:: #mythtv

Daily chat history

Current users (79):

aca20031, aloril, Anssi, Beirdo_, bobweaver, brfransen, Captain_Murdoch, cesman, Chutt, clever, coling, Cougar, danielk22, danielk221, dblain, dekarl, ElmerFudd, fetzerch, foobum, foxbuntu`, ghoti, Gibby, GreyFoxx, IReboot, J-e-f-f-A, jams, jarle, jarryd, jhall_, jheizer, joe_____, joki, jpabq, jpabq_, jpharvey, jst, jwhite, jya, jya_, kc, kenni, kormoc, KungFuJesus, kurre2, kwmonroe, laga, MavT, Moeabm, moparisthebest, MythBuild, MythLogBot, nephyrin, neufeld, Nothing4You, poptix, purserj, rsiebert_, seld, Sharky112065, skd5aner, SmallR2002, sphery, sraue, stuarta, stuartm, superm1, Tobbe5178, toeb, tonsofpcs, tris, unforgiven512, wagnerrp, wahrhaft, whoDat, wolfgang2, XDS2010_, xris, _charly_, _nyloc_
Monday, June 24th, 2013, 00:12 UTC
[00:12:23] jpabq_: jya_: It says "Unknown option: gitsrc"
[00:12:41] jpabq_: I git a "git pull". Is "git update" different?
[00:12:45] jya_: sorry
[00:12:51] jya_: git pull yes
[00:13:01] jya_: oops.. -gitdir
[00:13:16] jya_: too hasty
[00:13:21] skd5aner: gigem: when you return, please check your email (and PMs) – thanks :)
[00:14:26] jpabq_: jya_: I also don't really need the backend. Is it okay to leave off the -enable-backend ?
[00:14:55] jya_: i enable the backend the first time, as it forces extra packages like libmp3lame and libx264
[00:15:07] jya_: but if you don't think you'll ever need those… sure
[00:15:31] jpabq_: okay, sounds like it would be good to do it the first time. Thanks!
[00:24:12] jya: usually, I compile the dependencies in universal mode ; just in case
[00:24:30] jya: though I may not bother building a universal package in the future
[00:25:33] jya: only did so because on my old mini (C2D @ 2GHz) with a nvidia 320M; the 32 bits version runs about 5% faster than the 64 bits one; which is just enough to make it decode some mpeg2 1080i
[00:26:25] jya_ (jya_!~jyavenard@mythtv/developer/jya) has quit (Quit: jya_)
[00:29:57] jya: jpabq: working okay so far?
[00:33:43] jpabq: jya: still building. Definately made it MUCH futher than before.
[00:33:59] jya: what machine do you have?
[00:34:20] jya: the first time on my MBA it takes about 30 minutes ; after that with ccache about 5
[00:34:29] jpabq: dual-core i7 macbook pro. Few years old.
[00:34:38] jya: though I only ever compile once with the packager
[00:35:18] jpabq: I didn't get ccache setup with clang before starting it, so unless the "configure" scripts detected ccache automatically, it isn't using it.
[00:35:20] jya: is your aim to debug on mac (which IMHO is much easier than linux thanks to the great xcode gdb/lldb graphic interface) ; or simply to build a frontend?
[00:35:53] jpabq: Right now, simply to build a frontend. I am curious to try the mac's debugging, though..
[00:35:57] jya: the packager install ccache, and then run configure to use it… so if you build via the packger; you're definitely using it
[00:36:11] jpabq: Ah. Good to know.
[00:36:19] jya: if you want to try the mac debugging, I can give you a .xcodeproj file makes it much easier
[00:36:31] jpabq: Cool. Thanks.
[00:36:37] jya: you can set breakpoints, watch variables. the git integration is great.
[00:38:55] jpabq: I usually run gdb inside emacs under linux. With (setq gdb-many-windows 1) it is not a bad environment.
[00:39:53] jpabq: I have tried netbeans, and such, in the past, but always ended up going back to emacs because those other IDEs hid too much, and I had better control of gdb under emacs.
[00:40:36] jpabq_: Just hit an error.
[00:40:46] jya_ (jya_!~jyavenard@mythtv/developer/jya) has joined #mythtv
[00:41:13] jya: in xcode you see the threads on the left hand side
[00:41:38] jya: it allows to switch context very easily… which I always found emacs or kdbg to be lacking
[00:41:44] jya: what's the error?
[00:41:56] jpabq_: jya: It bombed, but I re-started it and it seems to be making progress.. looks like it is starting to actually build myth now.
[00:42:13] jpabq_: some osx packager error
[00:42:30] jya: should not bombed… it only crashes halfway for me when I build Qt itself
[00:42:50] jya: the WebKit module in the Qt source code, doesn't like parallel make much
[00:43:04] jya: do you have a copy of the error?
[00:45:15] jpabq_: Sorry, I should have cut&paste it before trying to restart the build.
[00:45:34] jpabq_: I was hoping iterm2 would allow me to cut the whole current buffer, but I don't see that option.
[00:46:13] jya: command-A should
[00:46:25] jya: you don't use the standard Terminal?
[00:46:56] jpabq: iTerm2 seems to work much better for me. Especially when I ssh into other machines and use emacs, etc.
[00:47:38] jya: never had an issue with Terminal.. been using it for years. I use it in combination of XQuartz so I can ssh -X to other machine
[00:47:54] jya: I set the history line to 5000 lines of history
[00:48:13] jpabq: Is XQuartz different than the X11 server that comes with OSX?
[00:48:28] jya: well, there's no more X11 server in 10.8
[00:48:34] jya: so I had little choice
[00:48:40] jpabq: Oh? Wow.
[00:49:30] jpabq: I use OpenNX a lot, but that requires an X server. Guess I should do some reasearch before upgrading the OS.
[00:50:03] jya: i had a great deal of trouble getting myth to compile with the latest xcode that shipped with 10.8
[00:50:20] jya: but it was a once off; now it's all in the myth source code and packager
[00:50:40] jya: i need to test the packager with Qt 5 at some stage
[00:51:51] jpabq_: [osx-pkg] Failed system call: " /usr/bin/make -j5 " with error code 2
[00:53:52] danielk221: jpabq: I wasn't able to get MythTV to compile on OSX. It got as far as starting to compile MythTV but then it complained that webkit was missing.
[00:55:57] danielk221: BTW Mark Spieth just got MythTV to compile on the virt I set up to be our new windows buildbot :)
[00:56:05] jya: danielk221: was that using Qt binary packages?
[00:56:38] danielk221: jya: No that had failed so I tried to use the script to build everything.
[00:56:41] jya: jpabq_: re-run the packager with -noparallel
[00:56:48] jya: it will be easier to track where it failed
[00:57:20] ** jya just updated the OS X packager readme **
[00:57:23] jpabq: jya: when/if it dies again, I will.
[00:57:25] danielk221: I should probably try the -noparallel option too...
[00:57:53] jya: danielk221: there's usually no need; parallel build only fail when compiling Qt
[00:58:00] jya: and I haven't had the issue with 4.8.4
[00:58:31] jya: though, with a -j5 (typical on quad-core); it's getting very hard to find the actual error in the console
[00:58:43] jya: oh jpabq better idea for you
[00:58:50] jpabq_: jya: I also set "CCACHE_CPP2=yes" to get rid of some of the clang warnings.
[00:59:03] jya: so you don't have to re-run the whole lot
[00:59:04] jpabq_: this time
[00:59:17] jya: simply go into the git/mythtv folder
[00:59:19] jya: and type make
[00:59:30] jya: from that point on, it will continue and you can see where the error is
[00:59:33] jya: you'll need to do:
[00:59:51] jya: export PATH=~/path/to/.osx-packager/build/bin:$PATH
[01:00:02] jya: so you have ccache in the path
[01:00:34] jpabq: Okay. I also have ccache in /opt/local/bin from macports
[01:01:25] jya: hum… it did just fail for me right at the end… oh, I know why.. I built Qt in debugging mode so I could step through...
[01:03:25] jpabq_: clang (LLVM option parsing): for the -stack-alignment option: may only occur zero or one times!
[01:04:00] jya: is that an error or a warning?
[01:04:15] jpabq_: Seems to stop it, so error. Trying to build filter_linearblend.c
[01:04:55] jpabq_: Stops same place every time. so definatly being treated as an error.
[01:05:10] jya: which version of clang are you using? which xcode?
[01:05:24] jya: i only ever use the latest version of XCode
[01:05:50] jya: during the last Apple WWDC, I showed some bugs of llvm when compiling the mythtv filters; they fixed it..
[01:05:53] jpabq_:
[01:05:59] jya: but it's only in xcode >= august '12
[01:06:20] jya: oh.. it is a very old version of xcode you are using
[01:06:30] jya: < 4.2
[01:06:31] jpabq_: xcode 4.2 I believe — comes with Lion.
[01:06:54] jya: yeah… the location of the SDK has now changed
[01:07:03] jpabq_: 4.2.1
[01:07:06] jya: can tell its old
[01:08:10] jya: I remember now that I had excluded some filters from the build
[01:08:22] jya: but as apple fixed the problem, I re-added them
[01:08:47] jpabq: Can I install a newer excode under Lion?
[01:08:56] jya: was just checking that
[01:09:37] jya: latest is 4.6.3
[01:09:58] jpabq_: I have the xcode 4 installer. Running it seems to have resulted in it downloading something...
[01:10:02] jya: nah it seems that 4.2.1 is the latest for lion
[01:10:12] danielk221: jpabq: You can usually update the OS for $20 to $40 these days.
[01:10:22] jya: oh, it does download the command line utilities and a new version of clang from time to time
[01:10:33] jpabq_: danielk221: Yeah, but I figured I would wait until the new version comes out in August
[01:10:34] jya: or you upgrade for free if you ask me :)
[01:11:12] jpabq_: jya: So you are saying I should start a timemachine backup ;-)
[01:11:33] jya: nah… i want you to build it now; not be offline for 10 hours while you're upgrading
[01:11:51] jya: but yeah, I can add you to my developer account, so you get download code
[01:12:38] jya: ok.. 4.6 runs in 10.7.5
[01:13:05] jya: if you start the App Store app; you'll get access to it
[01:13:38] jya:
[01:14:43] jya: better you compile with it; old version of clang had issues with the h264 code ; was crashing for no reason from time to time
[01:16:02] jpabq_: Says it is installing. I have two copies of "xcode installer" sitting in Launchpad. How do I get rid of the old ones?
[01:16:33] jya: you can have them all side by side now
[01:16:42] jya: they are self-contained bundles
[01:16:46] jya: no more /Developer business
[01:16:49] jpabq_: Be nice if I could tell which was which
[01:17:02] jya: if you just click on it
[01:17:04] jpabq_: So, I need to start this build from scratch with new arguments?
[01:17:08] jya: the version will show in the finder
[01:17:12] jya: or command-I
[01:17:46] jya: jpabq_: unfortunately, yes.. because the path to the compiler will change from /Developer/SDK/blah to /Applications/
[01:19:21] jpabq_: Looks like it is going to take a LONG time to download the new xcode.
[01:21:40] jya: it's 3+gig
[01:21:58] jya: I download at about 10MB/s here from apple's download, so I don't worry too much anymore
[01:23:39] pvr4me: jya, jpabq_: After installing XCode, don't you also need to install the "Command Line Tools"?
[01:23:51] pvr4me: . . . d-line-tools
[01:24:04] jya: he's using 4.2
[01:24:17] jya: there's no command line tools with 4.2; everything is installed for go
[01:24:20] jya: first go
[01:24:20] pvr4me: Yeah, but updating to 4.6, no?
[01:25:22] jpabq_: So, I should nuke .osx_packager. Then rebuild with " -qtbin /usr/bin -qtplugins /Developer/Applications/Qt/plugins -verbose", and it will automatically pick up the new xcode location?
[01:25:32] jya: jpabq_: don't need to
[01:25:50] jya: the existing packages compiled before are re-usabl
[01:26:01] jya: it's all in the README
[01:26:36] jya: xcode-select -switch /Applications/
[01:26:48] jpabq_: Cool.
[01:27:08] jya: but you can do all of that using the xcode gui
[01:27:13] SteveGoodey (SteveGoodey! has joined #mythtv
[01:27:30] jya: Xcode preferences -> click on the last tab: Locations
[01:27:34] jpabq_: Progress bar for the download pretty much hasn't moved
[01:27:54] jya: in the command line tools, select the xcode version you want…
[01:28:15] jya: its equivalent to running xcode-select, excet it's a tad more visual and you can see which version you're going to use
[01:29:06] pvr4me: jpabq_: you can use the "Activity Monitor" application (/Applications/Utilities/) to check on the download. See the Network tab
[01:30:23] jpabq: Looks like about 200kb/sec
[01:30:38] jpabq: I have a 12MBit DSL
[01:32:12] jpabq: I paused it, and re-started it — download jumped up to 800kb/sec
[01:32:37] jpabq: now up to 1mb/sec
[01:33:07] jpabq: and now down to 200kb/sec
[01:33:34] jya: damn… my irc client have removed stuartm link on how to write a jpeg viewer windows...
[01:33:59] jya: are you using the app store to download?
[01:34:06] jpabq: Yeah
[01:34:11] jya: should be using your closest akamai mirror
[01:34:19] jya: you're not using something like opendns are you?
[01:35:27] jpabq: No, worse — it is hitting a VPN host first. I will fix that and see if that helps.
[01:35:35] jya: well, in the mean time, you can simply amend the and avoid compiling the one that fail
[01:35:41] dblain (dblain!~dblain@mythtv/developer/dblain) has quit (Read error: Connection reset by peer)
[01:36:04] jya: edit filters/
[01:36:30] jya: and add to the line SUBDIRS -= greedyhdeint linearblend
[01:37:14] jpabq: I have other things to work on, so I will wait.
[01:37:19] jpabq: jya thanks for all your help!
[01:37:42] jya: you're welcome.. does show something is still not quite right with the mac compiler :(
[01:37:57] jya: either the tool is too ocmplicated, or the doc sucks
[01:37:57] pvr4me (pvr4me! has quit (Read error: Connection reset by peer)
[01:38:03] jya: or both
[01:38:32] pvr4me (pvr4me! has joined #mythtv
[02:08:24] dblain (dblain!~dblain@mythtv/developer/dblain) has joined #mythtv
[02:11:13] _nyloc_ (_nyloc_! has joined #mythtv
[02:11:35] SteveGoodey (SteveGoodey! has quit (Quit: Konversation terminated!)
[02:11:55] nyloc (nyloc! has quit (Ping timeout: 246 seconds)
[02:20:06] jpabq_: jya: Looks like I now make it as far as danielk221 did: Error! QtWebkit headers not found
[02:20:35] jya: when you installed the Qt binaries; did you install the headers?
[02:21:42] jpabq: Where do I get them?
[02:21:50] jya: it's part of the binaries
[02:22:08] jya: when you run the installer, you're asked to check for them or not
[02:23:09] jpabq: I don't remember it asking me. I will run it again.
[02:23:19] jya: i thought it was by default
[02:23:44] jya: when you go into /Library/Framework/QtWebKit
[02:23:53] jya: there should be a Headers folder in there
[02:23:58] jya: if not that could explain
[02:24:19] jya: maybe they forgot to include the headers in 4.8.4, last I compiled with packages was for 4.8.3
[02:24:31] jya: i'll have a look in a few minutes once I'm done here
[02:24:51] jpabq: Yes, there is a Headers folder in there.
[02:25:20] jya: is that during compilation of mythplugins?
[02:44:39] jpabq_: I tried running "./configure --runprefix=../Resources --enable-libmp3lame --disable-lirc --disable-distcc --prefix=/Volumes/Mercury/Users/john/src/Myth/.osx-packager/build --cc=/Applications/ ain/usr/bin/clang --cxx=/Applications/ hain/usr/bin/clang++ --qmake=/usr/bin/qmake --extra-libs=-F/Library/Frameworks --firewire-
[02:44:39] jpabq_: sdk=/Volumes/Mercury/Users/john/src/Myth/.osx-packager/build/lib --enable-libx264" by hand and it complains about "ERROR: libmp3lame >= 3.98.3 not found"
[02:47:08] jpabq_: jya: it complains about QtWebkit doing that configure step during the normal build process.
[02:47:13] jpabq_: during
[02:47:24] jpabq_: or doing. I think I am tired.
[02:47:57] jya: jpabq_: yes, libmp3lame and libx264 would only have been compiled if you asked to build the backend
[02:48:27] jya: what do you call normal build process ? is that when compiling mythtv or mythplugins ?
[02:48:52] jpabq_: -qtbin /usr/bin -qtplugins /Developer/Applications/Qt/plugins -gitdir ~/src/Myth/mythtv -verbose -nohead -enable-backend -noparallel
[02:50:00] jya: then libmp3lame would have been compiled
[02:50:26] jpabq_: That builds everything up to myth, the does that ./configure which spits out "Error! QtWebkit headers not found"
[02:50:31] jya: you need to amend your LFLAGS to have .osx-packager/buil/lib in
[02:50:54] jya: well, I'll try with a clean environment a little bit later and see what's going on
[02:51:07] jya: will upgrade to 4.8.4
[02:51:22] jpabq_: It only complains about libmp3lame if I try to run that ./configure by hand. If I use osx-packager, then it complains about QtWebkit instead.
[02:52:03] jya: there's no even a need to manually check for header in qtWebKit
[02:52:26] jya: checking for the proper location of qtwebkit in a shell script is going to be particularly difficult on a mac
[02:52:35] jya: as it can be deep within a framework
[02:54:19] jpabq: Okay. I am done for tonight. I will pick this up when I have some more time. I appreciate your time in getting this stuff to work!
[02:54:38] jya: sorry for the problems...
[02:56:54] jpabq: NP. If I was trying to figure this out from scratch, THEN I would probably be feeling a bit frustraited. ;-)
[02:58:05] stichnot: jpabq: I could be wrong about _first_keyframe and -1, I don't usually follow mythbackend in the debugger. What logging flags do you use?
[02:59:00] pvr4me (pvr4me! has quit (Quit: pvr4me)
[03:34:30] Guest68228 (Guest68228!~dblain@mythtv/developer/dblain) has joined #mythtv
[03:35:29] dblain (dblain!~dblain@mythtv/developer/dblain) has quit (Ping timeout: 248 seconds)
[03:44:30] Captain_Murdoch: anybody around that knows if we said that people transcoding to .mkv should not generate seektables? it sounds vaguely familiar and I have a ticket with a user complaining about playback ending early after they created a seektable for their .mkv.
[03:49:03] gary_buhrmaster (gary_buhrmaster!~gtb@2001:470:80e4:3::36) has joined #mythtv
[03:49:40] gary_buhrmaster: CM: There are various (on user list) references that mkv and seektables should not be used together, and ticket #7986 supposedly disabled their creation.
[03:49:40] ** MythLogBot **
[03:50:18] gary_buhrmaster: CM: Of course, so many users, so many tools, at least someone is probably creating seektables with mkv files.
[03:50:58] fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has quit (Ping timeout: 256 seconds)
[03:51:48] fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has joined #mythtv
[03:56:03] Captain_Murdoch: yeah, it looks like the seektable rebuild code sets the length in the DB to be the length as of the last keyframe and he's complaining that playback stops 50 seconds prior to the end, so I'm thinking his keyframes are that far apart. if what he is saying is true though, that if he fixes the length, playback will play till the end, it means we're stopping playback when we think we hit the end, not necessarily when we do, which d
[03:56:04] Captain_Murdoch: oes seem like a bug if true.
[03:56:08] Captain_Murdoch: [23363]
[03:56:08] MythLogBot: SVN 23363: (branch master)
[03:57:00] Captain_Murdoch: looks like that fix only prevents mythtranscode being used to generate .mkv seektables, this guy is using mythcommflag.
[03:57:53] Captain_Murdoch: the mythtranscode way needs to be dropped, especially with all the attention/effort that using the mythcommflagplayer.cpp version is getting.
[04:05:19] gary_buhrmaster (gary_buhrmaster!~gtb@2001:470:80e4:3::36) has quit (Ping timeout: 264 seconds)
[04:11:23] joki (joki! has quit (Ping timeout: 240 seconds)
[04:16:29] joki (joki! has joined #mythtv
[04:41:23] stichnot: Captain_Murdoch: I agree with gary_buhrmaster that all the lore says you should not create a seektable for mkv files. I've never dug into it, but supposedly mkv files have their own internal seektables that ffmpeg uses, and trying to use the myth seektables conflicts with that.
[04:42:09] Captain_Murdoch: ok, I'm going to comment and ask him what happens if he doesn't create a seektable. if that cures it, then that confirms it. :)
[04:42:19] Captain_Murdoch: thx
[04:55:21] jya: danielk221: jpabq_ I've fixed the compilation issue in regards to QtWebKit missing; proper fix would be to change it in the mythtv configure ; but for the time being ; I've added a work around in the mac packager
[04:56:50] jya: Captain_Murdoch: I was playing with mythavtest yesterday trying to fix an issue playing http://mp4 file ; I accidentally played a mp3 file instead. there too, playback stopped a several seconds before the end.
[04:57:06] jya: was very noticeable on a song that last only 3 minutes
[04:57:16] jya: mythmusic played it well
[04:58:10] Captain_Murdoch: 5–10 I could believe, but 50 would lead me to think that we're stopping at the length in the DB not at actual EOF. I couldn't find anything indicating that though, and without a sample, I decided to go onto other bugs in my queue. :)
[04:58:19] Captain_Murdoch: how off was yours?
[05:04:57] jya: hum… I'd say a good 30s, if not more
[05:05:24] jya: i missed the whole ending of the song :)
[05:05:44] jya: could actually time it
[05:12:00] stichnot: Captain_Murdoch: I see that the ticket refers to 0.25. I can barely remember how seektables, duration calculations, etc. worked in 0.26, much less 0.26, but I do know that every release has had substantial differences.
[05:12:21] stichnot: Hopefully not generated the seektable will make it go away...
[05:12:44] Captain_Murdoch: yeah, I figured I'd ask for info first and if he says it's still broken, then point him to v0.26 to test.
[05:12:46] Captain_Murdoch: :)
[05:58:23] SteveGoodey (SteveGoodey! has joined #mythtv
[06:05:08] SteveGoodey (SteveGoodey! has quit (Quit: Konversation terminated!)
[06:11:36] FabriceMG (FabriceMG! has joined #mythtv
[06:30:47] toeb (toeb! has quit (Quit: leaving)
[06:31:13] toeb (toeb! has joined #mythtv
[06:32:08] toeb (toeb! has quit (Client Quit)
[06:32:36] toeb (toeb! has joined #mythtv
[06:36:07] toeb (toeb! has quit (Client Quit)
[06:39:44] toeb (toeb! has joined #mythtv
[06:46:57] toeb (toeb! has quit (Quit: leaving)
[06:47:13] toeb (toeb! has joined #mythtv
[06:51:09] toeb (toeb! has quit (Client Quit)
[06:51:23] toeb (toeb! has joined #mythtv
[06:58:47] toeb (toeb! has quit (Quit: leaving)
[06:59:02] toeb (toeb! has joined #mythtv
[06:59:15] toeb (toeb! has quit (Client Quit)
[06:59:50] toeb (toeb! has joined #mythtv
[07:30:25] Guest68228 (Guest68228!~dblain@mythtv/developer/dblain) has quit (Ping timeout: 246 seconds)
[07:32:51] dblain (dblain!~dblain@mythtv/developer/dblain) has joined #mythtv
[07:36:11] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Remote host closed the connection)
[08:42:31] dblain (dblain!~dblain@mythtv/developer/dblain) has quit (Ping timeout: 256 seconds)
[08:43:08] dblain (dblain!~dblain@mythtv/developer/dblain) has joined #mythtv
[09:13:41] wilmoore-misc (wilmoore-misc! has quit (Remote host closed the connection)
[09:18:19] gregL (gregL! has quit (Ping timeout: 276 seconds)
[10:03:14] IReboot (IReboot! has quit (Remote host closed the connection)
[10:20:22] paul-h (paul-h!~Paul@ has joined #mythtv
[11:06:21] jya_ (jya_!~jyavenard@mythtv/developer/jya) has quit (Quit: jya_)
[11:07:34] jya_ (jya_!~jyavenard@mythtv/developer/jya) has joined #mythtv
[11:17:55] jya_: stuartm: where do you think i should put the code for the airplay picture thing? in libmythtv with the airplay code, or in libmythui?
[11:22:00] jya_: seeing the others, i guess it will be mythtv
[11:29:22] paul-h: jya_: there are a couple of audio related Coverity defects that need your attention 700240 and 700241
[11:30:14] jya_: yeah, I saw those.. I know nothing about alsa input.. those were written even before I knew myth existed I think
[11:30:20] jya_: can have a look of course
[11:30:52] paul-h: OK thanks
[11:33:24] jya_: there's not much to do there… there's no point checking for the return value; because that they fail or not change nothing to how the whole thing would run
[11:36:12] jya_: i would mark them as on purpose in the coverity page
[11:44:34] paul-h: Coverity does say we check the result 22 out of 24 times so why not those two?
[11:45:17] paul-h: Anyway thanks for looking I've changed them to intentional, ignore
[11:45:40] jya_: because the first one ; is just a call to alsa to free global configuration tree; there's nothing else to do if it errors, you continue as if nothing happened
[11:45:52] jya_: and the other is closing the device… error or not, it's the last intstruction
[11:47:04] jya_: so the AlsaBad already does what is required to be done: display an error/warning…
[11:47:45] paul-h: OK I see
[11:48:46] paul-h: Can't you put (void) AlsaBad(....) to silence the warning or is it not worth it
[11:49:52] jya_: stuartm: the image I have is in a QByteArray, storing it on disk, to load it again sounds like a waste of resources… would there be another solution to create the image?
[11:50:09] jya_: paul-h: i guess we could… will coverity recognise the difference?
[11:51:01] stuartm: jya_: yeah, you can use SetImage(QImage) instead, QImage will take a QByteArray
[11:51:28] paul-h: jya_: Don't know you're the professional programmer I'm the amateur you tell me :)
[11:51:54] jya_: SetImage expects a MythImage ; can I give a qimage instead?
[11:52:10] stuartm: in the example I had it load from disk since that's what your originally asked about
[11:52:17] paul-h: If nothing else it would tell anyone else who looks at it we can ignore the return value
[11:52:21] jya_: paul-h: ahah; I have no idea about coverity and what it checks or complain about :
[11:52:27] jya_: good point
[11:53:21] stuartm: jya_: right MythImage is derived from QImage, so you can create a MythImage instead and still feed it a QByteArray
[11:53:56] jya_: ok good...
[11:54:13] jya_: will look into it, and how I can pass it what type of image it is
[11:54:40] sphery: Captain_Murdoch: regarding generating seek tables for mkv, see: and, unfortunately, Janne's comment was before MythLogBot logged #mythtv, but I quoted him at . . . -28:19:30:20
[11:55:25] stuartm: jya_: hmm, no, sorry it _used_ to derive from QImage or maybe I'm just mis-remembering but now you have to assign a QImage – . . . 76afc764e20f
[11:55:46] jya_: stuartm: that's just as good...
[11:57:06] stuartm: so QImage from the bytearray, then assign the QImage to MythImage
[11:57:11] sphery: Captain_Murdoch: also, it's a known issue that playback will end early for those who set "Action on playback exit" (PlaybackExitPrompt) to prompt due to the way the isNearEnd() check works. It's supposed to prompt at 2s before the end of the video, but--unfortunately--in some video types (such as AVI, definitely), we may think we're 2s from the end when we're much more--up to 6 minutes
[11:57:58] jya_: stuartm: . . . 6975ea911d92
[11:58:15] jya_: doesn't show anything about taking a QImage
[11:59:18] jya_: ah, I create a MythImage and assign it a QImage
[11:59:24] stuartm: yeah
[12:00:06] stuartm: in theory we could reduce the steps there by adding an Assign(QByteArray) to MythImage ...
[12:00:10] jya_: what MythPainter parent do I give to the MythImage
[12:02:35] sphery: Captain_Murdoch: I don't know if this was fixed for all video types, but there were some changes made that fix it for at least the recording-type video formats. See #6974 , including comment:6 and and (where the last makes it sound like the issue isn't fixed for people with prompt enabled)
[12:02:35] ** MythLogBot **
[12:03:32] stuartm: jya_: you need to ask the current painter to create the image for you, MythImage *img = GetMythMainWindow()->GetCurrentPainter()->GetFormatImage();
[12:03:46] jya_: agh… I think I'm starting to get it :)
[12:04:14] sphery: Captain_Murdoch: is seems like that fix is only in master, though
[12:04:27] jya_: so I load the theme, this will create the object, i find the MythUIImage element I want to put my picture into, and pass it the QImage
[12:04:42] sphery: stichnot: ^^^ could your work on #6974 fix the issue Captain_Murdoch was talking about?
[12:04:42] ** MythLogBot **
[12:05:36] stuartm: that will ensure the painter returns a MythImage that is appropriately configured and formatted for display by that particular painter (pixmap vs image, colour standard etc)
[12:06:03] stuartm: jya_: yes, exactly
[12:10:10] jya_: stuartm: the coordinates given to a text area, can this be a percentage ?
[12:10:32] jya_: i guess I should read the mythui doc :)
[12:12:33] stuartm: jya_: yes it can, although that's obviously less precise – . . . n_Attributes
[12:13:40] sphery: Captain_Murdoch: and, fwiw, the information about the 2-second issue (which is magnified for some video formats) is in #6148 and #6448
[12:13:40] ** MythLogBot **
[12:13:40] ** MythLogBot **
[12:15:51] stuartm: the documentation is actually pretty good, not entirely complete and possibly lacking explanation in some areas but ...
[12:18:38] jya_: the argument you give in the MythScreenType constructor, the name
[12:19:26] jya_: is that the name of the xml that will be loaded?
[12:19:40] paul-h (paul-h!~Paul@ has quit (Read error: Connection reset by peer)
[12:24:54] stuartm: no, that's really only for internal reference and debugging
[12:25:25] stuartm: although it is also used by the telnet interface to tell the user which screen is currently being displayed
[12:26:23] jya_: should i insert the xml in one of the existing xml in themes/default; or I create a new one ?
[12:26:30] stuartm: the name of the xml that will be loaded is the one passed to LoadWindowFromXML(xmlfile, name, this)
[12:27:52] stuartm: jya_: good question ... I don't think it has a natural home in any of the existing files, so I'd create a new one
[12:28:23] jya_: ideally, I'd like to extend this feature set to allow some feature more alike to xmbc
[12:28:37] jya_: like a notification centre stuff
[12:29:05] jya_: either have it full screen ; or display as a banner on the side
[12:29:06] stuartm: yeah, maybe stick it in a file called notify-ui.xml ?
[12:29:32] dekarl (dekarl! has joined #mythtv
[12:30:14] stuartm: or if this is just for music, I guess music-ui.xml would be good in the future when mythmusic is merged into the frontend (can't be done until then)
[12:31:00] jya_: myth music would be a good candidate for this when it's playing in the background
[12:31:07] dekarl1 (dekarl1! has quit (Ping timeout: 264 seconds)
[12:31:14] jya_: show the name of the song playing when changing etc...
[12:31:33] jya_: need to think about it more , something extendable
[12:31:36] stuartm: the mythmusic mini player does something pretty similar, a popup shown anywhere that includes the artwork and track details
[12:31:56] jya_: haven't seen that one
[12:32:44] jya_: I like notify-ui.xml
[12:33:57] jya_: good name :)
[12:34:17] stuartm: the mythmusic miniplayer can display pretty much all related metadata, but this old screenshot from Metallurgy shows it in use –
[12:35:16] stuartm: I can't actually find a more recent screenshot
[12:35:25] ** jya_ wonder what will happen if I start airplay when myth music is playing in the background **
[12:37:30] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Remote host closed the connection)
[12:38:04] stichnot (stichnot! has joined #mythtv
[12:38:05] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[12:38:05] stichnot (stichnot! has quit (Changing host)
[12:38:28] stuartm: it would be nice to tie it into mythmusic, have mythmusic display the current track playing and those previously played
[12:38:43] stuartm: does airplay have any next-track metadata?
[12:44:08] jya_: let me find you what airplay i receive
[12:54:23] jya_: ah, here it is...
[12:54:24] jya_:
[12:54:54] jya_: I'm recompiling myth, will have an example of what myth receive
[12:59:47] stuartm: hmm, lot more there than I imagined, and it suggests that it's a two way protocol not just a push from the device?
[13:01:18] stuartm: i.e. MythTV could request information about playlists and available tracks from, and maybe even control, some devices?
[13:01:37] jya_: it is..
[13:02:15] jya_: we definitely receive title, artist, album, format, duration, number of times already played, album artwork
[13:02:23] jya_: those are the only one I print in the long
[13:03:21] jya_: that's right… DMAP is a pretty powerful protocol, that's what you can also use to communicate with an iPod.
[13:03:29] stuartm: I don't own any apple hardware, but it looks much more interesting than I'd be lead to believe :) I think there's the basis for a good mythmusic interface there
[13:03:40] stuartm: just not something for 0.27, obviously
[13:03:44] jya_: as for iTunes, that's what iTunes is using to communicate with another iTunes: retrieve the list of files, albums, playlist etc
[13:04:17] jya_: when you take an iPhone or iPad, it will display a list of all the iTunes running on your network
[13:04:35] jya_: from there, you can browse the content, perform search and play it
[13:05:17] jya_: it very well integrated.
[13:05:52] jya_: i have an iPod on a dock here, it has no local music, it streams it from iTunes on my iMac and play it on mythtv via airplay
[13:11:46] stuartm: does the ipad/iphone also act as a server? So you can query i* devices on the network, then play music from them on a client device or software?
[13:11:54] stichnot: sphery, Captain_Murdoch: Yes, it's conceivable that there is some 6974 involvement. I never had any experience with "very early" end-of-playback exiting on non-recording file types, though.
[13:13:14] jya_: stuartm: you can, but only via USB or BT interface, and that's subject to DRM
[13:14:21] jya_: for BT, this is over the iAP profile (Accessory Protocol).. and requires Apple's own hardware encryption chip… so not going to happen easily
[13:15:34] stuartm: ah, shame, would be a nice touch if you could store all your music on the ipod and have mythtv play directly from it – as opposed to mythtv being the dumb client with the ipod being the point of control
[13:17:12] jya_: well, we could do that for any iTunes running on the network
[13:17:16] stuartm: still if we can interface directly with itunes, that would at least allow the possibility of removing the ipod from the loop in your example
[13:17:23] jya_: i believe there's an open source library for this too
[13:17:52] jya_: i don't think it would be too much work to integrate this into mythmusic
[13:18:11] jya_: so you wouldn't need your local music library; instead browse the one on another machine
[13:18:38] stuartm: yeah
[13:19:58] jya_: should suggest this to paul-h
[13:23:56] stuartm: anyway, given the metadata that's available the current music 'mini player' looks like a good fit – a initial simple implementation might only re-use the window definition, but in the long term it could be properly integrated so mythmusic plays the tracks
[13:24:56] stuartm: that would save you the trouble of creating the theme xml, just load the 'miniplayer' window from music-ui.xml and plug in your own values
[13:25:07] jya_: yeah, I just want to wrap something quickly
[13:25:31] stuartm: would work with all existing themes
[13:25:39] jya_: integrating AirPlay in myth music however is going to be tricky
[13:25:54] jya_: it's all UDP based , similar to RDP protocol
[13:26:57] jya_: let me check the mini player… does it have room for showing a picture?
[13:27:11] stuartm: yes it does
[13:27:19] stuartm: can also display the visualiser
[13:27:39] stuartm: although in your case that won't be used (yet)
[13:28:17] jya_: interesting...
[13:28:31] jya_: there's no default theme for it ?
[13:28:42] jya_: can only find it in terra and mythcenter-wide
[13:28:51] stuartm: jya_: it's under mythplugins/mythmusic/theme/default{-wide}/
[13:29:31] jya_: ah… so we have themes in mythtv/themes that defined a theme for myth music ? a bit weird
[13:29:47] stuartm: default version shows artist, album, rating, length, progress, coverart and volume info
[13:30:20] jya_: seem I could be able to use this
[13:31:05] stuartm: jya_: it's just the way it works, not ideal but it was felt that it was better to keep themes as a single package instead of having themers distributing multiple parts for the plugins
[13:32:04] stuartm: by 0.28 mythmusic should have been merged into mythfrontend just as mythvideo was, it will be part of the core functionality of the frontend
[13:32:21] jya_: i thought paul-h didn't like the idea?
[13:34:38] stuartm: I can't remember exactly what he said, except maybe something about not seeing the need, but it's been on the roadmap for years now and unless the plugin API is completely re-written it's really the easiest way to provide even tighter integration
[13:36:20] stuartm: and by integration I mean smooth transitions from one media to another, with DVB radio channels you could have the mythmusic 'streaming' interface and OSD shown instead of re-implementing something very similar just for that purpose
[13:37:14] stuartm: and with storage groups you have to have a lot of the current plugin functionality in the backend anyway
[13:38:07] stuartm: at some point you have to ask whether music isn't something a 'media centre' should do out of the box?
[13:39:06] stuartm: anyway, I'm AFK for a couple of hours or more
[13:44:31] jya_: no worries; thanks for all the guidance
[13:45:19] jya_: one thing that would be nice is a python plugin interface like xbmc, even something where we could use their plugin in myth
[13:48:14] stichnot: danielk221: After my change feb2e0fae2dc7e163cffd004da92d7654765fa1f (support externally-generated .srt files in real time for in-progress recordings), I'm seeing problems with leaking connections on the frontend. Every time the frontend loads the latest version of the .srt file from the backend via RemoteFile, the number of file descriptors in the mythbackend process increases by 7, as...
[13:48:16] stichnot: ...measured by "ls -1 /proc/`pidof mythbackend`/fd | wc -l". Looking at /proc/<pid>/fd, it appears that one fd is for the .srt file, 2 are sockets, and 4 are pipes. These fds stay around until the mythfrontend process exits. Running mythfrontend in the debugger shows that the RemoteFile destructor is being called. Any ideas on how to debug this?
[13:49:30] stichnot: The RemoteFile object is a local in TextSubtitleParser::LoadSubtitles() so its destructor is always executed on function return.
[13:50:58] stichnot: Eventually the backend presumably runs out of fds, and more or less rejects new connections until the frontend quits.
[13:54:18] jya_: stichnot: the frontend never close the connection?
[14:03:39] stichnot: jya_: it looks that way. I'll have a look a the fd count on the frontend process.
[14:06:12] stichnot: yes, the frontend fd count increases by 4 each time and never drops down again.
[14:08:35] jya_: going to be a tad easier to debug if the issue is in the frontend
[14:13:22] stichnot: for me, an order of magnitude easier :)
[14:13:36] stichnot: since the live tv backend is the production system
[14:24:20] jya_: stuartm: when you use MythUIImage::SetImage(MythImage *) ; who is the owner of the image?
[14:26:42] jya_: will the MythUIMage that receive the MythImage takes ownership and i don't need to worry about deleting the image?
[14:28:22] danielk22 (danielk22! has joined #mythtv
[14:28:36] jya_: ah it seems to be reference counted.. perfect
[14:40:26] peper03 (peper03!~peper03@mythtv/developer/peper03) has joined #mythtv
[14:43:25] peper03: stichnot: There are no audio issues if a still is being shown, as there can be no audio. We just display the last frame and start a timer (assuming the timeout is not indefinite). No more data is read until the timer expires or an outside trigger occurs (e.g. selecting a menu option).
[14:43:42] sraue (sraue!~stephan@xbmc/staff/sraue) has quit (Read error: Connection reset by peer)
[14:44:00] sraue (sraue!~stephan@xbmc/staff/sraue) has joined #mythtv
[14:44:44] peper03: Still frames with audio are encoding differently. There, there's just a single video frame to be displayed until the next one is detected. In that cause, you continue reading data, so there could be audio. I took care of/added support for that scenario in #11579
[14:44:44] ** MythLogBot **
[14:48:50] stoffel (stoffel! has joined #mythtv
[14:55:36] FabriceMG (FabriceMG! has quit (Ping timeout: 246 seconds)
[14:56:40] FabriceMG (FabriceMG!~Thunderbi@ has joined #mythtv
[14:57:23] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 240 seconds)
[15:07:30] wilmoore-misc (wilmoore-misc! has joined #mythtv
[15:34:11] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[15:37:31] FabriceMG (FabriceMG!~Thunderbi@ has quit (Quit: FabriceMG)
[15:38:02] stichnot: peper03: For years I've wanted to be able to skip through menus, still frames, and other DVD junk, but I couldn't figure it out, so any way you can make that happen will be greatly appreciated. :)
[15:42:28] peper03: stichnot: I'll see what I can do :)
[15:45:05] peper03: stichnot: BTW, you wanted (for a given value of 'want') to apply the 2nd patch of #11371 to 0.26-fixes once it had baked in master for a bit. If you (or anyone else) doesn't have any objections, I can apply it now.
[15:45:05] ** MythLogBot **
[15:46:34] peper03: Then it's just the question of whether to close the ticket as the first patch (or at least the scenario of timestamp discontinuities) may still be of interest to someone.
[15:47:55] stichnot: I have no objections, whatever is reasonable and easy to maintain should be good.
[15:53:55] peper03: Ok. Have to be afk for a bit. I'll do it when I'm back.
[15:58:33] stoffel (stoffel! has quit (Ping timeout: 268 seconds)
[16:05:48] Steve-Goodey (Steve-Goodey! has joined #mythtv
[16:07:34] Captain_Murdoch: stichnot, RemoteFile leak possibly related to #11341 which I haven't had a chanced to look at yet.
[16:07:34] ** MythLogBot **
[16:08:38] stichnot: Captain_Murdoch, danielk22, jya: yeah, I just independently reached that same conclusion.
[16:09:41] stichnot: In my case, I will create a RemoteFile with an empty filename, then use ReOpen() for the real open.
[16:09:54] rsiebert_ (rsiebert_! has joined #mythtv
[16:10:26] stichnot: Alternatively or in addition, RemoteFile could cache the Open() result in the constructor.
[16:10:31] stichnot: (back in a few minutes)
[16:12:36] rsiebert (rsiebert! has quit (Ping timeout: 246 seconds)
[16:14:53] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 240 seconds)
[16:15:12] Captain_Murdoch: should be a simple fix in ::Open(). I think I took ownership of that file and tickets on a "you touched it last" rule at one point. :)  ::ReOpen() is for switching to a new file on an existing connection.
[16:25:23] jya_: stuartm:
[16:30:48] bobweaver (bobweaver!~bobweaver@ubuntu/member/bobweaver) has joined #mythtv
[16:33:41] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[16:34:05] stichnot: Captain_Murdoch: Why even have an Open() method if it's already being called in the constructor?
[16:35:07] Captain_Murdoch: I didn't write it. :) I think I took it over at some piont under the "you touched it last" rule. I was just checking to see if there was another way to set the url outside the ctor.
[16:35:28] Captain_Murdoch: yeah, you can call the ctor without a url and then call SetURL.
[16:35:32] stichnot: I wasn't trying to blame, but rather debate :)
[16:35:39] jya_: stuartm: the window doesn't appear to be floating on top of the other ; keyboard now does nothing in moving the cursor, like my new window has focus. I can use the myth to click on menu item. When I changed the theme ; the window disappeared; but when the song changed and I tried to update that window I got a crash. i;'m guessing the window got deleted but with no way of knowing ; i still had a reference to it
[16:35:40] Captain_Murdoch: you should really call isOpen() in your code, not Open().
[16:35:45] SteveGoodey (SteveGoodey! has joined #mythtv
[16:35:52] Captain_Murdoch: yeah, wasn't taking it as blame. :)
[16:36:05] stichnot: d'oh! I didn't see the isOpen method
[16:36:12] stichnot: thanks!
[16:36:47] Captain_Murdoch: simple fix is to check isOpen() at the top of Open(), but I have a more complete fix idea that will close if only halfway open, etc., if you just want to switch to isOpen() for your code for now.
[16:37:46] Captain_Murdoch: Open() issue is a bug though so I'll keep the ticket and get a fix in soon.
[16:37:57] stichnot: I will do that tonight, since I basically broke Master for that huge contingent of people who watch Live TV on the HD-PVR using the external .srt capture script.
[16:37:57] bobweaver: Hello there I have some simple questions. They are all about the service api DVR/AddRecordSchedule . what are the different strings that I can use for "Type" the only one that I can figure out is "single" , Other questions that I feeel silly asking.... what is AutoTranscode and Transcode ? and AutoCommflag is that for cutting out commercials ?
[16:39:04] jya_: oh, just crashed in MythUIText::LayoutParagraph when I changed the text
[16:42:02] ** jya_ would the Exiting slot what i need ... **
[16:44:06] Captain_Murdoch: bobweaver, look in libs/libmyth/recordingtypes.h at recTypeFromString(). not sure if it's documented anywhere, might be on the wiki if Robert documented that when he wrote it. the AutoTranscode and AutoCommflag says run those jobs automatically after any new recording is made. the Transcoder is the transcode profile to use I think, not 100% sure on that one.
[16:44:19] SteveGoodey (SteveGoodey! has quit (Remote host closed the connection)
[16:46:09] stichnot: jpabq: my misuse of RemoteFile::Open(), with the backend running out of file descriptors, may be why I wasn't seeing captions on HDPVR live TV after channel changes. When I looked at logs this morning, I could see that the REC_STARTED_WRITING event was indeed being dispatched after a channel change. (Sorry for the noise.)
[16:47:23] bobweaver: thanks Captain_Murdoch
[16:48:04] bobweaver: I was just searching through /libs/mythservicecontracts thanks for pointing me in the correct direction :)
[16:48:12] SteveGoodey (SteveGoodey! has joined #mythtv
[16:48:27] Captain_Murdoch: yw
[16:48:54] bobweaver: Captain_Murdoch, there it is line 29 -46
[16:49:16] stuartm: jya_: I'll fix it swallowing keypresses, that's a bug but I can see where the problems lies
[16:49:23] bobweaver: woops 74 — 94 :/
[16:49:55] jya_: stuartm: i can commit my changes if you have iTunes ; easy to reproduce
[16:51:07] jya_: quick question… The Exiting signal; is that only emitted if something else close it (like changing the theme) or will it be emitted if I do like m_screen->GetScreenStack()->PopScreen(m_screen, true, true);
[16:51:27] Steve-Goodey (Steve-Goodey! has quit (Ping timeout: 246 seconds)
[16:51:33] Steve_Goodey (Steve_Goodey! has joined #mythtv
[16:52:13] SteveGoodey (SteveGoodey! has quit (Client Quit)
[16:52:42] jya_: I guess, it swallowing the screen could make sense.. I see that if I press ESC the little window disappear (until of course i re-create it during the next song)
[17:00:56] stichnot: Captain_Murdoch: I would suggest making RemoteFile::Open() private. The two instances in metadatadownload.cpp should actually be calls to isOpen(), and then everything builds.
[17:04:15] jya_: stuartm: just noticed in my log:
[17:04:23] jya_: interestingly: it works regardless
[17:04:57] jya_: do you have an advice on how to create my screen while i'm in a particular thread (the raop device one)
[17:07:23] SteveGoodey (SteveGoodey! has joined #mythtv
[17:07:45] gigem: bobweaver: For what version of MythTV are you writing? The *RecordSchedule services prior to 0.27 (aka current master) were poorly designed and won't be supported going forward.
[17:08:00] bobweaver: Oo
[17:08:27] bobweaver: gigem, what is going to happen to them ? are you all just going to remove them ?
[17:08:27] jya_: stuartm: I'm guess I'll have the same issue closing the window as i'm completely in a different thread
[17:08:43] bobweaver: gigem, I am using 0.25 atm
[17:09:47] ** bobweaver starts to tear his hair out **
[17:10:15] gigem: bobweaver: There are newer versions in 0.27. I'm not claiming they are perfect, but they stand a much better chance of maintaining compatibility in the future.
[17:11:06] bobweaver: gigem, yeah I would think so. I guess I should update. and write statements to check what version end user is using and make my models off of that
[17:13:34] bobweaver: er
[17:13:37] bobweaver: ~/Templates/Mythbuntu-QML$ find . -type f | wc -l
[17:13:37] bobweaver: 9947
[17:13:48] bobweaver: a lot of that is pictures ^^
[17:13:56] stichnot: jya_: I wasn't paying attention to exactly what you're trying to accomplish, but one possibility is to dispatch a MythEvent from your thread and then put a handler in the event loop of the UI thread. Only problem is if you have to add a handler to many different event loops.
[17:14:43] stuartm: jya_: you need to send an event (mythevent would probably do) with a handler in the UI Thread – see MythMainWindow::customEvent()
[17:14:44] jya_: stichnot: I think i''m simply going to add a new connect/slot to run my function in a particular thread...
[17:15:01] stichnot: The MythUI stuff is deliberately not thread-safe, so you need to do all the creation/destruction in the UI thread
[17:15:02] stoffel (stoffel! has joined #mythtv
[17:15:14] jya_: few UIs are
[17:15:31] stuartm: send the metadata with the event and create the screen in the UI thread
[17:16:07] jya_: stuartm: that sounds like a much greater change than what I originally intended to do
[17:16:54] jya_: could I just connect(my_object, UI_thread, "signal()", slot()) ?
[17:17:54] stuartm: I think you can, but danielk221 has warned in the past about using signal/slots between threads like that – maybe they were less robust in the past
[17:19:09] jya_: well, i already use it a fair amount to signal when an airplay connection start, to tell the player to stop etc.
[17:19:13] Captain_Murdoch: stichnot, makes sense. I'll do that. if Open() and Close() are private, then SetURL() can go away since it's not even used according to git grep.
[17:20:47] stichnot: Captain_Murdoch: ok. I'll go ahead and commit my Open() -> isOpen() correction right away so it doesn't get it your way.
[17:21:35] stuartm: using events is pretty simple fwiw, a custom event can be pretty quick to write – the ExternalKeyCodeEvent class is just 11 lines long, the code to actually send and receive the event is maybe another 10 lines, the rest is just a copy/paste of what you've already written
[17:21:52] stuartm: BUT if you prefer signals instead then I see no problem with that
[17:22:10] Captain_Murdoch: stichnot, ok, I'll take a look at getting the change in during lunch in a couple hours. should be pretty quick.
[17:22:10] jya_: stuartm: I guess sending an event and have the UI respond to it makes sense too from an architecture point of view
[17:23:27] stuartm: jya_: Exiting() signal should be sent however the window is closed/deleted
[17:24:11] stuartm: I don't have itunes to test with, but there is another dialog I can test with anyway in mythmusic
[17:24:17] jya_: stuartm: maybe it's because i do it in a different thread; but when i call Popscreen(), I never receive the exiting signal
[17:25:04] jya_: anyway, 3:25AM, off to bed, will finish tomorrow
[17:25:25] stuartm: jya_: the signal is sent in the destructor
[17:25:57] jya_: weird then that I don't receive it… i know because i put a non-recursive lock in there, and never got a deadlock
[17:28:20] stuartm: actually it's not so weird, signal/slots created between two threads use a queued connection rather than direct, since the sender is deleted immediately after the signal is sent it's entirely possible that QT removes the signal from the queue before it's received
[17:29:57] stuartm: if that's the case it works when the sender/receiver are in the same thread because there a signal/slot are direct, the slot is called before control returns to the destructor and the screentype is actually deleted
[17:33:20] stuartm: fwiw – . . . lt-arguments
[17:35:19] stuartm: I can't see anything explicit about QT removing unsent signals from the queue when the sender is destroyed, and the 'destroyed()' signal seems to suggest that 'Exiting()' should also work
[17:45:25] gregL (gregL! has joined #mythtv
[17:48:09] stuartm: fwiw, the pointer should be null after deletion, but with multiple threads that's only a small help since things can change between the null pointer check and the actual usage of the pointer
[17:51:35] NightMonkey (NightMonkey!~NightrMon@ has joined #mythtv
[17:51:35] NightMonkey (NightMonkey!~NightrMon@ has quit (Changing host)
[17:51:35] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv
[18:15:35] gregL (gregL! has quit (Quit: Leaving)
[18:44:39] peper03: Should #11371 be closed? The patch for flattening timestamps on DVD playback is in master and 0.26-fixes. The original patch took a more generic approach as the problem of timestamp discontinuities could (and presumably does) occur in non-DVD streams.
[18:44:39] ** MythLogBot **
[19:00:33] wilmoore-misc (wilmoore-misc! has quit (Read error: Connection reset by peer)
[19:04:44] stichnot: peper03: sounds good to me. In any case, I kicked ownership off to you.
[19:05:03] peper03: stichnot: Why thank you ;-)
[19:05:21] stichnot: :)
[19:06:00] natanojl (natanojl! has joined #mythtv
[19:06:00] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has joined #mythtv
[19:06:00] natanojl (natanojl! has quit (Changing host)
[19:15:59] stoffel (stoffel! has quit (Ping timeout: 255 seconds)
[19:23:23] peper03: stuartm: Didn't you say a while back that you'd created a patch to set the default DVD language (menu, audio, subtitle)? Was it complete?
[19:24:02] wilmoore-misc (wilmoore-misc! has joined #mythtv
[19:26:57] Moeabm (Moeabm!~moeabm@ has joined #mythtv
[19:27:54] stuartm: hmm, I can't remember
[19:28:16] Moeabm (Moeabm!~moeabm@ has left #mythtv ()
[19:28:36] Moeabm (Moeabm!~moeabm@ has joined #mythtv
[19:29:13] peper03: Ok, not too important. I think it's fairly straightforward. Just thought it would save me the effort if you already had something.
[19:31:34] bobweaver: huh for some reason every time that I add a recording and then remove it after I remove it mythbackend crashs :/
[19:31:43] bobweaver: from the services api ^^
[19:31:45] stuartm: I do remember that I was disappointed with the result, I thought it would prevent dvds asking for the language before proceeding to the menu, but at least on the discs I tried that didn't happen
[19:33:58] peper03: I think the DVDs that ask aren't the ones that read the defaults. I'm pretty sure the default at the moment is always set to English, so it's not as though there is no default.
[19:35:05] peper03: Yep, it's set to 'en'.
[19:40:01] bobweaver: that is something that I have been worrying if the locals are different in the xml that the services bring back. That would kill my app :(
[19:40:57] sphery: bobweaver: gigem had fixed some problems with the services api scheduling stuff and that sounds familiar, so might work fine in master/0.27-pre
[19:42:13] bobweaver: that is the only thing that I can not translate on the fly is listmodels and what not. I was thinking if that is the case then I am goign to have to make drastic overhauls like having xmllistmodels that onstatus is ready make different table or use xmlhttprequest with javascript and have a altering engine that reads and writes to these xml files
[19:43:14] bobweaver: sphery, Yeah I am going to compile it all right now. I am on 0.25 but I might as well move up I just hate changing all my frontends and everything that comes along with changing db versions
[19:43:58] stuartm: peper03: found this patch in my stash –
[19:44:45] bobweaver: here is a example of the backend crashing on 0.25 with services api for reording and removing it in the video you can see that the backend crashs after removing a recording
[19:44:50] bobweaver:
[19:49:28] peper03: stuartm: Yep, that's pretty much what I came up with, although MythLanguage() in globalsettings.cpp checks for an empty string from gCoreContext. I don't know whether that can actually happen, but it probably doesn't hurt.
[19:52:03] peper03: I think setting the region code could be problematic as there are a handful of region 1 discs that pretend to be region free but then read the player's region code during playback and display a map of the world with the various DVD regions shown if the player isn't region 1. RCE it's called, if I remember correctly.
[19:52:06] stuartm: it should only possible if the user has directly altered the language field in the database
[19:53:49] peper03: Ok, then it's probably safe to leave that check out. Actually, libdvdnav will catch an empty string, so the extra check on our side is redundant.
[19:54:56] stuartm: peper03: that was added based on some reading I did where it was suggested that the absence of the region code could similarly result in some discs failing to play and since it seems more likely that users are trying to play discs from their region it seemed like the right thing to do
[19:55:35] stuartm: that said, in my patch I've hard-coded it which we definitely do not want to do in a final version
[19:56:16] peper03: Yes, it could be problematic. Anything that reads the region code during playback can use whatever logic it wants.
[19:57:12] peper03: One solution would be to allow the user to specify a region code and use that, or if they say 'region free' or 'none' or whatever, we try to read the region code from the disc and set it to that.
[19:57:55] peper03: By default, they wouldn't set anything but if they have a problematic disc, they have the ability to set a specific region code.
[19:59:19] peper03: My Star Wars DVD is intelligent! It respects all three language settings (menu, audio and subtitle) independently. That's how it's supposed to work :)
[20:02:07] stuartm: yeah, it's probably the only way to do it, but I don't like it for a couple of reasons – 1) Additional settings 2) Most users will set it to 'region free' thinking that has to be the 'best' option, they aren't going to put 2 and 2 together to figure out that's the reason their disc won't play
[20:03:29] stuartm: 3) If the user has to wade into the settings screens to toggle setting values on a disc by disc basis then they are going to get fed up pretty quickly
[20:04:31] stuartm: I'd be inclined to set the region mask correctly according to country – so no setting, then just see what happens, if it breaks playback of discs for many people we can rethink it?
[20:05:06] peper03: I can't think of any way round having an additional setting as there's no way to automatically cover all bases. You could call it 'Auto' rather than region free.
[20:05:18] stuartm: it all depends on how many people are importing discs from other regions
[20:05:54] stichnot: For playback settings, I prefer the "live setting" approach, such as what is used for allowing forced subtitles, and for subtitle zoom factor.
[20:06:05] peper03: Yep. I have a couple. A friend of mine used to buy R1 discs quite regularly before they started appearing at about the same time worldwide.
[20:06:53] stuartm: peper03: have you tried them when setting the region mask, just to see what happens?
[20:07:11] stuartm: stichnot: in this case, by the time we're playing it's too late
[20:07:17] peper03: Setting the region based on country could be fun too. We'd have to create a mapping of all the countries->region.
[20:07:38] peper03: stuartm: I haven't but hang on and I will.
[20:08:10] stuartm: that is actually my idea of fun :)
[20:09:18] peper03: You're sick ;-)
[20:09:37] stuartm: an alternate approach would utilise the locale configs, but since we've not persuaded everyone to provide one for their country/language it wouldn't work nearly as well
[20:09:46] stichnot: stuartm: good point.
[20:17:35] peper03: Ok, three R1 DVDs tested with the VM hard-coded to region 2. Not a problem with any of them.
[20:18:16] peper03: So the other option is just not to bother fiddling with the region code at all at the moment until someone complains :)
[20:18:23] stuartm: yeah
[20:19:12] peper03: I would imagine it's really only going to be discs that query the region code in the VM that could be affected.
[20:19:44] peper03: Actually, that makes sense. Most DVDs don't need to query it during playback as the player normally does the check.
[20:20:17] peper03: It just refuses the play the disc if the disc's region doesn't match its own. We're looking at it the wrong way round :)
[20:21:58] peper03: There's a list of RCE protected discs here:
[20:22:41] peper03: If anyone happens to have one of those (region 1), it would be interesting to know whether they play.
[20:30:11] jheizer: I have at least one.
[20:30:45] jheizer: Could I test it under .26-fixes? And if so is there an easy way to confirm my disk does infact have it?
[20:32:31] peper03: jheizer: 0.26-fixes should behave the same as master in this respect. You should a log entry like this from the frontend:
[20:32:33] peper03: libdvdnav: DVD disk reports itself with Region mask 0x00000000. Regions: 1 2 3 4 5 6 7 8
[20:33:15] jheizer: k, give me 2 minutes
[20:33:20] peper03: If you try to play it, it wouldn't surprise me if you get the world map.
[20:34:52] jheizer: I built this drive with a optical drive in it just because you hav ebeen working on it so much
[20:35:12] jheizer: haha before I never had any luck and said FEs always had a stand alone player there anyway
[20:36:01] peper03: Nice to know :)
[20:36:51] jheizer: intro, menu, and warning screens good so far
[20:37:25] jheizer: no libdvdnav in FE log
[20:37:36] peper03: Ah, thought so. There's an explicit message from libdvdnav if the DVD tries to read the region code: libdvdnav: Suspected RCE Region Protection!!!
[20:38:27] jheizer: I need to enable higher debugging level maybe?
[20:39:04] peper03: Just wondered that, but doesn't appear so. Do you not see any line with 'DVD disk reports itself with Region mask'?
[20:39:21] jheizer: grep for egio returns nothing either
[20:39:47] stuartm: actually, I don't think we're catching the log output from libdvdnav, you might need to run it from the console and look for the output there
[20:40:13] peper03: Ah, yes. Sorry, it's just a printf.
[20:40:24] jheizer: ah ok, on it
[20:40:49] jheizer: mask is all 0000000
[20:41:50] peper03: Ok, that should be 'all regions', which is what I would expect. Do you see the 'Suspected RCE' message?
[20:42:42] jheizer: oh
[20:42:48] jheizer: no the first time it printed the mask
[20:42:53] jheizer: but the second time yes, it is there
[20:42:58] jheizer: *not the first time
[20:43:47] jheizer: let me try it again as that time I stopped before the main title
[20:45:04] ** jheizer grumbles stupid full screen format movie **
[20:46:02] jheizer: Yeah the Suspected RCE message appears a few times but the main title is playing just fine
[20:46:38] peper03: :) The kids were given 'The Little Mermaid'. That had a 4:3 video, which was then letterboxed. Effective vertical resolution was something like 230 lines :-o
[20:47:01] jheizer: oooooouch
[20:48:15] peper03: There was even an intro extolling the virtues of the clarity of DVDs but it looked like it was a VHS capture! I kept waiting for the 'wipe' to show the difference between VHS and DVD, but no, that *was* 'DVD' quality.
[20:48:45] peper03: Ok, I just looked in the libdvdnav code where it initialises the region register. The comment says 'region free', the value says '1' (i.e. region 1). That's why it plays.
[20:49:06] jheizer: yeah sitting here next to the tv this one looks horrible
[20:49:09] jheizer: we are so spoiled
[20:49:33] jheizer: I don't seem to have any more to double check. One other was an HDDVD
[20:49:57] peper03: RCE was only used, as far as I know, on region 1 DVDs. So setting the player's region code to 1 makes sense. Only region 1 DVDs will query it, others won't care.
[20:51:12] jheizer: All that and my ZFS array is still destroying
[20:51:17] peper03: So after all that, we know nothing needs to be done :)
[20:51:19] jheizer: workign from home has its perks
[20:54:33] peper03: stuartm: Do you want to commit your patch or shall I do it? The only thing I added was a debug LOG (and removed the region code stuff).
[20:58:17] stuartm: peper03: you can do it
[20:58:53] Oleg_ (Oleg_! has joined #mythtv
[20:59:02] peper03: Done.
[20:59:28] Oleg_ (Oleg_! has left #mythtv ("Leaving")
[21:04:30] peper03: jheizer: Thanks for the test!
[21:06:37] jheizer: No problem
[21:06:59] bobweaver (bobweaver!~bobweaver@ubuntu/member/bobweaver) has quit (Quit: Leaving)
[21:07:51] stuartm: Disney mastered dvds are notoriously bad, great for finding all those dvd playback bugs but not if you're the owner of them
[21:09:51] jheizer: lol
[21:10:06] peper03: Yep. Doesn't matter if 99.9% of all DVDs play back properly if you've got one of the 0.1%
[21:12:02] stuartm: I'm not really sure Disney have produced much of value for a very long time, but then I don't have kids and my opinion isn't going to stop the multinational goliath Disney from churning out weak, saccharine crap while simultaneously tarnishing classic children's stories with their interpretations
[21:12:33] jheizer: but but ____ ____ 6 jsut came out!
[21:13:51] peper03: High School Musical 6?
[21:13:57] stuartm: I do however think it's adding insult to injury when they then flog very badly produced DVDs to children (well their parents)
[21:14:15] jheizer: I wonder how many more years it will be till you can reliably put a BR into a PC and play it back with out crossing your fingers and toes.
[21:15:26] jheizer: lol are there really 6 of those? I just picked a random huge sequel number.
[21:15:35] stuartm: it will probably be a while, I mean my bluray drive is currently crippled and won't play anything (until I can find a good host key) because of their nasty DRM
[21:16:01] peper03: I have no idea. Thankfully I've been spared any of them. Just seeing a still makes me cringe.
[21:16:15] jheizer: haha
[21:16:45] jheizer: My latest FE was a zotac with a BR drive. I know damn well it will never see a BR disk during its lifetime most likely.
[21:18:41] peper03: I would expect it will take quite a while too. Assuming someone eventually finds a way around the DRM, from the little I've seen of what's possible and how much time I've spent getting to grips with DVDs, it looks like a frightening amount of work.
[21:19:49] peper03: And it's not like I've been implementing lots of features. More just tweaking the existing code.
[21:20:03] sphery: peper03 / stichnot : +1 on live settings, even if it means tearing down the player and restarting from scratch (i.e. for something like a region)... IMHO, far more important than reducing settings is reducing the number of playback-related settings in the hard-to-find, hard-to-access, and makes-me-have-to-point-out-existence-location-and-description-of-settings-on-use rs-list frontend setup area. stuartm's 2) and 3) are exactly why that's so ...
[21:20:09] sphery: ... bad. :) (Applies to your discussion of prohibited uops settings as well as region settings. And feel free to move any relevant settings to live settings.)
[21:20:44] sphery: peper03: BTW, welcome! I'm glad to have you on board--especially as you've done far more for MythTV recently than I have in the last year.
[21:21:33] peper03: sphery: Thank you! You've spent a lot more time helping on the -users list than I have.
[21:21:39] sphery: that is feel free to move any already-existing relevant settings to live settings :)
[21:22:53] sphery: Well, I've been distracted by real life this year, but I'm hoping to get back into helping with MythTV (and am slowly wading back in). And it's definitely good that we've got someone else who's getting up to speed on the tough player stuff (as I know I won't be touching that :).
[21:22:55] stuartm: live settings aren't a panacea, especially 'sticky' settings, just because a user changes something during playback doesn't mean they expect it to stay that way for the next recording/video they watch
[21:23:19] sphery: right... something like the region or language needn't be sticky, but should be accessible during playback
[21:23:33] sphery: some things make sense as sticky settings
[21:24:02] sphery: but burying the setting in frontend setup where the user won't remember it exists, let alone where to find it, doesn't help anyone
[21:24:36] peper03: Prohibited uops might make sense as a live setting but then it's the question of on or off by default? And sticky or not?
[21:25:37] sphery: peper03: btw, seems a user in #-users is having a problem compiling due to an ld error on ../../libs/libmythtv/ undefined reference to `dvdnav_set_state' ... See
[21:25:56] sphery: not sure if it's something you've changed or one of stuartm's, but figured you guys might know which -l we're missing
[21:26:03] sphery: has the error
[21:26:48] peper03: Yes, that function is from me. It's a new one in libdvdnav.
[21:27:20] peper03: Jumping over...
[21:27:53] sphery: yeah, for prohibited uops, I think most would prefer off by default, but we could always punt and make it sticky
[21:28:14] stuartm: distclean, if that fails, blow away the ccache er, cache (-C)
[21:29:52] Steve_Goodey (Steve_Goodey! has quit (Quit: Konversation terminated!)
[21:30:40] SteveGoodey (SteveGoodey! has quit (Quit: Konversation terminated!)
[21:34:35] stuartm: sphery: annoying as it sometimes is, people accept that it works that way by design – all commercial dvd players respect prohibited operations – but they are likely to be less forgiving when ignoring prohibitions causes unexpected behaviour with a DVD and if it was my decision alone I'd say we should always respect prohibited ops, following the principle of least surprise
[21:35:23] sphery: that makes perfect sense to me, so I'd support the approach
[21:36:34] IReboot (IReboot! has joined #mythtv
[21:37:17] stichnot: hmm, all commercial DVRs don't let you flag commercials and create cutlists, should we follow that?
[21:37:24] peper03: stuartm: Didn't you make a change recently to how libmythdvdnav was linked?
[21:38:56] stichnot: My main point is that there should be a way to disable prohibited uops. It doesn't have to be a persistent setting, but it should be in the playback menu somewhere.
[21:39:28] stuartm: peper03: 1588286b
[21:39:37] stuartm:
[21:39:55] peper03: Thanks!
[21:40:08] stuartm: not to the way they are linked, I just stopped it installing the .a file which shouldn't be required
[21:41:41] sphery: stichnot: yeah, I think stuartm was saying that the default should be to respect prohibited uops, but have a live setting to disable them
[21:47:34] stichnot: ok, I was reacting to "we should always respect prohibited ops"
[21:49:29] peper03: My personal preference would be to make the setting sticky. Having to navigate throught the playback menu every time I start playback would annoy me. I also wonder if some users wouldn't be equally annoyed at upgrading and then finding they're restricted.
[21:49:44] peper03: I understand the both sides of the argument, of course.
[21:49:52] peper03: s/the//
[21:51:19] stichnot: I would put the live setting at the top level of my "custom playback menu". And the option wouldn't even show up unless you're playing a DVD.
[21:52:21] peper03: I could probably live with that.
[21:53:55] peper03: Would that mean, though, that all users would have to create their own custom playback menu if they want quick access?
[21:54:15] peper03: Just trying to judge how I would feel as a user in that situation.
[21:54:21] sphery: yeah, making it stickyt probably makes sense--especially if we respect it by default (since those who disable it are probably the kind who will want it always disabled)
[21:55:10] sphery: and, fwiw, you can also create an action for a key binding, so you could make it easy to toggle (regardless of how it's done in the menu)
[21:56:01] peper03: sphery: That's true, although I'm not sure I have any free buttons left on my remote :)
[21:57:55] sphery: hehe, yeah--though if it's sticky, you only have to navigate MENU|Video|... once (or once in a blue moon if you have an occasional DVD that needs them enabled)
[21:58:20] sphery: so even if it's not a default part of the custom playback menu, it wouldn't be too bad
[21:58:35] jpabq: sphery: I don't know if you saw, but Daniel pointed out that having the "lock timout" happen in the signal monitor causes problems for EIT. Moving it into TVRec is very wasy to do (and that is what I should have done to begin with), but there it would use "channel_timeout" to determine when to give up on the Lock. Current definition of channel_timeout is "Maximum time (in milliseconds) MythTV waits for a channel lock when scanning for channels
[21:58:35] jpabq: during setup, or for issuing a warning in Live TV mode."
[21:58:56] jpabq: That definition would change to cover recording as well — Which if the value is too low could cause recordings to fail...
[21:59:53] peper03: sphery: Yep, that's why I'd personally prefer it to be sticky. I wouldn't care too much then where the setting is.
[22:00:04] jpabq: Since you deal a lot with "configuration issues", I thought I would ask if you have any idea how to handle that.
[22:00:39] jpabq: s/wasy/easy/
[22:01:34] sphery: jpabq: yeah, I saw some discussion of it (and it was above my level of player/tuning knowledge :)... Does that mean we wouldn't be using signal_timeout, anymore? Or is the one we wouldn't be using one you added specifically for the signal monitor?
[22:03:14] jpabq: The capturecard table has both signal_timeout and channel_timeout. We could use either, but right now channel_timeout is always greater than signal_timeout, so that seemed safer. Right now, those values are not used during the recording process at all — they are purely informational during channel *scanning* and LiveTV.
[22:03:51] sphery: ok, so you wouldn't be changing those... you had said something about a "global" value, so it was one you added?
[22:04:53] jpabq: No, I was looking at the wrong thing. Ignore that comment.
[22:05:14] sphery: so what are we using now for the signal monitor?
[22:05:26] jpabq: There is a "global" timeout that the scheduler uses, but that is not appropriat for this.
[22:05:49] jpabq: Some hard-coded timeouts, that were selected per tuner type.
[22:06:13] sphery: so maybe do a db check update and if any channel_timeout is < the value you had used, increase it?
[22:06:53] jpabq: That should work, although I don't really need to bump the db schema.
[22:06:57] sphery: I do think that many users have too low a channel and signal timeout--and, in fact, I think our defaults are too low--so it's likely we'll have to keep telling them to check them, whether you change to use channel_timeout or not
[22:08:18] jpabq: I believe that the DB schema was updated very recently. I could add this logic to that db check, and leave it up to the "devs" to fix it for themselves...?
[22:08:31] stuartm: yesterday
[22:09:26] jpabq: Hmmm, backendhousekeeper.h is failing to compile
[22:09:56] sphery: yeah, the schema version bump is an unfortunate side effect of having a single variable for schema and data versioning in there... but then again, that approach makes it much easier when we have multiple components talking (where sometimes the data change needs to be in place for something else in the system, so we can't really say, "needs at least data version Y," because that also assumes any Y+n will work, and that may not be the case)
[22:10:26] sphery: I'd say just do another version bump... we shouldn't run out of numbers any time soon :)
[22:10:41] jpabq: Okay
[22:13:06] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Ping timeout: 246 seconds)
[22:15:21] sphery: jpabq: btw, thanks for thinking of the support side of the change before pushing it, too :)
[22:15:39] jpabq: :)
[22:15:44] sphery: (and for explaining it to me so I can try to help users who have issues)
[22:19:22] NightMonkey (NightMonkey!~NightrMon@ has joined #mythtv
[22:19:22] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv
[22:19:22] NightMonkey (NightMonkey!~NightrMon@ has quit (Changing host)
[22:23:53] peper03: stichnot: Just playing around with skipping/menus etc. I changed the definition of a menu to be 'NumButtons > 0'. Skipping works, but on the current test DVD, which has a fairly long intro before the menu, skipping forwards 30s jumps into the middle of the menu. In itself not a problem, but we've skipped over the packets containing the menu highlight information. You can navigate, but you can't see anything :(
[22:25:58] peper03: It may be possible to detect that scenario (NumButtons > 1 and it was just == 0 but we've no subpicture), but I'm not sure if it's possible to work out where the previous subpicture packet was to be able to jump there.
[22:27:12] peper03: Oooh. Looks like it may be available... That would be handy.
[22:27:24] jya_: stuartm: was just thinking…. I could make a UI that is just like miniplayer unless defined otherwise… So the question is: how can I define a theme that by default inherit entirely from mini player, though can be fully redefined without conflicting with miniplayer?
[22:36:11] bobweaver (bobweaver!~bobweaver@ubuntu/member/bobweaver) has joined #mythtv
[23:10:28] stichnot: peper03: too bad you can't "fast-forward really fast" instead of skipping so you don't miss the key packets. (But I have no idea how to think about that for skipping backward.)
[23:16:57] peper03: stichnot: I don't think it's too bad. Each NAV packet gives you a relative sector offset to the NAV packet containing the subpicture packets to be shown. If we detect we have buttons but no subpicture and the subpicture packet is behind us, we can work out exactly which sector to jump to. The user won't notice the extra jump.
[23:17:20] peper03: That's the theory, anyway. My head is getting in the way of the practice at the moment :(
[23:32:01] peper03: Yep, it works! If you try to jump into the middle of a menu, it's detected and you jump back to start of the menu. Perfect!
[23:33:03] peper03: All a bit hacked together at the moment, but it shows it works so I can clean it up and check there are no other side effects tomorrow or whenever.
[23:33:23] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 240 seconds)
[23:33:57] peper03 (peper03!~peper03@mythtv/developer/peper03) has quit (Quit: Konversation terminated!)
[23:34:15] stichnot (stichnot!~stichnot@ has joined #mythtv
[23:34:16] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[23:34:16] stichnot (stichnot!~stichnot@ has quit (Changing host)
[23:37:05] stichnot: peper03: that sounds really cool if it works out.
[23:38:45] wilmoore-misc (wilmoore-misc! has quit (Remote host closed the connection)
[23:40:08] jya_: now we need the same amount of effort going into the bluray player !
[23:40:14] bobweaver: barghh 0.27 is giving me so much troubles. backend crashs on start and when I run it manual it stays open still I scan or do something that is needed it seg faults again
[23:40:39] jya_: bobweaver: backtraces...
[23:55:46] bobweaver: jya, yeah there is just so much going on. I am going to compile myself and not use apt.
[23:56:07] jya_: for backtraces, there's no need to compile it yourself
[23:56:19] jya_: there's a dbg package that includes the debugging symbol
[23:56:49] sphery: bobweaver: info on debug packages at
[23:57:28] bobweaver: jya, nah I am saying that the logs and running the backend verbose just returns error after error apport is freaking out and I think That I would spend more time fixing it then I would if I built my self
[23:57:39] bobweaver: thanks sphery and jya
[23:58:14] jya_: without getting any details about the errors themselves, it's hard to give any proper advice

IRC Logs collected by BeirdoBot.
Please use the above link to report any bugs.