:: #mythtv

Daily chat history

Current users (86):

aloril, Anssi, anykey_, Beirdo, brfransen, brtb, CaCtus491, cattelan_away, cesman, Chutt, clever, coling, Cougar, damaltor, danielk22, dekarl, dlblog, Dr{Wh0}, ElmerFudd, eugo, foobum, foxbuntu, ghoti, gregL, GreyFoxx, Guest74934, hank, highzeth, J-e-f-f-A, j-rod|afk, jarle, jcarlos, jeff999, joe_, joki, jpabq, jpabq-, jstenback, justinh, jwhite, jya, k-man, kenni, knightr, kurre2, kwmonroe, laga, lolcat`, mag0o, Malard, MaverickTech, mrand, MythBuild, MythLogBot, mzanetti, noahric, peitolm, pheld, poptix, purserj, rhpot1991, rsiebert_, skd5aner, Slasher`, sphery, sraue, stuarta, stuartm, superm1, sutula, tgm4883, TheAsp, ThisNewGuy1, tomimo, tris, Unhelpful, vallor, wagnerrp, wahrhaft, wseltzer, xavierh, XDS2010_, xris, ybot_, zCougar, _charly_
Sunday, March 4th, 2012, 00:04 UTC
[00:04:46] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has quit (Ping timeout: 272 seconds)
[00:08:06] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has joined #mythtv
[00:11:19] stichnot (stichnot!chatzilla@mythtv/developer/stichnot) has quit (Remote host closed the connection)
[00:12:38] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has quit (Ping timeout: 265 seconds)
[00:14:39] stichnot (stichnot!chatzilla@mythtv/developer/stichnot) has joined #mythtv
[00:19:56] xavierh_: I am about to convert Recordings profile, I never really modified anything in those settings, so I am not sure how it does work
[00:20:53] xavierh_: First of all I can create a new profile, choose name , card type and hostname, but the new profile never appear in the list ... ?
[00:21:34] xavierh_: convert to MythUI ^^^
[00:22:24] xavierh_: Is it a bug or an expected behaviour I don't understand
[00:31:37] Captain_Murdoch: stuartm, sphery, nothing uses the channel icons SG yet that I know of, so I think we can change the path on that I think.
[00:32:18] Captain_Murdoch: stuartm, the 'one socket per file' is the way the myth:// file transfer code has always worked.
[00:33:50] Captain_Murdoch: the change I made for HLS mode was for writing files, there are gotchas for reading files. writing doesn't have the readahead buffer, etc. so I have it limited to changing files during writing currently.
[00:34:39] ** Captain_Murdoch sees the later discussion on the file transfer code... **
[00:38:58] jya: xavierh: I've found with the playback profile or recordings profile.. when you create a new profile, it doesn't show up until you go save, and go back into the settings
[00:39:03] jya: been like that for a long time
[00:40:26] stuartm: xavierh_: you can call it a bug (there is an open ticket) or an incomplete feature (more accurate description)
[00:40:49] stuartm: xavierh_: fwiw, I'm aiming to get a fix in for 0.25
[00:41:24] Captain_Murdoch: Beirdo, I think that there's an issue with audio timecodes, the audio has glitches when streaming to my iPhone. I'm looking around now.
[00:41:47] stuartm: not that anyone really needs than many recording profiles, it's not even clear why for DVB etc you'd want more than the two (livetv/other)
[00:48:59] xavierh_: jya: It does not create the profile because I choose card type which I don't have
[00:49:32] xavierh_: stuartm: should I allow to chose the card type only from the "installed" card ?
[00:49:53] stichnot: Has anyone recently experienced a problem where playback continues into the final cut region of the recording?
[00:50:56] xavierh_: Also once created, the new profile cannot be deleted, can it? Should I make it deletable? what would be the consequences ?
[00:54:05] knightr: In case anybody wonders that code removal in my last commit was OKed by sphery...
[00:55:58] stichnot: I'm pretty sure I tracked down an error in . . . 5863d5f659b6 that might have caused the problem, but I can't reliably reproduce the problem.
[00:57:07] Seeker`: Captain_Murdoch: pause / unpause and/or skip tracks usually fixes skipping
[00:57:16] Captain_Murdoch: Beirdo, I think the issue I'm seeing has to do with different audio codec framesizes.
[00:57:45] Seeker`: Captain_Murdoch: bah, ignore me
[00:57:52] Seeker`: thought you were talking about airplay stuff
[00:57:54] Captain_Murdoch: Seeker`, OK. :)
[00:58:13] Captain_Murdoch: nah, HLS using a test patch by Beirdo to fix an issue in mythtranscode fifo code that was broken by the original HLS patch.
[00:59:13] Seeker`: I'm mostly concentrating on my pathtracer atm, so only caught what you said out fo the corner of my eye
[01:19:58] xavierh_ (xavierh_! has quit (Remote host closed the connection)
[01:23:02] taylorr: jpabq: here is the git repo diff -> . . . 928ddf3d9b3e
[01:35:49] natanojl (natanojl! has quit (Ping timeout: 244 seconds)
[01:47:48] zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[02:05:58] davide (davide! has quit (Remote host closed the connection)
[02:06:19] davide (davide! has joined #mythtv
[02:46:10] noahric (noahric! has quit (Quit: noahric)
[02:50:28] sphery (sphery!~mdean@mythtv/developer/sphery) has quit (Quit: leaving)
[03:07:12] noahric (noahric!~noahric@ has joined #mythtv
[03:16:32] ** jya waiting for mythbuildbot with blamelist: Jean-Yves Avenard < > **
[03:18:46] MythBuild: build #2029 of master-linux-ppc is complete: Failure [failed compile core] Build details are at . . . /builds/2029 blamelist: Jean-Yves Avenard < >
[03:18:51] jya: ahah
[03:19:22] MythBuild: build #3268 of master-linux-64bit is complete: Failure [failed compile core] Build details are at . . . /builds/3268 blamelist: Jean-Yves Avenard < >
[03:19:43] jya: ahhh damn.. last minute change forgot the git rename :(
[03:20:15] jya: need to learn how to use git am better
[03:21:13] jya: and now we've lost previous history of util.h.. fuck1
[03:23:37] MythBuild: build #3013 of master-linux-32bit is complete: Failure [failed compile core] Build details are at . . . /builds/3013 blamelist: Jean-Yves Avenard < >
[03:23:42] MythBuild: build #2051 of master-freebsd-64bit is complete: Failure [failed compile core] Build details are at . . . /builds/2051 blamelist: Jean-Yves Avenard < >
[03:29:55] MythBuild: build #472 of master-osx-snow-leopard is complete: Failure [failed compile] Build details are at . . . d/builds/472 blamelist: Jean-Yves Avenard < >
[03:30:37] jya: you plan things carefully, re-check everything 10 times, yet you end up with a screw up
[03:34:16] Captain_Murdoch: if you revert that original commit, would it bring back the history? if so, then you could recommit with the rename.
[03:48:01] jya: Captain_Murdoch: I tried that
[03:48:09] jya: but it still shows as if it had no history
[03:48:48] jya: I guess I could try to revert completely the two commits , and redo it all once again
[03:48:48] stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has quit (Remote host closed the connection)
[03:49:44] jya: actually, I'm going to do just that, revert the whole lot
[04:03:41] MythBuild: build #3271 of master-linux-64bit is complete: Success [build successful] Build details are at . . . /builds/3271
[04:09:37] noahric (noahric!~noahric@ has quit (Quit: noahric)
[04:10:44] jya: the master-linux-64bits is significantly faster than all the others...
[04:11:50] MythBuild: build #2054 of master-freebsd-64bit is complete: Success [build successful] Build details are at . . . /builds/2054
[04:12:56] MythBuild: build #3016 of master-linux-32bit is complete: Success [build successful] Build details are at . . . /builds/3016
[04:13:16] jya: wagnerrp: are you managing the FreeBSD buildbot machine ?
[04:13:20] wagnerrp: strictly speaking, master-linux-64bit and master-freebsd-64bit are about the same
[04:13:36] wagnerrp: linux is quadcore, freebsd is dualcore, but the build is being done with -j1
[04:14:14] wagnerrp: except the freebsd buildbot lives on a ZFS mirror, on a system with woefully inadequate memory, and 6yr old hard drives
[04:14:19] jya: I see that it has detected Dcraw on the new freebsd build. I completely change the way dcraw is being checked… Could you see if the dcraw utility is installed ?
[04:15:13] jya: give me the output of which dcraw and hash dcraw
[04:15:20] jya: thanks!
[04:15:32] saintdev (saintdev!~saint@unaffiliated/saintdev) has left #mythtv ("Leaving")
[04:15:35] wagnerrp: i see no dcraw package
[04:15:51] wagnerrp: nor any dcraw executable
[04:16:25] jya: hum...
[04:16:35] MythBuild: build #474 of master-osx-snow-leopard is complete: Success [build successful] Build details are at . . . d/builds/474
[04:16:44] jya: it tested on mac and linux, and it behaves properly
[04:17:11] jya: . . . 1/logs/stdio didn't see it there either
[04:18:34] jya: hum… hash on FreeBSD sh always returns an exit code of 0
[04:18:37] MythBuild: build #2031 of master-linux-ppc is complete: Success [build successful] Build details are at . . . /builds/2031
[04:18:40] jya: that's no good
[04:19:52] jya: luckily, dcraw module will compile fine if it's not there, the mythgallery plugin just do a system
[04:21:17] MythBuild: build #1782 of master-debian-stable-64bit is complete: Failure [failed compile core] Build details are at . . . /builds/1782 blamelist: Jean-Yves Avenard < >
[04:21:58] jya: that's the slow one catching up
[04:26:52] wagnerrp: huh... commit timestamps at the time you run the command, not when you finish editing the commit message
[04:27:52] jya: wagnerrp: I mean it's the debian buildbot finishing building from the earlier commits for which all others have failed already
[04:28:22] ** jya wonder how to test if an exe exists that works across-platforms **
[04:28:48] wagnerrp: no, i committed something, took 6 minutes writing up a commit message, and the commit was timestamped when i first ran the command
[04:30:37] MythBuild: build #852 of master-vista-mingw-32bit is complete: Failure [failed compile] Build details are at . . . t/builds/852 blamelist: Jean-Yves Avenard < >
[04:37:58] MythBuild: build #1783 of master-debian-stable-64bit is complete: Success [build successful] Build details are at . . . /builds/1783
[04:47:23] Beirdo: BTW, that rename merits an ABI version bump
[04:47:30] Beirdo: and requires make distclean
[04:47:54] wagnerrp: what?
[04:48:23] Beirdo: renaming util.h to mythutil.h changed the ABI presented by libmythbase
[04:48:59] wagnerrp: ah, not me
[04:49:06] wagnerrp: well i bumped it anyway, so no worries
[04:49:12] Beirdo: K :)
[04:54:15] wagnerrp: wow, nearly half an hour for that one
[04:54:36] wagnerrp: i guess i touched a lot of files with it
[05:00:33] jya: Beirdo: how does it change the ABI number, no functions has been renamed
[05:01:08] jya: actually, the name mythutil.h isn't a good one, there's already a mythutil.h in program/mythutil
[05:01:48] jya: if I had renamed the util.cpp function I would agree
[05:02:17] Beirdo: because part of the ABI is the footprint of the headers themselves
[05:02:33] jya: seeing it that way… okay :)
[05:02:46] jya: but no existing binaries would break in any ways
[05:03:01] Beirdo: Hmm, I'd hope not :)
[05:03:55] jya: so what's best… rename util.h to mythutil.h and in mythutil/mythutil.cpp I link to ../libmythbase/mythutil.h ; or rename the util.h file mythmiscutil.h ? (IMHO the later)
[05:04:39] Beirdo: or mythbaseutil.h
[05:07:03] jya: there's already a mythbaseutil.h :(
[05:07:10] jya: that was my 2nd choice
[05:07:27] Beirdo: gah
[05:07:53] Beirdo: sounds like something that needs serious rethought after release, doesn't it?
[05:08:26] jya: uniquenamethatwontconflict.h would do
[05:09:01] clever: what if everybody has that idea?
[05:09:13] clever: could use /usr/bin/pwgen !
[05:09:42] Beirdo: I guess mythmiscutil.h is a half-decent thought then
[05:09:55] Beirdo: can't think of anything better
[05:09:56] jya: thanks :P
[05:26:28] jya: Beirdo: where is the ABI defined again?
[05:26:48] Beirdo: libmythbase/mythversion.h
[05:27:12] jya: thanks.. I always confuse it with dbcheck in libmythtv
[05:27:27] Beirdo: heh, not too surprising :)
[05:27:51] jya: it's not the most logical place to do dbcheck btw
[05:31:01] Beirdo: yeah
[05:34:33] jya: allright.. attempt #2
[05:36:42] jya: allright, be back in 30' to make sure everything is okay...
[05:40:47] MythBuild: build #853 of master-vista-mingw-32bit is complete: Success [build successful] Build details are at . . . t/builds/853
[05:47:48] Goga777 (Goga777! has joined #mythtv
[05:47:54] xris: odd. irc client says I have 3 messages mentioning me, but then it doesn't show them.
[05:49:57] Beirdo: heh
[05:53:06] jya: Beirdo: out of curiosity. How does buildbot determine the ETA of a build?
[05:53:15] sphery (sphery!~mdean@mythtv/developer/sphery) has joined #mythtv
[05:55:13] sphery: jya: one option for dcraw: rather than detect a working binary, reimplement it as proper functionality using
[05:55:18] sphery: :)
[05:58:11] Beirdo: don't even start with that, sphery, it's already on the list :)
[05:58:28] Beirdo: jya: based on previous performance, I believe
[05:58:30] Goga777 (Goga777! has quit (Ping timeout: 260 seconds)
[05:59:22] jya: Well, my next task is fixing #10370
[05:59:27] sphery: hehe, had a feeling you'd notice
[06:03:33] Goga777 (Goga777! has joined #mythtv
[06:04:13] jya: wagnerrp: new configure worked for dcraw and freebsd...
[06:17:33] Goga777 (Goga777! has quit (Quit: Leaving)
[07:47:04] wagnerrp: well this is interesting... . . . roduct_id=17
[07:47:18] wagnerrp: did someone write an openmax video renderer for us?
[08:09:18] kenni: jya: I saw #10403 last week on one of my frontends when I upgraded from 0.24-fixes to git master. It was because the Escape key had multiple key bindings attached in the Main Menu context. Once I ensured that it was bound to EXITPROMPT rather than EXIT, it worked again.
[08:09:47] kenni: interestingly, I didn't have the issue on another frontend of mine, which should be identical.
[08:15:08] MythBuild: build #768 of cppcheck-master is complete: Failure [failed git] Build details are at . . . r/builds/768 blamelist: Gavin Hurlbut < >
[08:31:47] ** Beirdo slaps his box **
[08:32:32] Beirdo: MythBuild: force build cppcheck-master now, maybe github's not borked
[08:32:32] MythBuild: Something bad happened (see logs): <type 'exceptions.ValueError'>
[08:32:41] Beirdo: MythBuild: force build cppcheck-master now, maybe github iss not borked
[08:32:42] MythBuild: build forced [ETA 11m35s]
[08:32:42] MythBuild: I'll give a shout when the build finishes
[08:32:48] Beirdo: stupid no '
[08:32:53] jya: kenni: It does it on a brand new frontend I had never used mythtv on… I'm pretty certain than last week there was the exit.. now, none of my frontends display them
[08:35:43] kenni: jya: ok, is the binding to the Esc key correct?
[08:36:13] jya: i don't know.. I never looked there provided I haven't changed a thing from default settings
[08:36:20] XChatMav (XChatMav! has quit (Ping timeout: 252 seconds)
[08:36:51] kenni: jya: neither have I, never went there before I experienced the issue with the lacking prompt
[08:36:52] MaverickTech (MaverickTech! has joined #mythtv
[08:37:37] jya: if it's something to do with borked settings, that's a bug anyway.. how do they get corrupted in the first place?
[08:37:50] jya: let me change the frontend identifier and try agaib
[08:38:07] kenni: I got a hint from the frontend log, which gave a warning about the multiple bindings on bootup
[08:39:21] jya: kenni: you're right.. on this brand new frontend, EXIT and EXITPROMPT are both bound to the same key
[08:41:10] kenni: so something, somewhere in master is apparently enforcing the EXIT keybinding even if EXITPROMPT is already set.
[09:12:55] jya: done… kids in bed… finally !
[09:14:20] MythBuild: build #769 of cppcheck-master is complete: Success [build successful] Build details are at . . . r/builds/769
[09:21:27] jya: kenni: I've deleted the EXIT keybinding, and it still exits no matter what..
[09:38:09] jya: [9220]
[09:38:09] MythLogBot: SVN 9220: (branch master)
[09:59:59] jarle (jarle! has joined #mythtv
[10:02:54] natanojl (natanojl! has joined #mythtv
[10:10:38] RogerM (RogerM!~chatzilla@ has joined #mythtv
[10:20:47] stuartm (stuartm! has joined #mythtv
[10:20:48] stuartm (stuartm! has quit (Changing host)
[10:20:48] stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has joined #mythtv
[10:34:52] stuartm: I really wish that Shepherd would pull it's act together and start being xmltv compliant
[10:34:58] stuartm: it's getting really old
[10:52:43] RogerM (RogerM!~chatzilla@ has quit (Quit: ChatZilla 0.9.88 [Firefox 10.0.1/20120208060813])
[11:05:02] Guest76506 (Guest76506! has quit (Remote host closed the connection)
[11:05:47] mike (mike! has joined #mythtv
[11:06:12] mike is now known as Guest74934
[11:10:53] AndyUbuntu (AndyUbuntu! has joined #mythtv
[11:18:23] natanojl (natanojl! has quit (Ping timeout: 245 seconds)
[11:18:44] jya: stuartm: that may be, but whatever they do works brilliantly… it hasn't missed a bit in years.. always up to date, and all by itself
[11:19:17] stuartm: much like every other xmltv compliant grabber
[11:20:23] stuartm: only those don't break whenever we change something in mythtv, because they stick to the standard
[11:24:13] jya: is shepherd broken right now?
[11:27:20] stuartm: jya: err, yeah, per the ticket you opened
[11:28:04] jya: which one ?
[11:28:15] stuartm: the one about channel icons not working in livetv
[11:28:55] jya: ah, but how is that got to do anything with shepherd? it just doesn't go and sick the images over the storage group
[11:29:09] jya: s/sick/get
[11:30:26] stuartm: jya: shepherd is downloading the icons itself and sticking them in a non-standard location, then modifying the database directly – when what it should be doing is defining the icon urls in the xmltv compliant xml and letting mythfilldatabase download them to the correct location
[11:30:41] jya: ahhhh
[11:30:42] stuartm: shepherd is going to break hard when we remove direct access to the database
[11:40:23] k-man: stuartm, have you ever mentioned to the devs your concerns with shepherd?
[11:40:24] stuartm: when I actually asked the Shepherd devs why it couldn't be written to be compliant the answer was basically 'because we don't want to' mixed with a little of 'things are different in Australia and you just don't understand' (which is crap, the situation is the same in many countries and xmltv/mythfilldatabase accounts for all of that)
[11:40:34] stuartm: k-man: yep ^^
[11:41:08] k-man: fair enough
[11:41:11] stuartm: they are wedded to their collection of scripts and cron jobs, they've no real interest in doing things properly
[11:41:54] k-man: when will direct access to the db be removed?
[11:53:08] stuartm: probably 0.27 which will be end of the year maybe, although sooner if you're using master obviously
[11:53:57] stuartm: /end of year/spring next year/
[11:54:12] aster1sk (aster1sk! has joined #mythtv
[11:54:17] jya: I'd love sooner.
[11:54:31] jya: I wouldn't mind working on that one
[11:55:10] stuartm: that's all it takes
[11:55:24] stuartm: someone willing to work on it that is
[11:56:06] aster1sk: Morning all, this picture is killing me, is there anything I can change to increase the quality — I'll provide details in a moment.
[11:56:37] stuartm: enable deinterlacing?
[11:56:52] aster1sk: It's Hauppage 1600 Analogue / Rogers Cable — 0.24.0+fixes.20110908.1de0431–0ubuntu1
[11:57:16] stuartm: aster1sk: you might be better off in #mythtv-users
[11:57:22] stuartm: that's the support channel
[11:58:18] aster1sk: Thank you for your help gentlemen.
[11:58:40] aster1sk (aster1sk! has left #mythtv ()
[11:59:46] AndyUbuntu (AndyUbuntu! has quit (Remote host closed the connection)
[12:37:11] RogerM (RogerM!~chatzilla@ has joined #mythtv
[12:42:52] anykey_ (anykey_! has quit (Quit: leaving)
[12:43:08] anykey_ (anykey_! has joined #mythtv
[12:49:17] stuartm: I thought we had changed the default capture resolution to 720x576, but when I create a new profile it's still setting it to 480x576, that's despite the help text warning of problems when setting a width less than 720 (or 768)
[12:49:35] stuartm: danielk22, gigem: ^^
[12:52:19] Goga777 (Goga777!~Goga777@ has joined #mythtv
[13:05:01] davide_ (davide_! has joined #mythtv
[13:05:02] davide_ (davide_! has quit (Changing host)
[13:05:02] davide_ (davide_!~david@mythtv/developer/gigem) has joined #mythtv
[13:05:56] davide (davide! has quit (Read error: Operation timed out)
[13:14:22] stuartm: seems that we still set 480x480 for v4l capture cards, even the new generation with hardware encoding, that seems wrong somehow
[13:34:42] jya: danielk22: could you have a look at this ? ; you were the original author of that code. This is to solve an issue where Qt plugins are loaded before the QCoreApplication has been created (ultimately failing on mac) …
[13:36:02] jya: I have doubts that declaring a variable as static const being thread-safe. What would happen if one thread call decode_text, initilisation start as it's the first time, and before it completes another thread call decode_text as well…
[13:37:23] jya: I don't even know if this is ever going to be a scenario that would appear in myth. The original ticket for that commit was [1420] and there are talks about the need to be re-entrant.
[13:37:23] MythLogBot: SVN 1420: (branch master)
[13:37:38] Goga777 (Goga777!~Goga777@ has quit (Read error: No route to host)
[13:38:10] jya: so I would appreciate your expert opinion on the matter, especially in regards to static const local variable used in a multi-threaded environment
[13:38:48] Goga777 (Goga777!~Goga777@ has joined #mythtv
[13:40:29] jya: #1420
[13:40:53] jya: is there a way to create a link automatically to a ticket number (here
[13:41:35] Goga777 (Goga777!~Goga777@ has quit (Client Quit)
[13:48:44] kth (kth!~kth@unaffiliated/kth) has joined #mythtv
[13:51:12] kth (kth!~kth@unaffiliated/kth) has quit (Client Quit)
[13:51:19] kth (kth!~kth@unaffiliated/kth) has joined #mythtv
[14:08:00] RogerM (RogerM!~chatzilla@ has quit (Quit: ChatZilla 0.9.88 [Firefox 10.0.1/20120208060813])
[14:11:37] stuartm:
[14:12:07] stuartm: ah, that's my local replace, won't work for you ...
[14:12:21] stuartm: MythLogBot: list
[14:12:55] stuartm: so no, but if you bend Beirdo's ear he might add a !ticket or similar
[14:15:14] stuartm: jya: I can't reproduce , are you sure that that frontend is still configured to show the exit prompt?
[14:15:53] danielk22: jya: Moving the iso8859_codecs declaration shouldn't cause any new run-time issues. But yeah, using a static const there means we only expect decode_text to be called from one thread, which doesn't seem like a good assumption to make.
[14:31:30] kenni: jya: Huh, I just removed the duplicate EXIT/EXITPROMPT binding, restarted the frontend and the problem was solved.
[14:40:33] kenni: stuartm: jya had the issue on a new frontend with default settings. I had the same issue last week, when I upgraded one of my frontends from 0.24-fixes to git master. After the upgrade the Escape key was bound to both EXIT and EXITPROMPT. In my case, removing the EXIT key binding solved the issue.
[14:42:08] stuartm: sphery: ^^ might want to look into that, seems like the DB update is going awry
[14:49:57] stoffel (stoffel! has joined #mythtv
[15:21:47] natanojl (natanojl! has joined #mythtv
[15:33:27] stoffel (stoffel! has quit (Ping timeout: 244 seconds)
[15:44:21] stuartm: we should add descriptions to the mythtv-setup menu explaining each step, that is AFAIK the only menu lacking <description> tags and yet it's probably the one which most needs them
[15:45:17] stuartm: it needs doing before we cut RC1, better still before the beta
[16:11:44] rsiebert_ (rsiebert_! has joined #mythtv
[16:15:10] rsiebert (rsiebert! has quit (Ping timeout: 276 seconds)
[16:23:56] wagnerrp: you know, at least some spammers attempt to hide their motives
[16:24:12] wagnerrp: but someone using an 'indianoceanseo' domain for their email? sheesh...
[16:26:53] Dave123 (Dave123! has quit (Quit: Leaving)
[16:33:08] natanojl (natanojl! has quit (Ping timeout: 240 seconds)
[16:33:14] wagnerrp: stuartm: you around?
[16:41:16] wagnerrp: stuartm: when ever you get in, any thoughts on the following entries in trac's spam monitor?
[16:42:15] wagnerrp: theres no URLs, theres nothing really objectionable, besides being completely irrelevant, theres even some other open source projects thrown in as keywords
[16:43:00] wagnerrp: they would have been passed through if it weren't for the additional email regex filter i put in place a year back
[16:59:28] mrand (mrand!~mrand@ubuntu/member/mrand) has quit (Ping timeout: 260 seconds)
[17:02:04] xavierh_ (xavierh_! has joined #mythtv
[17:06:00] mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv
[17:08:49] stuartm: wagnerrp: they are probing, trying to find sites that accept their posts and test the limits of the spam filters, my guess is that their script gives up quickly if even innocuous stuff gets blocked
[17:10:21] stuartm: i.e. it's not worth a human crafting a post to get around the spam filter if even scripted innocent looking stuff is blocked or moderated
[17:11:22] stuartm: it's elaborate, but I guess that's what they feel they have to do as more and more sites implement moderation or filtering of some sort
[17:14:30] stuartm: jpabq: should the channel group editor be in the frontend or mythtv-setup? The groupings are global no?
[17:25:17] stuartm: jya: audio/audiooutputpulse.cpp:245: warning: unused variable ‘cstate’
[17:26:26] stuartm: Beirdo: mpeg2fix.cpp:1774: warning: 'end' may be used uninitialized in this function mpeg2fix.cpp:1774: warning: 'start' may be used uninitialized in this function
[17:28:28] warped (warped!~piotro@ has joined #mythtv
[17:28:52] stuartm: dblain: symbol visibility issue? (libmythservicecontracts/datacontracts/programAndChannel.h) – . . . nings%20(47)
[17:29:12] jpabq: stuartm, I believe it should be on the frontend. Personally, I would be annoyed to have to fire up mythtv-setup to made a channel group adjustment.
[17:29:21] jpabq: to make
[17:29:35] jpabq: Assuming you are talking about "favorites"?
[17:30:53] wagnerrp: jpabq: what about having to fire up a web browser to do it instead?
[17:32:39] jpabq: wagnerrp, that could be OK.
[17:43:27] stuartm: jpabq: it's really about where they need to be if you're not using the frontend, if we expose channel groups to third party frontends then they should be managed on the backend – that's where you configure channels after all
[17:45:25] jpabq- (jpabq-! has quit (Ping timeout: 248 seconds)
[17:45:37] abqjp (abqjp! has joined #mythtv
[17:45:50] jpabq (jpabq!~jpabq@mythtv/developer/jpabq) has quit (Ping timeout: 265 seconds)
[17:45:50] abqjp is now known as jpabq
[17:45:51] jpabq (jpabq! has quit (Changing host)
[17:45:52] jpabq (jpabq!~jpabq@mythtv/developer/jpabq) has joined #mythtv
[17:46:37] jpabq- (jpabq-! has joined #mythtv
[17:49:36] taylorr: danielk22: any opinion on preferring dts over pts? seems like dts historically has been more reliable – looking at the latest ffmpeg source the best effort timestamp code prefers pts now
[17:58:02] stuartm: jpabq: arguably we need to be able to mark a channel as a favourite from the guide
[17:59:20] stuartm: that's the simplest UI, being able to bring up the guide and toggle something as a favourite (or move to another group) just as you can change a recordings recgroup from PBB
[17:59:41] jpabq: stuartm, completely agree.
[18:04:48] davide_ (davide_!~david@mythtv/developer/gigem) has quit (Remote host closed the connection)
[18:05:11] davide_ (davide_!~david@mythtv/developer/gigem) has joined #mythtv
[18:09:56] danielk22: taylorr: In what context? You would get pretty awful video playback if you used DTS as the presentation timestamp.
[18:20:19] kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has joined #mythtv
[18:41:28] xavierh: stuartm: I just converted channel group to MythUI ;( also I aggree it should be done in the program guide instead
[18:42:43] xavierh: Before I convert recording profiles to MythUI, any thought ? :)
[18:43:19] stuartm: xavierh: don't bother
[18:43:34] stuartm: I did try to warn that some of these screens probably wouldn't survive to 0.26, either they end up as part of backend setup or they are absorbed into other bits of the UI
[18:44:00] xavierh: Dont bother to convert ? how/where is it going to be done ? Service ApI ?
[18:44:17] stuartm: right
[18:44:49] stuartm: it's already been moved to the backend
[18:44:51] skrock (skrock! has joined #mythtv
[18:44:54] stuartm: s/backend/mythtv-setup/
[18:45:00] xavierh: Ok, what about chqnnel grouping, should I delete my changes?
[18:45:50] stuartm: xavierh: keep a copy handy, our plans to move that functionality into the guide might not work out
[18:46:38] stuartm: I doubt it, but there's no point throwing work away until it's actually happened
[18:46:41] xavierh: stuartm: Ok I will leave it like this, but I won't tidy it then. and concentrate in polishing the rest
[18:49:18] xavierh: stuartm: Does mythtv-setup will still use QWidget for 0.26 ?
[18:51:25] stuartm: if the http setup stuff is complete then I think the plan is to remove mythtv-setup entirely
[18:51:54] stuartm: at some future date someone (probably me) will write a new mythui setup application using the new API
[18:58:05] xavierh: that's a good news. Does any of the dev use 0.25 in production? I am using a old version (0.22-????) on ubuntu 8.04. I have been hold back by driver problem and lazyness. When I was about to upgrade, 0.25 was on sight.
[19:03:17] kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has quit (Quit: kormoc)
[19:03:26] stuartm: xavierh: I do
[19:05:22] xavierh: You just gave me the green light :)
[19:10:02] stuartm: just don't blame me if it all goes horribly wrong
[19:10:43] stuartm: there is an issue of some sort with the db upgrade and certain versions of mysql, might be best to check with sphery about that first
[19:20:28] knightr: Is there anything else beside mythmessage that uses the UDP notify port? (there used to be mythtvosd (which has been replaced by mythmessage) and mythudprelay)
[19:21:17] wagnerrp: thats the only one i saw when i recently redid all that stuff
[19:21:42] wagnerrp: technically, mythmessage is due for removal, and included in mythutil
[19:21:53] knightr: only one I saw too...
[19:22:01] knightr: but not for 0.25 though?
[19:22:45] wagnerrp: mythmessage has not existed in any release version yet
[19:23:06] knightr: wagnerrp, the reason I ask is that we have this string in the setup: "MythTV will listen for connections from the "mythtvosd" or "mythudprelay" programs on this port. For additional information, see ."
[19:23:12] wagnerrp: it was created during this cycle, and deprecated during this cycle
[19:23:25] wagnerrp: i *think* capt'm plans to remove the binary before we actually release
[19:23:40] wagnerrp: mythnotify, mythtvosd, and mythudprelay are all dead
[19:23:55] wagnerrp: the interface has changed, and they are no longer usable
[19:24:43] knightr: yep, I saw that they were dead and was getting ready to replace them with mytmessage but I didn't know it was included in mythutil...
[19:26:25] knightr: So the best bet would be to ask Captain_Murdoch what he intends to do with mythmessage I guess before modifying that message (or only mention mythutil...)
[19:27:52] knightr: Does mythutil still use the UDP notify port though? I haven't checked...
[19:28:35] knightr: yep, it does still use it...
[19:31:22] taylorr: danielk22: actually, from what I've read, when the decoder releases a frame it's pts should equal to the dts of the last packet sent to the the decoder :)
[19:31:50] taylorr: and we used to explicitly do pts = pkt->dts which seemed to always be spot on wrt av-sync
[19:32:25] taylorr: these issues always make it seem like we are stuck between a rock and a hard place
[19:33:03] Captain_Murdoch: yeah, we need to get rid of mythmessage. /me makes another note of something to do pre-0.25.
[19:34:29] taylorr: it just seems like more often that the pts is busted or missing and dts is usually solid
[19:35:42] wagnerrp: knightr: it uses the same port, but the message syntax is different, significantly simplified
[19:36:07] taylorr: danielk22: check these links out -> . . . accurate-pts and it references
[19:36:10] wagnerrp: mythnotify was much more capable, with the fatal flaw it only worked during playback
[19:36:43] natanojl (natanojl! has joined #mythtv
[19:37:21] taylorr: danielk22: that ffmpeg tutorial mentions "ffmpeg reorders the packets so that the DTS of the packet being processed by avcodec_decode_video() will always be the same as the PTS of the frame it returns"
[19:38:17] taylorr: danielk22: so basically the decoder doesn't release a frame with a certain pts until a packet with a matching dts enters the decoder
[19:39:01] knightr: wagnerrp, thank you! I have changed the message to only mention mythutil. There's no use in mentionning mythmessage if it's going to die...
[19:39:34] wagnerrp: i need to fix the help message for mythutil --message
[19:39:45] wagnerrp: its a bit screwy, and doesnt at all explain how to use it
[19:39:57] ** knightr will try to fix his backend adter getting these little fixes in... **
[19:40:27] knightr: wagnerrp, will you add a wiki page like the old MythNotfy stuff had too?
[19:40:45] wagnerrp: you have to use '--message_text' to tell it what to send
[19:41:12] wagnerrp: and the only way you know to do so would be to print out the '--print-template', and make the unintuitive leap thats the way it works
[19:43:29] knightr: hmm, you're right it's not that clear...
[19:46:02] wagnerrp: Captain_Murdoch: do you know if there were ever any plans to allow arbitrary theme code to be passed over the udp listener?
[19:46:16] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has joined #mythtv
[19:46:25] wagnerrp: as it exists now, the static message and timeout is very restrictive
[19:50:37] stuartm: wagnerrp: arbitrary, probably not, but a better range of predefined templates ...
[19:52:35] wagnerrp: perhaps something to hook into the plugin api, allow them to define new templates?
[19:56:30] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has quit (Ping timeout: 252 seconds)
[20:31:18] k-man: jya and stuartm, by what mechanism will direct access to the database be prevented in future versions?
[20:32:06] stuartm: the database will be embedded, no more mysql server
[20:32:17] k-man: stuartm, ah thanks
[20:32:30] k-man: stuartm, will it be embeded mysql?
[20:32:49] k-man: or some other db?
[20:33:33] noahric (noahric!~noahric@ has joined #mythtv
[20:33:58] stuartm: yes, for performance reasons and also because it's simpler to stick with mysql instead of having to re-write tables and dozens of queries to conform to whatever version of SQL other DBMS might implement
[20:35:37] k-man: thanks stuartm
[20:35:45] Beirdo: stuartm: where did those warnings come from?
[20:36:12] xris: stuartm: we could switch to an ORM! *ducks*
[20:36:13] stuartm: Beirdo: gcc 4.6.x, the BSD buildbot I think
[20:36:28] Beirdo: ahh
[20:36:43] skrock (skrock! has quit (Quit: WeeChat 0.3.6)
[20:37:15] stuartm: . . . nings%20(28)
[20:37:34] Beirdo: hmmm
[20:37:39] TheAsp (TheAsp! has joined #mythtv
[20:37:46] Beirdo: looks valid too
[20:38:21] TheAsp: I was supposed to remind someone that I had a patch for MPEG4 over firewire but I forget who... It's in this ticket:
[20:38:43] Beirdo: argh. how to fix.
[20:38:54] stuartm: xris: and then we can replace that with whatever happens to be in vogue next week
[20:38:59] Beirdo: gonna need to think a touch on that one :)
[20:39:16] stuartm: TheAsp: danielk22?
[20:39:17] xris: stuartm: oooh! that means we replace the backend with node.js.  :)
[20:39:21] Beirdo: stuartm: let's use /dev/null for our database. It's web-scale :)
[20:39:33] Beirdo: extremely fast writes
[20:39:39] Beirdo: forget about reads though
[20:39:48] xris: Beirdo: they're very fast, too
[20:39:53] Beirdo: hehe, true
[20:40:01] TheAsp: stuartm: Yeah, i thought it started with "d"
[20:40:33] TheAsp: I'll ask him when he wakes up
[20:42:53] k-man: stuartm, would you be able to point out to me the main areas where shepherd causes issues with mythtv? I'd like to understand the problem a little
[20:45:21] stuartm: k-man: well the most recent one is channel icons, it's downloading them itself and sticking them in a custom directory where mythtv in master cannot find them, it should be putting them in the xml and letting mythfilldatabase download them
[20:46:08] k-man: stuartm, ah yeah i saw that ticket
[20:46:53] stuartm: k-man: it's run via a cron job and inserts xml by calling mythfilldatabase --xmlfile instead of allowing mythtv to schedule the run (that's not causing problems right now, it's just not how it should work)
[20:47:49] stuartm: it's operating directly on the database for some stuff, e.g. the icons (I've not checked to see what else, but I doubt that's the only time it does it)
[20:49:07] k-man: stuartm, i use shepherd here, and the cron job is how i run it too
[20:49:16] stuartm: AFAIK it's no-where close to being an xmltv compliant grabber meaning it won't be able to support apiconfig when that's added, so the user will always have to do additional setup/configuration outside mythtv
[20:49:36] k-man: stuartm, i see
[20:50:10] Beirdo: stuartm: OK, I'll have a fix in for that warning shortly. Just rearranged the loops/ifs to get the same logic but differently.
[20:50:15] MaverickTech (MaverickTech! has quit (Ping timeout: 260 seconds)
[20:50:45] stuartm: k-man: might be worth reading through this thread for more –
[20:51:07] k-man: stuartm, thanks for the link
[20:51:42] stuartm:
[20:52:49] stuartm: that post is actually a reply to that comment in the ticket, but sphery responds pointing out why some of the assumptions being made by the Shepherd dev are wrong
[20:56:14] stuartm: k-man: mostly the thing that bothers me is that everyone else works with the xmltv project and within the spec to produce grabbers, there are something like 30+ different xmltv grabbers for 20+ countries distributed through the official xmltv release alone and countless more through unofficial sites
[20:57:05] k-man: right – i wonder why shepherd devs haven't gone down that route
[20:57:49] stuartm: my guess? NIH (not invented here)
[20:59:25] k-man: well, i posted a polite pointer to the mailing list that direct access to the db will be going away
[20:59:36] k-man: lets see the responce
[20:59:41] k-man: the shepherd mailing list that is
[21:01:32] stuartm: someone might like to point out the icon issue, that it should be using <channel id="{xmltvid}"><icon>{blah}</icon></channel>  ;– per
[21:02:21] stuartm: instead of downloading the icons to a custom location and inserting the urls directly into the database
[21:03:36] k-man: right
[21:10:29] MaverickTech (MaverickTech! has joined #mythtv
[21:15:23] k-man: stuartm, so do you intend to fix #10404? (the icons)
[21:18:43] stuartm: k-man: I can't decide, if I build in a workaround for Shepherd then there's no reason for the Shepherd devs to start doing things properly and that seems backwards
[21:19:13] k-man: stuartm, imho you shuold just say won't fix
[21:19:35] k-man: i beleive the shepherd devs will make shepherd work with that
[21:22:19] k-man: hah!
[21:22:57] k-man: anyway, gotta have a shower
[21:23:07] k-man: thanks for the info stuartm
[21:23:58] wagnerrp: Beirdo: you going to be around for a bit?
[21:24:02] eugo (eugo!~none@unaffiliated/eugo) has joined #mythtv
[21:26:38] wagnerrp: well im about to replace the freebsd buildbot... hopefully all goes well
[21:30:51] stuartm: new hardware, or just a software upgrade?
[21:31:10] wagnerrp: update the jail its installed to
[21:31:26] wagnerrp: freebsd 8.1 base --> freebsd 9.0 base, plus a bunch of missing dependencies
[21:34:00] Beirdo: yeah, I'll be here
[21:34:19] wagnerrp: good, because i need you to add a '--disable-vdpau'
[21:35:19] Beirdo: in the config for freebsd?
[21:35:51] wagnerrp: yeah, theres an ambiguous function overload issue that doesnt need to be resolved just this moment
[21:36:14] wagnerrp: compatibility issue with abs()
[21:36:26] Beirdo: heh
[21:36:43] Beirdo: that's core only, not in the plugins, right?
[21:36:48] wagnerrp: right
[21:36:59] Beirdo: K, one moment
[21:38:00] wagnerrp: ill be a couple minutes yet
[21:39:40] MythBuild (MythBuild! has quit (Quit: buildmaster reconfigured: bot disconnecting)
[21:39:53] Beirdo: there :)
[21:39:55] MythBuild (MythBuild! has joined #mythtv
[21:40:01] MythBuild: build #865 of master-vista-mingw-32bit is complete: Exception [exception setproperty] Build details are at . . . t/builds/865 blamelist: Gavin Hurlbut < >
[21:40:20] Beirdo: hehe, sorry vista
[21:40:44] Beirdo: MythBuild force build master-vista-mingw-32bit oops
[21:41:00] Beirdo: MythBuild: force build master-vista-mingw-32bit oops
[21:41:01] MythBuild: build #866 forced
[21:41:01] MythBuild: I'll give a shout when the build finishes
[21:41:01] MythBuild: build #866 of master-vista-mingw-32bit is complete: Exception [exception setproperty] Build details are at . . . t/builds/866
[21:41:04] wagnerrp: changing a property on one bot forces a rebuild on all the rest?
[21:41:07] Beirdo: oO
[21:41:12] Beirdo: I restarted the bot
[21:41:26] wagnerrp: oh, you have to restart the master for it to take effect
[21:41:28] Beirdo: I meant to do reconfig
[21:41:32] Beirdo: and I did restart
[21:41:33] Beirdo: oops
[21:42:38] Beirdo: MythBuild: force build master-vista-mingw-32bit oops
[21:42:39] MythBuild: build #867 forced
[21:42:39] MythBuild: I'll give a shout when the build finishes
[21:42:48] Beirdo: and now restarted that fragile slave
[21:43:08] wagnerrp: im going to move the build directory over to one of the 2TB disks
[21:43:38] wagnerrp: not surprisingly, even the slow 2TB green is considerably faster than the 2x320GB 7200RPM mirror its currently building on
[21:44:03] wagnerrp: what clock is your i7?
[21:44:27] Beirdo: 2.8GHz
[21:44:57] wagnerrp: interesting to see how it compares, 2.8 i7 vs. 3.3 PII
[21:45:28] Beirdo: as we are compiling -j1, they are likely gonna be fairly close
[21:45:29] wagnerrp: once its not taking several minutes just to do a git pull
[21:45:37] wagnerrp: yeah
[21:48:24] stuartm: PII?
[21:48:33] wagnerrp: phenom II
[21:48:43] wagnerrp: !pentium
[21:48:48] Beirdo: hehe
[21:48:56] stuartm: ah, well that's not the least bit confusing ;)
[21:49:03] Beirdo: didn't think Intel made 3.3GHz PII
[21:49:21] stuartm: they didn't, hence the confusion :[
[21:49:34] Beirdo: even so, if they did, same general results as we are compiling on a single core
[21:49:45] stuartm: highest speed PII was 450Mhz according to wikipedia
[21:49:52] Beirdo: OK, I need drugs. stupid headaches
[21:50:58] stuartm: Paramax works for me, nothing fancy, just paracetamol mixed with something else that causes it to be absorbed much quicker
[21:52:41] natanojl (natanojl! has quit (Ping timeout: 244 seconds)
[21:53:06] stuartm: although since I've been on drugs for my blood pressure the migraines have almost entirely gone
[21:58:55] warped (warped!~piotro@ has quit (Quit: warped)
[21:58:57] Beirdo: I should get back on the pressure meds, they really did help
[22:04:15] wagnerrp: Beirdo: i should be able to copy this buildbot.tac, and it will handle everything else?
[22:07:13] mrand (mrand!~mrand@ubuntu/member/mrand) has quit (Ping timeout: 260 seconds)
[22:11:57] wagnerrp: MythBuild: force build master-freebsd-64bit
[22:12:00] MythBuild: build #2071 forced
[22:12:00] MythBuild: I'll give a shout when the build finishes
[22:12:18] wagnerrp: crap, no... didnt set up ccache yet
[22:12:20] MythBuild: Hey! build master-freebsd-64bit #2071 is complete: ?? [build successful exception slave lost]
[22:12:20] MythBuild: Build details are at . . . /builds/2071
[22:14:07] MythBuild: Hey! build master-vista-mingw-32bit #867 is complete: Success [build successful]
[22:14:08] MythBuild: Build details are at . . . t/builds/867
[22:17:16] MythBuild: build #2072 forced
[22:17:16] MythBuild: I'll give a shout when the build finishes
[22:18:39] Beirdo: hah, it took both
[22:19:25] Beirdo: your slave tac should just work yeah
[22:19:55] wagnerrp: took both?
[22:20:15] Beirdo: I got confused and thought both those were vista builds
[22:20:19] Beirdo: I think I need a nap
[22:21:04] wagnerrp: does it use a separate repo for 0.24 and fixes? or is it smart enough to share one between them?
[22:21:37] wagnerrp: looks like separate for each
[22:21:41] Beirdo: it will check it out twice
[22:22:05] Beirdo: oh, I think I only made that change in master
[22:22:46] Beirdo: no, it's shared code
[22:22:48] Beirdo: never mind
[22:23:16] wagnerrp: does it internally do a 'new-workdir' or something?
[22:25:20] Beirdo: not sure off hand, but I think it just clones and fetch/pull
[22:27:34] wagnerrp: ive got all the bits needed for lame/faac/xvid/x264/vpx if you want to enable them at some point
[22:29:01] Beirdo: K. I think I'll take a nap first, let this headache bugger off
[22:34:56] Dr{Wh0} (Dr{Wh0}! has joined #mythtv
[22:35:28] Dr{Wh0}: Is it possible to have mythfrontend automaticly start playing live tv when it starts and not begin in the menu?
[22:36:54] Dr{Wh0}: hmm think i found a solution using xsendkeycode
[22:38:42] kth (kth!~kth@unaffiliated/kth) has quit (Quit: Leaving.)
[22:40:24] jya: danielk22: thanks for your answer.. would it be worthwhile to add a mutex/lock just around the initialisation of the array?
[22:42:59] jeff999 (jeff999! has joined #mythtv
[22:46:28] xavierh_ (xavierh_! has quit (Ping timeout: 276 seconds)
[22:54:27] joki (joki! has quit (Ping timeout: 252 seconds)
[22:54:58] joki (joki! has joined #mythtv
[23:04:06] MythBuild: Hey! build master-freebsd-64bit #2072 is complete: Success [build successful]
[23:04:06] MythBuild: Build details are at . . . /builds/2072
[23:06:22] highzeth (highzeth! has quit (Ping timeout: 260 seconds)
[23:08:00] highzeth (highzeth! has joined #mythtv
[23:10:02] wagnerrp: MythBuild: force build master-freebsd-64bit
[23:10:08] MythBuild: The build has been queued, I'll give a shout when it starts
[23:11:02] danielk22: jya: no the mutex lock needs to be around the use of the array's members.
[23:12:23] jya: ok… would that add too much overhead? I'm guessing the likelyhood of a problem to be quite low… as once the variable has been fully intialised (which only occurs once, and is rather speedy)
[23:12:39] MythBuild: build forced [ETA 27m41s]
[23:12:39] MythBuild: I'll give a shout when the build finishes
[23:12:47] mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv
[23:14:18] jams (jams! has quit (Ping timeout: 272 seconds)
[23:18:22] danielk22: jya: Don't worry about this. It's only used to decode the short name of the DVB channel when recording debugging is enabled or during channel scan. During channel scan it's always safe. During debugging a lot of other things can go sideways anyway.
[23:19:37] taylorr: danielk22: did you catch my earlier messages?
[23:20:07] danielk22: TheAsp: We're in feature freeze right now.. also jpabq knows more about the H.264 keyframe detection than I do.
[23:20:13] MythBuild: Hey! build master-freebsd-64bit #2074 is complete: Success [build successful]
[23:20:13] MythBuild: Build details are at . . . /builds/2074
[23:20:17] jya: ok, so I can just move the code… If you're interested the issue is discussed in ticket 9447. The only aim here was to delay the first call to QTextCodec::codecForName until after the QApplication has been created.
[23:22:13] danielk22: jya: yeah. delaying the first call is fine. At some point we'll need to worry about thread-safety, but at this point the function is little used.
[23:22:27] jya: ok… cool
[23:22:58] wagnerrp: MythBuild: force build .024-freebsd-64bit
[23:22:59] MythBuild: no such builder '.024-freebsd-64bit'
[23:23:27] wagnerrp: MythBuild: force build 0.24-freebsd-64bit
[23:23:28] MythBuild: build #124 forced
[23:23:29] MythBuild: I'll give a shout when the build finishes
[23:25:39] TheAsp: danielk22: Awesome... Know when it's over?
[23:29:36] danielk22: taylorr: DTS is the decode order, so a video like so IBP will have a decode order of IPB. So the pts and dts of those frames would be something like {10,11,12} and the dts would be {8,10,9}
[23:30:27] danielk22: taylorr: If you are only looking at the I frames DTS values are great because they can be generated even when absent in the file. But if you are looking at P and B frames then the DTS values are not monotonic.
[23:32:09] danielk22: TheAsp: April 3rd
[23:35:27] danielk22: taylorr: I think we should try to use PTS.. but if it looks bogus fall back on DTS. Using one or the other will just result in different files failing.
[23:38:14] danielk22: taylorr: PS I'm assuming here dts/pts in libav correspond to dts/pts values in MPEG... This should be true with MPEG files and simulated with other file types.. but I honestly don't know the current state of the libav we're using.
[23:41:00] danielk22: wagnerrp: Isn't f5aa5e835 a bit too large of a change for the eve of the beta release? What was this fixing exactly? Should we be delaying the beta a few weeks for this to settle?
[23:42:27] wagnerrp: danielk22: i know, i was a bit leery committing it
[23:42:43] wagnerrp: functionally, theres very little difference between the two, its more of a move
[23:43:04] wagnerrp: but the issue it resolved was that mythmessage was designed to operate off broadcast communication
[23:43:23] wagnerrp: however there was no mechanism to actually tell it what the broadcast address to listen to was
[23:43:45] wagnerrp: as a result, it would just blindly use, which wont even let you bind to it
[23:43:49] wagnerrp: at least not on freebsd
[23:45:18] wagnerrp: there shouldnt be any side effects from it, aside from what was already performed by the initial version two weeks ago
[23:46:13] ThisNewGuy1 (ThisNewGuy1! has joined #mythtv
[23:46:27] TheAsp: danielk22: awesome. i'll probably put a new patch in to match the current repository then
[23:46:29] danielk22: wagnerrp: so all it really does is provide a way to say in IPv6 ?
[23:47:00] wagnerrp: danielk22: right now, my network is a 10.254/22
[23:47:22] wagnerrp: this tells mythudplistener to listen on, rather than trying and failing to listen on
[23:47:30] ThisNewGuy (ThisNewGuy! has quit (Ping timeout: 252 seconds)
[23:50:10] wagnerrp: as for ipv6, there is no real broadcast address
[23:50:19] danielk22: wagnerrp: what's you confidence level? has anyone else been running this?
[23:50:48] wagnerrp: the server pool stuff is all the same, which has been in use for 2 weeks
[23:51:04] wagnerrp: this is just the routine that runs once on startup and selects the addresses to use
[23:51:06] danielk22: wagnerrp: I thought there was a defacto broadcast address everyone listened on w/IPv6
[23:51:58] wagnerrp: there is no broadcast at all, just multicast
[23:52:14] wagnerrp: there is the 'all nodes' multicast at ff02::1, which is a potential option
[23:53:02] danielk22: so you need to subscribe to it like IPv4 multicast? And hosts don't subscribe to ff02::1 by default?
[23:53:21] wagnerrp: not sure
[23:53:35] wagnerrp: it wasnt using ipv6 before, and i havent added it myself yet
[23:56:22] danielk22: ok. I think we should delay the beta until tomorrow at least so everyone gets a chance to run this, but it sounds pretty safe. We'll worry about IPv6 broadcast/multicast later.. I doubt many are running IPv6 only locally yet.
[23:57:24] wagnerrp: ok, as far as confidence is concerned, im confident in saying that any bugs in there currently were in there before the commit
[23:58:02] wagnerrp: its the same selection code, it just adds an outer loop so the associated broadcast address is saved as well
[23:58:38] wagnerrp: since instead of looping through addresses directly, it has to loop through addresses associated with interfaces

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