:: #mythtv

Daily chat history

Current users (84):

abqjp, aloril, Anduin, anykey_, Beirdo, ben1066, brfransen, brtb, CaCtus491, Captain_Murdoch, cattelan, Chutt, clever, coling, Cougar, damaltor, danielk22, davide, dekarl, dlblog, ElmerFudd, ertyu-m, foobum, foxbuntu, frankster, ghoti, gigem, gregL, GreyFoxx, highzeth, J-e-f-f-A, j-rod|afk, jams, jarle, jcarlos, joe_, joki, jstenback, justinh, jwhite, k-man, knightr, kormoc, kurre2, kwmonroe, laga, mag0o, MavT, mike|2, mrand, MythBuild, MythLogBot, mzanetti, peitolm, pheld, poptix, purserj, rhpot1991, skd5aner, Slasher`, sphery, sraue, stuarta, stuartm, superm1, sutula, taylorr, tester68, tgm4883, The_Ball, ThisNewGuy1, tomimo, tris, Unhelpful, vallor, wagnerrp, wahrhaft, wseltzer, xavierh, xris, Yancho, yb0t, zCougar, _charly_
Wednesday, March 14th, 2012, 00:05 UTC
[00:05:01] stichnot (stichnot!chatzilla@mythtv/developer/stichnot) has quit (Ping timeout: 244 seconds)
[00:10:36] noahric (noahric!~noahric@ has quit (Quit: noahric)
[00:25:04] jya: lots of issues with the theme downloader still… Yet to manage to properly download the new Sphere theme. Selecting the first time, you get an error about the backend being unable to download. if you try again, you get a "downloading" progress bar and you're stuck there. Must force quit to restart the frontend…
[00:26:41] xavierh (xavierh! has quit (Remote host closed the connection)
[00:34:35] danielk22 (danielk22!~danielk@ has quit (Ping timeout: 244 seconds)
[01:06:25] davide (davide! has quit (Remote host closed the connection)
[01:06:47] davide (davide! has joined #mythtv
[01:17:52] Cougar (Cougar! has quit (Ping timeout: 244 seconds)
[01:20:44] Cougar (Cougar! has joined #mythtv
[01:20:53] stichnot (stichnot!~chatzilla@mythtv/developer/stichnot) has joined #mythtv
[01:22:57] Captain_Murdoch: jya, can you test a simple patch for me?
[01:23:03] jya: sure
[01:23:32] jya: Captain_Murdoch: note that the issue I'm experiencing is when i download a theme from a remote frontend
[01:24:57] Captain_Murdoch:
[01:25:20] Captain_Murdoch: that will do the download on the FE itself rather than asking the MBE to do it then downloading from the MBE.
[01:25:41] Captain_Murdoch: if this works, I may commit it for 0.25 and deal with it after release.
[01:26:33] Captain_Murdoch: the patch works for me on a shared FE/MBE but I haven't had a chance to troubleshoot the issue on a remote FE.
[01:43:10] jya: Captain_Murdoch: where will the theme be located after downloading?
[01:43:20] jya: only on the frontend or will that be shared (not that it matters)
[01:43:34] Captain_Murdoch: only on the FE. ~/.mythtv/themes/THEMENAME
[01:43:51] jya: why not having something that attempt on the BE first, then on the FE ?
[01:44:00] Captain_Murdoch: with the patch that is. without the patch, the .zip will also exist inside the QNetworkDiskCache's cache on the MBE.
[01:44:05] guysoft42 (guysoft42!~guysoft@ has quit (Quit: Leaving)
[01:44:23] jya: I'm just finishing compiling Qt, and I'll try it shortly after
[01:44:35] Captain_Murdoch: that's what the current code does. it downloads on the MBE first. then the FE uses a URL like myth://Temp@MBE/ to download from the MBE.
[01:45:27] Captain_Murdoch: I think that I did see your 'hangs while downloading' issue once the other day. it almost looked like the MythDownloadManager hung. could be related to all the other work that has gone on inside there recently.
[01:46:41] jya: I always see the hang after it tells me that there was an error downloading from the BE
[01:46:53] jya: when I try again, that's when the hang occurs
[01:47:45] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has joined #mythtv
[01:53:06] jpabq__ (jpabq__!~jpabq@mythtv/developer/jpabq) has joined #mythtv
[01:55:38] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has quit (Ping timeout: 246 seconds)
[01:56:41] abqjp (abqjp! has quit (Ping timeout: 246 seconds)
[01:56:57] abqjp (abqjp! has joined #mythtv
[02:03:51] abqjp: taylorr, Attribute 'GPUCoreTemp' 94
[02:05:08] abqjp: That is 1080i H.264. Not timestretched, though. That seems a bit high, but the card seems to run perfectly. Only time I have any problem is trying to do H.264 1080i at 2.0 timestretch.
[02:05:44] abqjp: That is in an original style Antec Fusion case, but with the side fans set to blow into the case, instead of out of it.
[02:07:10] jya: abqjp: 94 degres is nothing, they are rated up to 120 degres
[02:07:41] abqjp: Good. Then I don't have to worry about it.
[02:09:39] abqjp: jya, that is with one of the Asus passive gt430 cards.
[02:10:18] abqjp: . . . DC_SLDI1GD3/
[02:16:00] jya: that is one big heatsink
[02:18:08] abqjp: Small enough to fit in my Antec Fusion case. Most passive heatsinks are too tall (Usually because of tubing).
[02:25:00] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has joined #mythtv
[02:28:32] abqjp (abqjp! has quit (Ping timeout: 246 seconds)
[02:28:34] jpabq__ (jpabq__!~jpabq@mythtv/developer/jpabq) has quit (Ping timeout: 252 seconds)
[02:28:58] abqjp (abqjp! has joined #mythtv
[02:34:06] noahric (noahric!~noahric@ has joined #mythtv
[02:39:30] noahric (noahric!~noahric@ has quit (Read error: Connection reset by peer)
[02:54:28] jya: Captain_Murdoch: okay.. finally finished compiling Qt.. I can test now..
[02:54:46] jya: but shouldn't we focus our effort in finding out why the backend isn't downloading ?
[03:33:51] sphery: If you figure that out (or why the frontend is unable to download from the backend--whatever part is failing), I'm sure he'll be even happier. I had promised him I'd look into it, but haven't gotten around to it, yet (which, I'm sure, is exactly where he is, too--and is why he's considering workarounds, now).
[03:57:47] taylorr: abqjp: ouch!
[03:58:55] taylorr: abqjp: that 120C is max.... running at a high temp will always affect the life of the chip
[04:00:27] taylorr: I noticed that the heatsink wasn't properly secured and the thermal paste didn't seem to be of the highest quality.... I applied a thin amount of thermal paste directly to die on the chip and resecured the heat sink and now at 65C
[04:04:19] taylorr: abqjp: max temp for GT430 is 98C -> . . . ecifications
[04:12:40] stichnot: In MythPlayer::NextCaptionTrack(), the order of captions/subtitles is (according to the comment): Text->TextStream->708/608->608/708->AVSubs->Teletext->NUV-> None. I would like to move AVSubs to the position just after TextStream and just before 708/608. The reason is that even when AVSubs are manually selected during playback, I find that occasionally there is some sort of a reset in the...
[04:12:41] stichnot: ...subtitle streams, like the selected stream goes away and comes back, and then the decoder turns captions off and back on again. This causes NextCaptionTrack() to go through the sequence again, but then if cc608 captions are present, they get selected because of their higher priority in the sequence.
[04:13:50] stichnot: I've tested this change on my DVD that has this problem, and whether I manually select cc608 or AVSubs, that selection remains in place.
[04:14:23] stichnot: Does anyone see any problems with changing this order?
[04:17:21] taylorr: when the decoder turns captions off and back on doesn't it still have the selected track info?
[04:18:30] stichnot: yes it does, but that information is superseded by the existence of the cc608 track.
[04:18:46] stichnot: the information and the decision process are in different classes
[04:19:15] taylorr: so the AVSubs are no longer available when the cc608 track gets selected?
[04:19:57] stichnot: let me dig up the sequence of events that leads to the problem...
[04:24:36] stichnot: First, something happens (I'm not sure what) to make the selected AVSubs track go away. Then DecoderBase::AutoSelectTrack() finds that wantedTrack[type] is once again available, and it calls m_parent->TracksChanged(type); at the end to reenable. MythPlayer::TracksChanged() sets enableCaptions=true, then MythPlayer::SetCaptionsEnabled() is called. This calls NextCaptionTrack(kDisplayNone) to g
[04:24:38] stichnot: o through that sequence of caption types and enable the first one available.
[04:26:48] stichnot: So I thought about getting NextCaptionTrack to figure out if DecoderBase's wantedTrack[type] is set and prioritize that if so, but then it just seemed simpler to unconditionally favor AVSubs over 608/708.
[04:29:00] taylorr: seems the proper solution would be to explicitly set to wantedTrack.... changing prioritization could break other scenarios
[04:29:20] taylorr: but you know more about this than most so your call
[04:44:41] kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has joined #mythtv
[04:45:25] sphery: stichnot: I recommend asking danielk22 about the order--pretty sure he set it. That said, it seems like it would be fine to put captions after av sub, primarily because "regular" TV--where we do want the caption tracks--won't have av subs (at least TTBOMK)
[04:59:19] ** Beirdo TT BONKs sphery on the head **
[04:59:22] stichnot: I couldn't come up with any scenarios involving breakage, but who knows. I can think some more about how to cleanly expose (and maintain) wantedTrack.
[04:59:45] stichnot: and also wait to hear what danielk22 has to say...
[05:09:11] sphery (sphery! has quit (Quit: leaving)
[05:09:30] sphery (sphery!~mdean@mythtv/developer/sphery) has joined #mythtv
[05:10:48] jya: Captain_Murdoch: your patched worked… Still had to download it twice to have it active, but that's much better
[05:11:55] jya: ohhh.. finally, Sphere theme working for me :)
[05:14:46] abqjp: Sphere? I don't see that one in my list even after refreshing the list...
[05:16:03] jya: sorry Steppes :)
[05:16:25] jya: bizarre, that only that one failed to install before
[05:17:09] abqjp: Oh... cool.
[05:17:25] stichnot: How can I view the "not shown" changes to NuppelVideoPlayer.cpp at . . . 5f9d4#diff-2 ?
[05:19:04] abqjp: jya, maybe only updates to pre-existing themes was working?
[05:19:32] jya: stichnot: you want to see it in github or you're happy from a command line ?
[05:19:37] sphery: stichnot: only way I've found is to not use github's viewer (i.e. use the git tool and repo)
[05:20:04] sphery: github is smarter than us, so it knows you don't realloy need to see that change
[05:21:06] jya: git diff 59f10b3982aec9ab228169c3935e94509ee5f9d4~ 59f10b3982aec9ab228169c3935e94509ee5f9d4 — libs/libmythtv/NuppelVideoPlayer.cpp
[05:21:51] stichnot: jya: sweet, thanks!
[05:22:07] jya: you have to be in the mythtv folder for that
[05:26:26] abqjp: jya, how is that different than "git diff 527fe8016..59f10b398" Is it only showing the changes IN NuppleVideoPlayer.cpp ?
[05:27:20] jya: one is change between two changeset
[05:27:26] jya: the other is change since this changeset
[05:27:39] abqjp: Ah. Thanks.
[05:27:48] jya: oh sorry, the line I provided give the changes on THAT file only
[05:27:52] jya: not the whole changeset
[05:28:15] jya: if you do git diff 59f10b3982aec9ab228169c3935e94509ee5f9d4
[05:28:23] jya: it will show all changes since 59f10b3982aec9ab228169c3935e94509ee5f9d4
[05:28:27] jya: that is.. a lot
[05:28:50] jya: (and that file NuppelVideoPlayer.cpp doesn't even exist anymore, it was renamed)
[05:28:51] abqjp: "git diff 527fe8016..59f10b398" is JUST that changeset, right?
[05:29:00] jya: no
[05:29:06] jya: ah sorry..
[05:29:32] jya: I didn't see the ..
[05:29:47] jya: that diff shows the changes between the two changeset
[05:29:58] jya: I don't like having to know to commit number
[05:30:05] jya: so I use the ~
[05:30:27] jya: stichnot wanted to see the change on one file only, not the whole lot
[05:30:47] abqjp: Yeah, I had to use git log, to look up the precessing hash.
[05:34:56] jya: damn… pressing a key just do beep thing on a mac occurs with Qt 4.8 as well
[05:34:58] jya: fuck.
[05:42:01] MavT (MavT! has joined #mythtv
[06:09:14] jya: sphery: could you give me the magic SQL queries I need to run to retrieve the video profile for one host?
[06:09:33] jya: I created the VDA profiles I want to automatifcally add later, just need to retrieve what was stored
[06:13:20] sphery: jya: ... Are you also adding one or more OpenGL Playback Profile Groups? I've been wanting to, but can't get anyone to provide a working OpenGL playback profile (let alone a group of them--even ignoring the desire to have a slim, normal, and high quality group).
[06:13:47] jya: I'm only looking at creating the VDA profile right now
[06:13:59] stichnot: OK, hopefully danielk22 can recall his reasoning as to the caption priority order established 6 years ago in NuppelVideoPlayer::ToggleCaptions(void), . . . 987fa0256223 :)
[06:14:04] jya: same High, Normal, Slim type.. Changing the quality of the deinterlacer
[06:14:28] sphery: In other words, if you have any ideas for useful opengl playback profiles and/or groups, we could really use some examples for them
[06:14:49] jya: it would be pretty similar to my VDA ones
[06:14:54] jya: except no vda decoding
[06:15:00] jya: in fact, it would be identical
[06:15:28] jya: But I thought that was High Quality was ?
[06:15:30] sphery: any I've created get terrible performance (allowing only 1x playback on my dual-core 3GHz CPU in performance mode with GT220--and not even 1X before my GT220), so I'm unwilling to come up with the profiles :)
[06:15:32] jya: an OpenGL profile
[06:16:08] jya: really? that's surprising, opengl performance with the GT220 is pretty good
[06:16:16] jya: having said that, why would you when ypu have vdpau ?
[06:16:29] sphery: High Quality is ffmpeg + Xv + softblend with linearblend for high-res and yadif for SDTV res
[06:16:46] wagnerrp: jya: better scalers than we provide for opengl
[06:17:01] jya: we should change that from Xv to OpenGL imho
[06:17:11] jya: nothing else would require change
[06:17:18] sphery: I use vdpau renderer with my gt220, but have been trying to come up with a working opengl profile so that people who use the wrong video cards can have an example :)
[06:17:51] jya: and changing vdpau -> standard, renderer: vdpau -> opengl, osd -> opengl2
[06:17:53] sphery: I don't think we should change existing Slim/Normal/High Quality... We should add OpenGL Slim/Normal/High Quality
[06:17:55] jya: didn't do the trick?
[06:18:21] jya: it would make no difference to people who already have the profiles
[06:18:24] jya: only on new install
[06:18:49] sphery: it would make it harder to figure out what's required to use that one versus just saying, "OpenGL Slim/Normal/High Quality"
[06:19:02] jya: do we really want peple to know though?
[06:19:11] jya: I mean, you want high quality, you get high quality
[06:19:20] jya: who needs to know it's xv or opengl underneath?
[06:19:24] jya: I don't think anyone care
[06:19:25] sphery: besides, we have slim/normal/hq and vdpau slim/normal/hq--and soon, vda slim/normal/hq--so why not opengl slim/normal/hq
[06:19:33] jya: and those that do would probably create their own profile anyway
[06:19:34] sphery: just makes it consistent
[06:19:54] sphery: and I think you have to know if it's xv or opengl because of the implementations of BrokenGL shipped by a specific vendor
[06:20:03] jya: because the way I see it, OpenGL is just part of what should be standard
[06:20:11] sphery: tell AMD/ATI
[06:20:17] jya: so you get: Standard/Sotware: high, normal, slim
[06:20:25] jya: then the hardware ones: VDA, VDPAU, XVDA
[06:20:28] sphery: I have to add the UI Painter setting back in because of broken AMD/ATI drivers
[06:20:47] jya: playback on my mac with openGL is brilliant
[06:20:52] jya: it's just under linux..
[06:20:58] jya: so really the card can easily do it..
[06:21:05] jya: if only they got their act together
[06:21:11] warped (warped!~piotro@ has joined #mythtv
[06:21:24] sphery: it works fine on mine--as long as I never use timestretch
[06:21:37] sphery: and only because I have a 3GHz dual-core cpu
[06:22:10] Beirdo: just think: you could use an Atom and an ION GPU... :)
[06:22:13] sphery: but, anyway, I want one of these people who keep telling me, "It works great," to give me one of these "great" profiles, at least
[06:22:57] sphery: as, obviously, it makes no sense for me to create one since I can't get it to work without max'ing out my system
[06:23:03] jya: I see.
[06:23:10] jya: well, I can create the default profile
[06:23:15] jya: subject to tweaks
[06:23:41] jya: I don't expect to get it perfect right away.. for the VDA, it 's great, I have the fastest mac available, and also the slowest
[06:23:46] jya: so testing that will be easy
[06:24:09] Beirdo: the slowest? Got an old powermac? :)
[06:24:23] jya: intel
[06:24:27] jya: macbook air
[06:24:30] Beirdo: ooh . hehe
[06:24:51] jya: it's easily the slowest intel ever. 1.4GHz core 2 duo
[06:24:56] sphery: I hope we can do the tweaking before pushing--since updating the definition should involve a db update to remove the old/changed definition--but it would be nice if you could come up with an example and others could test it
[06:24:58] Beirdo: I do have a G3 and G4 still around, not doing much work though
[06:25:10] jya: sphery: you don't need to change the db at all
[06:25:22] jya: the way the profiles are made is that if the profiles don't exist, they are created
[06:25:25] jya: if not, they aren't
[06:25:38] sphery: not if we want the people who upgraded before we fixed it to continue having broken ones in their databases
[06:26:02] warped (warped!~piotro@ has quit (Client Quit)
[06:26:08] sphery: no, I mean the "subject to tweaks"... get those tweaks done before we add the new profiles so that we don't save incorrect definitions to the database
[06:26:11] jya: well, that's the beauty of using beta isn't it ? :)
[06:26:24] Beirdo: well, use it quick :)
[06:26:36] jya: that's exacrltly how I created the vdpau profiles..
[06:26:38] Beirdo: RC1 is scheduled for this weekend, is it not? ;)
[06:26:41] jya: nothing complained about it
[06:26:50] Beirdo: :)
[06:26:54] jya: and we added default options later (like skipchroma etc)
[06:26:55] sphery: you didn't change the vdpau profiles after initial commit
[06:27:04] jya: sphery: I did..
[06:27:06] jya: more than once
[06:27:08] sphery: oh, then we should have fixed them
[06:27:19] jya: like when the vdpau filters change or got added
[06:27:29] jya: Like the colorspace option
[06:27:36] jya: got renamed a few times
[06:27:40] sphery: anyway, whatever, I guess that people can say, "I'm using <whatever>" and I can think they're using one definition, but they're using something else... Makes support fun that way.
[06:27:58] jya: yeah… well, I much prefer to have to delete the profiles to recreate them
[06:28:08] jya: than upgrading the DB
[06:28:15] jya: I personally *hate* db upgrades..
[06:28:26] jya: I think we should always try to do without
[06:29:10] sphery: well, I personally *hate* db's with broken data
[06:29:27] jya: profile wouldn't be "broken" data :)
[06:29:39] jya: one thing I never understood
[06:29:46] jya: is say you have a vdpau profile
[06:29:53] jya: and vdpau is suddenly not available
[06:30:01] jya: what profile does it default to?
[06:30:10] jya: if you have one rule with vdpau and another that is ffmpeg
[06:30:16] jya: does it skip the vdpau one ?
[06:30:19] sphery: anyway, the point is I would like to have example Xv-based, vdpau-based, opengl-based (and, I suppose vda-based) playback profile groups available in 0.25
[06:30:56] Beirdo: sphery: I think we are too late for that
[06:31:05] Beirdo: that sounds very much like a new feature to me
[06:31:23] Beirdo: for 0.25.1, sure ;)
[06:31:28] sphery: yes, if profile 1 is a vdpau one and profile 2 in that same playback profile group is an ffmpeg one--and both are a match for the show's resolution comparisons), it will notice that vdpau isn't available, so it will go to the next profile
[06:31:30] Beirdo: assuming we make one
[06:32:33] sphery: I don't see an example playback profile group configuration as a feature, but if we are calling it one, the opengl playback profile groups can wait until 0.26
[06:33:07] sphery: we haven't had them before (and, really, I think opengl rendering is useless, anyway, so... I was just hoping someone could prove me wrong)
[06:33:40] Beirdo: well, convince others :) I think we are kinda getting really late in the game for anything but pure bug-fixes, and that's not really a bug-fix...
[06:33:51] sphery: true
[06:33:55] Beirdo: my thinking anyways
[06:34:30] sphery: I did ask a few months ago--and a couple times since then--but still no one has brought me any working opengl-based playback profiles
[06:35:12] jya: Beirdo: it's not new feature as such, all you add are preset settings
[06:35:41] jya: and right now, on many mac for example, the change to opengl makes it unable to play h264 content without stuttering
[06:35:44] jya: so fix it is :)
[06:36:03] Beirdo: I don't see that as a bug fix, personally
[06:36:45] jya: I added VDPAU profiles exactly in the same lifecycle last time (for 0.22)
[06:36:46] Beirdo: that's stretching it... I'd suggest asking on -developers to make sure we have concensus on it (and so we can understand the ramifications, etc)
[06:36:58] jya: do we ever achieve concensus?
[06:37:15] jya: I mean my backport, so we have you thinking it's a great idea, and stuartm thinking it's not
[06:37:23] jya: from that point on: nothing
[06:38:04] Beirdo: well... I hear ya
[06:38:35] Beirdo: but as far as feature vs. bug-fix... I think we need to use the official communications channels
[06:38:56] jya: very well
[06:39:00] Beirdo: it doesn't sound like the right time for it to me, but I also don't completely understand it...
[06:39:03] Beirdo: so...
[06:39:27] jya: basically what the code does, is check if a profile of a given name exists
[06:39:40] jya: if it doesn't, create it with preset values
[06:39:42] Beirdo: less than a week to RC1, then 2 weeks to full release... if this can possibly be adding more bugs or issues, I'd rather we don't
[06:40:01] jya: I'll post the changeset when I'm done
[06:40:09] jya: it will alleviate your concern, trust me
[06:40:27] jya: the VDA are mac only for a start
[06:40:42] Beirdo: heh, fair enough.
[06:41:10] jya: one thing I would have loved , is change that damn help for the decoder
[06:41:11] Beirdo: hey, my TV deals with 50fps pretty well
[06:41:11] jarle (jarle! has quit (Ping timeout: 265 seconds)
[06:41:21] jya: it still only mentions mpeg2 only
[06:48:01] Yancho (Yancho!~mpulis@unaffiliated/yancho) has joined #mythtv
[06:53:19] jarle (jarle! has joined #mythtv
[07:09:16] gregL (gregL! has quit (Read error: Operation timed out)
[07:18:09] Chutt (Chutt! has quit (Read error: Connection reset by peer)
[07:18:17] gregL (gregL! has joined #mythtv
[07:33:20] The_Ball (The_Ball!~The_Ball@ has quit (Read error: Connection reset by peer)
[07:40:18] The_Ball (The_Ball!~The_Ball@ has joined #mythtv
[07:50:22] warped (warped!~piotro@ has joined #mythtv
[07:51:18] GreyFoxx (GreyFoxx!~greg@mythtv/developer/GreyFoxx) has quit (Ping timeout: 245 seconds)
[08:01:43] GreyFoxx (GreyFoxx! has joined #mythtv
[08:27:53] kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has quit (Quit: kormoc)
[08:54:12] warped (warped!~piotro@ has quit (Quit: warped)
[09:09:49] stuarta: sphery: thanks for knowing about the ticket
[09:21:12] xavierh (xavierh! has joined #mythtv
[09:50:07] jya: sphery: I remembered the testing I had made of all the deinterlacers many years ago. and markk was kind enough to point me some of the benchmarks he had run
[09:50:08] jya: openglkerneldeint
[09:50:13] jya: . . . #post1814111
[10:05:03] Guest19466 (Guest19466! has quit (Remote host closed the connection)
[10:05:52] mike|2 (mike|2! has joined #mythtv
[10:42:34] _charly_ (_charly_! has quit (Ping timeout: 246 seconds)
[10:50:47] zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv
[10:56:18] zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[10:56:42] _charly_ (_charly_! has joined #mythtv
[11:02:13] MavT (MavT! has quit (Ping timeout: 244 seconds)
[11:10:01] MavT (MavT! has joined #mythtv
[11:11:43] joki (joki! has quit (Ping timeout: 245 seconds)
[11:12:07] joki (joki! has joined #mythtv
[12:04:34] davide (davide! has quit (Remote host closed the connection)
[12:04:55] davide (davide! has joined #mythtv
[12:09:39] seeker (seeker!~seeker@unaffiliated/seeker) has joined #mythtv
[12:11:20] cattelan_away (cattelan_away! has quit (Ping timeout: 244 seconds)
[12:17:09] cattelan_away (cattelan_away! has joined #mythtv
[12:20:01] seeker (seeker!~seeker@unaffiliated/seeker) has quit (Quit: Colloquy for iPhone -
[12:36:51] danielk22 (danielk22!~danielk@ has joined #mythtv
[13:08:38] kenni: jya: If you're reluctant to fix the help text for the decoder due to the string freeze, don't be. We're still in the "soft string freeze", which is basically just a way of telling the translators that the strings are close to being final. There're still 4 more days to fix strings before we hit the hard string freeze with RC1.
[13:08:56] jya: ah cool
[13:09:04] jya: yeah, that's precisely why I didn' change it
[13:13:05] stuartm: right, with thousands of strings requiring translation the 'soft freeze' just tells translators that they should get to work if they haven't already done so on the bulk of the translations, the handful of strings changed during the soft freeze is a drop in the ocean
[13:22:21] zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv
[13:36:52] danielk22: kenni: thanks.. I found some unhelpful setup help text the other day and didn't change it because of the string freeze. I'll see if I can locate that again & fix.
[13:45:59] jya: sphery: if you want to try.. Here are the VDA and OpenGL new profiles (combined patch, but they are different commits)
[13:47:25] jya: For OpenGL High and Normal, I've set the number of processors to two. I'm guessing anyone with a graphic card powerful enough these days to do OpenGL would be using at least a dual core processor… It's pretty safe, I believe, to assume everyone has at least a dual core processor…
[13:49:32] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has quit (Quit: jpabq_)
[13:50:23] jya: anyone with a better suggestion that the woefully inadequate "Decoder to use to play back MPEG2 video."
[13:51:54] wagnerrp: which one is that?
[13:52:04] jya: TV playback profile
[13:52:08] jya: decoder to use
[13:52:21] jya: eg. the decoder to use to decode any content
[13:52:36] wagnerrp: oh, just the general description of that field
[13:52:41] jya: yep
[13:52:47] wagnerrp: i thought you were implying the libmpeg2 option somehow made it back in
[13:53:43] jya: according to the current doc, the decoder is only used for mpeg2
[13:57:54] xavierh: I was wondering if the help text for the VideoDirectory's type of setting should mention that the Storage group is the preferable method ?
[13:58:27] jya: xavierh: that's a very wise suggestion
[13:58:42] jya: (that option should even be removed IMHO)
[13:58:56] xavierh: Is it still used ?
[13:59:22] jya: it is.. if you have defined it, it will list content from there, along side any storage group
[13:59:57] j-rod|afk is now known as j-rod
[14:00:19] jya: or at least that was like that for 0.24.. it may have been completely obsolete, in which case it definitely needs to be removed
[14:00:28] wagnerrp: i would go so far as to put in a warning not to use it, as it will be removed from the next release
[14:01:10] wagnerrp: the last thing that was really keeping it in there was the issue with encrypted ISOs
[14:01:33] wagnerrp: but since the video player in 0.25 will attempt to access storage group content from the local filesystem if available
[14:01:37] wagnerrp: it sidesteps that whole issue
[14:01:44] jya: "was"? I know that was on markk todo list, but he did ever get to do it ?
[14:02:10] xavierh: Is the Default storage group used for the recordings ? I did not see any Recording storage which puzzled me
[14:02:16] wagnerrp: TBH, im not sure
[14:02:25] wagnerrp: i was under the impression that happened like a year ago
[14:02:54] wagnerrp: xavierh: correct
[14:03:12] jya: nah… there were talk just a few weeks ago on how to make myth:// available as a wrapper
[14:03:32] wagnerrp: right, for external players
[14:03:33] xavierh: Is it only used for the rcording, or other storage will use default storage if not defined ?
[14:03:46] wagnerrp: thats independent from the SG/ISO/CSS issue
[14:04:22] wagnerrp: even if the content is available locally, mythvideo will still feed the myth:// URI to an external player
[14:04:38] wagnerrp: and no external players (besides mythavtest) understand how to use that
[14:05:52] stichnot: danielk22: did you happen to read the discussion last night about priority order when cycling through captions types, and my proposal to move AVSubs before 608/708?
[14:07:16] xavierh: If the Default Storage group is only used for recording, it should be rename as Recordings Storage group
[14:09:12] jya: xavierh: my understanding is that it will be used for things like backup, fanarts etc.. if no other SG are defined
[14:10:37] xavierh: jya: would have been better to have a Recording storage group and a Default storage group. Also the Default is not used in all situation ad you need to define a Video storage group for the "MythVideo" to work
[14:13:17] taylorr: danielk22, Captain_Murdoch: do you guys remember any changes to the ringbuffer that would allow the player to start so fast that it's almost at the very end of the file from the start – when I start livetv I'm only 3 seconds behind real-time – this causes glitches in playback – skipping back 5 secs alleviates all problems
[14:14:11] taylorr: also, different issue, while seeking to the end of an in-progress recording it will exit back to the main menu
[14:14:24] taylorr: s/main menu/PBB/
[14:14:35] j-rod is now known as j-rod|afk
[14:19:16] j-rod|afk is now known as j-rod
[14:21:48] danielk22: stichnot: No I missed that. Ideally we'd keep AVSubs if those were selected and 608 when those are selected. If you are deaf you generally prefer the 608 captions which tell you '[loud thud]' '[suspenseful music]' etc.
[14:23:13] jya: taylorr: haaaa, that could explain why I get a lot of errors lately in LiveTV (all fine once it's recorded) with some stutter and in the log I see about missing data
[14:24:07] cattelan_away (cattelan_away! has quit (Ping timeout: 244 seconds)
[14:24:20] danielk22: taylorr: It probably wouldn't be the ringbuffer, it would be the MythPlayer::IsNearEnd() and IsReallyNearEnd() handling. And I vaguely recall changes there by Mark, but not if they were before or after 0.24.
[14:26:00] stuartm: does anyone remember why we prefer the audio track with the highest number of channels instead of logical order? I know it was a workaround for a particular source where the track ordering was incorrect but I can't remember what that source was
[14:26:04] danielk22: taylorr: There have also been changes to 'speed up playback start' which could have caused this. Again Mark was making the changes so I didn't worry too much about it as I figured he'd be dealing with the fallout.
[14:26:28] j-rod is now known as j-rod|afk
[14:27:23] jya: stuartm: that's a change I made… Preference is based depending on number of channels, codec
[14:27:45] jya: reason being otherwise it would never select say TrueHD or DTSHD, as AC3 is usually the first track
[14:28:18] jya: though there has been a similar mechanism for quite a long time, I only extended it to include the new codec and change a bit the codec priorities
[14:28:52] stuartm: it breaks some DVDs where the commentary track has more channels/better codec
[14:29:07] cattelan_away (cattelan_away! has joined #mythtv
[14:29:24] danielk22: stuartm: There is a long explanation of the logic in the doxygen docs for AvFormatDecoder::AutoSelectAudioTrack(void).
[14:30:15] jya: stuartm: it may break some DVD, but I guarantee you that doing otherwise would break most of the usual DVD / BD playback
[14:30:18] j-rod|afk is now known as j-rod
[14:30:43] danielk22: stuartm: Auto select should really exclude directors commentary tracks. Are they labeled in some way we can see?
[14:31:13] stuartm: jya: I can't see how, DVD doesn't include an audio hint that I'm aware of, so commercial players would always use logical track order?
[14:31:29] jya: e.g. it's far better like this (requires very little user input) . Otherwise, you would need to change pretty much all the time the audio track
[14:31:47] stuartm: danielk22: not that I'm aware of, but I'm no authority on DVD authoring/metadata
[14:31:56] jya: i'm not sure how commercial player do it… they always get it right :)
[14:32:38] stuartm: afaik they always play track #1 – but again I don't claim to know everything
[14:32:47] jya: having said that, I'm pretty certain that giving pririty to the track with the highest number of channel as been there for quite a while
[14:33:05] danielk22: jya: The probably run some code provided in the DVD, DVD's little executables to run in a VM.
[14:33:23] jya: I think the choice is something like language > number of channels > codec
[14:34:24] jya: i know for all my bluray, it required constantly to go into the menu, audio, change audio track, select the TrueHD one.. bloody pain (especially as our menu is far to be the most convenient)
[14:35:25] jya: stuartm: the default language has the top priority. Eg , if language is set to spanish as default, if you have 2 tracks, #1 english, #2 spanish: it will use spanish
[14:36:27] jya: ok.. I have my todolist for tomorrow, fix the bug in the VDA code… what would have probably takes markk 2 hours will take me the rest of the week :)
[14:36:38] jya: gotta learn one day
[14:36:40] stuartm: isn't the principle of least surprise best here? If track #1 is always the feature audio and that seems to be the case, then isn't it preferable to have the ac3/mpeg1 audio instead of getting the commentary track?
[14:37:06] stuartm: at least for DVDs
[14:37:35] stuartm: I consider getting the wrong audio worse than getting a stereo track
[14:37:42] jya: that may be true for the very rare cases where the commentary tracks have more channels than the default
[14:37:47] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has joined #mythtv
[14:38:17] jya: but while you would correctly handle one case, you would make it worse for the majority of others
[14:40:34] jya: anyhow, if you want to change the behaviour (and I haven't heard anyone complaining about it so far), feel free to add another user settings, disabled by default that ignore the audio track automatic selection and always select the first one
[14:41:07] jya: but surely there must be a way
[14:41:21] jya: I have a few DVD with a stereo, AC3 and DTS track (in that order)
[14:41:29] jya: and it always start with the DTS one
[14:41:37] jya: (my DVD player that is)
[14:41:39] stichnot: danielk22: Playing my problematic DVD (Winnie the Pooh 1977, main-title rip) with my caption reorder change — enabling AVSubs via OSD menu or keybinding locks AVSubs for the rest of the playback, and enabling cc608 also locks cc608 for the rest of playback. The only visible change is that if you have captions enabled by default, or toggle subtitles, you'll see AVSubs first on a DVD that...
[14:41:40] stichnot: ...has both.
[14:42:09] ** jya off to be **
[14:42:17] ** jya off to bed that is **
[14:42:19] stichnot: btw, on this and other DVDs I've seen, there's corruption or missing captions all over the place for cc608, but that's a bug for a different day...
[14:44:04] stichnot: danielk22: maybe I should open a ticket and upload an excerpt of the DVD? (though the excerpt will be pretty large as the ISO is 3.6GB for a 74-minute movie)
[14:46:22] danielk22: stichnot: I don't mind the change.
[14:46:58] stuartm: fwiw, I don't agree that getting a stereo track instead of 5.1/7.1 is worse than getting the directors commentary instead, I don't think 'granny' cares about getting surround sound but would be baffled if she got completely the wrong audio, those that care enough about getting the best audio track are technically minded enough to know how to change the track
[14:48:54] danielk22: stuartm: How many DVDs are actually mastered this way? Seems odd to have stereo for the main title and 5.1 for the director's commentary.
[14:49:27] stichnot: stuartm: I agree, as I've had a couple of such DVDs at home that have had waf impact.
[14:50:49] stichnot: Granny probably only has stereo audio equipment. Can that be taken into account for the auto track selection?
[14:52:28] stuartm: danielk22: I've no idea on the precise figures but I don't think it's as uncommon as might be first suspected, many TV series and old films were only recorded with stereo audio, when the commentary tracks were added years (or decades) later they were done in fully equipped studios and likely no-one would think twice about using just two channels
[14:53:02] Chutt (Chutt! has joined #mythtv
[14:54:03] stuartm: I'm trying to find out whether any audio metadata or hints are available in the DVD spec, but I'm not getting very far, google search results get dumber by the day
[14:56:05] taylorr: Beirdo, wagnerrp: why not keep the --logfile option so that people who don't want to use --logpath can have all there logs into a single, known location?
[14:56:50] wagnerrp: i just did the command line stuff
[14:57:40] wagnerrp: although the idea was to eliminate the issue weve been having trying to parse out what goes where when people submit logs with multiple applications logging to the same file
[14:57:46] stuartm: ah-hah
[14:57:48] wagnerrp: although since the application is named now...
[14:58:15] wagnerrp: potentially its more to do with difficulties of synchronizing multiple applications to the same file
[14:58:25] taylorr: danielk22: ok, I'll see if I can spot what changed in IsNearEnd() and IsReallyNearEnd() – hopefully we can solve both with a single shot
[14:58:35] stuartm: byte 5 of the audio attribute from the IFO, indicates the track type, 3 is commentary
[14:59:13] taylorr: wagnerrp: ok, it's always nice to have options if it's possible
[15:00:46] wagnerrp: taylorr: i think part of it was the fact that the new myth_system actually handles stdout/err
[15:00:57] wagnerrp: and if you dont handle it, it gets shunted to /dev/null
[15:01:05] wagnerrp: rather, if you dont tell it you want it
[15:01:45] wagnerrp: which reminds me, it would probably be a good idea to configure it to automatically handle logging coming from external applications
[15:02:22] wagnerrp: something to add back in for 0.26
[15:02:23] taylorr: you mean external apps like channel changing scripts?
[15:02:43] wagnerrp: yes
[15:03:00] taylorr: that will be missed
[15:03:03] wagnerrp: previously, they would tie into the same stdout as mythtv itself used
[15:03:22] wagnerrp: so if you used --logfile, they would have been nulled previously anyway
[15:03:47] wagnerrp: the only way you would have captured them in the past was to have the shell redirect stdout to a file
[15:03:54] taylorr: my channel changer script always sent logging to my backend logfile
[15:04:10] taylorr: wagnerrp: do think I did
[15:04:17] taylorr: s/do/don't/
[15:04:26] wagnerrp: ill have to check
[15:04:42] wagnerrp: but im pressure sure the previous logger would not do anything to stdout
[15:04:48] wagnerrp: s/pressure/pretty sure/
[15:05:23] taylorr: my channel changer is a perl scripts and it's "print" statements were logged for me
[15:08:21] stuartm: this looks kinda fun, there's a range of DVD functionality we currently ignore but could make use of
[15:10:24] xavierh: stuartm: Are thinking Karaoke mode :D
[15:10:58] seeker (seeker!~seeker@unaffiliated/seeker) has joined #mythtv
[15:11:08] stichnot (stichnot!~chatzilla@mythtv/developer/stichnot) has quit (Ping timeout: 240 seconds)
[15:11:10] stuartm: err, maybe as a long-term goal ;)
[15:12:59] xavierh: stuartm: Please don't, I want to be able to watch TV, If myth was supporting Karaoke, my wife would keep the remote control to herself :)
[15:13:50] wagnerrp: taylorr: looks like the whole application just got shifted over, and external applications were along for the ride... . . . ers.cpp#L275
[15:14:37] stuartm: right now I'm thinking about describing tracks better in the track selection dialog and fixing it so that we don't auto-select the commentary track, but after that I'll look harder at what we're not using and how it might improve the UI/UX
[15:15:46] wagnerrp: taylorr: well its something i can put on the list to do for 0.26
[15:16:15] wagnerrp: making it support the --logpath behavior would be simple enough
[15:16:19] stuartm: there's some stuff in there about 'parental management' controls which seems like it might allow titles to be hidden/disabled for children and maybe more
[15:16:41] wagnerrp: but wrapping it for database and syslog operation is considerably more intensive
[15:18:54] wagnerrp: not impossible, and i know how i would do it, but it would involve adding something like this... . . .
[15:19:00] wagnerrp: or adapting the existing ringbuffer code
[15:20:02] wagnerrp: right now, stdout/stderr from anything in myth_system is dumped into a QByteArray
[15:20:13] wagnerrp: where it remains until that object is destroyed
[15:21:59] cattelan_away (cattelan_away! has quit (Ping timeout: 244 seconds)
[15:28:34] cattelan_away (cattelan_away! has joined #mythtv
[15:45:45] zombor_ (zombor_!~zombor_@kohana/developer/zombor) has joined #mythtv
[15:48:27] zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Ping timeout: 244 seconds)
[15:57:20] zombor_ is now known as zombor
[15:57:38] cattelan_away (cattelan_away! has quit (Ping timeout: 244 seconds)
[16:01:17] cattelan_away (cattelan_away! has joined #mythtv
[16:02:19] gregL (gregL! has quit (Read error: Connection reset by peer)
[16:05:23] cattelan_away (cattelan_away! has quit (Ping timeout: 244 seconds)
[16:14:34] rsiebert (rsiebert! has joined #mythtv
[16:15:14] rsiebert_ (rsiebert_! has quit (Ping timeout: 246 seconds)
[16:19:29] gregL (gregL! has joined #mythtv
[16:27:42] davide: jpabq_, abqjp, jya: you don't need to look up the hash of the previous commit to see a diff. "git show <commit> [file...]" will do it.
[16:29:48] cattelan (cattelan! has joined #mythtv
[16:33:11] seeker (seeker!~seeker@unaffiliated/seeker) has quit (Quit: Colloquy for iPhone -
[16:40:06] stichnot (stichnot!~chatzilla@ has joined #mythtv
[16:40:06] stichnot (stichnot!~chatzilla@mythtv/developer/stichnot) has joined #mythtv
[16:40:06] stichnot (stichnot!~chatzilla@ has quit (Changing host)
[16:52:06] stichnot: stuartm: I'm still seeing the issue where entering frontend standby mode from the Watch Recordings screen causes it to ignore all keypresses except EXIT and jump points.
[16:53:45] stichnot: danielk22: after further testing, I'm not happy with just swapping the priority order between AVSubs and 608/708, because I still see situations where manually selecting 608 can later lead to AVSubs showing up instead.
[16:56:36] stichnot: I'd rather keep track of the specific selection within the MythPlayer class and prefer that one when turning on captions.
[16:57:36] stichnot: and then let the existing DecoderBase code continue to decide which AVSubs track is preferred when AVSubs are selected.
[17:02:11] seeker (seeker!~seeker@unaffiliated/seeker) has joined #mythtv
[17:04:42] danielk22: hmm, my mythfrontend appears to be completely locked up showing a "MythTV is in standby : Backend is recording" dialog. any ideas?
[17:05:25] stuartm: press 'Escape'
[17:05:33] stuartm: stichnot: I'll look into it
[17:05:34] danielk22: stuartm: doesn't work..
[17:06:05] davide (davide! has quit (Remote host closed the connection)
[17:06:26] davide (davide! has joined #mythtv
[17:07:02] stuartm: danielk22: maybe a backtrace would help, something is obviously blocking input and/or locking up the UI thread
[17:07:05] stichnot: danielk22: also try sending keypresses through the network control interface, I suppose
[17:07:26] stuartm: yeah, that would rule out a deadlock
[17:11:46] danielk22: stuartm: It unfroze after a minute or so. If I see it again I'll try to get a backtrace. Is there something I can do to speed up the reappearance?
[17:14:19] stichnot: Set FrontendIdleTimeout=1 in the Settings table — is that what you mean?
[17:14:31] stuartm: it should exit 'standby' immediately on any input (mouse, keyboard or lirc), although stichnot has indicated that it randomly will only respond to ESCAPE, this is the first report of it locking up and I'm unable to think of a reason for that at the moment, the 'standby' screen is no different to any other
[17:15:21] stuartm: 0 will disable it, 1 will set it to time out after one minute
[17:15:41] stuartm: which is likely not what is desired
[17:15:42] stichnot: for me it's not random. any time it enters standby from the pbb, it only responds to Escape and jump points. anywhere else, it responds to any key
[17:16:01] stichnot: 1 is good for debugging/reproducing, which is what I think danielk22 wanted
[17:16:10] stuartm: stichnot: hmm, I wonder if it's a bug in the jump-point code then
[17:16:42] stuartm: stichnot: ahh, right, yeah, I mis-understood what danielk22 was asking
[17:17:15] stichnot: I just tried it from one of the settings pages, and like pbb it also wouldn't respond to normal keypresses
[17:19:13] danielk22: I probably pressed some other key before I pressed escape, most likely either space or and up or down arrow; I wonder if the escape got queue up behind the other key. I was in the main menu though, not pbb.
[17:20:27] danielk22: s/space/shift/ — I often use shift to wake up a sleeping screen because it is fairly unlikely to do anything other than wake up the screen.
[17:21:37] stichnot: stuartm: you're suggesting that the jump-point code that brings it to the main menu before the standby screen is displayed, might have a bug? interesting
[17:22:56] stuartm: stichnot: it's one possibility, it's not the only one, what does the frontend log say?
[17:29:14] stichnot: nothing shows up in the rsyslog logs until it comes out of standby, then both "entering standby" and "leaving standby" messages appear
[17:29:30] stuartm: wtf ...
[17:29:48] kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has joined #mythtv
[17:30:16] stuartm: when does the "Entering standby mode after x minutes of inactivity" appear?
[17:31:05] stuartm: the syslog delay might just be a weird side effect of the new logging stuff, I'm not using syslog for mythfrontend
[17:32:19] stichnot: yeah, may have syslog and NFS delays going on here...
[17:32:22] warped (warped!~piotro@ has joined #mythtv
[17:32:29] stichnot: let me retry with --logpath
[17:33:42] dblain (dblain!~dblain@mythtv/developer/dblain) has quit (Ping timeout: 265 seconds)
[17:35:37] stuartm: underneath it all all 'standby' does is toggle the socket to the backend from blocking to non-blocking, nothing which should lock up the frontend or logging :/
[17:36:48] dblain (dblain! has joined #mythtv
[17:36:48] dblain (dblain! has quit (Changing host)
[17:36:49] dblain (dblain!~dblain@mythtv/developer/dblain) has joined #mythtv
[17:38:30] stuartm: the UI screen bit was just to appease those who wanted to keep the mythwelcome feel to things, so that they could see when a frontend wasn't going to block the backend shutdown and get the countdown when it was shutting down
[17:39:24] stichnot: from pbb, I get "Entering standby mode after 1 minutes of inactivity", followed by "Leaving standby mode" followed by "Entering standby mode". When I finally press Escape, there is a single "Leaving standby mode" message.
[17:40:45] stichnot: from the main menu, I get just "Entering standby mode after 1 minutes of inactivity". After a random keypress, "Leaving standby mode".
[17:40:51] taylorr: danielk22: I still think something is busted in the ringbuffer on the backend or the frontend since even if we are waiting for data it shouldn't result in a corrupted image – at most frame drops by the player
[17:41:25] stuartm: stichnot: ok, I now know why it's only responding to Escape, and it's probably related to the jumppoints in some way but I'll tackle it in the timeout code rather than risk new bugs by playing with the jumppoint stuff
[17:41:30] taylorr: one change we made was to enable the ringbuffer on the backend – we used to not do that – I think it's just a single line of code to change if I can find it
[17:41:40] stichnot: stuartm: thanks :)
[17:43:11] taylorr: danielk22: basically we get errors in the decoder if we play near the end of an live recording
[17:43:33] stichnot: stuartm: please add a note to #10407 if you think it's jumppoint related.
[17:45:54] Beirdo: taylorr: using one logfile is still doable via syslog. The real issue is mixing of all of the logs from all of the children into one file. If someone really wants that behavior (which makes debugging quite a lot more difficult), they can still use syslog.
[17:46:49] Beirdo: anyways, it's on my list for post-0.25 release to look at maybe making a way to log the stray outputs rather than trashbinning them, but that will have to wait for release
[17:47:07] taylorr: Beirdo: what keeps all these logs from building up – everytime we run a new process it creates a new file
[17:47:28] Beirdo: you can get logrotate to do it (or other such tools)
[17:47:45] Beirdo: but realistically, unless you hit a really nasty bug, it won't get bad
[17:48:09] Beirdo: on my backend, I haven't cleaned it out since July or so, and it has about 4G of logs
[17:48:24] taylorr: 4G is a LOT!
[17:48:32] Beirdo: which, on a box with 12TB of disk... is nothing
[17:48:33] stichnot (stichnot!~chatzilla@mythtv/developer/stichnot) has quit (Ping timeout: 252 seconds)
[17:48:55] taylorr: that's good for you but not someone who has a small system partition
[17:48:59] Beirdo: yes, more than most will ever see, but that's running max debug more than an average user
[17:49:12] taylorr: how would logrotate know what to do since the filenames are different each time
[17:49:12] stuarta: you can tell who is a sysadmin
[17:49:24] Beirdo: it can do wildcards
[17:49:40] Beirdo: like you can tell it to rotate /var/log/mythtv/myth*.log
[17:49:58] taylorr: ok, that sounds good
[17:50:10] Beirdo: we likely will need more tweaking
[17:50:28] Beirdo: and maybe including a commandline to enable per-pid logging
[17:50:35] Beirdo: but for now...
[17:51:06] Beirdo: it would still be per-application (in file mode) though
[18:02:05] Beirdo: I need to start taking notes before I forget some of these ideas :)
[18:12:42] Beirdo: done
[18:20:40] cattelan is now known as cattelan_away
[18:21:38] Beirdo: getting an ever-growing list of things to look at tweaking for 0.26 :) Cool.
[18:28:58] Beirdo: stuartm: you put in a way of changing the timeout (and disabling) the standby in the frontend, right?
[18:29:47] stuartm: yes
[18:29:59] Beirdo: I know you had mentioned it, but didn't remember if it went in or not :)
[18:30:16] Beirdo: cool, I'll go dig it up then
[18:30:29] stuartm: same place as the exit/shutdown settings
[18:30:44] Beirdo: cool
[18:31:11] Beirdo: A logical place too
[18:31:12] Beirdo: heh
[18:32:29] stuartm: I'm seeing a lot of these lately –
[18:32:46] Beirdo: eek
[18:33:02] Beirdo: sounds like we are freeing something we shouldn't be
[18:33:12] Beirdo: that's in normal playback?
[18:33:18] stuartm: I thought it was the driver, but I've just upgraded from 280.x and it's still occurring
[18:34:08] Beirdo: it seems to be in the driver by the backtrace, yeah, but maybe we free something the driver expects to be able to free?
[18:34:09] stuartm: Beirdo: all over the place, that time it was frontend startup, several times when exiting mythmusic, once or twice during playback and even just browsing around the frontend
[18:34:17] Beirdo: ick
[18:35:07] Beirdo: I'm still using 195.x
[18:35:30] Beirdo: on my frontend box.. I don't think I've seen one of those for a while
[18:35:31] stuartm: I can't rule out a hardware issue, possible memory problem (IGP, so shared memory)
[18:36:25] seeker (seeker!~seeker@unaffiliated/seeker) has quit (Remote host closed the connection)
[18:36:35] stuartm: nothing else seems to trigger it though, it's always mythfrontend that crashes and not other opengl using apps e.g. Stellarium
[18:36:53] Beirdo: yeah, that's gotta mean something.
[18:37:19] seeker (seeker!~seeker@unaffiliated/seeker) has joined #mythtv
[18:38:34] danielk22: The nvidia lib is a red herring. It's a double free or insufficient locking bug somewhere in the app.. valgrind can probably find it.
[18:38:35] stuartm: I'll grab a backtrace next time, they haven't meant much to me in the past but that's still when I thought it was the driver version
[18:39:14] Beirdo: yeah, if we can narrow it down, maybe we can fix it... or maybe not ;(
[18:39:56] seeker (seeker!~seeker@unaffiliated/seeker) has quit (Remote host closed the connection)
[18:40:27] stuartm: anyway, back to excluding commentary and AD tracks from the automatic track selection
[18:40:45] Beirdo: fun stuff. :)
[18:40:47] seeker (seeker!~seeker@unaffiliated/seeker) has joined #mythtv
[18:58:34] ThisNewGuy (ThisNewGuy! has joined #mythtv
[19:11:26] stuartm: it really is fun, although because of the feature freeze I'm having to keep it simple and maybe more of a hack than I would like
[19:12:34] Beirdo: yeah, but only 3 weeks (hopefully) and you can finish it right ;)
[19:13:06] Beirdo: we should start looking at remaining tickets again, I guess
[19:15:06] stuartm: just so nobody panics, I've created a new fork of MythTV on github just in case I need somewhere to push the work I'm unable to commit during the feature freeze
[19:16:13] Beirdo: heh, you can always push a branch into the main repo
[19:16:39] stuartm: it's called sporc ...
[19:16:43] Beirdo: like the utc and rtp ones danielk has and the gpu-commflag one I have :)
[19:16:53] Beirdo: but up to you how ya do it :)
[19:17:23] stuartm: Beirdo: true
[19:17:23] sphery: taylorr: a couple of logrotate configs: . . . nd.logrotate and . . . .logrotate.d
[19:17:59] stuartm: I wonder how long before sporc makes Phoronix
[19:18:08] Beirdo: but at least we know, which will keep people from going WTF?!?!?
[19:18:30] Beirdo: sporc. nice.
[19:19:26] ** stuartm wants to see a quote along the lines of "sporc is like torc, but with all the code which was ripped out added back in" **
[19:19:27] stuartm: ;)
[19:19:39] Beirdo: heheh
[19:20:28] sphery: taylorr: also, regarding "playing too close to the end", you may want to look at Mark S's mythtv-play-to-end.20120306.patch on ... Not sure if it would help, but it was meant to allow us to play to the end of the file (instead of exiting whenever we hit EOF), so might be useful? (And note that mythtv-6974-IsNearEnd_margin_for_end_of_recording_prompt.patch is a hack to work around issues in the prompt at ...
[19:20:31] natanojl (natanojl! has joined #mythtv
[19:20:34] sphery: ... end of recording code--and should, ideally, be redone as ), so you can ignore it.
[19:20:59] sphery: not sure how it will work with a file whose EOF keeps changing, but could be a good start
[19:29:09] stichnot (stichnot!chatzilla@mythtv/developer/stichnot) has joined #mythtv
[19:33:07] stichnot: Beirdo: would you be able to / interested in adding some sort of a "Refresh" button to the bottom of the IRC log page? That would make it considerably less painful to read the IRC logs on an iPhone.
[19:33:49] Beirdo: heh, that could be done.
[19:34:21] kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has quit (Quit: kormoc)
[19:35:40] stichnot: Sorry for the FRWOAP, but I'm html-illiterate
[19:36:17] Beirdo: heh, could you send me an email so I don't forget?
[19:37:20] stichnot: will do.
[19:37:23] kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has joined #mythtv
[19:37:32] Beirdo: thanks :)
[19:37:42] Beirdo: my memory sucks.
[19:45:00] davide: sphery (or anyone else): do you know of an easy way to remove seriesids when xmltv is used? the reporter of #10449 is the victim of bogus seriesids being provided by his grabber.
[19:47:12] stuartm: davide: xmltv doesn't supply a series id, we generate it based on the title in xmltvparser.cpp
[19:47:13] sphery: only way I know of is direct DB edits... dekarl may have some xmltv-based ideas, though?
[19:48:21] stuartm: if it's still demonstrably an issue with the grabber then it's better to get that fixed upstream by the grabber author than try to apply hacks which might break other grabbers supplying valid info (as I hope tv_grab_uk_rt will start doing in future)
[19:50:14] sphery: yeah, seems for xmltv we only have generated seriesids--generated in XMLTVParser::parseProgram() with QString seriesid = QString::number(ELFHash(pginfo->title.toLocal8Bit().constData()));
[19:50:42] davide: i agree it's better to fix the grabber, i was just hoping to give the guy an alternative in case fixing the grabber can't be done or would take awhile.
[19:51:16] sphery: which would imply that either the titles are the same or that ELFHash() doesn't provide sufficient coverage to be used to generate unique seriesids
[19:51:27] davide: if we're creating the seriesids, then this could be considered a mythtv bug. we should be creating hash coliisions that easily, should we?
[19:52:25] stuartm: I was looking at this the other day, the one significant thing missing from xmltv data that is present in the SD info is series id, which I'm going to discuss with the xmltv folk in the hope of adding it
[19:53:19] stuartm: sphery: it's going to suck if we change the hashing used and break backwards compatibility with hashes already in the database
[19:54:36] sphery: yep... for something like that, we'd want a db update--and, ideally, re-hash all existing oldrecorded/recorded/record (and other?) entries
[19:54:42] davide: stuartm: they could be fixed with a schema update. i'm sure jya would volunteer to do it! :)
[19:54:55] sphery: that said, I have no proof that ELF is the problem here...
[19:55:28] sphery: davide: 544 is /not/ a valid ELF hash
[19:55:36] stuartm: sphery: should be easy to write a quick test app, ELFHash() the two titles which appear to be colliding
[19:56:38] warped (warped!~piotro@ has quit (Quit: warped)
[19:56:54] davide: i was about to say i was going to run a test app on the titles he posted. iff 544 isn't valid, though, then something else is wrong.
[20:00:28] stichnot_ (stichnot_!~chatzilla@mythtv/developer/stichnot) has joined #mythtv
[20:00:58] stuartm: davide: is he actually using a proper xmltv grabber, or something that inserts directly into the database?
[20:01:02] stuartm: or even EIT
[20:01:40] stuartm: EIT will insert seriesid, although it should include the default_authority where that's available
[20:02:02] davide: he said he's using TVxb (
[20:03:08] stichnot (stichnot!chatzilla@mythtv/developer/stichnot) has quit (Ping timeout: 240 seconds)
[20:03:09] stichnot_ is now known as stichnot
[20:04:34] stuartm: hmm, not quite a 'proper' grabber, but it purports to use the official xmltv format and assuming that he's inserting the data through mythfilldatabase he should be OK
[20:13:48] sphery: davide: by my test program, the seriesid in row 1 is correct (79256886) and all the 544 ones are incorrect. has test output and test program (where ELFHash() was copied from mythtv code)
[20:15:02] davide: sphery: thanks for checking.
[20:15:15] stuartm: gigem: possibly still truncation issue in . . . ngs%202.png, since card ids can be > 9 (all of mine are 30+)
[20:15:20] davide: unless i'm reading the code wrong, it looks like seriesid is only populated by us when xmltv is used. so where could the 544 be coming from?
[20:15:49] sphery: good question
[20:15:56] stuartm: very good question
[20:16:24] sphery: and, yeah, with SD, we use the value provided and never generate a value, so we only generate values when using XMLTV
[20:18:01] sphery: does EIT do seriesid, too? perhaps he's got both xmltv and EIT on the same channel?
[20:18:28] davide: in the QString::toLocal8Bit() docs i see at, it says the results are undefined if any character can't be converted.
[20:18:51] davide: sphery: he's using analog.
[20:18:56] davide: so no eit.
[20:19:17] sphery: ah, yeah, guess that would be... (I asked in #mythtv-users a bit too quick, then.  :)
[20:20:25] davide: i've got a short status meeting i need to prepare for and attend. i'll be back in about 25 minutes.
[20:20:33] sphery: davide: my test was pure C, so I didn't test the qstring conversion... that could be it (though I'd think those chars should work fine)
[20:22:17] stuartm: sphery: does ELF hash always return a numerical value?
[20:22:37] stuartm: err, I guess yes, since it's returning 'int'
[20:22:42] stuartm: nm
[20:24:32] stuartm: sphery: I'm not actually sure why we're doing 'toLocal8Bit()' that seems pointless since it's already been converted to utf16 we risk losing information if converting down to utf8 or worse
[20:25:05] sphery: yeah, we're not actually using it with the underlying OS, so it seems unnecessary
[20:27:16] stuartm: won't do any harm to remove that and see what happens, at best it fixes the issue, at worst nothing happens
[20:31:13] seeker (seeker!~seeker@unaffiliated/seeker) has quit (Remote host closed the connection)
[20:31:37] ThisNewGuy (ThisNewGuy! has quit (Read error: Connection reset by peer)
[20:33:17] sphery: stuartm: wonder if that would cause problems because of how we're iterating over characters (using *name++ on const unsigned char *name)
[20:33:53] sphery: looks like maybe we're getting the 8-bit representation because of that?
[20:34:17] sphery: in . . . rser.cpp#L33 itself
[20:36:36] zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[20:37:30] cattelan_away is now known as cattelan
[20:37:46] stuartm: eww, yeah, converting to char isn't going to help, it doesn't handle unicode values
[20:38:11] Beirdo: I think we should force the rest of the world to use en_US.
[20:38:12] Beirdo: :)
[20:38:25] Beirdo: or en_GB (you guys already had a chance though)
[20:38:38] ** Beirdo is kidding of course **
[20:40:30] stuartm: and we're masking just the first 8 bits, hence why we were using toLocal8Bit I guess
[20:42:54] stuartm: sphery: seems that QT use ELF for hashing translation strings in the their qm files, maybe we should be looking at their implementation and comparing
[20:46:37] stuartm: sphery: maybe we should be using this –
[20:48:14] stuartm: uses uchar from the start instead of going via char which is pretty important
[20:48:30] ThisNewGuy (ThisNewGuy! has joined #mythtv
[20:50:31] stuartm: it doesn't squash ulong into int either
[20:55:14] sphery: I'm thinking, though, that--in this case--the problem isn't our ELF hash algorithm, but something else is either directly editing his DB or corrupting data for him.
[20:55:15] stuartm:
[20:55:45] sphery: as I did get the same hash value for the first record as his DB had, but the others were all corrupted to 544
[20:55:46] stuartm: sphery: right, I'm not sure it's our algorithm but I'm not entirely happy with it all the same
[20:56:23] sphery: yeah, might be worth looking at/improving (but maybe after 0.25?)
[20:56:24] stuartm: now that I look at it there is scope for error, the QT version is tidier
[20:57:10] davide: stuartm: double-digit cardids displayed fine when i made the change to do so several months ago. i'll see if if still works for me.
[20:58:23] davide: stuartm: would you like me to have the user try your patch?
[20:59:21] dekarl: davide, stuartm, sphery: there was a request for adding seriesid/programid to the DTD somewhere but I can only find the CRID<item> and CRID<series> request right now . . . ;atid=424138 but with good old DTD the extensions isnt as easy as with XSD ;)
[20:59:21] dekarl: btw, for xmltv we generate seriesid and programid. And otoh TVxb is payware/donationware (excemption for personal use) while explicitely mentioning that it only works because the guide sites tolerate the ToS violation...
[20:59:29] stuartm: davide: can't hurt, although let me upload a version which actually compiles ...
[20:59:44] davide: :)
[21:03:33] davide: dekarl: i don't see any bundled xmltv grabber for greece, so the guy might not have any choice but TVxb.
[21:05:26] dekarl: davide: thats like saying "for country X he has to use <the ms guide downloader and to xmltv converter programme that shall not be named>" ;)
[21:06:17] davide: dekarl: i know.
[21:06:19] dekarl: looking at the posted sql dump he'd likely get better data.
[21:07:17] davide: oops, i misunderstood you. it didn't occur to me the mce might work in greece too.
[21:07:46] stuartm:
[21:08:01] dekarl: But I don't get what is so hard about starting your own guide service. He can use a premade framework in Perl or Java for free.
[21:08:20] ThisNewGuy1 (ThisNewGuy1! has joined #mythtv
[21:08:43] dekarl: He's not getting the good editorial database others might know from TMS, but he's not getting that at the moment either
[21:09:30] stuartm: doesn't epgdata or whatever it was called cover Greece?
[21:10:09] ThisNewGuy (ThisNewGuy! has quit (Ping timeout: 245 seconds)
[21:10:40] stuartm: hmm, maybe not, last I heard they were aiming for EU-wide coverage but that's not happened
[21:11:45] dekarl: stuartm: I don't know, its not so easy to find at their site... but fiddling with the subscription link to point to Greece doesnt show an offer
[21:11:48] stuartm: at least judging by the fact that their site is only available in English, German, Spanish or French
[21:12:41] dekarl: says they have italy, netherlands and belgium, too
[21:12:51] stuartm: dekarl: even that much is broken here, there is no drop-down below "TV listings data for" when I suspect there should be
[21:14:26] dekarl: trying the links it looks like netherlands has been pulled, oh well
[21:14:30] sphery: dekarl: and, really, it's not the people in countries without any available XMLTV service using mc2xml that bothers me--it's the people in North America using mc2xml
[21:15:22] sphery: granted, it's still a violation of MCE ToS, but at least they're not doing it because they think stealing is better than paying $25
[21:16:57] dekarl: Well, its not as if some student/pupil couldn't set up such a service and make some bucks, then start his own business... I think Joakim is heading that way in Sweden and I like it.
[21:17:26] sphery: nice
[21:17:57] j-rod is now known as j-rod|afk
[21:24:34] dekarl: might be an idea to suggest extending musicbrainz to tvbrainz/moviebrainz as a GSoC project, are we a mentoring organization?
[21:25:52] stuartm: I'd start with musicbrainz support in mythtv :)
[21:26:28] kormoc: I doubt identifying recordings to figure out metadata about them is the best way to go about it
[21:26:46] dekarl: stuartm we already have some bits of musicbrainz support :D mythmusic looks at mb's tags (e.g. albumartist) and the ripper can use it via the cddb gateway (might be a nice default)
[21:26:48] kormoc: if it takes 15 minutes of high cpu + disk io, that seems of limited use
[21:26:58] stuartm: Paul has added support for artist/album/genre artwork, but only album artwork read from tags is really available until we add some grabbers
[21:27:57] dekarl: cool, who wants to write a artistphoto grabber that goes to flicker (and similar) to pull pictures via mbids? (not there yet as only 2–3 people tag their photos, so its a chicken'n'egg problem)
[21:28:06] stuartm: dekarl: the tags are misleading, we support one musicbrainz custom tag to indicate that the album is a compilation
[21:28:51] dekarl: with albumart i'd wait until the MB/InternetArchive Cover Collection takes off
[21:28:56] stuartm: as for the lookups, I didn't know there was a freedb gateway to musicbrainz
[21:37:07] superm1 (superm1!u4318@ubuntu/member/superm1) has quit (Remote host closed the connection)
[21:38:24] davide: okay, next stumper botd (bug of the day). does anyone have any idea what's going on with #10444? the behavior i see is exactly as expected. the guy seems intent on giving me all the proof i need except for the step by step instructions and logs i asked for. at this point, all i can do is guess at obscure things like bad builds, mismatched binaries/libraries, etc.
[21:38:38] superm1 (superm1!u4318@ubuntu/member/superm1) has joined #mythtv
[21:44:19] taylorr: danielk22: it seems like when the decoder reports issue is when it reads a chunk near where the ringbuffer is writing data
[21:50:36] dekarl: davide: the rule looks strange ... AND abs(datediff xyz) < 60%4 <- in my head the resolves to AND abs (x) < 0 which is never true
[21:51:14] stuartm: danielk22:
[21:52:00] stuartm: Beirdo: too since it seems the logging thread is involved ^^
[21:53:20] dekarl: davide: and it would be nice to see which rule "Monster Man" comes from, might be completely unrelated :)
[21:54:08] Beirdo: hmm
[21:54:31] Beirdo: you still have that open?
[21:55:01] Beirdo: oh, that's a sigint? you killed it... gotcha
[21:55:04] davide: dekarl: yeah, i noticed that too. he said it was added from mythweb as a canned rule, so i let it go. even if won't match anything it, it should record anything then. i'll make one more attempt to ask for detailed steps and logs.
[21:56:05] davide: stuartm: i confirmed the doublt-digit cardids don't get shown for me either with MythCenter-wide. i'
[21:56:17] davide: i'll try to figure out what's changed.
[21:56:44] Beirdo: stuartm: I'd try an export MALLOC_CHECK_=2
[21:56:47] Beirdo: and try again
[21:56:53] Beirdo: that should abort at the issue
[21:57:18] Beirdo: it looks like you are in =1 mode
[21:57:48] Beirdo: . . . hecking.html
[21:57:51] Beirdo: for reference
[22:02:15] superm1 (superm1!u4318@ubuntu/member/superm1) has quit (Remote host closed the connection)
[22:03:25] stuartm: 2 just causes the frontend to abort on startup in QT code
[22:04:00] Beirdo: it shouldn't unless that's what's causing the double-free
[22:04:04] Beirdo: odd
[22:04:30] Beirdo: got a backtrace of that? (curiosity)
[22:05:49] stuartm:
[22:06:54] Beirdo: nasty!
[22:07:49] Beirdo: wonder what it's not liking there.
[22:08:43] Beirdo: oh wow.
[22:08:46] Beirdo: what Qt?
[22:08:55] Beirdo:
[22:09:10] Beirdo: seems it might be a Qt bug for that
[22:12:01] Beirdo: looks similar at least
[22:15:28] stuartm: 4.7.4
[22:15:55] Beirdo: doesn't seem to be the exact same issue then.
[22:16:22] Beirdo: but that's making it impossible to use the malloc checker to find the other double-free issue
[22:16:43] Beirdo: valgrind might find it, if you have time to wait for it :)
[22:17:12] stuartm: not tonight, but maybe later in the week
[22:17:18] Beirdo: yeah. :)
[22:17:48] Beirdo: I'm gonna try running with =2 instead of =1 tonight, see what happens on my setup
[22:19:33] natanojl (natanojl! has quit (Ping timeout: 248 seconds)
[22:21:23] stuartm: mythdownloadmanager.cpp:1429:24: warning: unused variable ‘oldJar’
[22:21:46] stuartm: that's new?
[22:21:59] Beirdo: yes
[22:22:04] Beirdo: I'll fix it
[22:23:00] Beirdo: yeah, I missed deleting a useless line :)
[22:24:23] Beirdo: fixed. Thanks. I forgot to check the buildbot (naughty boy!)
[22:25:19] Beirdo: the mythuibuttontree warning should be easy to fix
[22:25:21] stuartm: spotted it while watching it build for the nth time tonight ;)
[22:25:32] Beirdo: change listid to int rather than uint
[22:26:14] Beirdo: heh, yeah, I often catch them flying by
[22:27:05] stuartm: yeah, that was my fault, although it was written when QT treated count() as uint instead of the strange int for a function that cannot return negative values
[22:27:31] Beirdo: heh :)
[22:27:39] stuartm: that was back with QT3
[22:27:58] Beirdo: yeah. would you like me to change that, or leave it for you? :)
[22:29:18] stuartm: you can fix if you want, I'm busy testing something else right now
[22:29:26] Beirdo: K.
[22:30:31] stuartm: I'm seeing a few more warnings with my version of gcc, although I think the bsd build is even more recent
[22:30:59] stuartm: . . . nings%20(47)
[22:33:22] Beirdo: OK, I'm putting in (uint) on the count to match the rest of the code or I'd have to change a lot more places, which isn't worht the trouble.
[22:34:58] Beirdo: oooh, no newline at end of file.
[22:35:20] Beirdo: the type-punned stuff is a bit more worrying
[22:37:32] Beirdo: haha, freebsd put FD_SETSIZE as unsigned?
[22:38:44] Beirdo: but I'm very happy that we've managed to knock out SO many warnings in this release cycle
[22:40:52] superm1 (superm1!u4318@ubuntu/member/superm1) has joined #mythtv
[22:46:00] ** Beirdo slaps teh 32-bit slave **
[22:46:58] Beirdo: no route to host.
[22:47:08] Beirdo: have a good sleep then, machine.
[22:52:55] danielk22: stuartm: would seem to indicate an error in the LoggingItem reference counting, but I can't find it.
[22:53:21] zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv
[22:53:39] stuartm: danielk22: what's easy_reader in StreamInfo?
[22:54:03] sphery: davide: do you actually have Droid Sans installed on your system? Most "doesn't fit in MythCenter-wide" issues are down to people without it installed getting DejaVu Sans as a substitute. Might want to check "fc-list | grep -i droid" and/or try downloading . . . ttf?raw=true and stick it anywhere in your MythCenter-wide directory (or in $MYTHTV_SHAREDIR/fonts) and ...
[22:54:07] stuartm: it's only referenced in the atscdescriptors by the look of it
[22:54:09] sphery: ... see if it fits.
[22:54:26] danielk22: stuartm:easy_reader are captions that don't use big words.. captions for those with limited english reading abilities.
[22:54:50] danielk22: stuartm: Sometimes seen on educational programming on the public broadcast station.
[22:55:42] sphery: davide: we should probably put droid sans into MythCenter-wide, just because this seems to happen a lot
[22:56:39] stuartm: danielk22: ok, I ask because in correcting a whitespace 'issue' in avformatdecoder I broke audio for dvb, when I started looking into it seems the incorrect number of arguments were specified for StreamInfo() when inserting an audio track, changing the whitespace changed which signature was matched
[22:57:10] danielk22: stuartm: Beirdo: I think perhaps the logging is just collateral damage.
[22:57:29] Beirdo: danielk22: I would agree (from what we're seeing there)
[22:57:37] danielk22: stuartm: how could changing whitespace change the function called??
[22:57:53] Beirdo: hard to say what it is so far though.
[22:58:36] Beirdo: if that were an abort, it would be a LOT easier to find the issue
[22:59:04] stuartm: danielk22: . . . .html#l02130
[23:00:08] stuartm: the sixth argument for that signature is easy_reader, why would we be setting that to true for an audio track?
[23:02:07] stuartm: then again, that doesn't explain the breakage of dvb audio, it's still strange
[23:02:36] taylorr (taylorr!~taylorr@unaffiliated/elmojo) has quit (Remote host closed the connection)
[23:03:27] stuartm:
[23:04:11] taylorr (taylorr! has joined #mythtv
[23:04:11] taylorr (taylorr! has quit (Changing host)
[23:04:12] taylorr (taylorr!~taylorr@unaffiliated/elmojo) has joined #mythtv
[23:05:50] davide: sphery: yes, i have the droid fonts. i reproduced and corrected the problem on my workstation at work, but i'm not using the myth approved 100dpi. i'll commit after i confirm on my tv at home. fwiw, MythCenter, default and defaul-wide all have the same problem.
[23:06:02] stuartm:
[23:06:38] davide: i think maybe one of jpabq_/abqjp's ui/clipping changes might have affected things. if i gave the "
[23:06:57] zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[23:07:37] davide: damn return key! if i gave the "card" area just a little more space, "..." would show up. if i then gave it a little more space, then a double digit card number would be shown.
[23:10:35] danielk22: stuartm: It looks like easy_reader was overloaded for the dual audio patch.. but that's not the cause of the wrong init being used
[23:12:23] stuartm: danielk22: yeah, I don't think the wrong init was being used now, it's late and it seemed like the only change which came even close to explaining the failure I was seeing
[23:13:36] danielk22: stuartm: The cause of that is that the second ctor used to have four ints before the first bool and the third used to have three, the types of the params led us to call the second ctor whenever there were four or more starting ints. Now the second ctor has 5 ints and the third has 6 ints..
[23:13:54] danielk22: stuartm: so the default ctor called when there are no bool params has flipped.
[23:14:57] danielk22: We should really get rid of the public ctor's aside from the first one and use factory methods.
[23:16:30] stuartm: aye, it's not the least bit helpful that the docs for that show args as 'a,b,c ... etc' rather than user-friendly names
[23:16:37] danielk22: of if that is too inefficient have some explicit StreamType that must be passed to the constructor.
[23:16:52] wagnerrp: Captain_Murdoch: you around?
[23:17:41] danielk22: stuartm: Yeah, it was never intended to be much more than a simple language->av_stream mapping.. but over time it grew...
[23:18:54] stuartm: I've just added another arg to deal with the need to 'type' audio tracks as commentary or audio-descriptive as opposed to 'normal'
[23:19:07] danielk22: The constructor was always kinda risky. That was my fault, I should have forseen that would be a problem later on.
[23:19:55] rsiebert (rsiebert! has quit (Ping timeout: 260 seconds)
[23:21:56] wagnerrp: danielk22, stuartm: either of you have an issue with me removing mythmessage?
[23:22:18] wagnerrp: someone just posted a page on the wiki documenting its usage
[23:23:00] Beirdo: that's covered by mythutil now, is it not?
[23:23:14] wagnerrp: correct, hence its planned removal
[23:23:47] ** Beirdo hands wagnerrp the nuke. Just need someone to type in the codes **
[23:25:54] wagnerrp: well let me rephrase... theres a page on the wiki just created of the same name, but its documentation is completely wrong
[23:26:13] Beirdo: heh. even better
[23:26:56] stuartm: danielk22: there's a comprehensive 'audioinfo' class in avformatdecoder.h, how about making that a subclass of StreamInfo and adding VideoInfo and SubtitleInfo subclasses? That would neatly compartmentalise the information which is specific to a particular stream type
[23:28:09] Beirdo: oh no. I detect thinking!
[23:30:46] knightr (knightr!~knightr@mythtv/developer/knightr) has quit (Ping timeout: 260 seconds)
[23:30:51] danielk22: stuartm: I dunno. StreamInfo was really supposed to be super simple, just enough to decide which stream to play. The problem with adding a lot of info is that it makes it harder to know what is actually initialized.
[23:31:29] knightr (knightr!~knightr@mythtv/developer/knightr) has joined #mythtv
[23:33:26] sphery: davide: 100dpi isn't required, anymore by any part of MythTV, so no worries about that
[23:34:49] stuartm: I understand, but at the same time having multiple structs/classes describing the same data isn't ideal unless there is a significant speed difference in their creation, at least for the audio stuff we appear to create the AudioInfo object later on anyway using some of the info from StreamInfo
[23:35:23] stuartm: anyway, it was just a thought, it's not really important to what I'm working on right now
[23:36:53] mrand (mrand!~mrand@ubuntu/member/mrand) has quit (Ping timeout: 248 seconds)
[23:37:52] danielk22: stuartm: Yeah, I just don't know without looking at it a lot more closely.
[23:40:05] mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv
[23:45:35] brtb (brtb! has quit (Read error: Operation timed out)
[23:45:40] brtb (brtb! has joined #mythtv
[23:50:05] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has quit (Quit: jpabq_)
[23:50:29] stichnot: If I add a protected field to MythPlayer, does it require a binary version bump?
[23:53:17] Beirdo: probably not, but I should have bumped it for my change with the cookie jars the other night and forgot. Ooops.
[23:54:02] stichnot: well, I can do it in the next day or so to be safe
[23:54:21] stichnot: assuming the fix I'm working on actually works :)
[23:56:37] Dave123 (Dave123! has quit (Read error: Connection reset by peer)
[23:58:29] ertyu-m (ertyu-m! has joined #mythtv

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