MythLogBot@irc.freenode.net :: #mythtv

Daily chat history

Current users (78):

aloril, amessina, andreaz, Anssi, arrrghhh, bhaak, brfransen, caelor, Captain_Murdoch, ChanServ, Chutt_, clever, coling, CyberJacob, dblain, dekarl, ElmerFudd, enyc, erhandsome, gary_buhrmaster, GhostOfRaven, ghoti, gigem, gregbert, gregL, GreyFoxx, Guest24265, Hydr0p0nX, J-e-f-f-A, jab416171, jams_, jheizer, jnylen, joki, jpabq, jpharvey, jst_, jya, kc, kurre2, lautriv__, libsci, markspieth, Merlin83b, MythBuild, MythLogBot, nephyrin`, peper03, poptix, pppingme, purserj, purserj_, rich0, saaki, Seeker`, seld, sheedy-away, skd5aner, sphery_, SpootDev, sraue, stuarta, stuartm, suffice, superm1, taylorr, tgm4883, Tobbe5178, tonsofpcs, tris, unforgiven512, vincent42, wagnerrp, Warped, XDS2010, xris, _charly_, _iwc
Sunday, January 10th, 2016, 00:57 UTC
[00:57:57] dizygoth_ (dizygoth_!~dizygothe@cpc15-stre5-2-0-cust147.1-1.cable.virginm.net) has quit (Quit: Leaving)
[01:18:53] cotterpin (cotterpin!~hp-mini@202-180-109-177.callplus.net.nz) has left #mythtv ()
[01:36:09] arescorpio (arescorpio!~arescorpi@90-239-16-190.fibertel.com.ar) has joined #mythtv
[02:08:46] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has quit (Ping timeout: 240 seconds)
[03:20:50] Hydr0p0nX (Hydr0p0nX!~hydr@66-191-154-200.dhcp.mtgm.al.charter.com) has quit (Ping timeout: 256 seconds)
[03:21:44] ShapeShifter499 (ShapeShifter499!~ShapeShif@unaffiliated/shapeshifter499) has quit (Remote host closed the connection)
[04:40:38] arescorpio (arescorpio!~arescorpi@90-239-16-190.fibertel.com.ar) has quit (Quit: Leaving.)
[04:44:46] joki (joki!~joki@p5B36DC0D.dip0.t-ipconnect.de) has quit (Ping timeout: 265 seconds)
[04:49:37] joki (joki!~joki@p5B36D868.dip0.t-ipconnect.de) has joined #mythtv
[05:29:19] Roklobotomy (Roklobotomy!~blah@ppp118-209-87-144.lns20.mel4.internode.on.net) has joined #mythtv
[05:41:39] Hydr0p0nX (Hydr0p0nX!~hydr@24-177-179-236.dhcp.leds.al.charter.com) has joined #mythtv
[06:16:30] Hydr0p0nX (Hydr0p0nX!~hydr@24-177-179-236.dhcp.leds.al.charter.com) has quit (Ping timeout: 256 seconds)
[08:05:41] dekarl (dekarl!~dekarl@mythtv/developer/dekarl) has quit (Quit: Leaving.)
[08:29:07] dekarl (dekarl!~dekarl@mythtv/developer/dekarl) has joined #mythtv
[08:43:56] amessina (amessina!~amessina@unaffiliated/amessina) has quit (Quit: Konversation terminated!)
[08:51:51] dekarl1 (dekarl1!~dekarl@mythtv/developer/dekarl) has joined #mythtv
[08:54:53] dekarl (dekarl!~dekarl@mythtv/developer/dekarl) has quit (Ping timeout: 276 seconds)
[08:57:02] dekarl1 is now known as dekarl
[09:14:52] skd5aner (skd5aner!~skd5aner@218.sub-70-198-71.myvzw.com) has quit (Ping timeout: 250 seconds)
[09:15:26] skd5aner (skd5aner!~skd5aner@218.sub-70-198-71.myvzw.com) has joined #mythtv
[09:43:08] Hydr0p0nX (Hydr0p0nX!~hydr@66-191-154-200.dhcp.mtgm.al.charter.com) has joined #mythtv
[10:01:18] SteveGoodey (SteveGoodey!~steve@host109-156-128-140.range109-156.btcentralplus.com) has joined #mythtv
[10:08:47] Roklobotomy (Roklobotomy!~blah@ppp118-209-87-144.lns20.mel4.internode.on.net) has quit (Remote host closed the connection)
[10:23:49] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has joined #mythtv
[10:28:51] natanojl: jya: VAAPI is broken for that VC1 sample in #12570 as well, but it does not crash. I don't know if it is the correct fix but checking for PIX_FMT_YUV420P in IS_VAAPI_PIX_FMT(enum PixelFormat fmt) allows playback
[10:28:51] ** MythLogBot http://code.mythtv.org/trac/ticket/12570 **
[10:29:21] jya: natanojl: the bug appears to be in the creation of the context, and that's our code
[10:29:33] jya: likely something simple, like constant name change
[10:29:36] jya: i'm on it
[10:29:59] jya: natanojl: doing that will bypass vaapi
[10:30:17] jya: falling back to sw decoder
[10:31:00] natanojl: jya: Ok.
[10:32:52] jya: problem my only available dev box is a poor mac mini with 600MB left on its 20GB SSD
[10:33:11] jya: can't even have the source code on it, so i mounted a disk over nfs
[10:33:15] jya: so damn so
[10:33:19] jya: slow
[10:34:58] jya: it's very weird that bug, in mythplayer videouOutput isn't set
[10:35:02] jya: how could that be
[10:36:52] natanojl: I checked the FFMPEG code earlier and it looked like ff_get_format is called from vc1_decode_init which is different from say H264
[10:37:19] jya: natanojl: the issue here is that the videoout hasn't been called yet
[10:37:33] jya: so when we ask to create the hardware decode context, it fails
[10:37:45] natanojl: jya: Yes, I realised that too
[10:37:50] jya: looks like ff_get_format is called much earlier because it's vc1
[10:38:18] jya: maybe i simply need to move the creating of the videoutout up
[10:38:23] jya: all that part is bad anyway
[10:38:42] jya: all other players create the video player withint ff_get_format
[10:38:59] jya: and that's where they decide to use vdpau or vaapi or whatever
[10:39:25] jya: we do things the other way round, so you must decide first what output we're going to do. and we can't have decoding fallback
[10:39:48] jya: if we properly used the ffmpeg API as its intended to be, we wouldn't even need video profiles
[10:39:54] jya: we could have all that automated
[10:40:19] jya: when I looked at fixing it, the task was so big it fell under my assumption that I would be better off rewriting the entire thing
[10:44:50] natanojl: jya: Then we came to the same conclusion about ff_get_format being called earlier. I noticed though that it would be called again at some point. Would it be enough for VDPAU to simply check avctx->hwaccel_contex before assigning render2?
[10:45:33] jya: natanojl: yeah, it's just being called during the scan stream process
[10:45:42] jya: I think just ignoring that will fix it
[10:46:09] jya: (that's exactly what my test is going)
[10:46:21] jya: that box is so slow... gotta wait to finish compiling
[10:46:28] natanojl: ok
[10:46:54] jya: likely the same for vaapi or any other places where we override get_format
[10:47:22] jya: yep that worked
[10:47:59] natanojl: great
[10:48:06] jya: dxva should be good, it doesn't attempt to do anything with the value returned
[10:48:31] jya: you're saying it fails with vaapi ?
[10:48:43] jya: that can't be the same .
[10:49:03] jya: does vaapi supports vc1 normally?
[10:51:26] natanojl: jya: Yes, it fails but VAAPI should support VC1, at least on my system. http://pastebin.com/BqjfqVSB
[10:51:56] jya: ok... going to reboot that laptop under linux then... it has an intel gpu.
[10:53:29] natanojl: no crash though, only that avcodec_open2 returns < 0 in AvFormatDecoder::OpenAVCodec()
[10:55:21] jya: ok.. thanks for the hint.. can you open a bug with the details you have and assign it to me ?
[10:56:04] natanojl: jya: It says it is using VAAPI if I allow PIX_FMT_YUV420P in IS_VAAPI_PIX_FMT(enum PixelFormat fmt) like I wrote before
[10:56:27] natanojl: jya: sure
[10:58:00] jya: natanojl: maybe it's a bug in ffmpeg then
[11:03:01] natanojl: jya: Perhaps
[11:24:05] jya: back on linux
[11:24:18] jya: much faster laptop
[11:26:12] natanojl: jya: It looks similar to the VDPAU problem. get_format_vaapi is called before the video player has been created so avctx->hwaccel_context is never set even though the format was PIX_FMT_VAAPI_VLD and therefore ff_vaapi_context_init fails
[11:27:04] jya: yes, but if it's anything like vdpau, it should be called again
[11:27:38] natanojl: it is
[11:27:49] jya: so it doesn't matter if it fails the first time
[11:28:03] jya: will tell when I'm done compiling
[11:35:52] jya: damn, no space left on device there too
[11:39:36] natanojl: Doh
[11:39:48] natanojl: jya: Created #12593 for you
[11:39:48] ** MythLogBot http://code.mythtv.org/trac/ticket/12593 **
[11:40:37] jya: at least I have a build of firefox I can clean, that will free 8Gig
[11:59:15] lautriv__ (lautriv__!~lautriv@funtoo/user/lautriv) has quit (Ping timeout: 240 seconds)
[12:04:35] jya: oh I see why returning that profile worked
[12:05:01] jya: in any case, the code is bad, because the ffmpeg documentation clearly states that you can't ever return PIX_FMT_NONE
[12:05:43] natanojl: hehe
[12:12:16] lautriv__ (lautriv__!~lautriv@funtoo/user/lautriv) has joined #mythtv
[12:28:14] dekarl: man, whats so hard in writing "php7 deprecates mysql, stillsupported is mysql access via option1 and option2"... wrt #12588
[12:28:14] ** MythLogBot http://code.mythtv.org/trac/ticket/12588 **
[12:28:34] dekarl: and http://php.net/manual/en/migration70.removed-exts-sapis.php
[12:29:24] dekarl: s/migration guide/note of removal/
[13:15:57] Warped (Warped!~Warped@unaffiliated/warped) has quit (Quit: ChatZilla 0.9.92 [Firefox 43.0.4/20160105164030])
[13:18:39] Warped (Warped!~Warped@unaffiliated/warped) has joined #mythtv
[14:05:31] dekarl: Does it make sense to build out VAAPI support when OpenGL wasn't detected by configure? http://paste.ubuntu.com/14458123/
[14:05:35] dekarl: s/out/our/
[14:08:30] kukks (kukks!~Guenter@samba/team/kukks) has joined #mythtv
[14:44:18] jya: dekarl: no, you can't do vaapi without opengl
[14:45:18] jya: right... any playback problem I should be looking at in master?
[14:50:43] gregbert (gregbert!24baec23@unaffiliated/gregbert) has quit (Ping timeout: 260 seconds)
[14:57:40] gregbert (gregbert!80630232@unaffiliated/gregbert) has joined #mythtv
[15:38:03] andreaz (andreaz!~andre_000@p4FC56A39.dip0.t-ipconnect.de) has joined #mythtv
[15:42:27] kukks (kukks!~Guenter@samba/team/kukks) has quit (Ping timeout: 255 seconds)
[15:50:20] kukks (kukks!~Guenter@ip5b42ce25.dynamic.kabel-deutschland.de) has joined #mythtv
[15:50:20] kukks (kukks!~Guenter@ip5b42ce25.dynamic.kabel-deutschland.de) has quit (Changing host)
[15:50:20] kukks (kukks!~Guenter@samba/team/kukks) has joined #mythtv
[16:12:20] skd5aner (skd5aner!~skd5aner@218.sub-70-198-71.myvzw.com) has quit (Ping timeout: 276 seconds)
[16:14:56] skd5aner (skd5aner!~skd5aner@54.sub-70-198-73.myvzw.com) has joined #mythtv
[16:52:23] peper03: Anyone know off the top of their heads why/if this line is really necessary? https://github.com/MythTV/mythtv/blob/master/ . . . ile.cpp#L809
[16:53:07] peper03: I haven't investigated it in detail but resetting the socket is apparently a very slow operation.
[16:54:36] peper03: If I try to play a DVD image from a storage group, it takes four seconds of seeking and reading before I get the playback popup asking me if I want to start from the bookmark or the beginning.
[16:54:57] peper03: If I comment out that line, it takes about 400ms!
[17:25:44] kukks (kukks!~Guenter@samba/team/kukks) has quit (Ping timeout: 256 seconds)
[17:37:43] gary_buhrmaster (gary_buhrmaster!~gtb@2601:647:4801:98e3:222:4dff:fe51:6728) has quit (Quit: Leaving.)
[17:40:10] gary_buhrmaster (gary_buhrmaster!~gtb@2601:647:4801:98e3:222:4dff:fe51:6728) has joined #mythtv
[17:43:58] amessina (amessina!~amessina@unaffiliated/amessina) has joined #mythtv
[17:55:15] jpharvey (jpharvey!~jpharvey@host81-159-19-182.range81-159.btcentralplus.com) has quit (Ping timeout: 240 seconds)
[18:01:08] natanojl: peper03: The MythSocket::ResetReal() code looks a little odd, but it probably doesn't matter. Why not use a while loop there? Now it could be that bytesAvailable() returns zero after the first waitForReadyRead() but we would try to read anyway
[18:05:52] peper03: natanojl: Yeah, I'm just looking at it now. The calls to waitForReadyRead are what's causing the slow-down. There's no data to be read anyway, so those two calls guarantee the function will take at least 60ms to return.
[18:06:31] peper03: Without those two calls, it takes 0.02ms to return.
[18:07:49] peper03: I don't really get why we should wait anyway. I suppose it's to handle the case of 'what about if data is just coming in?', but then, what happens if data comes in 31ms after the last call? Same problem.
[18:09:00] peper03: And I don't understand why it's necessary to call 'reset' anyway after seeking.
[18:09:52] natanojl: So how could it take four seconds if there is no data to be read? Is it that the socket thread is busy and the QMetaObject::invokeMethod blocks?
[18:10:24] peper03: No, it's because 'seek' is called many times. Every time it's called, it incurs a 60ms delay.
[18:11:46] peper03: Ah, ok. Now I see why it's being called. The 'SEEK' command is sent over the control socket, but the 'reset' call is performed on the data socket.
[18:11:58] peper03: That makes sense (even if the 30ms delays don't necessarily).
[18:14:31] natanojl: Ah, right
[18:20:11] paul-h (paul-h!~Paul@94.6.148.224) has joined #mythtv
[18:33:45] paul-h: Well that didn't last long LVR has quit
[18:36:44] knightr (knightr!~knightr@mythtv/developer/knightr) has quit (Ping timeout: 272 seconds)
[18:39:29] jpharvey (jpharvey!~jpharvey@host109-151-98-192.range109-151.btcentralplus.com) has joined #mythtv
[18:43:00] knightr (knightr!~knightr@69-165-170-178.dsl.teksavvy.com) has joined #mythtv
[18:43:00] knightr (knightr!~knightr@mythtv/developer/knightr) has joined #mythtv
[18:43:00] knightr (knightr!~knightr@69-165-170-178.dsl.teksavvy.com) has quit (Changing host)
[18:44:12] peper03: Because of the question of whether to support old versions of QT?
[18:49:27] tgm4883: paul-h: really?
[18:53:06] natanojl: paul-h: He contacted you directly? I did see that he removed some of his branches
[18:54:00] paul-h: Yeah I'll forward his email to the developers list in a minute
[18:54:16] kukks (kukks!~Guenter@samba/team/kukks) has joined #mythtv
[18:56:59] natanojl: meh
[19:03:40] gary_buhrmaster: I apologize if my comments in that (supported old versions) thread pushed lvr out.
[19:04:49] knightr: gary_buhrmaster, I agree with what you said
[19:05:04] knightr: I also agree with some of the things he said too...
[19:05:45] knightr: if people don't want to upgrade, run an old MythTV version, it won't stop running the day we release a new one...
[19:06:46] knightr: and one you are ready to upgrade they will be able to (at least on the MythTV side), some of us made sure of that...
[19:07:13] natanojl: gary_buhrmaster: I also agree with your comments
[19:08:02] gary_buhrmaster: knightr: It is rare someone is completely right or wrong in non fact based discussions. I did not disagree with the desire for support, just the realities seemed wrong to me.
[19:09:24] knightr: gary_buhrmaster, at least he was willing to try to improve those areas which some of us said could not be improved...
[19:09:36] knightr: a fresh new look from someone else is good sometimes...
[19:10:14] knightr: he could have found something the others didn't think of...
[19:12:08] knightr: we don't have the manpower to do that
[19:12:11] gary_buhrmaster: knightr: In most of the patches (which I just skimmed), I do not recall anyone ever said that the area could not be improved, just that there were no resources. lvr provided some resources. Which (usually) has value (as long as the basic philosophy of the project is maintained).
[19:12:15] lincoln-r (lincoln-r!~lincoln-r@203-114-182-185.dsl.sta.inspire.net.nz) has joined #mythtv
[19:12:46] knightr: gary_buhrmaster, some of us said some things had to be redone from the ground up...
[19:14:34] gary_buhrmaster: And some things should (likely) be redone from the ground up. But that is (admittedly) a bigger effort which is even more un-resourced. I have had the same challenge myself from time to time ("this stuff sucks, time to rewrite", and that never gets done).
[19:16:39] knightr: yes, but not now...
[19:23:44] lincoln-r: In an effort to retain contributors in the project, is it worth emailing to try convince him not to just quit?
[19:25:16] lincoln-r (lincoln-r!~lincoln-r@203-114-182-185.dsl.sta.inspire.net.nz) has quit (Quit: Bye)
[19:25:23] kukks (kukks!~Guenter@samba/team/kukks) has quit (Ping timeout: 264 seconds)
[19:25:33] lincoln-r (lincoln-r!~lincoln-r@203-114-182-185.dsl.sta.inspire.net.nz) has joined #mythtv
[19:26:34] kukks (kukks!~Guenter@ip5b42ce25.dynamic.kabel-deutschland.de) has joined #mythtv
[19:26:34] kukks (kukks!~Guenter@ip5b42ce25.dynamic.kabel-deutschland.de) has quit (Changing host)
[19:26:34] kukks (kukks!~Guenter@samba/team/kukks) has joined #mythtv
[19:29:57] knightr: lincoln-r, considering how fast he decided to leave I am not quite sure... if he had stayed it would not be the first nor the last time some people would have disagreed with him... everybody is entitled to his own opinions and it's not all black and white, there are gray areas...
[19:44:15] lincoln-r: Yes you're probably right there.
[19:49:26] kukks (kukks!~Guenter@samba/team/kukks) has quit (Ping timeout: 276 seconds)
[20:11:08] stuartm: for any project to work there has to be give and take, and quitting at the first sign of resistance is not really indicative of the right temperament
[20:12:53] taylorr: stuartm: I think perceived attitude probably is more of an issue than just a disagreement
[20:13:26] stuartm: of his stated reasons, it really comes down to the discussion over maintaining QT4 support in 0.28 (even though support was dropped a year ago)
[20:15:40] taylorr: stuartm: what is your opinion of splitting the development of the backend and frontend? it seems that kodi and other users of the services api can release new versions and not be tied to a backend version
[20:18:25] stuartm: well, they can and they can't – at least in the past some of xbmc's backwards compatibility came from simply not using those features which had changed between releases or worse faking support for newer protocol versions by repeatedly guessing at the correct protocol version number until it established a connection
[20:19:06] taylorr: good point
[20:19:08] stuartm: I'm not really sure what splitting frontend and backend development would really mean
[20:19:47] stuartm: potentially it would tell backend developers that they no longer needed to keep the frontend compatible with new features, which would only make things worse
[20:23:17] stuartm: the services API is supposed to be more stable that the internal protocol, so in some respects at least, transitioning parts of the frontend to use it might help users running older frontends with new backends but maybe not the other way around, all depends how new frontend features are implemented i.e. whether they are required to check for support in the backend they are connected to
[20:23:29] lincoln-r (lincoln-r!~lincoln-r@203-114-182-185.dsl.sta.inspire.net.nz) has quit (Ping timeout: 245 seconds)
[20:24:24] stuartm: which btw, I'm all in favour of as something we do going forward so that 0.29 could be compatible with 0.28 and vice versa
[20:25:24] stuartm: that's the sort of backwards compatibility I wholeheartedly support
[20:26:14] stuartm: just not the kind where I required Blu-rays to be playable on my DVD player because I only bought my DVD player three years ago
[20:28:10] taylorr: that's a whole other area that needs to be revisited at some point – removal of antiquated technologies that only complicated moving forward
[20:31:31] taylorr: stuartm: given your expertise on the gui side of things, is getting rid of qt painter and only supporting opengl make things much simpler?
[20:34:08] stuartm: it wouldn't hurt, there are issues with the QT painter that we've neither the time nor inclination to work on fixing, it doesn't get in the way of things in the same way that the XVideo support does (complicates so much of the OSD rendering and playback code)
[20:35:00] stuartm: but if we switched to only supporting OpenGL we'd not have to worry about breaking QT painter support even further
[20:41:06] taylorr: cool, good to know it's one more thing that could be removed – hopefully it never turns into the Xv removal drama so that people could use their ancient hardware at the expense of progress
[20:43:00] dekarl: taylorr: going opengl only (and some related things) allow people to run "the real mythfrontend" on 35$ hardware as opposed to an order of magnitude more today. So hardware deprecation isn't such an issue anymore (at least IMHO)
[20:43:02] gary_buhrmaster: Was the XV code ever deleted? I thought that was supposed to happen with .026, and since then it has just languished.
[20:43:38] taylorr: I don't think it was removed
[20:45:39] taylorr: dekarl: it's amazing what cheap hardware is capable of these days – I've got a Asrock Beebox N3150 and it's video capability is incredible (only $100 plus memory and drive)
[20:55:26] Roklobotomy (Roklobotomy!~blah@ppp118-209-87-144.lns20.mel4.internode.on.net) has joined #mythtv
[21:02:54] jpharvey (jpharvey!~jpharvey@host109-151-98-192.range109-151.btcentralplus.com) has quit (Remote host closed the connection)
[21:29:04] kukks (kukks!~Guenter@samba/team/kukks) has joined #mythtv
[21:30:32] knightr: jpabq, paul-h was there any new addition to the themes since 0.27 which I need to take into account for their translation (like a new XML file, etc...)?
[21:30:34] knightr: Thank you!
[21:38:00] paul-h: knightr: there is the new gallery theme file in themes/default/image-ui.xml
[21:39:08] paul-h: There are few new windows in MythMusic but they are all in existing theme files IIRC
[21:59:55] SteveGoodey (SteveGoodey!~steve@host109-156-128-140.range109-156.btcentralplus.com) has quit (Quit: Konversation terminated!)
[22:14:45] SteveGoodey (SteveGoodey!~steve@host109-156-128-140.range109-156.btcentralplus.com) has joined #mythtv
[22:20:11] SteveGoodey (SteveGoodey!~steve@host109-156-128-140.range109-156.btcentralplus.com) has quit (Ping timeout: 246 seconds)
[22:20:13] Steve-Goodey (Steve-Goodey!~steve@host109-156-128-140.range109-156.btcentralplus.com) has joined #mythtv
[22:22:17] knightr_ (knightr_!~knightr@69-165-170-178.dsl.teksavvy.com) has joined #mythtv
[22:22:17] knightr (knightr!~knightr@mythtv/developer/knightr) has quit (Read error: Connection reset by peer)
[22:22:28] Steve-Goodey (Steve-Goodey!~steve@host109-156-128-140.range109-156.btcentralplus.com) has quit (Client Quit)
[22:26:36] knightr__ (knightr__!~knightr@69-165-170-178.dsl.teksavvy.com) has joined #mythtv
[22:27:16] knightr_ (knightr_!~knightr@69-165-170-178.dsl.teksavvy.com) has quit (Ping timeout: 272 seconds)
[22:27:40] jpharvey (jpharvey!~jpharvey@host109-151-98-192.range109-151.btcentralplus.com) has joined #mythtv
[22:30:10] knightr__ is now known as knightr
[23:01:17] paul-h (paul-h!~Paul@94.6.148.224) has quit (Quit: Konversation terminated!)
[23:10:02] Roklobotomy (Roklobotomy!~blah@ppp118-209-87-144.lns20.mel4.internode.on.net) has quit (Remote host closed the connection)
[23:53:23] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has quit (Ping timeout: 255 seconds)
[23:57:40] cybrNaut (cybrNaut!cybrNaut@unaffiliated/cybrnaut) has quit (Ping timeout: 260 seconds)

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