:: #mythtv

Daily chat history

Current users (82):

aarcane, allesmueller__, aloril_, amessina, Anssi, Beirdo, benklop, brfransen, caelor, Captain_Murdoch, cesman, Chutt, clever, coling, Cougar, dblain, dekarl, eharris, ElmerFudd, fetzerch, ghoti, Gibby, gigem, gregL, GreyFoxx, Guest59384, jams, jarle, jarryd, jheizer, joe_____, johanbr, joki, jpharvey, jst, jwhite, jya, kc, kenni, knightr, kurre2, kwmonroe, laga_, materdaddy, MaverickTech, moparisthebest, mrand, MythBuild, MythLogBot, Nothing4You, peper03, poptix, purserj, rhpot1991, robink, rsiebert, Seeker`, seld_, Sharky112065, skd5aner, sl1ce, SmallR2002, sphery, sraue, stichnot, stuarta, stuartm, superm1, taylorr, tgm4883, Tobbe5178, toeb, tonsofpcs, tris, wagnerrp, wahrhaft, warped, wolfgang4, XDS2010, xris, _charly_, _nyloc_
Friday, October 4th, 2013, 00:39 UTC
[00:39:30] gigem (gigem!~david@mythtv/developer/gigem) has quit (Quit: WeeChat 0.4.1)
[00:40:25] gigem (gigem! has joined #mythtv
[00:40:25] gigem (gigem! has quit (Changing host)
[00:40:26] gigem (gigem!~david@mythtv/developer/gigem) has joined #mythtv
[00:52:11] rsiebert (rsiebert! has joined #mythtv
[00:54:25] knightr (knightr! has quit (Read error: Connection reset by peer)
[00:54:32] knightr (knightr!~Nicolas@mythtv/developer/knightr) has joined #mythtv
[00:55:01] rsiebert_ (rsiebert_! has quit (Ping timeout: 245 seconds)
[01:14:43] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Quit: Body blow! Body blow!)
[02:09:16] allesmueller__ (allesmueller__! has joined #mythtv
[02:13:15] allesmueller_ (allesmueller_! has quit (Ping timeout: 248 seconds)
[02:13:33] _nyloc_ (_nyloc_! has joined #mythtv
[02:14:48] nyloc (nyloc! has quit (Read error: Operation timed out)
[02:53:24] joki (joki! has quit (Ping timeout: 240 seconds)
[02:55:23] peper03 (peper03!~peper03@mythtv/developer/peper03) has quit (Ping timeout: 248 seconds)
[02:56:31] peper03 (peper03!~peper03@mythtv/developer/peper03) has joined #mythtv
[02:59:19] joki (joki! has joined #mythtv
[03:32:25] stichnot (stichnot! has joined #mythtv
[03:32:26] stichnot (stichnot! has quit (Changing host)
[03:32:26] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[03:57:28] fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has quit (Ping timeout: 264 seconds)
[03:58:19] fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has joined #mythtv
[04:18:14] benklop (benklop!~quassel@2001:470:f400:47:ae81:12ff:fe31:668b) has joined #mythtv
[04:34:38] xibalba (xibalba! has joined #mythtv
[05:29:58] FabriceMG (FabriceMG!~Thunderbi@ has joined #mythtv
[05:35:50] jams (jams! has quit (Ping timeout: 240 seconds)
[05:44:07] jams (jams! has joined #mythtv
[06:00:16] SteveGoodey (SteveGoodey! has joined #mythtv
[06:11:34] SteveGoodey (SteveGoodey! has quit (Quit: Konversation terminated!)
[06:19:20] jams (jams! has quit (Read error: Operation timed out)
[06:31:18] jams (jams! has joined #mythtv
[06:46:51] warped_ (warped_! has joined #mythtv
[06:48:00] warped_ (warped_! has quit (Client Quit)
[06:51:03] jams (jams! has quit (Ping timeout: 252 seconds)
[07:22:44] stoffel (stoffel! has joined #mythtv
[07:27:54] cesman (cesman!~cesman@pdpc/supporter/professional/cesman) has quit (Quit: Konversation terminated!)
[07:31:06] cesman (cesman! has joined #mythtv
[07:31:21] cesman (cesman! has quit (Changing host)
[07:31:21] cesman (cesman!~cesman@pdpc/supporter/professional/cesman) has joined #mythtv
[07:38:14] jams (jams! has joined #mythtv
[08:00:14] Merlin83b (Merlin83b! has joined #mythtv
[08:06:55] jya: in libmyth/mythcontext.cpp in MythContextPrivate::FindDatabase, config.xml was properly loaded and as a valid database detected as defined in config.xml… Yet, it looks at the default backend set for mythfrontend and use UPnP to detect it… so that information is overridden and instead of starting my backend with the database specified; it tries to use the database of the backend found by upnp...
[08:07:11] jya: why would that code ever be used when we're launching a backend ???!!
[08:07:49] jya: looks like just deleting MythFrontend><DefaultBackend> from config.xml would do the tric...
[08:08:52] jya: danielk: that behaviour seems to have been introduced in your change in 77b68d51fec1
[08:09:07] jya: "Resend UPnP search request when waiting for one backend to respond."
[08:09:39] jya: except here we're starting a backend, why would you search if another backend is running.. same when starting mythtv-setup
[08:10:52] stuarta: fun fun fun
[09:18:31] doev (doev! has joined #mythtv
[09:25:23] jya: yep… that did it… deleting the default backend set for the frontend allow the backend to start.
[09:25:56] jya: my guess is that default backend was added when I was investigating paulh bug when starting mythfrontend with -p
[09:36:51] stuarta: at least you found it
[09:40:47] jya: sure did… I reckon we should lift user requirements so that it is expected for user to be able to read the C++ code to understand why their myth install doesn't work as expected.
[09:41:12] warped_ (warped_! has joined #mythtv
[09:42:27] warped_ (warped_! has quit (Client Quit)
[09:50:27] warpme (warpme! has joined #mythtv
[09:53:31] dekarl1 (dekarl1! has joined #mythtv
[09:55:07] dekarl (dekarl! has quit (Ping timeout: 248 seconds)
[09:55:16] jya: wagnerrp: looking into bug #11878… Previously, the setting for the auto metadata lookup was per recording rule… now it's not checked at all and is global.
[09:55:35] jya: even if I fix that, I still think the behaviour of setting it per recording rule is broken
[09:55:52] jya: no more auto-link to trac?
[09:56:59] doev (doev! has quit (Quit: Verlassend)
[09:57:02] jya: the autometadata lookup was only defining the default value of what would otherwise be set when defining a recording rule. you could then change that value…
[09:57:32] jya: with the move to simply having a "task" running at regular interval, all this behaviour seems broken...
[09:57:34] jya: what do you think?
[09:58:50] jya: going offline to install 10.9, it's just reach GM..
[09:59:17] jya (jya!~jyavenard@mythtv/developer/jya) has quit (Quit: jya)
[10:03:35] FabriceMG (FabriceMG!~Thunderbi@ has quit (Quit: FabriceMG)
[10:04:55] Merlin83b: Now there's an early adopter!
[10:17:10] gregL (gregL! has quit (Ping timeout: 240 seconds)
[10:17:54] jya (jya!~jyavenard@mythtv/developer/jya) has joined #mythtv
[10:21:59] stuartm: #11878
[10:21:59] ** MythLogBot **
[10:23:15] stuartm: jya: when iamlindoro added that stuff, I complained that it was too difficult to disable and it is, at that time there were at least three different settings that needed to be flipped to turn it off completely
[10:23:53] stuartm: one in mythtv-setup on the housekeeper screen, two somewhere in the frontend
[10:24:03] gregL (gregL! has joined #mythtv
[10:25:01] stuartm: it should be just the one setting master setting, with the option to disable (not enable) per-recording rule
[11:15:52] ghoti_ (ghoti_! has quit (Quit: leaving)
[11:16:26] ghoti_ (ghoti_! has joined #mythtv
[11:17:07] ghoti_ (ghoti_! has quit (Client Quit)
[11:20:44] jya_ (jya_!~jyavenard@mythtv/developer/jya) has joined #mythtv
[11:32:51] jya: ok... so if I simply test where the mythmetadatalookup is started by the backend the value of the autometadadownload that will be good enough?
[11:33:35] jya: i don't believe disabling it in the recording rule will do anything... the setting stored per recording rule isn't tested anywhere anymore
[11:39:51] dblain: jya: sent while you were offline...
[11:40:02] dblain: [06:21] <stuartm> jya: when iamlindoro added that stuff, I complained that it was too difficult to disable and it is, at that time there were at least three different settings that needed to be flipped to turn it off completely
[11:40:02] dblain: [06:22] <stuartm> one in mythtv-setup on the housekeeper screen, two somewhere in the frontend
[11:40:02] dblain: [06:23] <stuartm> it should be just the one setting master setting, with the option to disable (not enable) per-recording rule
[11:40:16] jya_: dblain: yes, I read that ... that was my answer to that post
[11:40:29] dblain: oh
[11:40:40] dblain: sorry for the noise
[11:40:43] jya_: my laptop was turned on...
[11:40:49] jya_: no worries.
[11:41:03] jya_: should have prefixed my message with stuartm
[11:43:00] dblain: btw: Rethinking my position on the external libraries for windows... As much as I don't like making it more complicated to setup a build environment, it does make sense to move that source out of the main repo.
[11:44:17] skd5aner (skd5aner! has quit (Ping timeout: 256 seconds)
[11:45:02] skd5aner (skd5aner! has joined #mythtv
[11:50:36] jya_: great...
[11:50:59] jya_: I'm happy for it to be either way... just that it is consistent with all the other platforms
[11:51:16] jya_: I do would like a proper build project in the main repo
[11:51:27] jya_: something I can just open in an IDE, build and debug
[11:51:54] jya_: so for VS2010, that would be a .vsproj file, for mac a xcode project etc...
[11:52:18] jya_: on xbmc, for building the dependencies, before you run configure, you call makedepend
[11:52:34] rkulagow (rkulagow!45f6d03f@gateway/web/freenode/ip. has quit (Quit: Page closed)
[11:52:39] jya_: once that's done, you use configure and then can do whatever in the IDE of your choice
[11:52:41] jya (jya!~jyavenard@mythtv/developer/jya) has quit (Quit: jya)
[11:52:42] jya_ is now known as jya
[12:04:45] warpme (warpme! has quit (Quit: warpme)
[12:11:37] stoffel_ (stoffel_! has joined #mythtv
[12:12:06] stoffel (stoffel! has quit (Ping timeout: 245 seconds)
[12:31:20] jya_ (jya_!~jyavenard@mythtv/developer/jya) has joined #mythtv
[12:49:04] warpme (warpme! has joined #mythtv
[12:50:45] purserj (purserj! has quit (Ping timeout: 252 seconds)
[12:52:46] purserj (purserj! has joined #mythtv
[12:57:44] purserj (purserj! has quit (Remote host closed the connection)
[12:58:00] purserj (purserj! has joined #mythtv
[13:00:14] jya_ (jya_!~jyavenard@mythtv/developer/jya) has quit (Quit: jya_)
[13:00:46] amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has joined #mythtv
[13:19:40] purserj (purserj! has quit (Ping timeout: 246 seconds)
[13:19:41] jya (jya!~jyavenard@mythtv/developer/jya) has quit (Quit: jya)
[13:20:06] jya (jya!~jyavenard@mythtv/developer/jya) has joined #mythtv
[13:27:29] warpme (warpme! has quit (Quit: warpme)
[13:29:54] Jordack (Jordack! has joined #mythtv
[13:30:23] stoffel_ (stoffel_! has quit (Ping timeout: 260 seconds)
[13:46:58] purserj (purserj! has joined #mythtv
[13:47:42] dblain: jya_: The approach I took was to have configure.ps1 run qmake to either generate a visual studio solution of a makefile (user choice). I didn't want to commit the solution since it would need to be maintained. I rather have qmake generate it based on the pro files.
[13:49:21] dblain: jya: I also agree that the state of our pro files is not good. I wanted to clean them up, but ended up isolating the win32 bits because I wasn't sure of the impact on all the various linux systems.
[14:28:14] amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has quit (Quit: Konversation terminated!)
[14:41:01] amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has joined #mythtv
[14:55:36] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 252 seconds)
[14:57:26] stuartm: gigem: I'm thinking about moving most of the shutdown related stuff out of the scheduler and into a new shutdown utility class, the reason is that various different bits of shutdown/wakeup logic exist, each of which work differently and which are incompatible
[14:58:14] stuartm: for example, if you shutdown a combined fe/master be from the frontend, it won't set a wakeup time
[14:59:35] stuartm: and the SHUTDOWN_ related protocol commands again are entirely unrelated to the scheduler shutdown, not only do they not set a wakeup time, but they don't trigger other shutdown related events to occur
[15:01:02] stuartm: and then you've got mythshutdown (and mythwelcome) which do their own thing entirely, in short it's a complete mess of different implementations written at different times which all exist side by side but with no shared code
[15:03:51] warped_ (warped_! has joined #mythtv
[15:14:39] stuarta: stuartm: sounds good to me
[15:16:29] stuartm: the only downside is that it's a distraction which I really don't need right now :)
[15:20:13] stuarta: haha
[15:20:29] ** stuarta can relate to this. too much todo and not enough time **
[15:24:13] dekarl1 is now known as dekarl
[15:37:40] xibalba (xibalba! has left #mythtv ("Leaving")
[15:50:39] warped_ (warped_! has quit (Quit: warped_)
[16:05:32] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[16:10:38] stuartm:
[16:11:39] stuartm: seems to be the re-occuring flaw with github, you put all your eggs in the same basket as a whole lot of other people, making one huge target for a bored script kiddie
[16:14:28] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 240 seconds)
[16:15:08] jheizer: Placing an open source project there is one thing but hosting your company's code? I can't even imagine trusting that to an outside service.
[16:17:20] stuartm: and that's a huge problem for them, they give free hosting to open source projects in order to get exposure (most open source devs have a day job), but they need the commercial clients in order to pay the bills (nevermind making a profit)
[16:17:49] jheizer: Yuup
[16:18:05] stuartm: incidents like this, and it's just one of many that have happened, will bankrupt them
[16:19:29] jheizer: The idea that they are profitable is just beyond me. I guess I looks at it that is you are a developer and can't set up your own gti/svn/tfs/whatever you must not be very good...
[16:19:48] jheizer: assuming you can type git correctly...
[16:22:25] stuartm: I don't understand how many modern web businesses make a profit, unless they are selling physical products, and the truth of course is that they aren't profitable – they are kept afloat by venture capitalists who are either very stupid or very clever
[16:23:37] stuartm: the very clever ones know that if they build up the apparent value of a company, they'll be able to sell it on at a profit to the stupid investors later on
[16:23:59] jheizer: Hahah I agree 1000%. Some how now a days jsut a good idea is enough to some how have a company even if there is no way to make $ in the near future.
[16:24:01] stuartm: but there's no source of income, no assets of worth
[16:24:17] jheizer: :cough: twitter :cough:
[16:24:29] jheizer: Though, as a non twitter user maybe there is a way now?
[16:25:06] warpme (warpme! has joined #mythtv
[16:25:33] stuartm: people lose their minds when a company like twitter goes public, buying thousands of worthless shares
[16:25:35] Merlin83b: There is – you can pay them to advertise at their users.
[16:26:17] stuartm: are bankers and traders more cautious, less cavalier since the financial meltdown as we're meant to believe, certainly doesn't seem that way to me
[16:27:37] jheizer: Makes you think back to all the "stupid" ideas you had over the years that you thought were worthless.
[16:27:52] Merlin83b: Traders and bankers don't have to care about what a company does, they only need take interest in the markets. They do not live on dividends, they live on tiny movements in share price.
[16:27:54] stuartm: Merlin83b: and then you'll lose the users – as other services have found, but it's worse for twitter, the adverts won't be passive, they'll be broadcast to people and they will really hate that
[16:28:18] Merlin83b: They have inserted them passively, it looks like just another tweet in your feed.
[16:29:16] stuartm: Merlin83b: but at the end of the say, someone is left holding the worthless stock, like a twisted game of pass the parcel where the one holding the parcel loses everything
[16:30:16] Merlin83b: Someone holds the stock for sure, but it's only worthless if someone will no longer pay for it. And traders can even make profits on a falling value share.
[16:30:34] Merlin83b: Financial markets and the way they are currently used are frightening.
[16:30:38] Merlin83b: Money from nothing.
[16:31:46] stuartm: technically, at a smaller scale, selling worthless stock by inflating it's true value is fraud – a fraud that everyone who trades in that stock is implicated in, but which probably won't ever be punished
[16:33:05] dekarl: stuartm: if we can agree on a way to have better inetrefs in the recordings <- recording rules <- guide <- xmltv then I can switch two or three countries over to supplying the data. If I get some excess spare time maybe even the UK ;)
[16:34:02] stuartm: Merlin83b: there's a terrifying idea there, that these companies will continue to flourish for years without making an income because investors are too scared to write off the stock as junk, they'll keep passing it around and everyone will maintain the pretence because the implications should it fail would be significant
[16:34:16] Merlin83b: Quite so.
[16:35:47] jheizer: Till finally MySpace sells the last time for $34.54
[16:35:49] stuartm: dekarl: how efficient is that? Are the matches done at runtime, and do they get cached?
[16:36:49] dekarl: stuartm: the matches are done upstream at the guide provider
[16:36:53] stuartm: the ultimate evolution of the age-old pyramid schemes
[16:37:13] stuartm: dekarl: oh really? that's interesting!
[16:37:43] stuartm: well that would definitely be worth doing
[16:38:08] dekarl: e.g. . . . e_2013-10-04 at the end there's two movies and over the day some episodes
[16:39:50] dekarl: currently the data just gets put into the generic url / web site attribute as I don't know whats preferred downstream. But that could be as simple as two new episode-num systems e.g. <episode-num system="">18274</episode-num>
[16:40:18] stuartm: dekarl: in reference to your comment to the ticket, the problem with storing urls is that they can be fragile, another 'redesign' of the site or API and suddenly they don't work
[16:41:06] stuartm: from a DBAs point of view, it's also not an efficient use of space
[16:41:36] dekarl: metabroadcast has signalled that they may be willing to host a new xmltv feed if the community maintains it and they appear to be interested in tapping into thetvdb for pictures, so it all comes down to someone writing a java exporter that pushes xmltv to amazon s3 storage...
[16:41:55] stuartm: personally I like the prefix idea, it's incredibly simple, portable and doesn't require large amounts of code to be re-written
[16:42:18] dekarl: I looked at the api here to move my site to it once traffic makes it a priority:
[16:43:05] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has joined #mythtv
[16:45:31] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[16:49:24] dekarl: stuartm: just adding a prefix sound like a solution thats good enough for now.
[16:49:56] dekarl: thinking about it, a way to signal "generic episode" is missing from xmltv currently IIRC
[16:53:58] Merlin83b (Merlin83b! has quit (Quit: Leaving)
[17:01:40] SteveGoodey (SteveGoodey! has joined #mythtv
[17:01:50] warpme (warpme! has quit (Quit: warpme)
[17:03:21] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 245 seconds)
[17:06:04] stichnot (stichnot!~stichnot@ has joined #mythtv
[17:06:05] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[17:06:05] stichnot (stichnot!~stichnot@ has quit (Changing host)
[17:20:15] warped (warped! has joined #mythtv
[17:29:33] NightMonkey (NightMonkey! has joined #mythtv
[17:29:33] NightMonkey (NightMonkey! has quit (Changing host)
[17:29:34] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv
[17:32:25] doev (doev! has joined #mythtv
[17:41:44] stoffel (stoffel! has joined #mythtv
[17:47:36] SteveGoodey (SteveGoodey! has quit (Quit: Konversation terminated!)
[18:02:15] shodan45 (shodan45! has quit (Quit: Konversation terminated!)
[18:21:30] warpme (warpme! has joined #mythtv
[18:23:02] stoffel (stoffel! has quit (Ping timeout: 264 seconds)
[18:34:04] stichnot (stichnot!~stichnot@ has joined #mythtv
[18:34:04] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[18:34:04] stichnot (stichnot!~stichnot@ has quit (Changing host)
[18:35:02] doev (doev! has quit (Ping timeout: 264 seconds)
[18:48:55] gigem: stuartm: Your shutdown plan is more than fine with me. In fact, you're welcome to move the slave wake/sleep logic out of the scheduler too.
[18:50:26] stuartm: gigem: great, thanks
[18:51:17] stuartm: might hold off working on it until a little later in the cycle, but it's definitely on my todo list now
[18:54:05] warpme (warpme! has quit (Quit: warpme)
[18:58:39] fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has quit (Remote host closed the connection)
[19:04:21] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has quit (Ping timeout: 245 seconds)
[19:04:44] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Read error: Operation timed out)
[19:17:27] natanojl (natanojl! has joined #mythtv
[19:17:29] natanojl (natanojl! has quit (Signing in (natanojl))
[19:17:29] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has joined #mythtv
[19:19:34] cesman (cesman!~cesman@pdpc/supporter/professional/cesman) has quit (Quit: Konversation terminated!)
[19:23:11] NightMonkey (NightMonkey! has joined #mythtv
[19:23:11] NightMonkey (NightMonkey! has quit (Changing host)
[19:23:11] NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv
[19:25:11] cesman (cesman! has joined #mythtv
[19:50:58] warped: Hi, I'm trying to find root cause for #11882. By comparing DB for good & bad channels I found that problem is in lack of type=2 markup records in recordedmarkup table. So it looks like change in DVB stream in bad channels causes that recorder is not producing type2 markups. Interesting is that only myth have problem with those channels. My SAT provider has 2.5M subs and their change in BBCHD from week ago don't have and comments on multiple
[19:50:58] ** MythLogBot **
[19:50:58] warped: forums...
[20:00:29] fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has joined #mythtv
[20:15:47] stichnot: warped: type=2 should be MARK_BOOKMARK, which are added by the player, not the recorder.
[20:16:45] stuartm: warped: yeah, that's a red-herring, as stichnot notes those are just user bookmarks
[20:19:04] warped: Hmm so this becomes really interesting: I have repeatable behaviour where first play always gives black screen+exit prompt, but second+ is OK. So state is somewhere remembered. I look on DB – only difference is type2 in recordermarkup….
[20:19:41] clever: warped: does the problem return after restarting the frontend?
[20:21:03] warped: Of course another place for state remembering is HW + SW state. But if I do test: play firs time=black screen; play 2nd time=OK; play third time=OK; delete type2 in DB; play=black screen+exit.
[20:22:20] stuartm: warped: do you have 'Wait for keyframe' enabled in the card settings?
[20:23:23] stichnot: warped: also, do you have any post-record job that might be adding a bookmark?
[20:24:38] stichnot: warped: you may have the setting enabled that clears the bookmark when playback starts, so the second playback doesn't try seeking to the bookmark
[20:24:41] warped: clever: yes. I do restart BE, FE. Problem is still present. Most intriguing for me is that problem arrive week ago. Before week ago I had PERFECTLY working system during last months. There wasn't any change on my side – so delta arrives from air (SAT provider). Issue is present on ALL channels on given molex. In other words: there was kind of change on TS level.
[20:25:16] warped: stuartm: yes
[20:25:29] clever: if restarting the frontend makes the problem happen again, then its unlikely to be database
[20:25:36] clever: its likely to be a global variable inside the frontend
[20:25:42] clever: or some state in a lib its using
[20:26:16] warped: stichnot: regarding post-record: no. No any jobs.
[20:27:48] warped: stichnot: I have to check it. But week ago (when issue pop-ups) I didn't do any changes in 3 home FE + 4th on OSX. Currently 30% my paid channels are dead...
[20:28:06] stuartm: stichnot: I'm thinking it's the reverse, first attempt actually sets a bookmark at a point somewhere beyond the start of the recording, this allows the player to skip the problematic frames at the start of the recording so playback succeeds
[20:29:41] warped: stuartm: this is good hipothessis. To test I can modify type2 to higher value and see is this helps. High value should prove concept...
[20:30:23] stuartm: I'd be willing to bet that if you cut the first few seconds from the file that it would playback just fine, something – whether it's corrupted data due to a driver bug or just intra-keyframe frames that the driver is mishandling
[20:31:07] SteveGoodey (SteveGoodey! has joined #mythtv
[20:32:09] stuartm: warped: might be worth disabling the 'wait for keyframe' setting, it's just possible that we're mis-identifying the keyframes, it's got to be worth a try at least – you'll need to record something new after changing the setting
[20:33:10] warped: stuartm: regarding driver bug: I have 21 mplexes. 18 of them with 80+channels are working perfect as before. Just 3 of them week ago stopped to work. Before week ago all was perfect. I don't believe it has anything to do with driver….
[20:33:19] stichnot: stuartm: yeah, that sounds plausible, but I can't remember the sequence of when the bookmark gets cleared
[20:35:50] stichnot: warped: it could be that a change to the broadcaster's encoder exposed a preexisting mythtv bug. This happened to me not too long ago – one broadcaster changed the way they encoded closed captions, such that it was still within the spec, but it broke something in mythtv.
[20:37:07] warped: stuartm: within few minutes I'll do tests with "disabled_wait_for_keyframe". Just give me few sec.
[20:37:35] warped: stichnot: this is EXACTLY my hipothessis!
[20:38:20] stuartm: hypothesis
[20:42:12] warped: stuartm: regarding bookmark threshold: just checked it: there is threshold value. below it playback fails. Above is OK. So indeed it looks like beg. of recording has issue with long lack of I-frames (maybe)
[20:48:37] aarcane (aarcane! has quit (Read error: Connection reset by peer)
[20:49:02] aarcane (aarcane! has joined #mythtv
[20:49:13] warped: stuartm: I just finished test with "dvb_wait_for_seqstart=0". No change: player shows "layer(1): Waited 102ms for video buffers AAAAAAAAAAAAAA". It looks like NC+ (my SAT provider) changed TS structure in a way that myth producing recordings which have king of "garbage" at beginning causing mythplayer (and LiveTV) to hang….
[20:49:46] warped: s/king/kind/
[20:50:12] stuartm: jpabq: you're our resident expert on H.264 keyframes, any thoughts?
[21:19:58] warpme (warpme! has joined #mythtv
[21:26:56] warped: I believe root cause might be: 1\recorder wrongly misses very first I-frames so resulting recording file has them at long places from beg. (too long for decoder) or 2\mythtv decoder has way too small tolerance for distance to first I-frame. I think 1\ is much more probable…
[21:42:58] SteveGoodey (SteveGoodey! has quit (Ping timeout: 240 seconds)
[21:54:03] stuartm: heh, so mythweb in 0.27 isn't using my preferred time format, it's correct in the settings but it's insisting on displaying everything as AM/PM instead of the 24hr clock
[21:57:19] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has quit (Ping timeout: 246 seconds)
[22:07:37] Jordack (Jordack! has quit ()
[22:10:36] jpabq: warped: if you look at the recordedseek table, what is the first entry for type=9 ? Is the offset not equal to zero?
[22:13:58] warped: jpabq: for type=9 I have entry with offset=0
[22:14:49] jpabq: That means that it thinks the first thing in the file, is a keyframe.
[22:15:12] warped: I have 204 entries of type=9
[22:17:05] warped: ok, then we can look into file. I can provide it to You via HTTP
[22:21:42] jpabq: Do you have an entries with type 6 or 7?
[22:22:27] jpabq: How long is this recording? 204 of type 9 is way to few — indicating it is not finding all of them, only some.
[22:23:19] warped: I have only 9 & 33
[22:24:08] warped: link to file is
[22:25:04] warped: recording is about few min. file has 240MB
[22:26:00] jpabq: Oh, OK. Then 204 keyframes might be about right.
[22:29:00] jpabq: Looks like it is going to take an hour to download that.
[22:29:48] warped: Oh sorry – give me sec. I'll halt other things :-)
[22:31:11] jpabq: Now it says 30 minutes.
[22:32:43] warped: :)
[22:33:42] warped: You know – this East in Europe. 3/30Mb is mid standard in HSIA
[22:39:41] warpme (warpme! has quit (Quit: warpme)
[22:41:26] jpabq: It is done.
[22:41:55] warped: perfect so I can resume my DW things :-)
[22:44:10] stuartm: 3/30 is actually pretty good, much better than the UK average, better even than I currently get (but exactly what I should be getting – bastards want to charge me for a new Docsis 3 compatible modem)
[22:46:50] warped: Heh – cable thing. Compared to xDSL – cable (where I am) looks good. Especially with DOCSIS 3.0 which gives OFDM in 500MHz band compared to annexD in DSL....
[22:47:41] warped: anyway – pleasure to talk with You guys even on 3/30 :-p
[22:54:45] warped: stuartm: go with DOCSIS – they want to prepare ground – so if with You will be clever enough and use power of loyalty programs – they will be "forced" to double DW speed for You annually :-). This is at leas picture in Poland in my case.
[22:57:19] stuartm: I'm only pissed because as an existing customer for over a decade they want to charge me for the hardware, but they give it away free to new customers, they've got a strange idea of rewarding loyalty
[22:59:56] jpabq: warped: It is not clear to me if this is a recording issue, or a playback. I am guessing playback, since mplayer has no problem with the file. I am good with the recording section of myth, but no so good with the playback. Every time I look at ffmpeg source code, I want to run away screaming.
[23:02:20] warpme (warpme! has joined #mythtv
[23:02:22] warped: Right. I exactly feel Your pain. Usually it is abut sending them service cancel letter. Not a call – but letter. They have quite aggressive retention programs (as customer lifetime value is critical for their top-line) so (if company has resumable marketing del) – they should return to You with proposal at least not worse that they proposals in acquiring programs (I hope). This is quite standard strategy I know from meeting major carriers her
[23:02:22] warped: Poland.
[23:04:37] warpme (warpme! has quit (Client Quit)
[23:07:13] warped: jpabq: right. Indeed, my second hypothesis was player as mplyer/VLC/OSX played recordings file OK. Fortunately capture sample which at end of day should simply start under myth player without any delays/black screens
[23:07:59] warped: s/capture/we have capture/
[23:19:13] warped: jpabaq: Heh – myth is so powerfull. There is always price such advancement. It is complication. All of You are doing is incredible job keeping this rolling. MarkK decided for scratch….. He can provide advantage from learning from lessons learnt – but price is time and adoption: who knows torc and what is it's advantage over XBMC…oh sorry – it is off topic. Forgive me.
[23:31:51] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 245 seconds)
[23:37:31] stichnot (stichnot!~stichnot@ has joined #mythtv
[23:37:31] stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv
[23:37:31] stichnot (stichnot!~stichnot@ has quit (Changing host)

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