Back to Home / #mythtv / 2003 / 01 / Prev Day | Next Day
#mythtv IRC Logs for 2003-01-10

00:00<mdz>oh, it came on 5 times in one set of program data
00:01<mdz>that's why so many checks
00:02<mdz>sent you an email with the particulars
00:02<Chutt>ok, thanks
00:02<mdz>I executed the query straight from the log and it worked fine
00:02<mdz>found both the old and the new row
00:03<Chutt>i wonder what's failing it, then
00:03<Chutt>pity there's no result from the query
00:04<mdz>that would make the query log even more ridiculously large
00:04<Chutt>yeah, but =)
00:04<mdz>maybe it doesn't delete it correctly if it's at a certain place in the list
00:05<Chutt>i'm fairly sure that that's right
00:06<mdz>I hate some of the STL iterator stuff
00:07<Chutt>some of the Qt stl-like classes cause g++ 2.95 to segfault during compile, though =)
00:07-!-aw [] has joined #mythtv
00:07<mdz>it would be a lot less ugly if you could do instance::iterator and get the iterator type and things like that
00:08<Chutt>that's why i'm using vector<> and list<> in some places
00:08<mdz>rather than having to typedef things or use ridiculous template types everywhere
00:08<mdz>I use vector<> when I don't care just so I can ignore iterators
00:09<aw>anybody try mythweather through a transparent squid proxy? mine isn't happy
00:09* bigguysighs
00:10<mdz>if anything behaves differently through your transparent proxy, it isn't very transparent :-P
00:10<aw>works on a browser
00:11<aw>I get "clientReadRequest: FD 16 Invalid Request" from squid
00:11<moegreen>aw: It's probably because I don't have it sending a hostname with the request
00:12<mdz>are the chanid/start/end in oldrecorded used by anything?
00:13<Chutt>moegreen, considered using the QHttp class for that, btw?
00:13<Chutt>mdz, no
00:13<bigguy>I hate writing. I always did poorly in English. Math and the Sciences always suited me better.
00:14<moegreen>Chutt: Not really, because the request is so simply and only returns one page. I might now because I'm planning on adding the radar map page
00:14<bigguy>but maybe people will get the point of my post
00:14<Chutt>ok =)
00:15-!-bbeattie [] has joined #mythtv
00:15<bbeattie>Has anyone here built a linux HTPC?
00:18<bigguy>bbeattie: well there are some in here that use linux for that purpose
00:18<bigguy>bbeattie: what exactly are you wanting?
00:19<bbeattie>I'm just wondering if, and how people have been able to get over the refresh/sync issues when using a computer to display video. They don't play nice, and I'm wondering how people solve that problem. I figure mythtv people have faced it, and maybe have a solution.
00:20<mdz>Chutt: think it would be unreasonable to have the scheduler print a message to stdout if it is pruning a recording as a duplicate?
00:20<Chutt>go ahead and add it
00:20<Chutt>might get kinda verbose, though
00:21<bbeattie>Chutt: why not make a log file for stuff? and allow it to be viewed via setup menu?
00:21<Chutt>because any log file that would be halfway useful would be a huge drain on resources
00:21<bigguy>bbeattie: displaying on a tv right?
00:22<bbeattie>biguy: projector actually.
00:22<bigguy>bbeattie: alot of people seem to favor using scan converters
00:22<bigguy>I've seen some talked about on the list
00:24<bbeattie>bigbuy for what type of display device? A projector can support 120hz, which would work well for most things, (120/24=5, 120/60=2) but 23.976, 29.97, and 59.94 don't match perfectly, and some judder will occur unless something is done.
00:24<bbeattie>bigguy, sorry.
00:25<bbeattie>mythtv records at 29.97 doesn't it? But that won't always display well on computer monitors, correct
00:25<bbeattie>Does mythtv just deal with judder? or does it try to compensate somehow to reduce it?
00:26<Chutt>it just does the best it can within the constraints of X
00:27<bbeattie>Have you heard anything about people trying to use specific refresh rates to sync to 59.94, or 119.88? To stop judder?
00:29<bbeattie>Alright. - Do you know of any other irc channels for people doing HTPC's with Linux?
00:30<bigguy>bbeattie: you might as on avs forums
00:30<bigguy>that is their linux HTPC forum
00:30<bbeattie>That's my homepage. :-)
00:31<bbeattie>none has really addressed the issue much, except a few trials, but no good results.
00:31<bbeattie>I'm going to do it write, then write a FAQ on how to do it right.
00:33<PeteCool>I never saw any problems with on my TV with my 23.976 movie rips
00:34<bbeattie>PeteCool: Are you using a tv-out on your video card? And if you look at, say credits of a movie, are they "perfectly smooth"? Or does it jump ever 2, 10, or 20 seconds?
00:34<Chutt>i always get a laugh outta reading the avsforum linux forum
00:34<bigguy>credits don't do so well in divx/mpeg4
00:35<Chutt>that ago guy is hilarious
00:35<bigguy>atleast from what I have seen
00:35<PeteCool>bbeattie: I trim the credits... I never watch them anyway
00:36<PeteCool>bbeattie: but, I don't use divx but xvid, which IMO, worked better a year ago... and divx didn't change much since then, but xvid quality increased much
00:36<bbeattie>bigguy You have to encode it at the same refresh as it was taken from, and display at the same refresh, otherwise it's off.
00:37<PeteCool>bbeattie: but the 23.976 fps scenes looks right both on my monitor (85hz) and on my tv, through my radeon 8500 tv-out
00:37<bigguy>I ripped alot of movies using divx5 at 23.976 some with mp3 and some with ac3 audio
00:38<bigguy>but I always encode the credits at a lower setting
00:38<PeteCool>what's so great|cool about the credit's anyway?
00:39<bbeattie>PeteCool: It's a good way to see how far off refresh really is. It's a good test bed.
00:39<bbeattie>PeteCool: Do the credits scroll perfectly smooth? It shouldn't be especially on a 85hz refresh.
00:40<bbeattie>85/23.976=3.5452118785 .. That's not very clean.
00:41<mdz>the rage theatre clocks the video refresh to the TV sync
00:41<bigguy>they seem smooth to my eyes
00:41<mdz>when TV-out is enabled
00:41<mdz>that's the whole point of doing TV-out on the video card rather than with a converter
00:42<mdz>now vektor is going to wake up and tell me why it sucks anyway
00:42<bbeattie>mdz: haha
00:42<PeteCool>and how does vektor output to tv?
00:42<PeteCool>with magic?
00:42<bbeattie>He's still working on perfecting it.
00:43<mdz>he wants to get an interrupt whenever the video card refreshes
00:43<PeteCool>His magic? :)
00:45<PeteCool>is he some sort of... pvrophile?
00:45<bbeattie>So ragetheater varries it's 15 pin D-Sub's refresh rate, to match whatever it's outputting? It can tell if it's a DVD at 23.976, 24, 29.97, 30, 59.94, 60 or HDTV at those same refresh rates?
00:46<Chutt>i think he meant it locks the tv output to 59.94
00:47<mdz>odd, on several recordings now I've gotten audio from the tuner with video from the s-video input
00:47<bbeattie>That's given, any ntsc has to be locked to 59.97 (S-video, composite, or component)
00:47<mdz>bbeattie: it outputs NTSC
00:48<mdz>Chutt: that switch is completely internal to bttv, right?
00:48<Chutt>mdz, what switch?
00:48<mdz>Chutt: tuner audio vs. line in audio for Television/S-Video
00:49<Chutt>i thought so, yes
00:49<mdz>it doesn't happen in live tv
00:50-!-knoppix [] has joined #mythtv
00:50<bigguy>their is that distro guy again
00:50<bbeattie>mdz: Right, any "video" signal is always 59.97. But when you use DVI, or 15pin D-Sub, video cards can't sync to "video" as well. -- But if you do use video-out, it is either not optimizing the output (just doing whatever was last written) or it's doing 3:2 reverse pulldown if we're lucky, but I don't know if it knows to do that, or could. -- That, is what I need to resolve for a linux HTPC done correctly.
00:51<knoppix>hmm, I'm running debian and having trouble with the Perl 5.6 to 5.8 upgrade ... mythtv wants the latter but some stuff doesn't want to install ... creating a yucky mess of my installation now ... can anyone suggest what to do now ? :)
00:51<mdz>bbeattie: DVI and D-sub have nothing to do with TV output
00:51<mdz>not analog TV anyway
00:51<bbeattie>mdz: They do when you are doing HDTV via RPTV or projection.
00:52<bbeattie>mdz: Right, :-) I'm refering to HDTV here.
00:52<mdz>then do progressive HDTV
00:52<mdz>not interlaced
00:52<knoppix>for starters, am I correct that mythtv needs Perl 5.8?
00:52<Chutt>no, mythtv doesn't use perl
00:52<mdz>knoppix: mythtv does not need perl at all, except insofar as xmltv needs it
00:53<mdz>and xmltv works fine with perl <5.8
00:53<knoppix>mdz: oh, but it appears that libxmltv-blah needs it
00:53<bbeattie>mdz: DTV/HDTV spec includes interlaced, plus the 5 difference refresh rates. That's what the current problem is.
00:53<mdz>knoppix: it needs whatever version of perl it was compiled against
00:54<knoppix>mdz: oh, and it appears in the debian package that that is now 5.8
00:54<mdz>or whatever version it's built for, I should say
00:54<Chutt>i actually approved a post to the mailing list from someone that wasn't subscribed
00:54<mdz>knoppix: of course, it is in unstable and unstable has perl 5.8
00:54<mdz>knoppix: there are debs for perl 5.6 on my site, but not the current version of xmltv
00:54<knoppix>mdz: sure, but everything was working for me with 5.6 until today :)
00:54<knoppix>dunno why
00:54<mdz>the xmltv maintainer has offered to maintain woody debs of xmltv
00:55<bbeattie>I'll come back when I have things figured out and start writing a faq.
00:55<knoppix>mdz: oh, so mythtv was using that I guess ...
00:56-!-bbeattie [] has quit ["where's a bed...."]
00:57<knoppix>mdz: libxmltv-perl that mythtv wants these days is 0.5.6-1 and that requires perl 5.8 so I guess I have no choice now but to make my system happy under perl 5.8?
00:58<Chutt>unless you use the older libxmltv-perl that's on his site, yes.
01:00<knoppix>Chutt: okay, I've got my work cut out for me tonight. Thanks for the confirmation. Apparently 5.6 to 5.8 transition has been tricky for other distros too.
01:00<Chutt>debian transitioned to 5.8 quite a while ago
01:00<Chutt>and there wasn't any big deal about it, iirc
01:00<PeteCool>yeah, it was easy
01:00<Chutt>everything got rebuilt at pretty much the same time
01:01<knoppix>Chutt: is there an easy way for me to update my system ... I installed debian the knoppix way :)
01:01<Chutt>i didn't notice that
01:01<knoppix>yeah, using the insall to hard-drive script ... oh duh, yeah :)
01:02<knoppix>I think I'm suffering from too many apt sources in my /etc/apt/sources.list
01:02<PeteCool>knoppix: apt-get dist-upgrade --show-upgraded
01:02<PeteCool>knoppix: then type y and enter
01:02<knoppix>I wish there was a nice GUI to set/play with apt's PIN system so I could control this process better
01:03<PeteCool>aptitude is easier imho
01:03<knoppix>PeteCool: I'll try again ... I think updated just fine (and got Perl 5.8) ... just can't get mythtv back installed
01:03<Chutt>dselect's better :p
01:03<Chutt>except for whoever the dumbass was that changed the keybindings
01:03<knoppix>Chutt: dselect I thought was "older" (and clunkier) maybe I should give it a spin again
01:04<Chutt>it's ancient =)
01:05<knoppix>okay, my problem is that mythtv won't install ... dist-upgrade worked fine ... it is that damn libxmltv-perl
01:06<bigguy>so has anyone seen the new "in the works" installer debian is supposedly gonna have ready for the next release?
01:06-!-Soopizzle [] has joined #mythtv
01:06-!-PeteCool [] has quit ["Client Exiting"]
01:13-!-knoppix [] has quit ["Client Exiting"]
01:30<-- aw( has left #mythtv ("Client Exiting")
01:30-!-moegreen [] has quit ["[BX] Become a BitchX Certified Systems Engineer today! Apply within!"]
01:54-!-moegreen [] has joined #mythtv
03:04<Chutt>think i may've fixed the remote playback issues
03:10-!-Vee2d2 [] has quit [Connection timed out]
03:11-!-Vee2d2 [] has joined #mythtv
03:12-!-Vee2d2 [] has quit [Client Quit]
03:21-!-moegreen [] has quit [Read error: 113 (No route to host)]
03:54-!-mythtv [] has joined #mythtv
04:10-!-Soopizzle [] has quit [Read error: 104 (Connection reset by peer)]
04:10-!-Soopizzle [] has joined #mythtv
04:25-!-Soopizzle [] has quit [Read error: 104 (Connection reset by peer)]
04:25-!-Soopizzle [] has joined #mythtv
04:59-!-nevertheless [] has joined #mythtv
05:09-!-Soopizzle [] has quit [Read error: 104 (Connection reset by peer)]
05:10-!-Soopizzle [] has joined #mythtv
05:30-!-Soopizzle [] has quit [Read error: 104 (Connection reset by peer)]
05:31-!-Soopizzle [] has joined #mythtv
05:38-!-Soopizzle [] has quit [Read error: 54 (Connection reset by peer)]
05:38-!-Soopizzle [] has joined #mythtv
07:54-!-mythtv_ [] has joined #mythtv
07:59-!-giuppi [~giuppi@] has joined #mythtv
08:03-!-mythtv [] has quit [Read error: 60 (Operation timed out)]
08:38-!-giuppi [] has quit ["Lost terminal"]
08:38-!-mythtv_ [] has quit [Read error: 104 (Connection reset by peer)]
09:52-!-Universe [] has quit [Read error: 110 (Connection timed out)]
09:55-!-moegreen [] has joined #mythtv
10:37-!-bigguy [] has quit [Read error: 54 (Connection reset by peer)]
10:46-!-moegreen [] has quit ["BitchX-1.0c19 -- just do it."]
10:49<hurdel>mdz: you here?
10:49<mdz_>hurdel: yes
10:51<hurdel>wondering about your mplayer patch
10:52<hurdel>can i use version MPlayer-0.90rc2.tar.bz2
10:54<hurdel>heh, i'll try it
10:54<mdz_>if it doesn't apply cleanly, you may have to adjust it
10:54<mdz_>but if it applies cleanly it will probably work
10:54<hurdel>not sure how to patch tho :(
10:55<Chutt>bruce markey sent me a patch that _really_ fixes the remote playback weirdness
11:06<Chutt>the backend stopped segfaulting when i installed over it
11:07<Chutt>it did awhile ago, but..
11:11<mdz_>maybe qt fixed it
11:12-!-moegreen [] has joined #mythtv
11:15<rkulagow>hurdel: read the HOWTO on patching mplayer
11:18<Chutt>everything's in the docs =)
11:18<hurdel>rkulagow : which howto is that? the mythtv one?
11:21-!-nevertheless [] has quit [Read error: 104 (Connection reset by peer)]
11:23-!-m0tion [] has joined #mythtv
11:23<m0tion>can you use any true type font with mythtv?
11:24<rkulagow>hurdel: yes.
11:26-!-jasongrichmond [~jrichmond@] has joined #mythtv
11:26<jasongrichmond>u here chutt?
11:28<jasongrichmond>web analytics are up
11:28<Chutt>oh? cool
11:29-!-bigguy [] has joined #mythtv
11:29--> Spark^( has joined #mythtv
11:29<Chutt>how do i get to it? =)
11:30<jasongrichmond>you have to ssh to and port forward local port whatever to remote 8080 (8080 is blocked at my FW, you have to port forward)
11:30<jasongrichmond>then http://localhost:8080
11:31<jasongrichmond>some of the stuff will take a few days to update
11:31<jasongrichmond>new reports are run 3am daily
11:31<Spark^>is anyone familiar with the channel changing code. i'm not able to change channels and am trying to do something about it. i've narrowed it down to a db problem (though the error message indicates that anyhow). inital problem is the default channel name being '3', which i don't have as a name (only a number). i can change this, and it works better, though i don't get the channel i've specified as default.
11:31<Spark^>i'm wondering how the 'digit entry' (ie. upper-left corner of screen) are supposed to work
11:32<hurdel>rkulagow: exactly what document are you talking about with mplayer patch description?
11:32<Spark^>in that it currently checks any digits entered against the channel *names* from what i can tell
11:33<hurdel>rkulagow : think i just found it
11:41<-- Spark^has quit ("using sirc version 2.211+KSIRC/1.1")
11:43-!-PhazE3 [] has joined #mythtv
11:49-!-Universe [] has joined #mythtv
11:54<hurdel>is somebody maintaining mythweb to keep up to date with the db changes in cvs?
11:55<lichen>trying :)
11:56<lichen>hopefully soon ill get some time to make it use the new record table
11:56<lichen>should simplify lots of the functions i think actually
11:57<Chutt>yeah, it should
11:58<hurdel>how does everybody maintain db integrity (no loss of data) when upgrading from cvs to newer cvs?
11:58<Chutt>there were instructions in mdz's email
12:01<hurdel>hmm, don't remember seeing that
12:01<hurdel>was that recently? (this week?)
12:02<mdz_>they were in the CVS commit message
12:02<mdz_>then I posted ANOTHER message to mythtv-dev
12:02<mdz_>and, of course, it's obvious from looking at 0-7-to-0-8.sql
12:06<hurdel>0-7-to-0-8.sql isn't useful if you are already at a previous cvs (0.8)
12:06<Chutt>yes it is.
12:10<hurdel>ok, i am a moron
12:10<jasongrichmond>enjoy the weekend and GO JETS! =)
12:10<-- jasongrichmond(~jrichmond@ has left #mythtv
12:21<-- Universehas quit ()
12:23<lichen>the sad thing is here at my desk they gave me a 2.5ghz machine for me to setup with linux for the SOLE PURPOSE of bridging two vlans
12:23<lichen>what a waste, i wanna take it home :(
12:26<bigguy>you could swap it out for a nice p233mmx box ;)
12:26<lichen>hah yeah really, one of my main servers at home is a 233
12:27<lichen>heck my workstation is a dual 400... and my myth box is a 1ghz im borrowing from a friend
12:27<PhazE3>My workstation at home is a k6-2 500 because im using my 1ghz for myth :(
12:28<lichen>hah sounds about like me.. but here at work i have a dual 700 and a 2.4ghz with a gig of ram... and two flat panels side by side
12:28<lichen>im still working on taking one of the flat panels home
12:29<PhazE3>I have a 1.6 ghz at work with 1/2 a gig :( You have me beat
12:29<bigguy>I have a couple of athlon xp machines, a cel 1.2 , a cel 533, 2x cel 300, p233mmx(gateway/router), ibm p166+ and several working boxes of lesser speeds
12:30<bigguy>I have 1gig pc3000 in my xp 2000+ machine
12:30<PhazE3>at home I also have a amd 200, 386 (router) and a p1 100
12:31<bigguy>I offloaded all my old 486's and lesser machines to a kid just wanting to play around with old crap
12:31<lichen>hah i love being a nerd
12:31<lichen>we have something like 10 computers (in use) at my apt
12:31<lichen>including our tivo
12:33<PhazE3>I live alone and have 5 running :) what does that say
12:33<bigguy>I still wish I had bought the dual alpha machines when I had the chance :(
12:33<lichen>hahah i just wish i had actual GOOD computers
12:33<PhazE3>I have an alpha 500 at work... its nice :)
12:33<lichen>im trying t oget my roommates to build a nice box with me for the mythbackend
12:34<lichen>and then everyone can fend for themselves to get a box to run mythfrontend :)
12:34<lichen>anyways.. i guess i should get some of these work orders done.. bbl
12:34* lichen&
12:34--> Spark^( has joined #mythtv
12:35<bigguy>PhazE3: from what I remember these were the fastest procs alpha made the guy that was gonna sell them to me(I know him and he hasn't lied to me before) got them at auction from a bankrupt company
12:35<Spark^>can anyone show me their channel db table (or at least part of it)
12:35<Spark^>i'm confused about what should be in each field
12:36* bigguygoes to decide what he wants to cook for lunch. some minute steak would be nice
12:37<Spark^>my channel.channum fields contain the first few chars of the xmltvid field, e.g. "midla". whereas GetCurrentChannelNum() seems to want a channel number (e.g. E2, or 49, etc.)
12:40<Spark^>hmmm, it finally works with it setup like that, and i even get a nice osd with description. neat. so does anyone know why the default tv channel is '3' - is this a common US frequency/channel?
12:41<Spark^>and i now understand how the three figure stuff works... i think, its the actual tv frequency rather than a channel number (28, 25, 49 etc. rather than 1,2,3,4,5)?
12:59-!-nevertheless [] has joined #mythtv
13:00<-- Spark^has quit (Read error: 104 (Connection reset by peer))
13:03-!-Universe [] has joined #mythtv
13:07-!-m0tion [] has quit []
13:07-!-lichen_ [] has quit []
13:07-!-SadMan [] has quit []
13:07-!-Chutt [] has quit []
13:07-!-rkulagow [] has quit [Read error: 104 (Connection reset by peer)]
13:08-!-m0tion [] has joined #mythtv
13:08-!-lichen_ [] has joined #mythtv
13:08-!-SadMan [] has joined #mythtv
13:08-!-Chutt [] has joined #mythtv
13:22<PhazE3>lichen: you are the one who maintains mythweb?
13:42-!-Disconnected (Connection timed out).
13:42--> Youare now talking on #mythtv
13:42-!-Topic for #mythtv is
13:42-!-Topic for #mythtv set by Chutt at Thu Jan 9 14:19:41
13:46-!-Soopizzle [] has quit [Read error: 54 (Connection reset by peer)]
13:46-!-Soopizzle [] has joined #mythtv
13:52<lichen>phaze, yeah.. for the most part
13:54-!-nevertheless [] has quit [Remote closed the connection]
14:08--> Spark^( has joined #mythtv
14:17<-- Spark^has quit ("using sirc version 2.211+KSIRC/1.1")
14:29-!-nevertheless [] has joined #mythtv
14:33<hurdel>that guy on the mailing list sure doesn't give up about mpeg2 support
14:35<Soopizzle>what is he saying?
14:36<hurdel>that there are lots of things that need/use mpeg2, not just the one card
14:36<Chutt>but for live tv stuff, it's only that card that needs it
14:36<Chutt>which is what i've been talking about
14:37<Soopizzle>hey chutt, how far are you from ATI HQ?
14:37<vektor>Which card?
14:37<Chutt>the wintv-pvr cards
14:37<vektor>Oh, ok.
14:37<vektor>Are there working drivers?
14:37<Chutt>well, there's reference drivers that were leaked
14:37<vektor>Oh, ok.
14:38<Chutt>they're hacky, and not allowed to be redistributed
14:38<hurdel>the last thing we need is myth itself becoming questionably legal, stay strong chutt
14:38<vektor>I'm curious though, does DScaler support it?
14:38<Chutt>doubt it
14:38<Chutt>it doesn't have a bt8x8 in it
14:39<vektor>Ah, ok this is interesting.
14:39<vektor>DScaler does not support the PVR-250.
14:39<Chutt>or the 350, i'd assume =)
14:39<vektor>It does support the 'WinTV-PVR' which has an 878 in it. :)
14:40<Chutt>the wintv-pvr is discontinued, now
14:40<Chutt>or will be soon, iirc
14:41<Chutt>so the guy's bitching about how mpeg2 support in the tv portion of the app is necessary to watch dvds and junk
14:41<Chutt>when, well, the tv stuff's for watching tv =)
14:42<Chutt>and the only reason it'd need mpeg2 support is for such cards that produce only mpeg2
14:42<vektor>Or for HDTV recording, eh?
14:42<Chutt>right, but no hdtv cards in linux =)
14:42<Chutt>and you wouldn't want to do software decoding of hdtv, anyway
14:43<Chutt>most of the hdtv cards can only encode or decode, too
14:43<Chutt>which sucks
14:46<vektor>Well, you only need input.
14:47<vektor>You can use the hardware MPEG2 decoding assistance on like the Geforce4.
14:47<vektor>There do exist HDTV inputs usable under linux.
14:47<vektor>A look at who's working on all the MPEG2 TS code in xine/mplayer will show you who has some. :)
14:47<vektor>It's not something I'm going to want any time soon though :)
15:19-!-moegreen [] has quit ["[BX] moegreen has no reason... just kidding :)"]
15:26-!-moegreen [] has joined #mythtv
15:27-!-Edgan [] has joined #mythtv
15:44<mdz_>heh, nice, somebody's mail from August 7th finally made it out of their mail queue to mythtv-dev
15:44<Chutt>i'd more believe that his clock was wrong =)
15:49<Chutt>yeah, since 0.7 wasn't out in august =)
15:50<Chutt>the berkeley encoder he's talking about does do it in semi-real time
15:50<Chutt>it's hackish, though
15:50<Chutt>does more work overall than a single process would do
15:50<Chutt>and uses original frames instead of the already compressed things for calculations needing previous frame
15:51<mdz_>how can that work?
15:51<mdz_>it doesn't have the original frame when decoding
15:51<Chutt>low quality
15:51<mdz_>sounds like lots of noise
15:51<Chutt>recalculates a lot of things
15:52<Chutt>almost got this damn live tv -> recording transition done
15:53<mdz_>hmm, too bad the source code isn't there
15:53<mdz_>I assume this is the project:
15:54<mdz_>heh, their performance results suck compared to what you can buy for $30 today
15:54<Chutt>it's really not necessary
15:55<mdz_>their fastest system did _almost_ one 320x240 frame per second
15:55<mdz_>oh, there is source there. just didn't see it
15:56<mdz_>Chutt: should I move that obsolete 0-7-to-0-8.sql stuff into cvs.sql?
15:57<Chutt>what obsolete stuff?
15:59<mdz_>well, I added a column to the recording tables
15:59<mdz_>and now the recording tables are gone
16:00<Chutt>just get rid of it entirely
16:00<mdz_>ok, and I'll change the table migration to skip the profile column also then
16:00<mdz_>if anyone was actually using it, they should know how to fix their database
16:04<Chutt>i suppose i should go shovel the driveway
16:07<Soopizzle>wait for our warm air to hit oyu
16:08<mdz_>why bother, it'll just snow again tomorrow
16:09<Soopizzle>that's what happened to us
16:09<Soopizzle>warm air set melted all the snow
16:09<Soopizzle>then it snowed 2 days straight
16:18<PhazE3>It has not snowed here in like 25 years
16:19<PhazE3>I love Texas
16:20-!-Edgan [] has quit [Read error: 54 (Connection reset by peer)]
16:22<Soopizzle>what fun is that
16:22<mdz_>hey that's a great idea
16:23<mdz_>just encode half the frame at a time
16:23<mdz_>I wonder why nobody thought of that
16:23<mdz_>and sarcasm
16:27-!-Edgan [] has joined #mythtv
16:53-!-Soopizzle [] has quit [Read error: 54 (Connection reset by peer)]
17:05<PhazE3>Is this hardware mpeg stuff every gonna die :( Im tired of people bitching
17:07-!-bigguy [] has quit [Read error: 113 (No route to host)]
18:01-!-rkulagow [] has joined #mythtv
19:05<lichen_>hmm that was weird, i hit forward and it goes back whiel watching a recording
19:20-!-TheAsp [] has joined #mythtv
19:20<TheAsp>grr, qmake is busted
19:21<TheAsp>(by design)
19:22<Universe>I think its a user error
19:27<TheAsp>no, it just needs to be slapped around every now and then
19:35<Universe>that means its a ID10T error
19:44<lichen_>man navigation is acting al lweird, its like i fast forward a little then it skips back, and trying to rewind like barely works
19:54-!-nevertheless [] has quit [Remote closed the connection]
19:57<hurdel>if i'm installing a fresh cvs, what order do i apply the sql files?
19:59<hurdel>anybody here
20:05<Universe>it would probably be good to apply mc.sql first
20:05<Universe>then the 07-08
20:06<Universe>because the latest changes are in there..
20:06<Universe>so mc first
20:06<Universe>unless mc has all the changes, which it wouldn't hurt applying 07-08...
20:16* TheAspwatches the countdown
20:20<TheAsp>I take it mythweb wont work with the new record table?
20:23<Chutt>theasp, nope, not yet
20:26<mdz>shouldn't be hard to fix
20:28<TheAsp>was thinking that
20:28<TheAsp>i may give it a try next week, if it's not already done
20:46<rkulagow>chutt, can you please sync web page docs with CVS docs. thanks.
20:54-!-rkulagow [] has quit [Read error: 104 (Connection reset by peer)]
21:37<mdz>dammit, my myth box has started hanging lately
21:37-!-moegreen [] has quit [Read error: 104 (Connection reset by peer)]
21:37<mdz>I hope it isn't bad RAM; it has one big stick
21:41<TheAsp>O M G
21:43<TheAsp>it stopped recording for no reason
21:43<TheAsp>no error
21:43<TheAsp>just switched
21:44<TheAsp>so, ah, *cough* what happened in the first 45 mins of john doe
21:51<-- Universehas quit ()
22:27-!-rkulagow [] has joined #mythtv
22:28-!-mythtv [] has joined #mythtv
22:32-!-Universe [] has joined #mythtv
23:00-!-mythtv [] has quit [Read error: 104 (Connection reset by peer)]
23:01<Chutt>so i've got the live tv -> recording -> live tv stuff working
23:01<Chutt>except it deletes the recording when you exit live tv, instead of the ringbuffer file =)
23:05<rkulagow>well, i convinced the wife that moving a PC into the living room isn't a totally crazy idea. i used mythgallery to show grandma all the digital pics we've been taking. weird thing is that every once in a while, mythgallery would act as if i had done a keypress and randomly jump back or forward a few pictures. (this was with the keyboard, not a remote). it wasn't queued up keycommands either. weird.
23:05<Chutt>was it in slideshow mode?
23:06<rkulagow>no, it wasn't. manual advance using the cursor keys.
23:06<TheAsp>chutt: i think i was moving around in the gui and the recording stopped earlier
23:07<TheAsp>just as i pressed esc
23:07<Chutt>with cvs?
23:07<TheAsp>few hours old
23:07<TheAsp>it didnt crash
23:07<TheAsp>it just stopped
23:07<Chutt>was the stop time in the database right?
23:11<mdz>hmm, scheduledrecording gets confused when chanid is null
23:11<mdz>it breaks the query
23:12<mdz>giving it a default should fix it
23:12<Chutt>when is it ever null?
23:14<mdz>when it's Unknown
23:14<mdz>SELECT recordid FROM record WHERE record.title = 'Unknown' AND ((record.type = 4) OR ((record.chanid = ) AND ((record.type = 3) OR (((TIME_TO_SEC(record.starttime) = TIME_TO_SEC('16:30:00')) AND (TIME_TO_SEC(record.endtime) = TIME_TO_SEC('19:00:00')) ) AND ((record.type = 2) OR ((TO_DAYS(record.startdate) = TO_DAYS('2003-01-15')) AND (TO_DAYS(record.enddate) = TO_DAYS('2003-
23:14<mdz> ) ) ) ));
23:14<mdz>unless something else is happening there
23:14<Chutt>that shouldn't even be hitting the db for that, though
23:14<mdz>it looks like it was when I was browsing unknown data in the epg
23:14<mdz>it will never call GetProgramRecordingStatus?
23:15<Chutt>well, it's not supposed to
23:15<Chutt>the unknown programs used to initialize their type so that it thought it wasn't being recorded
23:15<Chutt>so it wouldn't look em up
23:15<mdz>I should avoid it in scheduledrecording anyway, it shouldn't break like that
23:17<mdz>I'll find out how it's getting called though
23:17<Chutt>you probably aren't initializing the type like before =0
23:17<Chutt>since the type lives in the recording class, no?
23:18<Chutt>and yeah, i just got the same thing
23:18<mdz>I don't remember seeing that though
23:18<mdz>I would have replaced it with the equivalent initialization
23:18<mdz>where did it happen?
23:20<Chutt>or, maybe i'm misremembering it
23:21<mdz>I looked in the old guidegrid and the old programinfo and I can't find anyplace where it's doing that
23:21<Chutt>the old code did it that way, looks like the faster stuff that that guy wrote doesn't
23:22<Chutt>so, now this other stuff works
23:22<mdz>what's the right way to check if it's unknown?
23:22<Chutt>i dunno
23:23<Chutt>could always test for title and category to be those two values set in the ui
23:23<Chutt>but, hmm
23:23<mdz>that seems to be the only way
23:23<mdz>unless you want to add a flag to programinfo, which seems unnecessary
23:23<Chutt>can you initialize a dummy scheduledrecording class for the unknown programs and just skip all of it that way?
23:24<mdz>yeah, it could
23:24<Chutt>heh, oops
23:25<mdz>it'll have the same effect once I fix scheduledrecording though
23:25<Chutt>can't exit the playback if it's gone from live-tv -> recording + watching
23:25<mdz>because it defaults to notrecording
23:25<Chutt>is it possible to select one of those unknown programs for recording?
23:26<mdz>should it prevent that?
23:26<mdz>did it before?
23:26<Chutt>i believe so
23:26<mdz>happen to remember how?
23:27<mdz>applyrecordingstatechange sure didn't
23:27<Chutt>maybe it didn't
23:27<TheAsp>either of you look at that perl script to fix show end times? would applying it regulary be safe?
23:27<Chutt>theasp, yeah, it should be
23:27<mdz>I took one look at it and ran the other way
23:27<TheAsp>i rewrote parts of it
23:27<Chutt>it'd be easy to do the same thing in mythfilldatabase, though
23:27<Chutt>and it'd be much simpler
23:27<mdz>oh, wait, that was the shell script which pruned the db
23:27<Chutt>same as that
23:28<Chutt>easier to do in the program
23:28<mdz>better to fix xmltv though
23:29* TheAspruns away from xmltv :P
23:29<mdz>at least report a bug about it
23:29<Chutt>they've discussed it on their list, iirc
23:31<Chutt>allright, i'm committing this junk to cvs
23:31<mdz>what's this?
23:31<mdz>live tv transition stuff?
23:31<Chutt>the live tv -> recording -> live tv
23:31<Chutt>that dialog that pops up and junk
23:31<mdz>working well?
23:31<TheAsp>chutt: aparently I have 50000 seconds until I can test it
23:31<Chutt>seems to be
23:32<Chutt>only thing that doesn't work is if you exit the player process while it's doing that
23:32<Chutt>i need to make the backend transition to none instead of back to live tv
23:32<Chutt>but, oh well, i'll do that later =)
23:33<Chutt>fixed a couple other bugs, too
23:33<Chutt>the secondary control sockets that get opened up for like file transfer and the live tv buffer were getting sent events
23:33<Chutt>and they should've been
23:34<Chutt>might've caused some network desyncing
23:34<mdz>did they know they were getting events, or do they only expect a data stream?
23:34<Chutt>they didn't know they were
23:35<mdz>so those events were just getting lost?
23:35<Chutt>no, the main control socket was getting them too
23:35<Chutt>these were getting sent data, and weren't reading it in
23:35<Chutt>so next time they'd read something, they'd get the old event, instead of what they were supposed to get
23:35<mdz>did they get garbage at the beginning the next time they did expect data?
23:35-!-bigguy [] has joined #mythtv
23:36<Chutt>it shouldn've happened very often, but it could've
23:36<mdz>my only regression from 0.7 is that stutter at the beginning of playback
23:36<mdz>dunno if that happens to you still
23:36<Chutt>it does
23:36<mdz>what a pain to find though
23:37<Chutt>i _think_ it's because the audio's not really playing back yet
23:37<Chutt>at the beginning
23:37<mdz>why's that different now?
23:37<Chutt>well, before
23:37<Chutt>it'd parse through the beginning of the file looking for the first audio frame
23:38<mdz>I checked in that scheduledrecording fix, so it doesn't blow up anymore
23:38<Chutt>and it'd use its effective dsp to start off with
23:38<mdz>I'm inclined to leave the guidegrid alone for now
23:38<mdz>that thing scares me
23:38<Chutt>to speed up startup, i just made it use the recording frequency as a starting point
23:38<Chutt>instead of looking for the frame
23:38<Chutt>but, i dunno if that caused it or not
23:38<Chutt>heh, what's wrong with the guidegrid? =)
23:39<Chutt>it was simpler, once
23:39<mdz>no matter how big I make my editor window, I can't get a big enough piece to actually see what's going on
23:39<mdz>I blame your brace style :-P
23:41<Chutt>made the scheduler class a qobject
23:41<mdz>is it really that much slower to just load all of the data for a day, instead of querying for lots of little pieces?
23:41<Chutt>so now it can get events broadcast to it
23:41<mdz>what sort of events?
23:41<Chutt>you mean, querying is that much slower than loading all of the data for a day?
23:42<Chutt>or what?
23:42<mdz>no, I assumed it loaded it in pieces so that it didn't takea long time to start up
23:42<Chutt>what are we talking about
23:42<Chutt>the epg?
23:42<Chutt>what pieces?
23:43<mdz>when I scroll each 30 minutes, it loads more data from the db
23:43<Chutt>i suppose it could load a whole day's worth
23:43<Chutt>and then just scroll through that
23:43<mdz>yeah, unless it took too long to start up
23:43<mdz>I don't think it would on my setup
23:43<Chutt>i don't think it would here, either
23:44<mdz>I should hope not, since yours should be way faster with the local db and local root fs and all :-)
23:44<Chutt>as for the events, i'm using it for the replies from the frontend to the 'can i record using this tuner' question for now
23:44<Chutt>but it'll also get used for when another backend comes online or goes away
23:45<Chutt>to tell it to redo all the scheduling
23:48<Chutt>stuff's in cvs now
23:49<Chutt>i said in the commit message to do a 'make distclean'
23:50<Chutt>i dunno how necessary that is, though
23:51<mdz>there are a couple of places in guidegrid.cpp where it looks eerily similar to other parts of the file but is subtly different :-)
23:52<Chutt>that's moegreen's code =)
23:52<Chutt>all the alternate layout stuff
23:52<mdz>a big cut and paste job?
23:53<mdz>I did find one place where it was loading the recording status and not using it
23:53<Chutt>the different arrow types are all really similar
23:53<mdz>but it probably used it later anyway
23:54<mdz>it'd be interesting to try loading all of the program info and recording status info in one big query
23:54<mdz>see if it's workable
23:54<Chutt>it'd cut down on a number of queries
23:54<mdz>then it wouldn't need to touch the database at all except to set up recordings
23:54<Chutt>and that's always good
23:54<mdz>should scroll very fast
23:55<Chutt>just don't trade off too much complexity =)
23:55<Chutt>slow code that's easier to understand is almost always better than horribly fast junk that you can't read
23:55<mdz>the one function to load all the info might be a little hairy
23:55<mdz>but it would simplify all the movement code
23:56<Chutt>all the movement code now is 'change start points, reload all'
23:56<Chutt>since it's 6 queries for the screen, not counting the recording types
23:57<mdz>the recording status currently does a query for every single programinfo
23:57<Chutt>which is another 30, max
23:57<Chutt>or so
23:58<mdz>which is better than the 3 per program it did before, but still not so good
23:59<mdz>there is quite a bit of stuff in guidegrid that I cannot read though :-)
23:59<Chutt>like i said, it was simpler, once =)
23:59<mdz> mins = 5 * (mins / 5);
23:59<mdz>like that