Back to Home / #mythtv / 2008 / 07 / Prev Day | Next Day
#mythtv IRC Logs for 2008-07-24

---Logopened Thu Jul 24 00:00:40 2008
00:06-!-foxbuntu [] has joined #mythtv
00:06-!-foxbuntu [] has quit [Remote closed the connection]
00:09-!-MavT [] has joined #mythtv
00:17-!-foxbuntu [] has joined #mythtv
00:24-!-superm1 [n=superm1@ubuntu/member/superm1] has quit ["Leaving"]
00:28-!-superm1 [n=superm1@ubuntu/member/superm1] has joined #mythtv
00:28-!-MaverickTech [] has quit [Read error: 110 (Connection timed out)]
00:48-!-lepsie [] has joined #mythtv
00:49-!-leprechau [] has quit [Read error: 104 (Connection reset by peer)]
01:01-!-foxbuntu [] has quit [Remote closed the connection]
01:02-!-lepsie [] has quit [Client Quit]
01:02-!-foxbuntu [] has joined #mythtv
01:02-!-leprechau [] has joined #mythtv
01:03-!-gnome42 [] has quit [Remote closed the connection]
01:04-!-briand [] has quit [Read error: 110 (Connection timed out)]
01:04-!-briand [] has joined #mythtv
01:20-!-foxbuntu [] has quit [Remote closed the connection]
01:22-!-foxbuntu [] has joined #mythtv
01:23-!-XChatMav [] has joined #mythtv
01:26-!-jhulst [n=jhulst@unaffiliated/jhulst] has quit [No route to host]
01:30-!-famicom [] has joined #mythtv
01:40-!-MavT [] has quit [Read error: 110 (Connection timed out)]
02:00-!-kali67 [] has joined #mythtv
02:02-!-foxbuntu [] has quit ["Leaving"]
02:08<xris>hmm, mythtv iphone app. wonder how many licenses that violates... (wishing I had an iphone to play with it)
02:15<clever>xris: turns the iphone into a mythtv remote?
02:16<Chutt>xris, shouldn't violate anything
02:16<Chutt>it's just a remote
02:17<clever>Chutt: i was thinking of making the same thing on my treo but the bluetooth code isnt compiling
02:17<clever>and one of the posts there mentions a treo one allready existing
02:23-!-|gunni| [] has joined #mythtv
02:24-!-_gunni_ [] has quit [Read error: 60 (Operation timed out)]
02:34-!-mattwire [] has joined #mythtv
02:37<xris>Chutt: from what I understand, iphone licensing doesn't allow open source software... and depending on how much mythtv code was *looked* at for reference, it could be considered a gpl derrivative work
02:37<xris>just an interesting concept
02:37<Chutt>that's a weird definition of derivative work
02:38<xris>lose definition of "could be" :)
02:38-!-edannenbe [] has joined #mythtv
02:38<xris>couldn't have the mythtv remote if you didn't have mythtv. one is derived from the other.
02:39<xris>would be nicer if apple just allows FOSS stuff on the iphone
02:39<Chutt>again, a weird definition of a derivative work
02:40<xris>speaking of which, it's now the top story on slashdot.. the licensing thing, not the remote app
03:05-!-xris [] has quit []
03:17-!-Netsplit <-> quits: iamlindoro, famicom, RaYmAn-Bx, Kevin`, Winkie, laga, til, phedny
03:17-!-Netsplit over, joins: Winkie
03:17-!-Netsplit <-> quits: phunguy, mattwire, meshugga, sutula, CDev, Cougar, compbrain, renato_, prg3, leprechau, (+66 more, use /NETSPLIT to show all of them)
03:17-!-Netsplit over, joins: famicom, phedny, iamlindoro, RaYmAn-Bx, til, Kevin`, laga
03:18-!-Netsplit over, joins: edannenbe, mattwire, |gunni|, kali67, XChatMav, briand, leprechau, superm1, noisymime, Beirdo (+66 more)
03:19-!-kali67 [] has quit [Remote closed the connection]
03:19-!-kali67 [] has joined #mythtv
03:38-!-MaverickTech [] has joined #mythtv
03:54-!-naeo [n=asl@] has joined #mythtv
03:55-!-XChatMav [] has quit [Connection timed out]
03:57-!-bex [] has joined #Mythtv
03:57-!-bex [] has left #Mythtv []
04:04-!-famicom_ [] has joined #mythtv
04:13-!-famicom [] has quit [Read error: 110 (Connection timed out)]
04:38-!-edannenbe [] has quit [Read error: 113 (No route to host)]
04:46-!-sphery_ [] has joined #mythtv
04:52-!-Dibblah [] has joined #mythtv
04:59-!-sphery [] has quit [Read error: 110 (Connection timed out)]
05:01-!-kali67 [] has quit []
05:59-!-dagar [] has quit [Read error: 110 (Connection timed out)]
06:01-!-dagar [] has joined #mythtv
06:07-!-Dibblah [] has quit [Read error: 113 (No route to host)]
07:13-!-edannenbe [] has joined #mythtv
07:17-!-naeo [n=asl@] has quit ["What's the point of going out? We're just going to wind up back here anyway"]
07:31-!-MaverickTech [] has quit [Read error: 110 (Connection timed out)]
07:35-!-joobie [] has joined #mythtv
07:55-!-Netsplit <-> quits: iamlindoro, RaYmAn-Bx, Kevin`, laga, til, phedny
07:57-!-phedny [n=mark@2001:610:656:0:0:0:0:115] has joined #mythtv
07:57-!-iamlindoro [] has joined #mythtv
07:57-!-RaYmAn-Bx [] has joined #mythtv
07:57-!-til [] has joined #mythtv
07:57-!-Kevin` [] has joined #mythtv
07:57-!-laga [] has joined #mythtv
08:00-!-joobie [] has quit [Connection reset by peer]
08:00-!-joobie [] has joined #mythtv
08:00-!-edannenbe [] has quit [Read error: 113 (No route to host)]
08:06-!-lyricnz [] has joined #mythtv
08:10-!-joobie [] has quit ["This computer has gone to sleep"]
08:34-!-MaverickTech [i=Maverick@] has joined #mythtv
08:55-!-Dibblah [] has joined #mythtv
09:11-!-Cardoe [n=Cardoe@gentoo/developer/Cardoe] has joined #mythtv
09:24-!-Wobbo [] has joined #mythtv
09:25<Wobbo>is there a UI developer channel?
09:26<Wobbo>simple question can't find documentation in
09:26<Wobbo>so where do i find list of the variable used in the xml docs like: %CHANNUM%
09:26<Wobbo>i am looking for channel name
09:33-!-jarle [] has quit [Remote closed the connection]
09:39-!-jarle [] has joined #mythtv
09:45<Cougar>Wobbo: doesn't existing OSDs give you any hint? You can always read a code too ;-)
09:47<Wobbo>no none of theme use channel name they all use the call sign
09:48<clever>how long till mythtv can send/recv on that:P
10:09-!-moodboom [] has joined #mythtv
10:23<Wobbo>so there is no documentation on that?
10:27-!-hyperactivecrond [] has joined #mythtv
10:27-!-hyperactivecrond [] has left #mythtv ["Leaving"]
10:30-!-gnome42 [] has joined #mythtv
10:47-!-MaverickTech [i=Maverick@] has quit [Read error: 110 (Connection timed out)]
11:02-!-edannenbe [] has joined #mythtv
11:06-!-Chutt [] has quit [Remote closed the connection]
11:20-!-foxbuntu [] has joined #mythtv
11:22-!-MrGandalf [] has joined #mythtv
11:27<stuartm>Wobbo: try "channel"
11:28*stuarta waves
11:29<stuartm>stuarta: hi, still hasn't arrived ...
11:29<stuarta>hmmm, will check with girly see when she posted it...
11:29<stuartm>at a glance I'm not sure it will work for all instances of the program_info OSD, but generally we use a programinfo map - see libs/libmythtv/programinfo.cpp ProgramInfo::ToMap()
11:32-!-edannenbe [] has quit [Remote closed the connection]
11:33<Wobbo>stuartm: "channel" is %CHANNUM% %CALLSIGN%
11:35-!-falstaff [] has joined #mythtv
11:36<falstaff>i would like to add some features to mythvideo. Im using Ubuntu 8.04, can i combine this with mythvideo from trunk?
11:37<stuarta>ubuntu != trunk
11:37<stuartm>%LONGCHANNEL% looks like the closest you'll get
11:37<laga>yeah, but ubuntu = release-21-fixes by default ;)
11:37<falstaff>yes i know
11:37<stuartm>without supplying a patch
11:37<stuarta>so you need to uninstall the packages from ubuntu
11:37<stuarta>and install svn head from source
11:38<falstaff>so i have to use whole mythtv from trunk?
11:38<stuartm>Wobbo: but the exact format can be changed in the settings - see ChannelFormat and LongChannelFormat
11:38<falstaff>i cant just compile mythvideo from trunk and use it with the packages from ubuntu?
11:39<stuartm>Wobbo: Settings > TV Settings > General (first page)
11:39<falstaff>stuarta, Can i download mythvideo source from ubuntu and send patches to the mailing list?
11:39<stuarta>there is no point patching 0.21 as things are being rewritten all the time
11:40<stuartm>mythvideo looks a lot different right now, even trunk isn't the right version to work against
11:40<falstaff>stuarta, even in mythvideo?
11:41<stuartm>falstaff: mythvideo has changed a lot as part of the port to mythui
11:41<falstaff>hm ok
11:41<falstaff>but when trunk isnt the right version, which one is?
11:41<stuartm>37 files changed, 5258 insertions(+), 10357 deletions(-)
11:41<stuartm>which hopefully in a week or so will be merged back to trunk
11:44<falstaff>okey thanks
11:56<falstaff>Do you develop on a virtual machine? does mythtv works on a virtual machine? It uses opengl right?
11:56<stuarta>it can use opengl
11:56<stuarta>doesn't have to
11:57<stuarta>it's hard to develop on a virtual machine, due to the hardware access required to capture content
12:00-!-huhlig [] has joined #mythtv
12:05<falstaff>from wiki: Use Myth for GUI tasks. Isaac is currently working on a UI layer that's independent of Qt, and patches that use native Qt widgets or extend the Myth wrappers around the Qt GUI will be rejected... => this is mythui right?
12:07<stuartm>although that advice is a little dated now, even the official MythUI wiki page is out of sync -
12:11<stuartm>falstaff: what sort of features are you planning?
12:11<falstaff>hm small improvments...
12:12<falstaff>what i would like to have is something who tells me which videos are new...
12:13-!-eharris_ [] has quit [Read error: 110 (Connection timed out)]
12:13<falstaff>so adding a "dateadded" field to videometadata and then something on the gui.. i thougt about a "faiding star", which does an alphablending, from full visible for new movies to invidible for movies older then x (x=3months..)
12:14<stuartm>ok, sounds good - there is a definite need to improve the metadata stored for videos, converging with what we store for recordings etc - might not be a bad idea to keep them similar in implementation so that mythvideo can be integrated into the frontend eventually using the same code
12:15<stuartm>star idea is nice, and easier with mythui too
12:16-!-stuartm is now known as gbee
12:17<falstaff>the video* tables are used by mythvideo only right?
12:18<falstaff>When does mythvideo generate this on the database? first loading?
12:18<gbee>statetype maybe then themers can use whatever they want, a fading star, a traffic light image, wilting plant ;)
12:19<gbee>very first time it's loaded by mythfrontend as a plugin
12:19<falstaff>Ah okey... And how is migration done?
12:19<gbee>database migration from one install to another?
12:20<falstaff>version to version...
12:21<gbee>oh, well we compare a database version number in the settings table to the one hardcoded in the version you are using, then a series of statements in dbcheck.cpp are run on your database until it's updated to the same version
12:23<gbee>updates from version 1015 to 1016, the current version number is changed at the top of dbcheck.cpp - "const QString currentDatabaseVersion = "1018";"
12:23<falstaff>ah cool
12:24<falstaff>and this version number is for all myth* applications?
12:24<gbee>no, the version is unique to the plugin
12:25<falstaff>But settings table is from mythtv not mythvideo right?
12:25<falstaff>DBSchemaVer right? on my db its set to 1214
12:26<gbee>no it's shared, but each plugin stores their version with a different name e.g. mythvideo.DBSchemaVer
12:27<gbee>DBSchemaVer is the main schema version for the frontend/backend, then each plugin has their own - GalleryDBSchemaVer, MusicDBSchemaVer
12:27<falstaff>ah okey
12:27<gbee>mythvideo seems to name it slightly differently for some reason
12:28<falstaff>So a new field for mythvideo leads to this changes: dbcheck.cpp, const QString currentDatabaseVersion = "1019"; and a upgradestatement like the one you sent (
12:30<falstaff>how do you develop? any ide?
12:30<clever>i do most of my code tweaking with plain old vim
12:31<stuarta>same here
12:32<clever>i recently found out i can do code folding in vim:P
12:32<clever>helps alot
12:33<Wobbo>stuartm: thanks
12:38<falstaff>i like vim :-)
12:40<falstaff>It leads to reading the framework documentations... which is never wrong :-)
12:41<falstaff>mythui_mythvideo is built by the makefile from mythtv trunk right?
12:44-!-xris [] has joined #mythtv
12:44-!-Wobbo [] has quit ["Ex-Chat"]
12:47<gbee>probably should have made it a full branch of trunk, but I didn't so ..
12:52<sphery_>falstaff: just a quick mention, though most video* tables are MythVideo, videosource is mythtv. I think Anduin had ideas for changing table names, so if you're doing a /lot/ of changes, might want to talk to him.
12:55<Anduin>I just like easily identifiable (why new mythvideo settings have mythvideo. before them), videosource is easy enough to remember that videotablename is unique enough for me.
12:57<danielk22>stuarta: Do you use MHEG at all? MHIContext needs to be converted to use QPainterPath instead of Q3PointArray...
12:57<Anduin>(though at one point I did consider changing all mythvideo tables to be more like mythmusic, with video_)
13:00-!-kormoc [n=kormoc@unaffiliated/kormoc] has joined #mythtv
13:00-!-eharris [] has joined #mythtv
13:07<stuarta>danielk22: i generally have it turned off, but it's easy enough to turn it on for testing
13:10<danielk22>It might be possible to just rewrite ::makeEllipse() if the conversion is difficult otherwise.
13:10<danielk22>I'm trying to eliminate as much Qt3 stuff in libmythtv as possible.
13:11<stuarta>would you like me to take a look?
13:11<danielk22>I'd love it :)
13:11<stuarta>nothing like the deep end theory
13:11<stuarta>cheers. i'm gunna need that
13:17<stuarta>UI101 - The basics of porting from qt3->qt4
13:17*stuarta enrolls
13:19*gbee graduated already
13:20<stuarta>you are already doing your masters
13:22-!-kali67 [] has joined #mythtv
13:22<gbee>I need a tree class in libmythui and I'm thinking of lifting GenericTree from libmyth, anyone want a copy to remain available in libmyth for non-ui purposes?
13:28<danielk22>it's used for the OSD stuff.. still UI, but not MythUI
13:28<stuarta>QObject: Do not delete object, 'unnamed', during its event handler!
13:28<stuarta>well that's nice
13:31<gbee>danielk22: my grepping missed that, or rather I saw OSDGenericTree but didn't see it's connection to GenericTree - I can't remove the libmyth version immediately even if I wanted to since some unconverted plugins/screens still depend on it
13:31<gbee>I'll use a renamed copy for now
13:32<gbee>stuarta: yeah, I'm seeing that two/three times on startup
13:32<gbee>doesn't help that the object hasn't been given a name :)
13:35<stuarta>wonder if there's a way of naming it based upon it's parent name
13:36<stuarta>man there is some crap on that satellit
13:37<stuarta>ooo. i found african soccer!
13:37<gbee>supposed to be some more channels getting EIT soon
13:38<gbee>and the FTA launch of Five
13:39-!-kali67 [] has quit [Remote closed the connection]
13:39-!-kali67 [] has joined #mythtv
13:39-!-jarle [] has quit [Remote closed the connection]
13:40-!-jarle [] has joined #mythtv
13:42<stuarta>hmmm, luxe tv hd (54206) gives me the following
13:42<stuarta>[h264 @ 0x7fa209d17f30]illegal aspect ratio
13:49-!-famicom_ [] has quit ["Leaving"]
13:58-!-famicom [] has joined #mythtv
13:58-!-iamlindoro_ [n=iamlindo@] has joined #mythtv
14:04<MrGandalf>stuarta: uh oh, police are coming now.. though that's not as bad as an illegal instruction I hear.
14:04<gbee>stuarta: it works though?
14:04<stuarta>but so does most HD on this box
14:05<gbee>plays smoothly for me with the skiploop patch
14:05<stuarta>may need to try that someday
14:07<stuarta>some eitfixup work needed on 52361 atm
14:09<stuarta>i am just wondering down the channel list before anyone accuses me of looking at *really* crap channels
14:09<stuarta>wandering even
14:14-!-jgarvey [] has joined #mythtv
14:37<stuarta>when this happens -> " Waited too long for decoder to pause" we need to do something other than wait
14:37<stuarta>even accepting input would be nice
14:44-!-dagar [] has quit [No route to host]
14:48<MrGandalf>can someone explain this "skiploop" patch? I heard something about it on the ffmpeg-devel list as well.
14:52<gbee>disables a cpu intensive de-blocking filter on the h.264 decoder, it's need for some streams but generally not for high quality broadcast stuff like the BBC
14:53<MrGandalf>ah, ok.. different than what's being talked about on the list then
14:53<gbee>disabling it reduces the CPU requirements for AVC with ffmpeg
14:53<MrGandalf>coreavc as well I believe
14:53-!-_gunni_ [] has joined #mythtv
14:53<gbee>not by much, but enough for some
14:54-!-|gunni| [] has quit [Read error: 60 (Operation timed out)]
14:54<gbee>coreavc allows it to be disabled as well, but the mythtv patch in question only works for ffmpeg
14:57<gbee>stuarta: there are a couple of conditions where we get 'stuck', that's one and I agree we need to accept input and even reduce the number of loops before give up
14:57<stuarta>glad i'm not the only one
15:12<gbee>fixing some of the underlying causes wouldn't hurt either, I'm not sure that the decoder hasn't paused when we print that error
15:13<stuarta>it is however definitely stuck
15:26-!-cattelan [] has joined #mythtv
15:39-!-cattelan_ [] has quit [Read error: 110 (Connection timed out)]
15:40-!-cattelan [] has quit [Remote closed the connection]
15:43-!-cattelan [] has joined #mythtv
15:43-!-jgarvey_ [] has joined #mythtv
15:54*stuarta suspects converting MHEG to QT4 isn't a 1hr job
15:55<danielk22>stuarta: just convert that one thing, I've already done the rest here.
15:56<gbee>I'm happy to test any patches
15:56<stuarta>if you don't hear of success by the end of the night, assume i didn't get it done
15:57-!-jgarvey [] has quit [Connection timed out]
15:58<gbee>not personally in the mood to do any work tonight, aside from testing patches
15:59<gbee>I need this one last widget for mythui before I can put the mythvideo port to bed, but it will have to wait
16:00<stuarta>it appears it's only used in MHIDLA::DrawOval, MHIDLA::DrawArcSector, MHIDLA::DrawPoly and that is it
16:00-!-jgarvey_ [] has quit [Read error: 60 (Operation timed out)]
16:00<stuarta>danielk22: that all you after?
16:01<danielk22>QPainterPath instead of Q3PointArray in MHIContext
16:02<danielk22>for the ellipse at least
16:05<stuarta>which file are you looking at libs/libmythtv/mhi.[h|cpp] ?
16:05-!-Chutt [] has joined #mythtv
16:11-!-iamlindoro_ [n=iamlindo@] has quit [Read error: 60 (Operation timed out)]
16:15-!-jgarvey_ [] has joined #mythtv
16:19-!-iamlindoro_ [n=iamlindo@] has joined #mythtv
16:39-!-iamlindoro_ [n=iamlindo@] has quit [Read error: 104 (Connection reset by peer)]
16:41<stuarta>anyone else's frontend take ages to shutdown (svn head)
16:41<gbee>been that way ever since the move to QT4, starts faster but takes ages to exit
16:42<danielk22>takes ages to fire up gdb on a coredump too
16:42<stuarta>never seen it take soooooo long
16:42<gbee>:( conflicts between the mythvideo branch and trunk
16:43-!-MrGandalf [] has quit [Remote closed the connection]
16:44-!-dagar [] has joined #mythtv
16:48<stuarta>looks like mheg is borked in head
16:49<stuarta>i don't get any of the background colours
16:49<stuarta>can't actually compare to 0.21 atm since it's busy recording on the wrong mux :(
17:00<gbee>mine's recording too, on all tuners - I'll check tomorrow
17:08<stuarta>unless i'm reading this wrong i just found a bug
17:08<stuarta>args to putPoints are arse about
17:08<stuarta>at mhi.cpp 1266
17:09-!-huhlig [] has quit [Read error: 110 (Connection timed out)]
17:19<stuarta>yup the mheg's definitely screwed in head, and working in 0.21
17:24-!-dekarl1 [] has joined #mythtv
17:31-!-dekarl [] has quit [Read error: 145 (Connection timed out)]
17:34-!-moodboom [] has quit [Client Quit]
17:45-!-PointyPumper [n=pintlezz@] has quit [Read error: 110 (Connection timed out)]
17:46-!-jgarvey_ [] has quit ["Leaving"]
17:47-!-lyricnz [] has quit []
17:48-!-dekarl1 [] has left #mythtv []
17:54<gbee>it would probably have to be QT4 related, but I wouldn't know where to start looking
17:57<stuarta>probably due to the qt3 compat classes being not quite up to scratch
17:57<stuarta>making progress
17:57-!-foxbuntu [] has quit ["Leaving"]
17:58<stuarta>up to the hard bit now
18:03<Chutt>if you're touching the mheg library
18:03<Chutt>would it be hard to port to qt4?
18:04<gbee>that's what he and Daniel are doing
18:04<stuarta>that's what i'm attempting :)
18:04<Chutt>ah, excellent
18:04<Chutt>i looked at that briefly
18:04<Chutt>but decided i couldn't test it so didn't =)
18:04<stuarta>i'm currently enrolled in UI101, the basics of qt3 -> qt4 conversion :)
18:04<gbee>in the process noticing that it's currently a little broken since the initial QT4 port
18:06<gbee>Chutt: thinking of copying GenericTree to mythui as the basis of the tree widget, does this fit with what you had in mind?
18:06<Chutt>unless there's something better in qt4
18:06-!-Cardoe [n=Cardoe@gentoo/developer/Cardoe] has quit ["Leaving"]
18:08<gbee>I've had a quick look, mostly browsing the list of classes for anything that has tree in the name ;) but only things I found were complete widgets, no generic tree stuff
18:08<Chutt>i thought they were more model/view type stuff now
18:08<Chutt>maybe not
18:10<gbee>and the more specific -
18:12<gbee>neither fits what I was looking for, but generic tree works well enough and it's already used for mythvideo/mythmusic etc so that simplifies things - I can drop their trees direct into the widget
18:13<gbee>SetTree(GenericTree video_list); - that sort of thing
18:16-!-dagar [] has quit [Remote closed the connection]
18:18-!-dagar [] has joined #mythtv
18:18<stuarta>what i'm trying to work out is why it does down to such a low level
18:19<stuarta>it appears to draw things point by point, line by line
18:23<gbee>coords use less bandwidth than images?
18:24<stuarta>seems to talk a lot about compositing an image and then alpha blending it
18:24<gbee>hmm, no idea
18:26<stuarta>it's building a QImage pixel by pixel
18:29<gbee>using vector based drawing makes a lot of sense for the bulk of what mheg does, since it scales well and like I said uses a lot less bandwidth, but that doesn't mean we're drawing it as efficiently as we can
18:29<stuarta>undoubtly not
18:29<gbee>but I really would have to defer to Dave Matthews on this
18:29<stuarta>i'm sensing i've opened a can of worms
18:30<stuarta>it seems the MHEG does describe stuff as vector graphics
18:30<danielk22>It doesn't need to be super efficient...
18:30<stuarta>step 1. make work :)
18:30<danielk22>it has vector graphics + videos
18:30<stuarta>step 2+. make work fast
18:30<gbee>and non vector graphics, for logos etc
18:30<stuarta>they are bitmaps
18:31<gbee>danielk22: no it doesn't have to be super efficient because once drawn those images are cached - though I'd like to see disk based caching too since it's very slow to initially load and draw when entering livetv
18:32<gbee>most of the artwork doesn't change, so it could be cached to disk and loaded much faster
18:32<falstaff>what is mheg?
18:32-!-cattelan [] has quit ["This computer has gone to sleep"]
18:32<stuarta>that interactive tv crap
18:32<gbee>similar to MHP used in Europe?
18:32<falstaff>hm, we dont have this in europe right?
18:33<stuarta>tho MHP is basically java
18:33<stuarta>or javascript
18:33<stuarta>they aren't the same
18:34<danielk22>They don't really have any of this in the US. The standard exists, but no one implements it. sort of like teletext in the US :)
18:34<gbee>it's a little similar to teletext but much updated and improved, used in the UK and NZ instead of teletext for digital tv
18:36<hads>It's annoying to parse EPG out of too.
18:36<stuarta>thats the understatement of the century
18:37<hads>We (NZ) have DVB-S which has the standard EIT stream and DVB-T which only has MHEG.
18:38*stuarta sp0rks NZ dvb-t
18:38<hads>Yah. They don't even broadcast DVB-T where I am so I don't get to play with it. Although the MHEG stream is in the DVB-S too.
18:38<hads>HD is only broadcast on DVB-T
18:42-!-joobie [] has joined #mythtv
18:51<falstaff>good nigh
18:51-!-falstaff [] has quit ["have to sleep :-)"]
18:57-!-cattelan [] has joined #mythtv
18:58-!-jhulst [n=jhulst@unaffiliated/jhulst] has joined #mythtv
19:06-!-Dibblah [] has quit [Read error: 113 (No route to host)]
19:21<Chutt>danielk22, for your qt3 removal patch, careful of the autodelete feature of q3ptrlist
19:21-!-cattelan [] has quit ["This computer has gone to sleep"]
19:26<gbee>or the lack of it in qt4
19:26<Chutt>nevermind, looks like it's mostly gone
19:26<Chutt>this is just removing the Q3 classes, though
19:28-!-PointyPumper [n=pintlezz@] has joined #mythtv
19:35-!-huhlig [n=huhlig@] has joined #mythtv
19:37<stuarta>well it's no worse
19:37-!-MaverickTech [] has joined #mythtv
19:38-!-mattwire [] has quit [Remote closed the connection]
19:41-!-compbrain [] has quit [Remote closed the connection]
19:42<stuarta>i can safely say i've made it no worse
19:42<stuarta>it's however still completely broken
19:49<danielk22>chutt: yup, that seemed to be one of the main reason people were using the Qt 3 containers.
19:50-!-joobie [] has quit [Read error: 104 (Connection reset by peer)]
19:50-!-joobie [] has joined #mythtv
19:50<stuarta>QRasterPaintEngine::begin: Unsupported image format (3)
19:52<stuarta>eh? "Note that this functionality is only available in Qt for Embedded Linux"
20:01-!-famicom [] has quit ["Leaving"]
20:01<danielk22>methinks they got a little crazy with 'cruft removal' in Qt4 :)
20:04<gbee>I think someone summed it up nicely when they called it busy work, "So we have to create a new version of QT and make it look like we've done enough work to convince paying customers to upgrade, what can we do?"
20:06<gbee>Cue worker bees renaming a whole bunch of methods and classes for no particular reason
20:08<stuarta>that sums it up
20:15<stuarta>nn all
20:20-!-cattelan [] has joined #mythtv
20:30-!-huhlig [n=huhlig@] has quit [No route to host]
20:41-!-azexian [] has joined #mythtv
20:42-!-azexian [] has left #mythtv []
20:43-!-susanna777 [] has joined #mythtv
20:44<susanna777>hello i have a question about myth tv
20:44<susanna777>hmm... i think i'm on the wrong room
20:44-!-susanna777 [] has left #mythtv ["Leaving"]
20:54-!-foxbuntu [] has joined #mythtv
21:18-!-jk1joel [] has quit [Remote closed the connection]
21:28-!-kormoc [n=kormoc@unaffiliated/kormoc] has quit []
21:48-!-kali67 [] has quit [Read error: 110 (Connection timed out)]
21:54<danielk22>Does anyone know what it is that makes Q3SocketDevice work under Qt4, or why a QUdpSocket replacement wouldn't work in Qt4? Does the class need to be created in a special thread, something else?
21:55<danielk22>Some GlobalFunctionQt4InitNetwork() needs to be called?
21:57<clever>danielk22: ive got it sending signals just fine in the example app the docs provide
21:57<clever>but i cant get it to do the same in mythtv
21:57<clever>i made a patch to check it for data in the event loop of that thread
21:57<clever>basicaly polling it
21:57<danielk22>have you tried doing a "new QUdpSocket()" inside the pthread thread?
21:58<clever>i beleive thats what it is doing
21:58<danielk22>clever: you patched Qt's event loop?
21:59<clever> qsd = new QUdpSocket(this);
21:59<clever>no i patched the custom mythtv event loop
21:59<danielk22>which one?
21:59<clever>that thread isnt using the qt event loop
21:59<clever>i allready posted the patch
22:00<danielk22>ticket# ?
22:00<danielk22>Anyway, so you didn't figure out what was wrong?
22:00<clever>nope but i managed to fix it anyway
22:00<clever>just bypassing the whole signal system
22:01<danielk22>Well if we can't get the signal system to work, I wonder if it's worth it to use the Qt classes at all for this.
22:01<clever>it still makes the stuff simple
22:01<clever>i find
22:01<danielk22>It seems like we're just asking for trouble using a half broken networking system.
22:02<danielk22>What if a different parts are broken on different platforms?
22:02<danielk22>Then we end up writing the posix socket code anyway.
22:03<danielk22>Which is probably more maintainable in the long run anyway by different programmers.
22:03<clever>i only made a few small tweaks and got it working
22:03<clever>some of them i didnt even need to do
22:03-!-jhulst [n=jhulst@unaffiliated/jhulst] has quit [No route to host]
22:04<clever>parts are leftover from trying to fix the signals
22:07<danielk22>Hmmm, it probably is the QThread thing ... All you are doing is emulating what the Qt event loop should be doing.
22:08<clever>some1 else tried giving it its own private qthread
22:08<clever>and it still refused to sendsignals:P
22:08<clever>but with the help of #qt i got it working
22:09<clever>the guy helping me found his own way into trac and dug thru other files to see where the event loop was
22:09<clever>and i couldnt even figured out where the thread was created
22:15<danielk22>heh, I think I'll just rewrite it to use polling.. I want to rewrite it to be thread-safe anyway.
22:16<clever>a lock or 2 arround my patch would probly do that
22:16<clever>if you cant get the lock just skip the socket
22:16<clever>the event loop repeats enough i suspect
22:17<clever>it feels like something thats repeated every frame
22:17<clever>from where it is in the code
22:20<clever>bbl sleep
22:22<Chutt>no polling
22:23<danielk22>chutt: it's blocking
22:23<Chutt>danielk22, all we need to do is change the TV play event loop to be a QThread that runs its own Qt event loop
22:23<Chutt>then hook up the signals from the udpsocket thingie
22:23<Chutt>ie, change the current event loop to be a timer or whatnot
22:24<Chutt>i had started doing that, but got interrupted by work
22:24<danielk22>how far along where you?
22:24<Chutt>not far enough
22:24<Chutt>it'd also fix all those timers you changed back on the original port =)
22:25-!-MavT [] has joined #mythtv
22:25<Chutt>it's the right way to do things, i think
22:25<Chutt>also a bit cleaner
22:26-!-XChatMav [] has joined #mythtv
22:28-!-MaverickTech [] has quit [Read error: 60 (Operation timed out)]
22:31<danielk22>k, it looks like new UDPNotify is always called from the RunTV() thread. So that might work...
22:32<danielk22>now how does one start up a thread specific Qt event loop, any URL handy?
22:33-!-SqWaw [] has joined #mythtv
22:34<danielk22>ah, QThread::exec();
22:44-!-MavT [] has quit [Read error: 110 (Connection timed out)]
23:05-!-dlblog [] has quit [Read error: 60 (Operation timed out)]
23:05-!-hank_ [] has joined #mythtv
23:06-!-hank_ [] has left #mythtv ["Ex-Chat"]
23:16<danielk22>heh, it works. Proof of concept patch on #5281. :)
23:17-!-danielk22 is now known as danielk_Zzzzzz
23:18<Chutt>it may be easier to make TV just inherit QThread
23:18<danielk_Zzzzzz>Yeah, it seems a little snappier too. But I think that's just the event loop lost some of it's QWaitCondition functionality in the Qt4 port..
23:19<danielk_Zzzzzz>Yeah, I just wanted to test the concept quickly, I'd like to make the other threads also use QThread if possible. And they can't all own the TV class :)
23:20<danielk_Zzzzzz>I dunno if Qt handles real time threads though.
23:20<Chutt>it does have thread priority stuff
23:20<Chutt>but not the same realtime
23:20<Chutt>i have a partial patch where i started libmythtv attached to a ticket somewhere
23:20<Chutt>didn't get too far :/
23:21<Chutt>did the recorder, though
23:21<danielk_Zzzzzz>Is there any documentation on how their priorities map to OS priorities?
23:21-!-jhulst [n=jhulst@unaffiliated/jhulst] has joined #mythtv
23:21<Chutt>but the source just queries the highest and lowest priority from pthreads
23:21<Chutt>and maps the Qt range to that
23:22<Chutt>i was thinking of just dumping the realtime stuff
23:22<Chutt>never liked that
23:22<danielk_Zzzzzz>So TimeCritialPriority would just be -19 ?
23:22<danielk_Zzzzzz>instead of -50 or whatever real time is?
23:23<Chutt>one sec
23:24<Chutt>it calls sched_get_priority_min/max()
23:24<Chutt>with 'policy' gotten from a call to pthread_getschedparam()
23:25<Chutt>or pthread_attr_getschedpolicy
23:26<Chutt>looks like those _can_ include SCHED_FIFO, but i don't know if they will
23:27-!-cattelan_ [] has joined #mythtv
23:27<Chutt>realtime threads are really just poor bandaids, though
23:28<danielk_Zzzzzz>yeah, I don't run with em, but they do make the difference for some people.
23:28<danielk_Zzzzzz>anyway, gotta go to bed soon.
23:44-!-cattelan [] has quit [Read error: 110 (Connection timed out)]
---Logclosed Fri Jul 25 00:00:09 2008