Back to Home / #mythtv / 2002 / 10 / Prev Day | Next Day
#mythtv IRC Logs for 2002-10-28

00:05<Chutt>libavcodec just got a wma decoder
00:07-!-marick [] has quit [Remote closed the connection]
00:15<vektor>kick ass
00:57-!-rcaskey_afk [] has quit [Read error: 110 (Connection timed out)]
01:44-!-Soopaman [] has quit [Connection timed out]
06:43-!-oztiks [] has joined #Mythtv
06:43-!-skitzo [] has quit [Read error: 54 (Connection reset by peer)]
09:44-!-Universe [] has joined #mythtv
10:00-!-Universe [] has quit [Read error: 104 (Connection reset by peer)]
10:11-!-tss [] has joined #mythtv
10:15-!-Tuscany [~username@] has joined #mythtv
10:16<tss>how's it going?
10:29<mdz_>Chutt: when you said the mplayer output and mythtv output looked identical...were you deinterlacing in mythtv, mplayer, neither or both?
10:31-!-mdz [] has quit []
10:31-!-mdz [] has joined #mythtv
10:32<mdz_>were you using them on a TV or on a monitor?
10:32<mdz_>when I playback the recordings in mplayer without deinterlacing, they look liney
10:32<mdz_>horizontal lines appearing in places with motion
10:33<Chutt>that's how it should be
10:33<mdz_>right, I have seen that with other recordings too
10:33<mdz_>I don't see it on the TV though, which makes sense I suppose
10:37<mdz_>I need a video camera so that I can record what I am seeing on the TV so you can see it :-)
10:37<mdz_>it looks sort of...dithered
10:37<mdz_>like MPEG does when it is at a low bitrate
10:38<Chutt>tried adjusting brightness/contrast on your tv?
10:38<mdz_>not yet, but I would not enjoy having to adjust every time I switch between mythtv and regular TV
10:38<Chutt>of course
10:39<mdz_>is it possible to set those things on the capture card via a v4l API, or does it have to be done post-capture?
10:40<Chutt>yeah, that's part of the v4l api
10:40<Chutt>i just don't have any of that hooked up
10:40<mdz_>I fixed up my mplayer patch to decode the X frame and set things up from it
10:41<mdz_>it could use some testing with different parameters; all my recordings are the same
10:41<Chutt>all mine are 640x480/mpeg4/mp3
10:41<mdz_>well great
10:41<mdz_>anybody out there NOT using exactly the same settings?
10:41<Chutt>doesn't help? =)
10:44<Chutt>what you can do fairly easily
10:44<Chutt>is just change things, watch tv for a few seconds
10:44<Chutt>and just try to play the ringbuf.nuv file
10:45<mdz_>it always starts from the beginning of ringbuf at the start of live viewing?
10:56<mdz_>are the nuppelvideo guys interested in adding the stuff that you need to the format spec?
10:56<Chutt>i need to email him again
10:56<Chutt>i did back in may, no response
10:56<mdz_>the web page has said that 0.6 will be out in 3 weeks for some time now
10:57<mdz_>the web server says it was last modified Oct 16th though
10:58<mdz_>if codec parameters and a seek table were added to the format, that would be handy indeed
10:59<Chutt>i'm probably going to add a seek table sometime this week
11:00<mdz_>my NFS setup will thank you profusely
11:01<Chutt>it will only really help with seeking forward, of course
11:01<Chutt>shall i add an option to only seek to the nearest keyframe while i'm at it?
11:01<mdz_>yeah, that would be great too
11:02<mdz_>then it wouldn't be too hard to implement rewind and fast forward
11:02<mdz_>but with fast seeking maybe I wouldn't miss them
11:02<Chutt>what, real rewind and fast forward?
11:03<mdz_>continuous, variable-speed
11:03<mdz_>rather than just skips
11:03<Chutt>it's a little harder than you'd think =)
11:03<Chutt>at least with the current setup
11:03<mdz_>probably :-)
11:03<mdz_>a while back, I got a crash most of the way through a multi-hour recording
11:03<mdz_>it would have taken like 15 minutes of skipping to get back to where it stopped
11:04<mdz_>so I changed the skip time to skip to the point I wanted in one skip
11:04<Chutt>have you tried the sticky seek keys?
11:04<mdz_>couldn't skip for the rest of the film though :-)
11:04<mdz_>I have not
11:04<mdz_>I was not aware of their existence
11:04<mdz_>does it make it toggle a loop?
11:04<mdz_>skip -> skip until stopped?
11:05<mdz_>displaying a single frame for each skip, or what?
11:05<Chutt>pretty much
11:05<Chutt>really depends on the speed of the box how many frames'll get displayed
11:06<mdz_>I see
11:06<mdz_>that would be no problem here, of course
11:06<Chutt>it just sends seek messages continually
11:06<mdz_>because each skip takes 5 seconds
11:06<Chutt>going to nearest keyframe only, and with a seek table
11:06<Chutt>that should help you considerably
11:07<mdz_>should make it as good as seeking AVI in mplayer
11:07<mdz_>which is plenty
11:07<mdz_>I have tried that, and it is quite usable in my setup
11:11<mdz_>if Mr. 162112 doesn't respond soon, he may find his ITP hijacked
11:12<Chutt>go for it
11:14-!-Tuscany [] has quit [Read error: 104 (Connection reset by peer)]
11:14<Chutt>heh, why bother waiting? :p
11:16<mdz_>well, compared to, say, #68256, it would seem premature
11:16<mdz_> 1982 days in preparation.
11:17<mdz_>yeah, that one's going to happen
11:17<Chutt>yeah, but it shouldn'tve taken all that long to package
11:18<mdz_>no modern program could take 5 years to package :-)
11:19<mdz_>the guy who ITP'd xmltv pasted a control file snipped that made it look like he was already done
11:19<Chutt>which is why it should be in the distro already
11:19<mdz_>I would really, really rather not maintain any more Perl code if I can help it
11:19<mdz_>especially not the kind of perl code that lurks in dark corners like xmlt
11:20<mdz_>but I may have to make a sacrifice in the name of mythtv
11:20* mdz_stares dramatically into the distance
12:11-!-tss [] has quit ["bbl"]
12:15-!-oztiks [] has quit []
12:15-!-nyquiljer [] has quit []
12:15-!-vektor [] has quit []
12:15-!-Chutt [] has quit []
12:15-!-oztiks [] has joined #mythtv
12:15-!-nyquiljer [jer@] has joined #mythtv
12:15-!-vektor [] has joined #mythtv
12:15-!-Chutt [] has joined #mythtv
12:26<Chutt>wonder how hard it'd be to whip up a nice tv-based display thingie for ebooks
12:50<mdz_>ebooks are awful
12:51<mdz_>and I bet the ebook people would make it as difficult as possible for you to whip up a nice display thingie
12:51<Chutt>the latest david weber book (published by baen)
12:51<Chutt>came with a bound-in cdrom
12:52<Chutt>with the full text of 20-some other books on it
12:52<Chutt>in 4 or 5 formats, all completely unlocked
12:52<Chutt>one of the formats being just plain old .rtf
12:53<Chutt>and most (all?) of baen's new books have been available like that for a fairly small monthly fee
12:55<Chutt>it's something like $15 a month, and you get all the books published that they published new that month
12:58-!-Universe [] has joined #mythtv
13:02<Chutt>qt 3.0.6 release a few days ago
13:06<Chutt>hrm, nothing interesting fixed, though
13:08<Universe>so... not worth updating
13:08<Chutt>not worth going out of your way for, at least
13:54-!-Tuscany0_ [~username@] has joined #mythtv
14:20-!-TechJosh [] has joined #mythtv
14:28<mdz_>ah, I wouldn't consider those ebooks
14:28<mdz_>ebook typically means evil copy protection crap
14:28<mdz_>which cripples legitimate use
14:31<Chutt>so should i take that guy up on his offer of a g200tv in exchange for supporting it?
14:52<Universe>he said he wanted it back after you get it working
14:52<Chutt>no, this is another guy
14:52<mdz_>what does he want you to support?
14:52<mdz_>capture from it?
14:53<Universe>go for it...
14:53<mdz_>I don't think the guys have gotten that to work
14:55<mdz_>they have some talk about v4l though
14:55<Chutt>well, he said the mjpeg capture works
14:56<mdz_>yeah, I am reading this now:
14:56<mdz_>whoa, matrox released the specs for the tvout chip??
14:56<mdz_>I might need to get myself one of those
14:57<mdz_>ATI refuses to do that
14:57<Chutt>ah well, i just emailed him and said i'd do what i could to get it supported
14:58<vektor>mdz: matrox didn't release specs for the tvout chip
14:58<vektor>unless they just did :)
14:58<TechJosh>Hi, I'm new to linux and havent tried installing MythTV yet, but I wanted to say that I am following this project with great interest... from the screenshots I've seen, this is the best PVR software available for linux.
15:01<vektor>mdz: There's nothing in that README that seems to indicate that they released tvout specs.
15:03<mdz_>vektor: under Extra
15:04<mdz_>it seems to imply that
15:04<mdz_>or maybe I'm misreading it
15:04<mdz_>it looked like question/answer
15:04<vektor>you're misreading it
15:05<vektor>there are specs for the G400 and i can dcc them to you
15:05<mdz_>but maybe he is just asking "Anyone know if the Maven TVOut spec is available anywhere?" and then going on about unrelated specs
15:05<vektor>there are no specs for the tv out chip
15:05<vektor>he's asking
15:05<mdz_>ok, I feel better now :-)
15:05<vektor>it's really shitty
15:05<vektor>see this:
15:05<-- TechJoshhas quit ()
15:05<mdz_>is it the same macrovision bullshit?
15:07<vektor>i think so
15:07<vektor>the driver in directfb kicks ass
15:07<vektor>and we can port it to X if you want
15:07<vektor>but it's only for the G400
15:07<vektor>not for G450/G550
15:08<vektor>however, we _may_ be able to get a driver for that
15:08<vektor>like, don't tell anyone, but someone has been passing around source for the HAL lib.
15:08<vektor>so that may be the key to get it all working.
15:08<vektor>matrox's binary driver
15:08<vektor>but you didn't hear that from me
15:09<mdz_>the GATOS stuff uses VBE to enable TV out currently
15:09<mdz_>but there is a guy who is reverse-engineering the registers to set it up right
15:10<mdz_>I haven't been following his work because the VESA stuff is good enough for me to worry about other parts of my system for a while
15:10<vektor>i have high quality requirements for my tv output
15:10<vektor>so i'm really really happy with the situation with the matrox card
15:10<vektor>since i have like perfect quality now
15:10<vektor>reverse engineering is difficult
15:10<vektor>i have the specs for some of the tvout chips used on nvidia OEM cards
15:11<vektor>but some of them are total crap
15:11<vektor>no interrupt per field!
15:11<vektor>so it's useless if you want to get the quality of like a hardware DVD player
15:11<vektor>you'll never make it
15:11<vektor>like that's kind of my quality requirement
15:11<vektor>if the TV out looks worse than my hardware dvd player, fuck it, i'll just go buy a player :(
15:11<vektor>but i really want to do my OSD and stuff, but i don't want to give up quality because of it
15:12<Chutt>that's the main reason i don't have a button call some dvd player program
15:13<vektor>you should try my dvd player
15:13<vektor>it kicks if your output is a CRT ;)
15:16-!-TechJosh [] has joined #mythtv
15:17<Chutt>techjosh, it's probably a bad idea to irc as root
15:17<TechJosh>yeah, i know
15:18<TechJosh>im just lazy
15:20<mdz_>vektor: how is it doing tv-out now?
15:20<mdz_>Chutt: what reason is that? quality? or OSD?
15:21<vektor>mdz: How is what doing tv-out ?
15:21<vektor>mdz: How is DirectFB doing it, or the HAL lib?
15:22<Chutt>mdz, my hardware dvd player's much bettter than any software player in linux that i've seen
15:22<Chutt>so, i can't be bothered to add a simple little button to mythtv for it just yet =)
15:23<vektor>Chutt: exactly.
15:23<vektor>well if your output is a TV I can believe that.
15:23<Chutt>even to a monitor
15:23<vektor>movietime (my DVD player) blows away my hardware DVD player if my 3:2 pulldown is working.
15:23<vektor>like, 24fps progressive, correctly amortized, is stunning.
15:24<vektor>all the windows dvd players seem to take the safe way out though and deinterlace to 29.97fps anyhting that is interlaced.
15:24<mdz_>vektor: directfb
15:25<vektor>mdz: well it just has its own driver code, plus it uses a small patch to the matroxfb kernel module to get an interrupt for the second crtc
15:25<vektor>like same as X would do, minus the interrupt code, which we can do our own kernel module for.
15:26<vektor>the other thing about directfb though is that your application calls the driver code directly, so it's completely unsafe but you can at least get a pointer to video memory.
15:26<vektor>and that really helps your speed
15:26<vektor>fucking X :)
15:28<mdz_>I'm not worried about anything better than TV quality, as my only video output is a TV
15:28<mdz_>I don't have a hardware DVD player to compare to, but DVD playback looks damn good
15:29<vektor>using what player?
15:29<vektor>i find it difficult to demonstrate what i mean by good or bad
15:29<vektor>i mean if you were here right now i could show you exactly what i mean
15:30<vektor>but trying to guess what it looks like for you is impossible
15:30<mdz_>likewise for you
15:30<vektor>so you're using tvout with your ati card?
15:30<mdz_>send me a VHS cassette demonstrating the difference :-P
15:30<vektor>and mplayer is outputting to that?
15:30<vektor>and you're playing DVDs?
15:30<vektor>if you're playing a dvd it's likely encoded at 24fps progressive
15:30<vektor>since i assume you're in NTSC-land
15:30<mdz_>I am
15:31<vektor>so then you're not likely to see much judder
15:31<vektor>the only issue is resolution which you probably also won't notice
15:31<vektor>so yeah, for most DVDs the quality in that setup should be 'ok' so long as the tv encoder doesn't tear and flips on 59.94fps boundaries
15:31<vektor>i'm talking about like quality on more difficult DVDs
15:32<vektor>like anything that's video-source
15:32<vektor>so, my Bjork music video DVD, or my copy of Transformers: The Movie
15:32<vektor>or even any of the documentaries on all the DVDs I have with 'making-of' specials.
15:33<vektor>if you play anything like that in your setup, you'll either see a) jerky motion, b) poor vertical resolution quality, or c) jumpy crap.
15:33<mdz_>If I want to watch videos, I have a VCR
15:33<vektor>similarily, if you have like a separate capture card from your tv output card and you try to do pass-thru it will also suck
15:34<Chutt>i wonder
15:34<vektor>Chutt: do you concur with my assessment?
15:34<Chutt>getting xmltv installed is the biggest hurdle to all this crap
15:34<mdz_>if I just want to pass-through, the ATI card does that nicely via XV and I don't need to mess with the capture card
15:34<Chutt>now, you _can_ compile perl code, right?
15:34<mdz_>Chutt: you _can_, but working is another story
15:34<mdz_>I don't know how it handles modules
15:35<Chutt>well, the windows xmltv stuff is distributed as an exe
15:35<mdz_>man perlcc
15:35<mdz_> The code generated in this way is not guaranteed to work. The whole
15:35<mdz_> codegen suite ("perlcc" included) should be considered very experimen-
15:35<mdz_> tal. Use for production purposes is strongly discouraged.
15:35<Chutt>yeah, but if i made it available, etc.
15:35<Chutt>say 'use only if you can't install xmltv normally'
15:36<vektor>so xmltv just sucks the bag in your opinion?
15:36<Chutt>it'd be nice if it were easier
15:36<mdz_>I don't see what the trouble is of installing XMLTV with CPAN
15:36<vektor>don't you just 'apt-get install xmltv'? i haven't really looked into how it works...
15:36<mdz_>you could pretty much script it
15:36<Chutt>i dunno what other people's problems with it are either =)
15:36<Chutt>but that's the #1 thing i get emails about
15:37<mdz_>use CPAN; install blahblah
15:37<vektor>mdz: so you think my search for high quality tv output is ridiculous?
15:38<mdz_>vektor: re: apt-get, not quite yet:
15:38<mdz_>vektor: I'm not judging anyone else's quest :-)
15:39<vektor>mdz: i'm just curious.
15:39<mdz_>vektor: of course not
15:39<Chutt>perlcc makes a 9.4MB stripped binary of tv_grab_na
15:39<Chutt>that's hilarious =)
15:39<mdz_>vektor: I hope to benefit from your obsessions :-)
15:40<mdz_>but personally, I only have two levels of concern for video quality
15:40<vektor>mdz: i guess i just find it annoying that some people seem to have 'awesome tv output' but that whenever i try it it looks like crap.
15:40<mdz_>digital all the way through to pixel-exact output
15:40<mdz_>and "good enough"
15:41<vektor>well like to me, it's not "good enough" if whenever there's a pan in a movie it jumps like you're on a bumpy busride
15:41<mdz_>what I have now is about as good as I expect from my television set for the kind of things that I watch
15:41<vektor>and i don't mean like little sutters that people don't notice, i mean big jumps
15:41<vektor>like, i don't see how _anyone_ can watch tv using xawtv for example
15:42<vektor>like i'm just wondering why everything looks like shit for me and everyone else seems to say 'ah, good enough'.
15:42<vektor>is it that my configuration is that wrong? should i go out and buy an ATI card?
15:42<vektor>or is it really that i just have high expectations for quality
15:42<mdz_>dunno, I have no idea how things look to you
15:43<mdz_>but I figure that if I want quality, I'll get a digital display
15:43<vektor>and how will you output to that?
15:43<mdz_>if and when they are reasonably priced
15:44<vektor>like is there some magic tv that has vga in and can multisync?
15:44<vektor>and if you had that, how would you do framerate conversion?
15:44<vektor>yes, for deinterlacing!
15:44<mdz_>DVI output
15:44<vektor>ok, dude, you have a 29.97 _interlaced_ DVD
15:44<vektor>how is your magic DVI output going to handle that?
15:45<vektor>i'm just curious :)
15:45<mdz_><vektor> and how will you output to that?
15:45<mdz_>I was just answering your question
15:45<vektor>sorry :)
15:45<vektor>anyway i know what you mean
15:45<vektor>tv output is supposed to suck ass
15:45<vektor>so who cares
15:45<vektor>i dunno
15:46<mdz_>yeah, I guess I just have low expectations for it
15:46<mdz_>but I'm a digital bigot
15:46<vektor>but you have alot (all?) the same issues with a digital output too
15:46<vektor>since you need to deinterlace
15:46<vektor>and it's fucking hard
15:46<vektor>and you still need vbi sync
15:46<vektor>yeah the vsync interrupt
15:47<vektor>whatever you want to call it
15:47<vektor>vertical blanking interval interrupt
15:47<vektor>however you like to think of it
15:48<mdz_>sync what against what?
15:48<vektor>well, what refresh rate will your digital output run at?
15:48<vektor>and how will you amortize your frames across it?
15:49<vektor>like right now, with tvtime, we have this problem.
15:49<vektor>my output is a CRT, it's not digital but close enough, and it's very hard to get high quality.
15:49<vektor>because of framerate conversion problem.s
15:50<vektor>dvd output to tv is really nice since you're guarenteed a very specific refresh rate from the output device
15:50<mdz_>my ideal digital output would translate video memory writes directly to the output signal, so there wouldn't be a refresh
15:50<vektor>(or if you're a hardware dvd player, it's even better, since you get to generate the vbis)
15:51<mdz_>turn on this pixel or that
15:51<vektor>well that would be ideal, wouldn't it.
15:51<vektor>i don't think dvi works that way :)
15:51<mdz_>I'm not sure how it works
15:51<vektor>oh well
15:51<vektor>i get your point though.
15:51<mdz_>or how its successor will work
15:51<TechJosh>have you tried running your signal through a time base corrector?
15:51<vektor>ideally we wouldn't need to think about this.
15:51<mdz_>or whenever digital video actually happens
15:52<vektor>digital video has happened.
15:52<vektor>1080i is here.
15:52<vektor>still need to deal with framerate conversion for PC applications though.
15:52<vektor>i is great in many cases.
15:52<mdz_>i sounds like a pain in the ass
15:53<vektor>it is if you're a PC
15:53<vektor>since you have no control over the output refresh
15:53<vektor>like if you want to output to a stupid LCD panel or a projector
15:53<vektor>or even with a consumer video card
15:53<vektor>it's just a design flaw
15:53<vektor>PC ahrdware isn't designed for video.
15:54<vektor>that's all.
15:54<Chutt>wc -l tv_grab_na.c
15:54<Chutt> 410053 tv_grab_na.c
15:54<Chutt>ah well
15:55<Chutt>mdz, if i just had .debs of everything... =)
15:55<mdz_>Chutt: if somebody would send me some hardware... =)
15:55<mdz_>the hardware bribery system has been successful in the past
15:55<mdz_>for getting software written
15:55<vektor>mdz: do you see what i mean or am i on crack?
15:55<Chutt>if i really, really wanted em, i'd just do debs myself
15:55<mdz_>vektor: I understand
15:56<vektor>ok ret
15:56<vektor>mdz: you should try tvtime sometime :)
15:56<mdz_>vektor: but the system as a whole seems so clunky to me that I don't complain if I think it could look metter
15:56<mdz_>vektor: I will
15:57<vektor>what's the autoconf variable for the toplevel source directory?
15:58<vektor>yeah nevermind :)
15:59<mdz_>vektor: $(top_srcdir)
16:00<vektor>mdz: thanks!
16:39<-- TechJosh( has left #mythtv
16:44-!-TechJosh [] has joined #mythtv
17:29<-- TechJoshhas quit ()
17:31-!-TechJosh [] has joined #mythtv
18:02-!-TechJosh [] has quit ["using sirc version 2.211+KSIRC/1.2.1"]
18:21<oztiks>Chutt: you bastardized my faq! ;P
18:23<Chutt>heh, yes, yes i did :p
18:53-!-skitzo [] has joined #Mythtv
18:53-!-oztiks [] has quit [Read error: 54 (Connection reset by peer)]
18:56-!-TechJosh [] has joined #mythtv
19:16-!-TardisX [] has joined #mythtv
19:36<Chutt>well, that was easy
19:36<Chutt>saves a seek table
19:36<Chutt>now i just have to parse it in
19:50<-- TechJosh( has left #mythtv
19:58-!-TechJosh [] has joined #mythtv
20:33<Chutt>mdz, so, me not being all that familiar with nfs
20:34<Chutt>will seeking to the end of the file to read in the seektable kill things?
20:35<skitzo>regarding NFS? assuming -i- understand your question, no
20:36<Chutt>like, it won't decide to transfer all the file or anything stupid like that
20:36<skitzo>no - it shouldnt =)
20:36<skitzo>'try and see' ;P
20:36<Chutt>well, i don't have enough room on my webserver to stick a large enough file
20:36<skitzo>if you can throw me a piece of code
20:36<skitzo>that'll do it
20:36<skitzo>i can run it real fast
20:36<Chutt>it'll be in cvs soon enough
20:43-!-TechJosh [] has quit [Read error: 104 (Connection reset by peer)]
20:51-!-kleetus [] has joined #mythtv
20:54<vektor>Chutt: i have some psycho code in tvime now for crazy PAL modes and stuff
20:54<vektor>Chutt: you might want to leech
20:54<Universe>psycho code doesn't sound like something usable..
20:55<Universe>phat code on the other hand...
21:07-!-skitzo [] has quit []
21:07-!-Universe [] has quit []
21:07-!-TardisX [] has quit []
21:07-!-nyquiljer [] has quit []
21:07-!-vektor [] has quit []
21:07-!-Chutt [] has quit []
21:07-!-Tuscany0_ [] has quit []
21:07-!-mdz [] has quit []
21:08-!-TardisX [] has joined #mythtv
21:08-!-skitzo [] has joined #mythtv
21:08-!-Tuscany0_ [~username@] has joined #mythtv
21:08-!-Universe [] has joined #mythtv
21:08-!-Chutt [] has joined #mythtv
21:08-!-vektor [] has joined #mythtv
21:08-!-nyquiljer [jer@] has joined #mythtv
21:08-!-mdz [] has joined #mythtv
21:24<mdz>Chutt: yeah, that's no problem re: NFS
21:25<Chutt>didn't think so, but
21:26<Chutt>this is weird
21:26<mdz>the elusive xmltv ITPer responded
21:26<Chutt>i've got an occasional hang on exiting a playback
21:26<mdz>he has an apt repository set up with some packages
21:27<Chutt>what bug # was that again?
21:27<mdz>man, we are going to hit 200k soon
21:28<Chutt>time for a big slashdot story
21:28<Chutt>"DEBIAN HAS 200,000 BUGS"
21:28<Chutt>i dunno
21:29<mdz>hmm, I've never uninstalled a module with CPAN before
21:29<mdz>does it clean up after itself?
21:29<mdz>is this playback hang new with the stuff you added today?
21:30<Chutt>no, it's new with the fix for the segfault
21:30<Chutt>it's _looking_ like the mutex lock i added is getting corrupted somehow
21:30<Chutt>heh, he's still using xmltv 0.5
21:33<Chutt>ah well
21:33<Chutt>if it works
21:33<mdz>is it even possible to uninstall this stuff? hmm
21:33<Chutt>i just deleted the cpan-installed stuff from the perl lib/ dir
21:33<mdz>it may be time to rm -rf /usr/local/lib/perl
21:34<Chutt>that's essentially what i did
21:34<mdz>it saves a packing list, no?
21:34<mdz>what's the point of that if it can't uninstall?
21:34<Chutt>i have no idea
21:34<Chutt>this is annoying
21:35<Chutt>i can't figure out how this thing's getting corrupted
21:35<mdz>what sort of corruption are you seeing?
21:35<Chutt>nothing can be holding the lock
21:35<Chutt>but the lock says its held
21:35<mdz>man, I'm having to upgrade this system to perl 5.8 now
21:35<mdz>I might as well just run unstable on it
21:35<Chutt>is x 4.2 in testing yet?
21:36<Chutt>unless pthread_mutex_trylock is messing up
21:38-!-Tuscany0_ [] has quit [Read error: 104 (Connection reset by peer)]
21:38<mdz>if it builds on arm it could go in tomorrow
21:38<mdz>ah, it did build on arm
21:38<mdz>on oct 20
21:39<mdz>it's not trylock that's hanging, is it?
21:39<mdz>just a plain lock?
21:39<Chutt>actually, i should just move the locking up a level
21:39<Chutt>wouldn't need the trylock then
21:41<mdz>how does mythfilldatabase decide which days to grab?
21:41<mdz>it has been weirding me out
21:41<Chutt>it always grabs the next day
21:41<Chutt>then goes through today through a week from now, and grabs that day if there's no data in the database
21:41<mdz>so it always gets each day's data twice?
21:42<mdz>once ahead of time, and once the day before?
21:42<mdz>ok, so this is normal :-)
21:42<Chutt>it's to catch any scheduling changes that happen
21:43<mdz>makes sense
21:43<mdz>confusing as hell though
21:43<mdz>I guess it's better to grab before midnight then, rather than after. I've started doing that anyway
21:44<Chutt>i've got it cronned for 11:30
21:44<mdz>yeah, I'm doing 11:00 right now
21:44<mdz>sometimes it's _slow_ and I didn't want it to run over
21:45<-- nyquiljer(jer@ has left #mythtv
21:45<mdz>especially since I wasn't sure if it actually had data for the next day, based on the output, and hadn't bothered to find out
21:45<Chutt>at least zap2it's not as slow as =)
21:45<mdz>never used them
21:45<mdz>are you still planning on making a 0.7 release soon?
21:46<Chutt>if thor updates mythweb, yeah
21:47<Chutt>if not, i dunno
21:48<Chutt>fairly soon, anyway
21:49<Chutt>this new locking seems to work
21:50<Chutt>and it's simpler, too
21:50<Chutt>hate doing something the hard way, then figuring out an easy way to do it later
21:50<Chutt>the seek table works
21:50<Chutt>just going to add the inexact skips
21:50<mdz>what's the granularity?
21:51<Chutt>30 frames, same as before
21:51<mdz>that should be a great improvement for me
21:51<Chutt>it adds 42 kB per hour to the file, though =)
21:52<mdz>oh dear
21:53<mdz>this bash is unhappy
21:53<mdz>anytime I type a command not in PATH, it gives a bus error
21:53<mdz>smells like glibc upgrade
21:53<Chutt>glibc update?
21:55<Chutt>i think i shall default exact seeks to off
21:55<mdz>oh man, this box is hosed
21:56<mdz>if anyone is thinking about upgrading glibc 2.2 -> 2.3 and then back down and back up again...
21:56<mdz>...well, don't
21:56<-- Universehas quit ()
21:58<Chutt>a reboot doesn't clear it all out?
21:58<mdz>I am having to do that now
21:58<mdz>I have no console on this machine really, so there's not much I can do
21:58<mdz>all I have are the remote control buttons, and not many unix commands can be typed with 0-9 and a-f
21:59<mdz>I have a spacebar bound for myth, though
21:59<Chutt>i was thinking earlier
21:59<Chutt>most of my users probably use it all more than i do
21:59<Chutt>i need to fix that
22:00<mdz>use what?
22:00<mdz>ah :-)
22:00<mdz>this NFS root stuff is tricky sometimes
22:00<mdz>for example, I can't use dhcp-client
22:01<mdz>because it brings down the network interface, does the protocol, then tries to run dhclient-script :-)
22:01<mdz>which of course is not available because the network interface is down
22:01<Chutt>couldn't stick that all in a ramdisk?
22:01<mdz>I could, I'm thinking about it
22:01<mdz>it would be nice if dhcp-client would actually configure the interface itself instead of using some damn shell script
22:01<mdz>I should use udhcpc or something
22:02<mdz>I'm going to try that as long as I'm bringing it down
22:02<mdz>it bugs me that my lease goes away and my dynamic DNS goes away
22:09-!-witten [] has joined #mythtv
22:10<witten>does mythtv record tv to disk?
22:11<kleetus>to hard disk, yes
22:13<witten>I'll have to try that out
22:18<Chutt>what would be the point otherwise? =)
22:18<witten>I didn't know if that feature had been implemented yet
22:19<mdz>Chutt: I have Kenneth's xmltv packages and my mythtv package running now
22:19<Chutt>mdz, neat?
22:19<Chutt>err, no ?
22:20<Chutt>what's the user have to do to set things up by themselves?
22:21<mdz>Chutt: 1. apt-get install mythtv mysql-server 2. create and populate the database 3. mythfrontend
22:21<mdz>pretty much
22:22<Chutt>can the db creation be in the packaging?
22:22<mdz>it can
22:22<Chutt>i dunno if there's precedent for that or not
22:22<mdz>but I'll have to add prompting and such for it
22:22<mdz>there is
22:22<mdz>a bunch of web apps and such do it, like bugzilla
22:22<Chutt>well, cool
22:22<mdz>there's no standard tool for it, though. I wrote one for work, maybe I'll package it up
22:22<witten>mdz: where does one apt-get install mythtv from?
22:22<Chutt>ack, brb
22:22<mdz>witten: nowhere yet, I'll be putting the packages up soon
22:22<mdz>witten: are you interested in testing?
22:23<witten>is there a debian directory in the tarball?
22:23<witten>mdz: sure
22:23<mdz>no there is not; I've been building preliminary packages though
22:23<witten>oh ok
22:24<witten>mysql only? no postgres?
22:28<mdz>currently I believe everyone is using mysql. I'm not sure whether there is anything mysql-specific, though
22:28<mdz>it's using QT's crazy SQL absraction layer that I know nothing about
22:28<mdz>but there is a postgresql plugin for it
22:28<witten>oh, heh
22:30-!-Universe [] has joined #mythtv
22:31<Chutt>i don't know if there's anything mysql specific in the queries
22:31<witten>will the deb eventually make its way into unstable?
22:31<Chutt>doubt it
22:31<mdz>Chutt: how would you feel about versioning libmyth by release, i.e. rather than
22:31<Chutt>i _hope_ to have it stable eventually
22:32<Chutt>but for now? sure
22:32<mdz>then it could live in /usr/lib
22:32<mdz>and I wouldn't have to use ugly wrapper scripts
22:32<Chutt>i'll make that change
22:32<mdz>great, thanks
22:32<Chutt>and i'll be checking the seek stuff in after CSI's over
22:33<TardisX>what does the seek stuff do for us laymen?
22:33<mdz>it makes things go
22:33<Chutt>just some speedups
22:33<TardisX>in seeking I guess?
22:33<Chutt>especially for people that are nfs mounting things =)
22:33<TardisX>will it make it easier for arbitary seeking in the recording?
22:34<TardisX>sweet++ :-)
22:34<witten>mdz: where can I get the debs to test?
22:34<Chutt>has a precalculated seek table, so you can seek to any second in the recording
22:34<mdz>witten: patience :-)
22:34<witten>mdz: ok. I gotta go.. email if you need any testers
22:35-!-witten [] has quit ["bbl"]
22:37<TardisX>re postgres support - its probably only minor changes to mc.sql and perhaps some queries....
22:38<TardisX>I have a feeling that myth probably uses mysql date functions which are mysql specific. I don't think there is an SQL standard for date manips, apart from normal < > =
22:41<TardisX>DATE_ADD(CURRENT_DATE, INTERVAL %d DAY) <-- mysql specific
22:41<TardisX>i think :-)
22:42-!-mdz [] has quit []
22:42-!-mdz [] has joined #mythtv
22:44<-- Universehas quit ()
22:47<Chutt>people that reply to unrelated mailing list messages instead of starting a new thread suck
22:54<mdz>Chutt: yes they do
23:04<Chutt>might get more visits to the website today than the last release day
23:07<Chutt>heh, part of that's the link on =)
23:08<TardisX>how many hits do you get?
23:10<Chutt>average for the month is 338 visits a day
23:10<Chutt>hits is hard to say, since i moved all the screenshots off of the site halfway through
23:11<TardisX>ah OK..
23:11<TardisX>Do you check referrers? I can't remember how I found it... I think perhaps from one of the other PVR projects.
23:12<Chutt>yeah, i do
23:12<Chutt>not all that often, but i get curious every once in awhile
23:12<TardisX>I use webalizer which has a nice breakdown of referrers, plus search engine keywords used to reach the site
23:15<Chutt>heh, finally broke 1000 downloads of the 0.6 release
23:16<skitzo>Chutt: whats the whole mysql seektable thing?
23:16<Chutt>mdz, your updated mplayer patch looks good, though it'll be out of date and won't work after my next checkin =)
23:16<Chutt>i'm changing the text string at the beginning to MythTVVideo
23:16<mdz>Chutt: you bastard
23:17<Chutt>skitzo, mysql seektable?
23:17<TardisX>oh yeah mean to ask something... how can I change the OSD font size?
23:17<Chutt>and upping the version # to 0.7 in the file format
23:17<Chutt>tardisx, edit the osd theme file
23:18<skitzo>[22:35:42] <Chutt> has a precalculated seek table, so you can seek to any second in the recording
23:18<skitzo>my bad
23:18<skitzo>i thought i had read mysql into that
23:20<TardisX>I can't wait to go home now and see if I can read all the descriptions of programs in the OSD :-)
23:21<Chutt>too small?
23:21<TardisX>other way :-)
23:21<TardisX>they almost always run off the bottom
23:21<TardisX>I get quite verbose descriptions from the TV guide
23:22<Chutt>you must get better descriptions that the us stuff gives, then
23:22<skitzo>where are the 'OSD theme files'
23:22<skitzo>(adding to faq)
23:22<TardisX>I would prefer they were smaller, but more consistent. Sometimes I get nothing..
23:22<skitzo>TardisX: welcome to my life with directv ;)
23:22<Chutt>or something like that
23:23<TardisX>might be worth a note since you are editing the faq, that to change the font you need to change it in settings.txt (which ttf file to use) and specify the same font in the osd theme file. I think that's right. :-)
23:23<mdz>Chutt: has anyone sent you a patch for page up / page down / back 1 day / forward 1 day in mythepg?
23:23<mdz>that takes forever to navigate
23:23<Chutt>yes, it does
23:23<skitzo>TardisX: okay
23:23<skitzo>Chutt: is what TardisX said accurate?
23:23<Chutt>added that to my todo list after setting up a week's worth of recordings before i left
23:24<Chutt>skitzo, yup
23:24<Chutt>mdz, want to add it yourself? =)
23:24<mdz>Chutt: it doesn't look too difficult to add
23:24<mdz>Chutt: but I bet you could do it faster (wink, wink)
23:24<Chutt>code that i don't have to write is good
23:25<skitzo>you are the only person i have ever seen use .txt as a config file extension ;P
23:25<Chutt>i use .xml for the menu configs, though
23:26<skitzo>thats a good thing =)
23:26<Chutt>i caved and used xml
23:26<skitzo>itd actually be kinda neat if the configs were xml
23:26<skitzo>*actually* it'd be even neater if the configs were all in mysql and all the static files had were mysql pointers.. ;)
23:27<Chutt>i just borrowed the existing config file code from work
23:27* skitzois just babbling.
23:27<mdz>Chutt: speaking of config files
23:27<mdz>Chutt: thought about moving them into /etc so I don't have to symlink them?
23:28<Chutt>i'd really rather not have em there
23:28<mdz>whyever not?
23:28<skitzo>as in /etc/?
23:28<Chutt>they're not hurting anything in share/
23:28<TardisX> /usr/local/etc at least...
23:28<skitzo>/usr/local/mythtv/etc/ ;)
23:28<TardisX>but I think the share is correct - at least in the context of the theme data
23:28<Chutt>and really, you shouldn't be editing the files that are in there
23:29<mdz>Chutt: they're pretty clearly configuration files
23:29<mdz>Chutt: and configuration files go in /etc
23:29<Chutt>yeah, but they're example config files
23:29<Chutt>if you want to mod em, you should copy them to ~/.mythtv/ and edit
23:29<mdz>Chutt: are they examples, or are they system defaults?
23:29<mdz>Chutt: if ~/.mythtv/blah exists, is share/mythtv/blah still read?
23:29<skitzo>oh, what is the deal with ~/.mythtv/ ?
23:29<mdz>Chutt: I rest my case ;-)
23:29<Chutt>they're both :p
23:29<skitzo>(what is the point?)
23:30<TardisX>ln -s /usr/local/share/mythtv /etc
23:30<TardisX>everyone's happy
23:30<mdz>TardisX: the font, themes and xml stuff belong in share
23:31<Chutt>the menu configs would be the same as the config files, imo
23:31<skitzo>(is mythtv meant to be configured for multiple users.. or what?)
23:31<Chutt>skitzo, no, its not, really
23:31<mdz>ah, agreed
23:31<skitzo>then whats the purpose of .mythtv/ ?
23:31<Chutt>but, when you do an install, it copies over the installed files
23:31<skitzo>oh psha
23:31<skitzo>wouldnt it be more logical to just change the makefile?;)
23:32<Chutt>so if you stash stuff in your home dir, you can keep some persistant settings
23:32<Chutt>because then you wouldn't catch changes in the config file
23:32<skitzo>/usr/local/share/mythtv/customsettings.txt ;)
23:33<mdz>this is what conffiles are for
23:33<Chutt>yeah yeah
23:33<Chutt>i like it in share, so :p
23:34<skitzo>anyway i think the .mythtv just adds potential confusion.
23:34<mdz>I'll symlink it so you can pretend it's in share :-P
23:34<Chutt>heh, ok
23:34<skitzo>particularly for morons like myself
23:34<Chutt>you are replacing the font, right?
23:34<mdz>skitzo: what's confusing? it's a common setup for unix software
23:34<mdz>Chutt: am I? is it not legal?
23:34<skitzo>it's a common setup for multi-user software, no?
23:34<mdz>oh, we had this talk already I think
23:35<skitzo>and .* generally denotes something not meant to be touched by humans.. or is trying to play incognito.
23:35<Chutt>there's a good case for it to be legal, but debian already removed it from other packages, so
23:35<Chutt>skitzo, err, no it doesn't
23:35<skitzo>i didnt say always
23:35<Chutt>it denotes something that you don't want cluttering up your 'ls' of your homedir
23:35<skitzo>anyway, i really don't give a shit, i'm giving my two cents. nothing more.
23:36<mdz>I dunno, the (C) Sun Microsystems 1999 is pretty explicit :-)
23:36<Chutt>well, yeah, but it was distributed as GPL
23:36<Chutt>by them
23:37<Chutt>so they can fire the person that included it in the openoffice distro, but other than that, they can't do much aside from say 'oh, we didn't mean for that to be in there'
23:37<mdz>I think any copyright notice on a particular file would override any license they had granted for the work as a whole
23:37<mdz>but IANAL
23:37<Chutt>a copyright notice isn't a license, though
23:38<Chutt>my biggest problem with the free ttfs that are recommended replacements are that they're like 3x larger
23:38<Chutt>and my .tar.gz is getting bloated as is
23:38<mdz>you should check out FreeSerif and see if it looks decent to you
23:38<mdz>or is that the one you're talking about?
23:38<Chutt>yeah, that's what i'm talking about
23:39<mdz>it's big, but it supports many character sets
23:39<mdz>it's quality
23:39<mdz>I don't care if you ship it in the tarball or not, I'll just depend on ttf-freefont :-)
23:39<skitzo>distribute fonts in a diff package.
23:39<skitzo>[on a server with bandwidth]
23:40<Chutt>it needs to be part of the distribution
23:40<Chutt>else then i'll get emails asking why the OSD isn't showing any fonts
23:40<mdz>if I have to repackage the tarball to upload to debian I will be unhappy
23:40<mdz>wow, the tarball is big
23:40<Chutt>i think there's junk i can cut out
23:41<Chutt>there's a .jpg of the lcd stuff in there
23:41<mdz>it also has a whole copy of libavcodec that I wouldn't use
23:41<skitzo>i dunno.. i dont think people are really expecting something this functional to be 200kb large.
23:41<Chutt>mdz, unless you're planning on packaging cvs libavcodec...
23:41<skitzo>(and if the fonts are in a seperate package they would only have to be downloaded once, and you could download the rest of the updates in the other package)
23:41<Chutt>i can cut out the archs in my copy of libavcodec, though
23:42<mdz>Chutt: Version: 0.4.6cvs20020821-0.2
23:42<mdz>Chutt: is that not recent enough?
23:42<Chutt>wouldn't even come close to working
23:42<mdz>I could bug marillat to package a newer snapshot
23:42<Chutt>api changes are fairly frequent
23:42<Chutt>and i statically link stuff in, anyway
23:43<mdz>please tell me you don't link the static library into your shared library
23:43<Chutt>'course not
23:47<Chutt>afraid of non-relocated symbol errors from lintian or whatever?
23:47<TardisX>must ... learn ... more ... c
23:50<Chutt>only way to learn is by doing
23:50<mdz>Chutt: afraid of not working on several architectures
23:51<Chutt>oh, don't worry about that
23:51<TardisX>I know the basics... I can do v. minor hacks. I have my hands full keeping up with perl though :-)
23:51<Chutt>this is x86 only :p
23:51<mdz>Chutt: I don't want to hear about how nobody will ever use it on those architectures, because some crazy person will :-P
23:51<Chutt>and i'll tell said crazy person to quit his bitching and buy a x86 box :p
23:52<Chutt>i'm doing the library name change now, btw
23:52<mdz>they will send you patches with assembler for weird architectures
23:52<mdz>stranger things have happened
23:52<mdz>Chutt: thanks
23:53<mdz>I curse maintainers that use wrappers like that
23:53<mdz>when I run gdb on /usr/bin/blah and it turns out ot be a shell script