:: #mythtv

Daily chat history

Current users (82):

aloril, Anduin, anykey_, Beirdo, ben1066, brfransen, brtb, CaCtus491, Captain_Murdoch, cattelan, Chutt, clever, coling, Cougar, damaltor, danielk22, Dave123, dekarl, dlblog, ElmerFudd, ertyu-m, foobum, foxbuntu, frankster, ghoti, gigem, gregL, GreyFoxx, Guest97643, highzeth, J-e-f-f-A, j-rod|afk, jams, jarle, jcarlos, joe_, jpabq-, jstenback, justinh, jwhite, k-man, knightr, kormoc, kurre2, kwmonroe, laga, mag0o, MavT, mrand, MythBuild, MythLogBot, mzanetti, natanojl, peitolm, pheld, poptix, purserj, rhpot1991, skd5aner, Slasher`, sphery, sraue, stuarta, stuartm, superm1, sutula, taylorr, tgm4883, The_Ball, ThisNewGuy, tomimo, tris, Unhelpful, vallor, wagnerrp, wahrhaft, wseltzer, xavierh, xris, yb0t, zCougar, _charly_
Friday, March 16th, 2012, 00:07 UTC
[00:07:48] stichnot (stichnot!~chatzilla@mythtv/developer/stichnot) has quit (Ping timeout: 255 seconds)
[00:33:49] zombor_ (zombor_!~zombor_@kohana/developer/zombor) has joined #mythtv
[00:44:06] knightr: stuartm, Hi!. Would you see any problem with me adding Chinese date/time formats in globalsettings.cpp? (I would get them reviewed of course) I have a (working) patch that let translators add new date formats at the translation level but the more I think about it the more I think language files might not be the best place for that (locale files might actually be better if their functionality was expanded). Since we`re in feature freeze and soon
[00:44:06] knightr: string freeze it would probably be best to add the date formats to globalsettings.cpp and try to address the issue once 0.25 is released. Would that be ok with you? Thank you!
[00:50:40] knightr: sphery, don't worry about it, we're not in string freeze yet...
[00:51:18] knightr: (hard freeze..)
[00:51:58] sphery: yeah, but it's still unfortunate that it went in after many had already gotten mostly up-to-date.
[00:53:17] knightr: it's not a few strings that are going to kill us, it's many at the last minute...
[00:54:30] knightr: I'll have a patch that will add more tr()s that you added with that commit...
[00:54:45] knightr: s/that/than what
[00:56:46] knightr: (and we wouldn't die anyway, it's just that we promised them the strings would stop changing at the hard freeze...)
[01:07:48] skd5aner: danielk22, sphery: I haven't yet upgraded, but I'm probably willing to do so – but I've got company staying at my house for a week, so won't be able to until sometime next week :)
[01:07:52] sphery: hehe, well, I suppose that since there were a couple others today, it's not necessarily a /lot/ more work for them
[01:08:17] sphery: skd5aner: ah, ok... thought you had just upgraded for testing, but I remember your having said that, now
[01:08:29] skd5aner: danielk22: any thoughts on #9868? That's the one I thought sounded a little like the segfault you fixed a couple of weeks ago (sorry, forget the ticket #)
[01:15:30] mrand (mrand!~mrand@ubuntu/member/mrand) has quit (Ping timeout: 252 seconds)
[01:16:56] stichnot (stichnot!chatzilla@nat/intel/x-gbljxvdoyxjeouie) has joined #mythtv
[01:16:57] stichnot (stichnot!chatzilla@mythtv/developer/stichnot) has joined #mythtv
[01:16:57] stichnot (stichnot!chatzilla@nat/intel/x-gbljxvdoyxjeouie) has quit (Changing host)
[01:21:31] danielk22: skd5aner: I believe #9868 is fixed. I can't find the commit in the commits e-mail list though. It may be a case of the commit hook not firing and closing the ticket, or I may have typoed the commit number, or (hopefully not) I may have forgotten to push the commit.
[01:22:19] skd5aner: danielk22: heh – ok... I'm hoping to have some time to go through the last week's worth of commits for the release notes – I'll keep an eye out
[01:22:52] skd5aner: as always, my family has managed to segfault myth 3 times already since being here...
[01:23:23] danielk22: skd5aner: Make sure ulimit -c unlimited has been set :)
[01:24:38] mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv
[01:30:17] OldEnK (OldEnK! has joined #mythtv
[01:36:47] stichnot: skd5aner: I assume they're trying to watch Live TV? :)
[01:37:11] skd5aner: stichnot: one of them at least was related to live tv, and one was related to watching a dvd iso
[01:37:16] skd5aner: don't recall the third
[01:38:17] skd5aner: Oh, I remember... it wasn't a segfault, but it's a pretty nasty bug in mythgallery that basically rendors video output completely borked in mfe when trying to playback video clips
[01:38:53] skd5aner: I've mentioned it before, and I don't recall if there is/was a ticket for it, but at this point I'm not going to submit one until I see if the issue exists in 0.25 too
[01:39:14] skd5aner: but, if it does exist in 0.25, it should definitely be looked at
[01:43:21] danielk22: skd5aner: There isn't much time left to report bugs against 0.25 pre-release...
[01:45:12] OldEnK (OldEnK! has left #mythtv ("Leaving")
[01:54:49] jya: stichnot: I've changed MythDateTimeToString to return a constant string.
[01:55:19] jya: 1.4s without your patch
[01:55:22] jya: from 6.9s
[01:55:33] jya: so the culprit is definitely that function
[01:56:44] jya: the data format is "Fri Mar 16 2012" showing in the settings
[02:00:40] jya: and short date is MM/DD
[02:01:55] jya: I've changed the settings for date to be DD/MM/YY and short one to "DD/MM", it's now back to 1.7s
[02:04:14] jya: When I change the date (long, not short) back to Friday, March 16 2012. it jumps again back to 6.7s… Which is bizarre beause looking at the recording screen, I don't see the long date format anywhere… So why would it matter? or is that theme dependent and as such is always calculated?
[02:15:29] jya_ (jya_!~jyavenard@mythtv/developer/jya) has quit (Quit: jya_)
[02:15:30] stichnot: jya: ProgramInfo::ToMap computes all possible strings, whether the theme uses them or not
[02:17:12] skd5aner: danielk22: I could submit it against recent 0.24-fixes. My guess is it's still a problem in 0.25, but I could definitely be wrong
[02:17:27] stichnot: so the whole point of that patch is to defer each item's ToMap() until it is first displayed on a page of recordings
[02:17:48] jya: stichnot: yes, I got that.
[02:18:00] skd5aner: anyone here have mythgallery installed and want to try and put a video clip (in my case, I was testing a .mov) in one of the direrctoies and see what happens when you try to play it?
[02:25:55] knightr: jya, I didn't know 2012 had so many months :-) (there's a typo in your post...)
[02:26:08] jya: no there isn't
[02:26:31] jya: ah sorry 030
[02:26:41] jya: :)
[02:27:40] knightr: I'm sure people understood what you mean anyway though...
[02:28:52] knightr: s/mean/meant
[02:28:58] stichnot: jya: that's wild. only 5 of the ProgramInfo::ToMap() strings use kDateFull. Something is seriously wrong with Qt there...
[02:37:14] sphery: stichnot / jya : isn't my ddd MMM d yyyy equivalent to ""Fri Mar 16 2012"
[02:37:25] sphery: if so, that's really weird since mine is fast with that
[02:38:05] jya: sphery: probably just the mac implementation of it
[02:38:29] sphery: I'm still thinking it's that crazy "spin around looking for /etc/localtime stuff" thing it was doing for some users before your MythSignalingTimer implementation
[02:39:30] sphery: jya: what happens if you put it to the long one and then start mythfrontend with: TZ='Whatever/New_Zealand' mythfrontend
[02:39:49] sphery: (but with proper time zone identifier)
[02:44:01] stichnot: sphery: that's definitely worth checking on, I'll have to dig up notes on how I found that
[02:52:32] XChatMav (XChatMav! has joined #mythtv
[02:53:41] mrand (mrand!~mrand@ubuntu/member/mrand) has quit (Ping timeout: 248 seconds)
[02:55:21] MavT (MavT! has quit (Ping timeout: 252 seconds)
[02:56:29] mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv
[02:56:50] mrand (mrand!~mrand@ubuntu/member/mrand) has left #mythtv ()
[03:11:30] mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv
[03:21:31] stichnot: sphery: I recall strace showed tons of stats of /etc/localtime (over NFS in my case), 70 per second, corresponding to the Pulse() rate, and setting TZ fixed this.
[03:23:19] sphery: yeah, I'm figuring if it's doing the same "check /etc/localtime each call" that would be a lot of file system access for 2000 (recordings) * 5 (strings per recording) in a short time.
[03:25:24] stichnot: also, to give proper credit/blame, MythSignalingTimer was danielk22's implementation of my "proof of concept"
[03:26:36] stichnot: 500 microseconds per stat? Plausible I guess.
[03:30:11] jya: sphery: I rpelied on the mailing list.. Starting with TZ="Australia/Melbourne" made no difference whatsoever
[03:31:31] sphery: hmmm, ok
[03:31:59] sphery: wonder why it's so slow for some, then
[03:32:58] Beirdo: hmm
[03:33:23] Beirdo: my frontend went into standby. I thought I set it to 0, disabling that
[03:33:41] jya: sphery: remember, what I'm seeing is on the mac..
[03:34:21] Beirdo: I'll chase it later
[03:42:53] zombor_ (zombor_!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[04:56:38] knightr: The main_settings.xml files still contains an entry that says "Pick the menu layout, font, size, and locale", shouldn't we remove font and size from there?
[04:59:33] ThisNewGuy1 (ThisNewGuy1! has quit (Read error: Connection reset by peer)
[05:11:55] ThisNewGuy (ThisNewGuy! has joined #mythtv
[05:22:16] sphery: knightr: ah, in the menu themes... yeah, definitely needs changing (though maybe it means window/screen size, but now that we have Screen Wizard, I suppose we can ignore that part)
[06:31:39] jya: jpabq-: you're the h264 expert I believe. Do we have a function that parse h264 sps and return the number of reference frames, if it's interlaced or not etc?
[06:33:36] jya: sphery: what a weird commit you've done
[06:34:23] jya: default to a setting that we fundamentally want to remove later, that you know is broken for most of the recent theme, just for the sake of making them change it
[06:34:46] sphery: well, was forced into it by broken GL implementations
[06:35:03] sphery: we can't default to auto, because too many implementations say they will work, but don't
[06:35:13] jya: so we pretty much borke it for everyone else because of ATI broken opengl
[06:35:17] sphery: we can't default to opengl (and, if we could, we should use auto)
[06:35:23] sphery: which means we have to use qt
[06:35:29] sphery: and, yeah, that's pretty much it
[06:36:07] sphery: users with AMD/ATI Catalyst drivers who have OpenGL painter get black screen on exiting playback
[06:36:19] sphery: and many users have issues even starting
[06:36:23] jya: that sounds like a not very thought change IMHO… I'm in favour of making it work out of the box for everyone but ATI users… up to ATI to fix their drivers, not us to break it for everyone else
[06:36:25] sphery: so, I didn't see much choice
[06:37:39] jya: and your commit text doesn't make it that appealing either, "unless they happen to notice the setting buried in frontend Setup, and
[06:37:39] jya: they happen to read the help text which says that the default value is
[06:37:40] jya: only for use on systems with broken OpenGL"
[06:37:44] sphery: well, I'm open to suggestions from others, but let's just say that I'm tired from swimming upstream for a year trying to make it do the right thing
[06:38:06] jya: well, i don't see how that should be your concern....
[06:38:30] sphery: anyway, at least for 0.25, we're better off leaving it like it was for 0.24 and below
[06:38:37] sphery: users from 0.24 will pick up their old setting
[06:38:42] sphery: which, presumably, worked
[06:39:21] jya: sure, but I still think this is a step backward in our ultimate goal
[06:39:23] sphery: and users who have it set to opengl will work--but if they use ssh forwarding for mythtv-setup or something, won't get auto-fallback to Qt, so will see a slow UI
[06:39:43] jya: why should we support users ssh-ing ?
[06:39:52] sphery: agreed, but I don't see any possible way to both "do the right thing" and actually work on all the broken setups users have :)
[06:40:01] jya: I know it affects me, because that's how I run mythtv-setup
[06:40:24] jya: but that doesn't mean it should be a supported way.. Nor that you should care about one user complaining that he has to install X libraries...
[06:40:29] stichnot (stichnot!chatzilla@mythtv/developer/stichnot) has quit (Ping timeout: 252 seconds)
[06:40:35] sphery: people are still buying AMD graphics based systems for MythTV with no intent of putting in discrete nvidia graphics and are being adamant that they shouldn't have to...
[06:41:03] jya: I think the message that ATI cards are shit under linux is pretty much something we've been saying lour and clear. it's the same mesage that comes back ever time
[06:41:16] jya: sure they shouldn't have to
[06:41:31] sphery: all I'm saying about the ssh forwarding is that users with opengl would be better off changing to auto so they get the "do the right thing" behavior
[06:41:40] jya: but that's up to ATI to fix… the more unhappy you make their user, the more likely ATI will finally fix it
[06:41:53] sphery: but those with broken systems need to use Qt in spite of their graphics card/driver saying they can do the job
[06:42:07] jya: that is make people buy another card until they fix teir drivers
[06:42:11] jya: sure
[06:42:25] jya: so those with broken system can go and change from OpenGL to Qt
[06:42:46] jya: but why make it more complicated for those that would have worked out of the box?
[06:43:07] sphery: well, I'm happy to put it back as it was--or to leave the setting and just change the default to auto--but I'll wait to see what other devs think
[06:43:08] jya: actually, selecting Qt as a painter is the least worry they are going to have with an ATI card anyway
[06:43:13] kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has quit (Quit: kormoc)
[06:43:17] jya: if that was the only setting required: then sure..
[06:43:32] jya: but it's not… even with Qt as painter it will crap out soon enough
[06:44:35] jya: so my theory is that, ATI drivers are broken… Even with Qt painter it will suck.. so we're better off making it working great for anyone not using borked drivers.
[06:45:03] jya: nvidia, intel: it will look nice...
[06:45:44] jya: if it doesn't work, I'm sure people will first think "@#$@#@% ATI card" then,oh my god, mythtv isn't pretty" :)
[06:46:18] jya: my only concern would be for new users starting mythtv-setup
[06:46:35] jya: but are new users ever going to run mythtv-setup over ssh ?
[06:46:52] jya: or would that only be existing, experienced user doing so?
[06:47:16] sphery: new users will definitely run mythtv-setup before mythfrontend on broken systems
[06:47:19] jya: (BTW, opengl painter works on my mac when I use OpenG:… not well, but it works and in the worse case it default back to Qt)
[06:47:26] Anduin (Anduin!~awithers@pdpc/supporter/professional/anduin) has quit (Remote host closed the connection)
[06:47:39] jya: when I use ssh I mean
[06:48:37] sphery: but I could put in an (ugly hack of an) exclusion for mythtv-setup so it always uses qt painter, and change the defaults to auto (and switch back to the not-the-same-as-0.24-fixes setting name so that existing users get the right default)
[06:49:06] sphery: at this point, I don't care, anymore, which way we go, but want consensus from the devs
[06:49:14] jya: i think until the settings are moved to mythui, using Qt or OpenGL in myhttv-setup makes no difference on how it looks
[06:49:41] jya: so using a default of Qt for mythtv-setup is fine by me
[06:50:47] stoffel (stoffel! has joined #mythtv
[06:53:16] sphery: at this point, though, it comes down to, do we want to use a setup that will almost definitely cause problems for users with ATI cards/drivers and users with broken installations of drivers (for nvidia or intel or ...) such that they get a failure and have to know to restart mythfrontend with -O ThemePainter=qt then go to setup and change the setting
[07:07:43] jya: and how would those users use XBMC or Torc then ? :)
[07:18:54] aloril (aloril! has quit (Ping timeout: 245 seconds)
[07:32:51] aloril (aloril! has joined #mythtv
[07:55:04] The_Ball_ (The_Ball_!~The_Ball@ has quit (Ping timeout: 260 seconds)
[08:01:53] The_Ball (The_Ball!~The_Ball@ has joined #mythtv
[08:19:08] gregL (gregL! has quit (Read error: Connection reset by peer)
[08:36:47] gregL (gregL! has joined #mythtv
[09:22:27] stuarta: danielk22: i've deleted all my watched dr who's which is the first couple of series of the new "reboot" as you call it
[10:05:02] mike|3 (mike|3! has quit (Remote host closed the connection)
[10:05:49] mike (mike! has joined #mythtv
[10:06:14] mike is now known as Guest97643
[10:28:28] xavierh: Where should I create a theme to see it in the menu ?
[10:28:54] xavierh: The wiki explain everything except this :)
[11:17:01] joki (joki! has quit (Ping timeout: 246 seconds)
[11:17:22] joki (joki! has joined #mythtv
[11:41:42] stuartm: to see it in the theme selection menu?
[11:42:40] stuartm: xavierh: ~/.mythtv/themes/ or /usr/{local/}share/mythtv/themes/
[11:42:57] stuartm: to make it available for download you need to talk to Captain_Murdoch
[11:57:34] Captain_Murdoch: xavierh, just send me a URL to a tarball or svn or git repo and I can make it available on the download site.
[12:01:39] seeker (seeker!~seeker@unaffiliated/seeker) has joined #mythtv
[12:08:53] xavierh: stuartm, Captain_Murdoch: thanks, it is just for testing, I want to understand what it cost to do a new complete theme and if thing could be improved somehow. I have it in /usr/local/share/mythtv/themes/ but it is a symbolic link, might be what the problem. I just create the themeinfo.xml. I assume it will use the default then, wouldn't it ?
[12:12:13] stichnot (stichnot!chatzilla@mythtv/developer/stichnot) has joined #mythtv
[12:16:17] stuartm: xavierh: at a minimum you need to copy over menu-ui.xml from another theme, the default theme doesn't provide one
[12:20:18] seeker (seeker!~seeker@unaffiliated/seeker) has quit (Quit: Colloquy for iPhone -
[12:22:43] rhpot1991 (rhpot1991!~rhpot1991@ubuntu/member/rhpot1991) has quit (Ping timeout: 260 seconds)
[12:22:48] rhpot1991 (rhpot1991!~rhpot1991@2001:4968:202:3:20f:eaff:fefc:ba0e) has joined #mythtv
[12:22:48] rhpot1991 (rhpot1991!~rhpot1991@2001:4968:202:3:20f:eaff:fefc:ba0e) has quit (Changing host)
[12:22:49] rhpot1991 (rhpot1991!~rhpot1991@ubuntu/member/rhpot1991) has joined #mythtv
[12:23:18] tgm4883 (tgm4883!~tgm4883@2001:4968:202:3:20f:eaff:fefc:ba0e) has quit (Ping timeout: 260 seconds)
[12:23:23] tgm4883 (tgm4883!~tgm4883@2001:4968:202:3:20f:eaff:fefc:ba0e) has joined #mythtv
[12:23:23] tgm4883 (tgm4883!~tgm4883@2001:4968:202:3:20f:eaff:fefc:ba0e) has quit (Changing host)
[12:23:24] tgm4883 (tgm4883!~tgm4883@ubuntu/member/tgm4883) has joined #mythtv
[12:31:57] danielk22: sphery: ATI drivers are awful, if the OpenGL detection works with Intel & nVidia cards then I think it is safe to leave in. I do think that while mythtv-setup remains it should use the Qt painter, there is really no benefit to it running the OpenGL painter and some pain.
[12:32:17] Anduin (Anduin!~awithers@pdpc/supporter/professional/anduin) has joined #mythtv
[13:07:42] zombor_ (zombor_!~zombor_@kohana/developer/zombor) has joined #mythtv
[13:16:53] jya: stuartm: I can now play your BBC HD video…
[13:30:18] stuartm: jya: if you'd like to try something a little more recent I can make it available
[13:30:58] stuartm: I'm told by someone with access to the spec (I assume) that the commentary/AD tracks aren't marked as such in the metadata available for blu-ray, which is a shame
[13:31:10] jya: this one didn't play because the h264 header wasn't something VDA understood (plus it was interlaced anyway). Both conditions I know test to default for ffmpeg
[13:32:02] jya: stuartm: I seem to recall that my bluray player marked some tracks with tags.. like visually impaired
[13:32:32] j-rod|afk is now known as j-rod
[13:32:40] stuartm: jya: the player, or the menus offered by the disc itself?
[13:32:48] jya: for the h264 file, if you have some progressive content, 720p or 1080p… even 576p.. that would allow me to test if decoding works well
[13:33:09] jya: and some interlaced ones to test that my interlacing detection works well
[13:35:31] jya: I'll need to check for the BD player
[13:35:31] j-rod is now known as j-rod|afk
[13:35:39] Jordack (Jordack! has joined #mythtv
[13:35:41] jya: I haven't used it in a while, I play everything via myth these days
[13:35:56] j-rod|afk is now known as j-rod
[13:36:02] jya: in any case, I'm not sure it matters much theses
[13:36:04] jya: days
[13:36:27] jya: a BD, must of a multi-channels LPCM stream, or TrueHD/DTS-HD
[13:36:43] stuartm: for now at least the 'experts' in #libbluray don't believe the information is there, so it's not going to be implemented any time soon
[13:36:47] jya: I doubt commentary track will ever be in that format. so are very unlikely going to be selected
[13:36:54] j-rod is now known as j-rod|afk
[13:36:56] j-rod|afk is now known as j-rod
[13:37:05] jya: talking about LPCM
[13:37:13] jya: I haven't added the test in there
[13:37:38] jya: so if we have a 5.1 PCM (so uncompressed audio) and a 5.1 AC3, it will select the AC3
[13:37:45] jya: I had forgotten about that one
[13:38:08] stuartm: right, I was more interested in proper labelling of the tracks for Blu-ray than concerned about the wrong audio track being selected, since as you point out that's less likely to happen for Blu-ray
[13:39:23] stuartm: with Blu-ray's offering a couple of dozen audio tracks labelling the commentary tracks would be even more useful than it is for DVD
[13:42:01] highzeth (highzeth! has quit (Ping timeout: 252 seconds)
[13:42:55] jya: stuartm: the worse with BD is finding the right video track to play.
[13:43:04] jya: I bought the planet of the apes collection
[13:43:25] jya: one of them have like 60 video tracks… and only one is good, all the other one are jumps into another
[13:52:57] highzeth (highzeth! has joined #mythtv
[13:57:36] gregL (gregL! has quit (Quit: Leaving)
[14:01:20] knightr: stuartm, Hi!. Would you see any problem with me adding Chinese date/time formats in globalsettings.cpp? (I would get them reviewed of course) I have a (working) patch that let translators add new date formats at the translation level but the more I think about it the more I think language files might not be the best place for that (locale files might actually be better if their functionality was expanded). Since we`re in feature freeze and soon
[14:01:40] knightr: string freeze it would probably be best to add the date formats to globalsettings.cpp and try to address the issue once 0.25 is released. Would that be ok with you? Thank you!
[14:02:36] knightr: (this is a repaste, it looks like I sent this when you were not connected yesterday..)
[14:02:52] knightr: I have to go now, ttyl..
[14:10:08] stuartm: knightr: no problem
[14:26:43] xavierh: stuartm: thankx for menu-ui.xml, strange that there is no default for that one, should it be added i.e. IF I do it will be considered ?
[14:30:16] stuartm: its assumed that every theme will provide at least that much, it's the initial landing point for anyone using the frontend, probably the screen they will see the most and a theme without it would make no sense – since default is there as a safety net there's really no need for it to have a menu-ui.xml
[14:38:43] zombor_ is now known as zombor
[15:03:11] gregL (gregL! has joined #mythtv
[15:08:29] stichnot (stichnot!chatzilla@mythtv/developer/stichnot) has quit (Ping timeout: 250 seconds)
[15:52:11] jya: wagnerrp: on my mac, I'm seeing mythavtest crash often at the end of playing a test video
[15:52:40] jya: it's always in the Comandlineparser.cpp
[15:52:54] jya:
[16:08:11] stichnot (stichnot!~chatzilla@mythtv/developer/stichnot) has joined #mythtv
[16:13:17] ** wagnerrp takes a look **
[16:17:05] wagnerrp: jya: this is easily repeatable?
[16:17:41] jya: not always… but most of the time
[16:18:01] jya: I usually interrupt playback with ctrl-C
[16:18:04] wagnerrp: in mythtv/libs/libmythbase/mythcommandlineparser.cpp
[16:18:13] wagnerrp: try swapping lines 1142 and 1143
[16:18:22] jya: but I'd say that most of the time (maybe all) I let mythavtest run to completion it crash
[16:19:08] jya: tryinh
[16:21:03] jya: Ok… ran mythavtest three times on that video, it has crashed twice, and 3rd time hang forever
[16:22:25] wagnerrp: its it possible its some very low level bug in qt?
[16:22:39] jya: well, swapping the lines looks good
[16:22:42] jya: no crash
[16:22:46] wagnerrp: im only thinking that due to the heavy optimization that would be using 'sse42'
[16:22:50] jya: 1st time.. starting 2nd time
[16:22:54] jya: yep no crash
[16:22:57] wagnerrp: oh, i thought you meant after swapping, it was still crashing
[16:23:08] jya: oh no… I was making sure I could reproduce it
[16:23:13] wagnerrp: ok
[16:23:22] jya: it 3:23AM, and my memory is a tad fuzzy :)
[16:23:41] jya: well, changing those two lines definitely made it not crash
[16:23:53] jya: I have still an error: Error: Not all threads were shut down properly:
[16:23:55] wagnerrp: yeah, bad ordering on cleanup prior to termination
[16:23:59] jya: Thread AppleRemote is still running
[16:24:04] jya: need to have a look at that one
[16:24:40] jya: what a weird way to look through the element
[16:25:00] wagnerrp: odd that ive never seen it crash before
[16:25:22] jya: Doesn't QMap default destructor do that normally?
[16:26:16] wagnerrp: theyre pointers, not objects
[16:26:33] jya: I mean rather than looping, QMap default destructor will call the destructor for all the elements already. so if cleanuplinks is in the destructor of the CommandLineArg class
[16:26:35] jya: ah right
[16:26:57] jya: with QMap you gain nothing using object vs pointers
[16:26:58] wagnerrp: all those ReferenceCounter objects would end up as 'lost memory' in the milliseconds between the commandlineparser destructor was run, and the application was terminated
[16:27:25] jya: QMap uses copy on write, so basically it will use a reference until it must duplicate the object
[16:27:57] wagnerrp: or if some application wanted to destroy that stuff after initial setup, you could potentially have whole tens of kilobytes of leaked memory
[16:28:07] jya: QMape don't deal with pointers very well. I've had plenty of weird crashes when I was using pointers in QVector or QMap.. I understand now why, but it took a while
[16:28:16] wagnerrp: i didnt want to suffer the wrath of Udo
[16:28:31] jya: that wouldn't be a leak
[16:28:47] jya: when an application is killed, any memory it has ever allocated is clear by the kernel
[16:29:07] wagnerrp: im saying if someone decided to destroy the command line parser without cleaning up all those command line options
[16:29:12] jya: when you quit an application, it can never leak (not with modern system that is)
[16:29:27] wagnerrp: and destroy it manually, while leaving the application running
[16:29:40] wagnerrp: i.e. extract the information you need, and destroy the parser
[16:30:28] jya: I understand what you are doing… What I'm saying is that instead of doing a QMap of pointers, you could actually use the objects, you wouldn't waste any more RAM nor lose performance. QMap doesn't copy the whole object like std:vect would. And then you wouldn't have to clean the ovbject manually, default destructor will do its magic
[16:31:05] wagnerrp: no, i need pointers because i want to edit those objects, and have the edits show up in every place the pointer exists
[16:31:21] wagnerrp: QMap would create duplicates on edits if i didnt use pointers
[16:31:22] jya: let me find you the article about the Q objcts and how they handle copy on write, so say even when you return a QMap on the stack, it actually doesn't copy the object on the stack. it passes references
[16:32:15] jya: where do you edit them? do you edit the object you store in the qmap, or outside, after you store it in the qmap?
[16:33:04] wagnerrp: both
[16:33:20] jya: oay
[16:34:04] wagnerrp: some pointers are in qmaps, some are in qlists
[16:34:11] wagnerrp: commandlinepointers reference other commandlinepointers
[16:35:32] jya: i see
[16:37:02] jya: this VDA is pretty efficient, but it's not working on interlaced content… so rather useless for watching TV :(
[16:37:32] wagnerrp: Fox and ABC would work fine!
[16:37:42] wagnerrp: thats most of what ive been recording the past few weeks
[16:38:03] wagnerrp: with CBS and NBC on mid-season breaks for several shows
[16:38:12] jya: they are all progressive channels?
[16:38:42] jya: like I have sample of the matrix bluray ; 5.6% CPU
[16:39:07] jya: but watching stuartm sample: it climbs to over 100% when deinterlcaing as well
[16:39:36] wagnerrp: yeah, corporate policy for FOX and ABC to be 720p
[16:41:10] stuartm: wagnerrp: that sucks
[16:42:24] wagnerrp: id much rather have the wonderful encoder my local FOX affiliate has at 720p, than the utter garbage my CBS affiliate runs at 1080i
[16:43:20] wagnerrp: although it may just be because all my recordings are actually 24p, and the engineers at FOX just understand how to use duplicate frames properly
[16:52:04] stuartm: jya: turning up an 1080p sample is more difficult that I thought, one or both of ITV1/BBC HD have shown progressive at some point for the odd film, but I don't know which if any of the hundreds of recordings I have those might be
[16:52:50] jya: ok
[16:52:57] jya: right now I'm going through all of those:
[16:52:57] jya:
[16:53:40] jya: so far of the one that failed, they also failed in VLC and xbmc but one: the negative_nalsize.mp4
[16:56:03] jya: hum: this one doesn't play… it cuts here and there. xbmc sees a discontinuity but recovers nicela
[16:59:02] jya: ah, actually, it's even worse when using ffmpeg!
[17:00:00] stuartm: the debug osd doesn't show whether it's interlaced or progressive, or bitrate, both would be useful
[17:00:18] jya: stuartm: well kinda
[17:00:25] jya: it would show 25fps
[17:00:41] jya: which on UK TV would be 50Hz interlaced
[17:00:56] stuartm: true
[17:01:13] jya: you go into menu -> playback -> playback data
[17:04:03] jya: taylorr: an interesting one… . . . %20sync).mkv it calculates its a 7.99fps
[17:04:56] jya: it's dropping fframes like crazy
[17:04:58] MavT (MavT! has joined #mythtv
[17:07:28] XChatMav (XChatMav! has quit (Ping timeout: 272 seconds)
[17:12:16] seeker (seeker!~seeker@unaffiliated/seeker) has joined #mythtv
[17:34:34] kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has joined #mythtv
[17:41:23] ** dblain isn't sure that anyone cares, but just in case... using vs2010 to compile on Windows 8 and Mythbackend runs. **
[17:43:06] jya: dblain: did you create a complete new VS project that simulates all the qmakefile ?
[17:43:41] jya: or are you only using the compiler?
[17:43:57] dblain: Used the Qt integration tools... the solution has been in the Packaging repository for 6+ months.
[17:44:17] jya: I'm thinking of creating an xcode mythtv project, would help greatly in debugging and profiling: and finally having a proper IDE for working on myth
[17:44:22] dblain: It offers full IDE integration... all projects listed and full debugging support
[17:44:42] jya: dblain: I've never compiled myth on windows, and I think the last time I tried was for 0.20 :)
[17:44:51] jya: ah cool...
[17:45:12] jya: I'll definitely have a look to check what you've done
[17:45:45] jya: because my guess is that it will be very similar to what is required with xcode
[17:45:52] dblain: It's a little bit of a pain to keep in sync with all the changes, but for me its been worth it.
[17:46:37] dblain: jya, not sure it will help. Qt has a Visual Studio addin... I just had to open each pro file using it and then tweek the project setting and some of the code.
[17:46:37] jya: i;m not surprised. I've found that ultimately the only way to debug myth is to add plenty lof log messages
[17:46:50] jya: so does qt on mac
[17:47:12] jya: you run qmake with a specific spec file, and it generates xcode project instead
[17:47:13] dblain: With VS, I have full source level debugging... one thing I like about microsoft is the quality of the dev tools.
[17:47:47] NightMonkey (NightMonkey!~NightMonk@pdpc/supporter/professional/nightmonkey) has joined #mythtv
[17:47:58] stoffel (stoffel! has quit (Read error: Operation timed out)
[17:47:59] jya: I'm with you… IMHO vistual studio is the best IDE out there… I worked on WP7 application in the past few months… it was a delight working with it
[17:48:03] dblain: yep... the only difference is that qmake is hidden from the developer... handled in vs project settings. I have an option to run it from the command line, but ultimately is wasn't worth it.
[17:48:32] jya: so do you run the qmake -> vs project stuff often
[17:48:41] jya: and do you have to do it manually ?
[17:49:31] jya: okay.. you gave me hope… I'll look into this next.
[17:50:15] dblain: When I was setting up the solution, I just used the Qt menu choice in the IDE that was "Open Qt Project" or something like that. It converted the pro file to a vcproj file and I've been manually maintainng it since. It would be nice not to constantly add/remove files as the main source changes, but what I have works and until I get fed up with it...
[17:50:56] jya: ah… so Qt adds an entry in VS..
[17:51:03] jya: yeah, that won't be as simple on a mac
[17:51:11] dblain: yes, but your qmake approach would also work
[17:51:51] jya: ok… i'll look into this too… great.. thanks
[17:51:55] jya: off to bed now
[17:52:15] dblain: If Visual Studio ever becomes a truely supported tool for myth developers, I would look into generating the solution using qmake... until then... have a good night.
[17:55:43] jya: just had a thought… how do you manage all the dependencies?
[17:58:46] dblain: There are only a few and I've found windows versions. The only thing I can't compile in VS is all the FFMeg based libraries which I use the minwg versions.
[17:59:09] natanojl (natanojl! has joined #mythtv
[17:59:20] dblain: so I'm cheating a little :/
[18:00:46] dblain: you can see all the dependencies I use in the packaging repo.
[18:24:50] ThisNewGuy1 (ThisNewGuy1! has joined #mythtv
[18:26:56] ThisNewGuy (ThisNewGuy! has quit (Ping timeout: 252 seconds)
[18:31:48] stoffel (stoffel! has joined #mythtv
[18:48:38] stichnot: I reread stuartm's comment at about my patch in . I was indeed able to make a simpler version of the patch, which basically just moves initialization code from PlaybackBox::updateRecList() into PlaybackBox::ItemVisible(), plus a few lines in MythUIButtonListItem to ensure that the initialization is
[18:48:40] stichnot: only done once. Is this a "bug" worth fixing for 0.25? Something to ask on the mailing list?
[18:53:48] stuartm: stichnot: ask on the mailing list, add the updated patch to the ticket too
[18:54:32] stuartm: I'm curious about the MythUIButtonListItem changes
[18:59:20] seeker (seeker!~seeker@unaffiliated/seeker) has quit (Quit: Colloquy for iPhone -
[19:06:45] sphery: danielk22: currently the opengl autodetection is still there, but is only used if you set the theme painter to Auto. However, since it fails with broken drivers (AMD/ATI) or broken driver installations (which users get quite often with nvidia and possibly with Intel), if I set Auto as the default for mythfrontend, users would have to know that if mythfrontend startup fails, they need to restart with -O ThemePainter=qt then go into setup to ...
[19:06:51] sphery: ... change the painter. Currently the default is Qt, which works for everyone--at least as much as the Qt painter works--but that means that users are unlikely to know that they're better off changing to Auto (or could change to DirectX or OpenGL, but Auto is better). So, either it fails for lots of users who are just trying MythTV and they either give up or report tons of "invalid" bugs, or we keep defaulting to Qt painter and have to ...
[19:06:57] sphery: ... somehow inform users to switch to Auto manually
[19:08:16] sphery: and, really, for AMD/ATI users, sometimes--depending on card/drivers--startup works fine and UI looks great, then after exiting playback, they get black screen and are stuck
[19:09:12] sphery: I'm thinking defaulting to Qt for at least one more release until we can come up with a better way of deciding is safest for now
[19:12:19] danielk22: sphery: Makes sense, especially if this isn't just affecting ATI users.
[19:18:08] sphery: danielk22: Also, I'm wondering if . . . 120306.patch might help with (and, possibly, the issue taylorr is seeing). Changes EOF handling (and could probably be extended to specify a type for EOF of in-progress recording). Might be too invasive for feature freeze, though.
[19:19:14] danielk22: sphery: It might but it is too likely to cause other problems.
[19:24:22] SteveGoodey (SteveGoodey! has joined #mythtv
[19:33:01] knightr (knightr!~knightr@mythtv/developer/knightr) has quit (Ping timeout: 276 seconds)
[19:33:23] knightr (knightr!~knightr@mythtv/developer/knightr) has joined #mythtv
[19:37:32] stuartm: if I'd know we'd be defaulting to qt for another release I might have made more of an effort to fix the QT clipping issues affecting Terra  :/
[19:47:22] sphery: well, we can do whatever you all want
[19:49:10] JayAge (JayAge! has joined #mythtv
[19:49:32] JayAge (JayAge! has left #mythtv ()
[19:54:02] ThisNewGuy1 (ThisNewGuy1! has quit (Ping timeout: 250 seconds)
[19:56:41] ThisNewGuy (ThisNewGuy! has joined #mythtv
[20:13:45] stoffel (stoffel! has quit (Remote host closed the connection)
[20:22:22] stichnot: stuartm: thinking about it, I can probably get by with no MythUIButtonListItem changes just by probing for existing content to indicate whether it's fully initialized. In any case, I'll update the ticket and ask the mailing list.
[20:33:21] skrock (skrock! has joined #mythtv
[20:42:51] NightMonkey (NightMonkey!~NightMonk@pdpc/supporter/professional/nightmonkey) has quit (Ping timeout: 244 seconds)
[20:46:39] stoffel (stoffel! has joined #mythtv
[20:48:57] Jordack (Jordack! has quit ()
[20:52:19] NightMonkey (NightMonkey!~NightMonk@pdpc/supporter/professional/nightmonkey) has joined #mythtv
[21:04:47] RogerM (RogerM!~chatzilla@ has joined #mythtv
[21:07:19] zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[21:09:44] j-rod is now known as j-rod|afk
[21:10:02] RogerM (RogerM!~chatzilla@ has quit (Quit: ChatZilla [Firefox 11.0/20120312181643])
[21:13:25] SteveGoodey (SteveGoodey! has quit (Remote host closed the connection)
[21:16:35] stoffel (stoffel! has quit (Remote host closed the connection)
[21:46:39] skrock (skrock! has quit (Quit: WeeChat 0.3.7)
[21:50:50] stichnot: Wow stuartm, using your advice, my latest lazystrings patch for speeding up pbb loading is 1/3 the size of the previous patch and only affects one file.
[21:56:25] stuartm: your welcome ;)
[21:58:30] sphery: nice
[22:02:52] stuarta: awesome
[22:03:29] stichnot: I'll upload it and consult the mailing list oracle after testing on the ION
[22:13:14] stuartm: any noticeable slowdown when scrolling on faster machines?
[22:16:40] stichnot: stuartm: on my way out, I'll answer later
[22:20:54] stichnot (stichnot!~chatzilla@mythtv/developer/stichnot) has quit (Ping timeout: 244 seconds)
[22:34:20] joki (joki! has quit (Quit: c ya!)
[22:38:24] knightr: sphery, I sent Roger an email to ask him if there was anything I could help him with.
[22:39:08] sphery: knightr: good, I was just a bit too slow to catch him
[22:39:18] sphery: thanks
[22:41:41] knightr: sphery, no problem... I doubt it's an actual translation question though, he most likely wants to know what one of the English sentences mean (it's sometimes difficult to figure out what a sentence means unless you see it in context which is not always easy to do since we don't always know under which conditions a message appears...)
[22:42:03] knightr: Thank you for trying though...
[23:18:33] NightMonkey (NightMonkey!~NightMonk@pdpc/supporter/professional/nightmonkey) has quit (Ping timeout: 250 seconds)
[23:26:55] NightMonkey (NightMonkey!~NightMonk@pdpc/supporter/professional/nightmonkey) has joined #mythtv
[23:26:58] zombor (zombor!~zombor_@ has joined #mythtv
[23:26:59] zombor (zombor!~zombor_@ has quit (Changing host)
[23:26:59] zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv
[23:27:09] zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[23:28:50] stichnot (stichnot!~chatzilla@mythtv/developer/stichnot) has joined #mythtv
[23:30:22] zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv
[23:34:01] zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[23:37:14] danielk22: stuartm: stichnot: cool! I really liked the idea but my only reservation was the size and invasiveness.
[23:38:21] sphery: knightr: yeah, that's why I thought I might actually be able to help
[23:40:54] zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv
[23:52:25] zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[23:56:38] NightMonkey (NightMonkey!~NightMonk@pdpc/supporter/professional/nightmonkey) has quit (Quit: Body blow! Body blow!)

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