need $500.00 worth of microsoft development infrastructure to make a compute do something for you
15:28<TheWildgoose>I have a hunch actually that .Net might actually be the real killer feature for linux... Just need Mono to mature
15:28<TheWildgoose>I think Visual Studio is more like $2000
15:29<TheWildgoose>Its about \xA31200
15:29<thor_>more than gcc
15:29<TheWildgoose>However, we are spending about \xA31million on a custom piece of accounting software right now
15:29<TheWildgoose>So 10 licences for VB don't really show up
15:29<TheWildgoose>And that is how most businesses think
15:30<thor_>working on a bit of "accounting software" myself
15:30<TheWildgoose>In the UK contractors earn perhaps \xA3500 per day, so if MS software makes them code the solution 2 days quicker then it pays for itself. The stuff about long term maintainability is somewhat overlooked
15:31<TheWildgoose>Yeah, was thinking about putting in a tender myself!
15:31<TheWildgoose>But I think that is the point. My company is quite small, 140 ish people
15:31<TheWildgoose>but we are spending perhaps \xA31-2m per year on IT.
15:31<TheWildgoose>So "fastest" is what counts (notice how I avoid saying "best")
15:32<TheWildgoose>Anyway, personally I am a linux fan, and most of my machines run that
15:32<thor_>I don't care about the financial arguments, VB generates hoards of people who write "code" but who never realize the (fairly) simple relationship between hardware, operating systems, source code, and applications.
15:32<TheWildgoose>I miss a few windows apps though such as Photoshop... (It's much better than the Gimp for lots of things)
15:33<TheWildgoose>Yep, agreed
15:33<thor_>'cause their way up in some rarefied Microsoft sandbox
15:33<TheWildgoose>I think what many linux people miss though is that they are a) better programmers, b) more efficient given a certain task c) perhaps cheaper, d) probably cleverer.... but....
15:33<thor_>connecting little coloured lines together
15:34<TheWildgoose>My company doesn't care about the technical solution. The boss just wants to be use the bit of software that he has just spec'd out to me....
15:34<TheWildgoose>If I can write it in xxxxx then fine. But get it done by the end of the day...
15:35<TheWildgoose>Also once you get on the MS bandwagon they use lots of lockin to keep you there
15:35<thor_>and bully for them
15:35<TheWildgoose>Personally I rather like a lot of MS software. I see no reason to bash it. The main thing which annoys me is the mis-use of that power
15:35<TheWildgoose>Anyway. Doesn't get us anywhere....
15:36<thor_>nope ;-)
15:36<TheWildgoose>OK, going to code this up quickly
15:36<TheWildgoose>Friends is on....
15:36<TheWildgoose>See ya. And thanks for pointing out this problem. Will get a fix in
15:37<TheWildgoose>By the way. Anyone from the UK here? I have been playing with integrating the tv_grab_uk_rt grabber. Means you can get 14 days of programs instead of 4 days
15:37<DogBoy>TheWildgoose, you can run photoshop on linux
15:37<TheWildgoose>...slow as hell though
15:37<TheWildgoose>Yes, I think a few people paid some money to get the crossover office to support it?
15:38<DogBoy>with just wine
15:38<TheWildgoose>Photoshop 8 is out though, and it has some very tasty stuff...! 16 bit all the way
15:38<TheWildgoose>RAW importer, etc
15:38<TheWildgoose>very handy
15:41<DogBoy>I kinda like fireworks myself
16:03-!-Justin_ [~justin@] has quit [Read error: 104 (Connection reset by peer)]
16:04-!-Justin_ [] has joined #mythtv
16:20-!-sc00p [] has quit ["Client exiting"]
16:20-!-atomix [] has quit ["ChatZilla 0.8.31 [Mozilla rv:1.4/20030903]"]
16:22-!-krenin [~krenin@] has quit [Read error: 110 (Connection timed out)]
16:30-!-overridex [] has joined #mythtv
16:31-!-SarahEmm [] has joined #mythtv
16:41-!-FryGuy [] has quit [Read error: 104 (Connection reset by peer)]
16:42* o_cee is away: sleeping
16:47-!-FryGuy [] has joined #mythtv
16:48-!-mecraw [~mecraw@] has joined #mythtv
17:00<SarahEmm>okay, i have btaudio all set up and working, but my original issue is still there, video is slow and jerky and i get constant 'delaying to next trigger' messages from mythfrontend, and some 'strange error flushing buffer' messages in mythbackend.log
17:01<sfr>ignore the strange error thing, that's always there.
17:02<SarahEmm>any ideas where to debug this?
17:03<sfr>except decreasing the resolution, no.
17:04-!-FryGuy [] has quit [Read error: 104 (Connection reset by peer)]
17:04<SarahEmm>hrm, i guess i can try that
17:05<SarahEmm>i'm running 480x480 now, tried both RT-JPEG and MPEG4
17:05<sfr>SarahEmm: what CPU do you have?
17:05<SarahEmm>celeron 1.7ghz
17:06-!-FryGuy [] has joined #mythtv
17:06<sfr>SarahEmm: iirc my 2x800MHz PIII box didn't handle that well. But i switched to a h/w base mjpeg card long ago.
17:07<SarahEmm>so i guess i'll try lowering it..
17:08<sfr>SarahEmm: but then, it could be something else. any messages on the backend?
17:09<sfr>SarahEmm: and how loaded is it during encoding.
17:09<SarahEmm>the only message on the backend is the 'delaying to next trigger' messages
17:09<SarahEmm>let me check how loaded it is..
17:10-!-accurate [] has quit ["Leaving"]
17:11<sfr>SarahEmm: err, that's a message from the frontend if i
17:11<sfr>if i'm not mistaken.
17:12<SarahEmm>err, yeah, sorry
17:12<SarahEmm>the only one from the backend is the strange error flushing buffer, which is aparantly normal
17:14<Captain_Murdoch_>sarahemm, any messages about XV on the frontend?
17:14-!-mecraw_ [~mecraw@] has joined #mythtv
17:15<SarahEmm>and direct rendering is on...
17:15<SarahEmm>hmm, is there a way to make sure it's using xv?
17:17<sfr>SarahEmm: there should be a message like: using XV port 55
17:17<sfr>SarahEmm: or simply run 'xvinfo' on the frontend
17:18<SarahEmm>what am i looking for in the xvinfo output?
17:18<sfr>something similar to: X-Video Extension version 2.2
17:18<sfr>screen #0
17:18<sfr> Adaptor #0: "Matrox G-Series Backend Scaler"
17:18<sfr> number of ports: 1
17:18<sfr> port base: 55
17:18<sfr> operations supported: PutImage
17:18<sfr> supported visuals:
17:18<SarahEmm>ahh, yeah, 'using xv port 60'
17:18<sfr> depth 16, visualID 0x23
17:18<SarahEmm>so it looks like XV is ok
17:19<SarahEmm>Cpu(s): 34.3% user, 3.6% system, 0.0% nice, 62.1% idle
17:19<SarahEmm>that's while i'm watching live TV
17:20<SarahEmm>so it doesn't look CPU-bound...
17:21-!-mecraw [~mecraw@] has quit [Read error: 60 (Operation timed out)]
17:27-!-lmatter [] has quit ["Client Exiting"]
17:29<kvandivo>tmk: you ever end up with one of the m179s?
17:30<TheWildgoose>thor - hi you around still?
17:31<TheWildgoose>I think I understand why the old code segfaulted now
17:31<TheWildgoose>and also why the new code doesnt' work...
17:33<TheWildgoose>Anyone here any good with QString?
17:34<sfr>does knowing what it is count?
17:38<kvandivo>what more than could you possibly want? ;)
17:38<sfr>working code?
17:40<kvandivo>ahh.. for that i would refer one to the mythtv source tree.. literally hundreds of instances of QString in action. (probably thousands)
17:41<TheWildgoose>kvandivo thanks
17:41<TheWildgoose>just serching for that
17:41<TheWildgoose>I have a char* string, but it isn't null terminated
17:41<TheWildgoose>need to be able to specify a length
17:42<TheWildgoose>it appears that there IS a constructor though which takes length as an arg
17:42<kvandivo>well.. if it isn't null terminated, then its not really a string, according to the standard vernacular.. a character array, I would buy..
17:42<TheWildgoose>ahh, rats
17:42<TheWildgoose>actually looking closer, no it doesn't
17:42<TheWildgoose>yeah, whatever,
17:43<TheWildgoose>I still need to read it into a qtstring though...
17:43<TheWildgoose>I do agree by the way
17:43<TheWildgoose>Any suggestions?
17:43<kvandivo>do you _know_ the length?
17:43<kvandivo>(I would hope so)
17:44<TheWildgoose>assume char* unterminated, and int length
17:44<kvandivo>and you can't make work?
17:45<TheWildgoose>hmm, but do I have a qchar* ?
17:45<TheWildgoose>I think not...?
17:47<kvandivo>i suspect you could, but at any rate, if speed isn't an issue you could always allocate a char * for length+1 and add the 0
17:48<TheWildgoose>i think that is best. Hang on let me try it
17:49<kvandivo>why in the world isn't your first string null terminated?
17:51-!-m0j0 [~m0j0@] has quit ["Client exiting"]
17:53<TheWildgoose>it comes from flacdecoder.cpp
17:53<kvandivo>ahh.. that..
17:54<TheWildgoose>stored as a char* plus a length column...
17:54-!-mattfelsen [] has joined #mythtv
17:57<TheWildgoose>dumb question. Arrays in c++ are zero based yes?
17:58<TheWildgoose>ie char* p = new[123]; p[0] = something;
17:59<TheWildgoose>I use so many prog languages these days I find it hard to declare an int sometimes....
17:59<sfr>isn't it the same in any (sane) language?
18:01<TheWildgoose>well sql uses declare
18:01<TheWildgoose>vb uses dim
18:01<TheWildgoose>c/c++ use int blah
18:01<TheWildgoose>pascal uses blah int
18:02<TheWildgoose>perl... well perl just doesn't
18:02<TheWildgoose>pl/1 I'm really rusty on...
18:02<TheWildgoose>just syntax though. basically the same thing
18:02<TheWildgoose>until you get onto the logic and functional languages... Miranda/prolog, etc
18:03<TheWildgoose>yeah cobol is pretty verbose...!
18:04<sfr>that is valid COBOL? me shivers
18:04<kvandivo>no.. i'm sure its not that bad.. it's just amusing to make fun of it
18:04<SarahEmm>cobol is ridiculously verbose heh
18:05<kvandivo>(this said from a guy that could have made 4 times as much as he was making back around y2k if he had wanted to code in cobol)
18:05<sfr>heh. you didn't?
18:06<kvandivo>one of the scariest ads I think i've ever seen was Microsoft selling a package called Visual Object Oriented Cobol.
18:12-!-Captain_Murdoch_ [] has quit []
18:12<TheWildgoose>yeah, Cobol ++
18:12<TheWildgoose>I think the MS compiler is still not a bad cobol compiler though?
18:13<TheWildgoose>Actually isn't it an intel compiler?
18:13<TheWildgoose>I think intel are still doing optimised compilers for c++ and cobol...
18:13<tmk>kvandivo: nope
18:14<TheWildgoose>I read VB overtook cobol for most lines of code written a few years back.... scary..
18:14<kvandivo>tmk: hmmm..
18:14<TheWildgoose>yeah, I'm sure the compiler is from intel
18:14<tmk>kvandivo: a couple of people have offered to donate one though
18:15<kvandivo>ok.. i ordered one last night. i figured i could throw you one for a while if need be. (I'd want it back, though. :)
18:15<tmk>yeah that's ine
18:15<tmk>i have 2 cards already anyways
18:15-!-[-ripp-] [] has joined #mythtv
18:15<kvandivo>i've only got a 250 for right now
18:15-!-[-ripp-] [] has quit [Client Quit]
18:16<kvandivo>well.. i've gotta get for now.. later, all
18:16-!-kvandivo is now known as kvandivo-away
18:21-!-warlord is now known as warlord-afk
18:24<yebyen>anyone use multiple machines, and the debian packages?
18:28<sfr>two machines but cvs
18:35-!-mattfelsen [] has quit [Remote closed the connection]
18:49-!-warthawg [] has joined #mythtv
18:56<thor_>TheWildgoose, sorry had to step away for a while .... you figure it out ?
18:57-!-ChaosExiguum [] has joined #mythtv
19:00<yebyen>sfr: do you need to edit the source, and change some "" backend address to the actual address of the backend?
19:01<tmk>did you read the documentation yebyen
19:01<sfr>yebyen: no
19:05-!-steelep [~signwatch@] has quit ["me is bugging out"]
19:14<warthawg>can i use mythtv to watch live OTA HDTV with a pcHDTV card?
19:18-!-mecraw_ [~mecraw@] has quit ["Trillian ("]
19:19<tmk>warthawg: probably
19:29-!-ChaosExiguum [] has quit [Read error: 110 (Connection timed out)]
19:41-!-josephk [] has joined #mythtv
19:43<josephk> 350 came:D
19:46<warthawg>is it good to have a 350?
19:46<SarahEmm>well, they do hardware encoding *and* decoding
19:46<SarahEmm>so i think yes :)
19:47<josephk>amazon got it to me in 2 days..standard shipping...weird, but I'm not complainging:)
19:47<josephk>or spelling
19:48<tmk>good luck to you
19:48<tmk>don't forget to extract your firmware
19:48<josephk>yeah I cvs look good?
19:48<tmk>decoder-alpha is the way to go
19:48<tmk>let me update that real quick
19:48<tmk>one sec
19:49<josephk>I should have it working by
19:50<tmk>ok, grab decoder-alpha.tar.gz
19:50-!-poo [] has quit ["Download Gaim:"]
19:50<tmk>linked off
19:50<josephk>you da man
19:51<tmk>head over to #ivtv-dev
19:51<tmk>don't wanna flood #mythtv
20:01<TheWildgoose>thor yes figured it out thanks. The reason for the old (ugly) code is simply that the comments stuff is not null terminated. Hence copying it to a buffer and then to a qstring
20:01-!-hfb [] has quit ["Client exiting"]
20:01<TheWildgoose>I have reverted to the original code and then fixed that for long strings by dynamically allocating the interim buffer
20:02<TheWildgoose>uses something like char* buffer = new char[element->len + 1]
20:02<TheWildgoose>instead of char* buffer = char[512]
20:02<TheWildgoose>patch sent to the list, but hopefully this is better than the orig patch.
20:02<TheWildgoose>over and out...
20:03-!-poo [] has joined #MythTV
20:16<thor_>thanks ... I'll give the original guy a day or so and then apply it
20:17<thor_>ah, he's already replied ...
20:17<thor_>I'll put it in
20:17-!-SarahEmm [] has left #mythtv []
20:38-!-treke [~ggilbert@] has joined #mythtv
20:47-!-treke [~ggilbert@] has left #mythtv []
20:51-!-sfr [] has quit ["Client exiting"]
20:51-!-tmk [] has quit ["Client exiting"]
20:59-!-warthawg [] has quit ["Client exiting"]
21:32-!-Timon [] has joined #mythtv
21:38-!-tmk [] has joined #mythtv
21:43<overridex>hey guys... quick question about the recording duplicates stuff
21:43<overridex>if i delete a recording, is info about it still kept in the db, or will it then record that episode again?
21:49<Timon>This blows goats, bought the awesome looking ATC-610 case, and the ABit NF-7, didn't realize the case was micro-atx, the mobo isn't.
21:57-!-kslater [] has joined #mythtv
22:03-!-warlord-afk is now known as warlord
22:03<warlord>Input #0, mpeg, from '/var/mythtv/video/1004_20031003200000_20031003210000.nuv':
22:03<warlord> Stream #0.0: Video: mpeg2video, 720x480, 29.97 fps, 16000 kb/s
22:03<warlord> Stream #0.1: Audio: mp2, 48000 Hz, stereo, 384 kb/s
22:03<warlord>Audio has changed: 8000hz mono
22:03<warlord>Audio has changed: 48000hz stereo
22:06-!-[DJ]HaCK [] has joined #mythtv
22:07-!-tmk [] has quit []
22:08-!-tdb30_ [] has quit [Read error: 110 (Connection timed out)]
22:42-!-Justin___ [] has joined #mythtv
22:42-!-Justin_ [] has quit [Read error: 54 (Connection reset by peer)]
22:43-!-Justin___ is now known as Justin_
23:03-!-bobnvic [] has joined #mythtv
23:14<[DJ]HaCK>anyone using mythgame ? I'm trying to use zsnes so I made a small shell script , all it does is exec /usr/bin/zsnes $6 but zsnes complain .. error opening file
23:15<[DJ]HaCK>if I /usr/bin/mythgamelaunch.snes /path/to/my\ rom\ name\ with\ spaces.smc it works
23:15<[DJ]HaCK>but when mythfrontend /usr/bin/mythgamelaunch.snes -st -alt -ne -soundquality 4 "/path/to/my rom name with spaces.smc" zsnes get the wrong filename
23:17<overridex>i've never used it but you could try setting exec /usr/bin/zsnes $6 to exec /usr/bin/zsnes \"$6\" instead
23:18<[DJ]HaCK>yay! that's what I tried earlier , but you made me think exec /usr/bin/zsnes "$6" worked :)
23:18<bobnvic>can anyone help me configure mythweb to use coverfiles for the videos? what should I put in conf.php video_img_path to refer to a location NOT on my webserver (somewhere in my /home directory)?
23:21<[DJ]HaCK>no idea , I'm not there yet , mythweb is next after dvd and games
23:38<bobnvic>HA! Figured it out. Just need a little symlink magic.
23:43-!-dopez [] has quit ["nn"]
23:45-!-warthawg [] has joined #mythtv
23:45<warthawg>i am using a pcHDTV card, it uses /dev/dtv not /dev/video0, how do i make that work with mythtv?
23:46<mikegrb>the easiest way would be symlink /dev/video0 to /dev/dtv
23:47<warthawg>k thanks
23:50<warthawg>now i can connect video source and connection, thanks again
23:51<mikegrb>no prob
23:53-!-warthawg [] has left #mythtv ["Client exiting"]
23:54<[DJ]HaCK>mmm pchdtv card , he gotta have a nice signal :)
---Logclosed Sat Oct 04 00:00:43 2003