| Wednesday, July 13th, 2011, 00:01 UTC | ||
| [00:01:38] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [00:04:05] | kc (kc!~Casper@unaffiliated/kc) has joined #mythtv | |
| [00:14:51] | _charly_ (_charly_!kroseneg@sunrise.schmidham.net) has quit (Excess Flood) | |
| [00:19:04] | andreax1 (andreax1!~andreaz@p57B94355.dip.t-dialin.net) has joined #mythtv | |
| [00:20:15] | andreax (andreax!~andreaz@p57B941F0.dip.t-dialin.net) has quit (Ping timeout: 276 seconds) | |
| [00:28:17] | andreax1 (andreax1!~andreaz@p57B94355.dip.t-dialin.net) has quit (Quit: Leaving.) | |
| [00:40:30] | davide_ (davide_!~david@mythtv/developer/gigem) has quit (Remote host closed the connection) | |
| [00:40:58] | davide_ (davide_!~david@host103.16.intrusion.com) has joined #mythtv | |
| [00:40:58] | davide_ (davide_!~david@host103.16.intrusion.com) has quit (Changing host) | |
| [00:40:58] | davide_ (davide_!~david@mythtv/developer/gigem) has joined #mythtv | |
| [00:53:26] | dekarl (dekarl!~dekarl@dslb-084-058-100-069.pools.arcor-ip.net) has quit (Ping timeout: 264 seconds) | |
| [00:54:57] | Mousey (Mousey!~wtfisme@ross154.net) has quit (Ping timeout: 255 seconds) | |
| [01:32:05] | jcarlos (jcarlos!~quassel@85.137.99.76.dyn.user.ono.com) has quit (Read error: Connection reset by peer) | |
| [01:34:04] | jcarlos (jcarlos!~quassel@85.137.99.76.dyn.user.ono.com) has joined #mythtv | |
| [01:52:16] | MaverickTech (MaverickTech!~MaverickT@dns2.arel.com.au) has joined #mythtv | |
| [02:00:33] | _charly_ (_charly_!kroseneg@sunrise.schmidham.net) has joined #mythtv | |
| [02:13:19] | quentusrex_ (quentusrex_!~quentusre@freeswitch/developer/quentusrex) has quit (Ping timeout: 258 seconds) | |
| [02:26:25] | quentusrex_ (quentusrex_!~quentusre@freeswitch/developer/quentusrex) has joined #mythtv | |
| [02:34:18] | kc (kc!~Casper@unaffiliated/kc) has quit (Ping timeout: 252 seconds) | |
| [02:36:52] | xris: | hmm, I think mythweb needs a *lot* of work on the video side..still queries the db directly for everything. |
| [02:47:30] | kc (kc!~Casper@unaffiliated/kc) has joined #mythtv | |
| [02:54:30] | xris: | any reason the video scanner imports pdf files? |
| [02:59:10] | iamlindoro: | it only imports the extensions you've defined |
| [03:00:46] | xris: | there. delete 100 lines of code, and it works better. :) |
| [03:00:55] | xris: | (mythweb video scanning, that is) |
| [03:01:21] | iamlindoro: | You may want to look at BackendServerIP:6544/Video/wsdl |
| [03:01:29] | iamlindoro: | to look at the XML/JSON video API |
| [03:01:40] | iamlindoro: | with which you can get all the files in the DB, delete, add, etc. |
| [03:03:32] | iamlindoro: | (or if you don't do wsdl (I don't), you can look at programs/mythbackend/services/video.h) |
| [03:04:07] | xris: | never configured anything. we must have some defaults somewhere? |
| [03:05:07] | iamlindoro: | any what, sorry? |
| [03:05:28] | xris: | I'm talking about SCAN_VIDEOS importing a bunch of pdf files |
| [03:05:50] | iamlindoro: | PDF isn't one of the defaults, the extensions imported are determined by the File Types dialog in the frontend |
| [03:06:11] | iamlindoro: | Setup->Media Settings->Video->File Types |
| [03:06:13] | xris: | yeah, I've never opened that, so it must have pulled them from somewhere. |
| [03:06:20] | iamlindoro: | Not the code |
| [03:06:29] | iamlindoro: | Someone put it in, 'cause it's nowhere in the code |
| [03:07:11] | xris: | know waht the db table is? frontend is busy being watched at the moment. |
| [03:07:26] | iamlindoro: | videotypes |
| [03:07:30] | xris: | yeah, found it. |
| [03:07:34] | xris: | odd. it has txt and stuff, too. |
| [03:08:08] | xris: | and no pdf |
| [03:08:43] | iamlindoro: | Ah, it's possible you have it set to show unknown filetypes |
| [03:08:49] | iamlindoro: | which affects the scanner too |
| [03:09:02] | iamlindoro: | Which means it will import EVERYTHING you point it at |
| [03:09:14] | xris: | ah, yeah. I have txt in as ignore, so that explains it. |
| [03:09:22] | iamlindoro: | VideoListUnknownFiletypes in settings |
| [03:09:25] | xris: | must have been left over from before you cleaned that stuff up. |
| [03:09:35] | iamlindoro: | specifically, for the host your MBE is on |
| [03:09:45] | xris: | only one box so that's easy |
| [03:10:03] | iamlindoro: | set that to 0 and I suspect your scanning will be more as expected |
| [03:10:39] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [03:11:03] | xris: | yeah |
| [03:11:46] | iamlindoro: | TBH I can't remember if that setting even still has a UI... definitely a good candidate for deletion, doesn't serve much purpose |
| [03:11:54] | iamlindoro: | besides breaking stuff |
| [03:12:22] | xris: | I vaguely remember a period where videotypes wasn't working right. or I was debugging something and needed it on. |
| [03:13:11] | iamlindoro: | Anyhoo, MythWeb may be able to use, eg, BackendServerIP:6544/Video/GetVideos?Descending=false&StartIndex=0&Count =50 (Returns XML or JSON for the first 50 videos) |
| [03:13:44] | xris: | yeah, but that requires more work than I have time for tonight. replacing the scanner was quick/easy. |
| [03:15:56] | xris: | mythweb really just needs to be redone from the ground up now that the backend has a real API |
| [03:16:40] | xris: | given more people working on the python stuff, doing so in python sounds more likely, despite certain earlier statements about "never" from people like me. |
| [03:20:51] | jpabq: | Latest master is giving me a lot of http://pastebin.com/jrDaViia |
| [03:22:29] | jpabq: | Heh, Michael Drons just posted the same problem to the list. |
| [03:23:37] | Beirdo: | yeah. I guess we should get that into a ticket so it can be tracked down. |
| [03:27:43] | MaverickTech (MaverickTech!~MaverickT@dns2.arel.com.au) has quit (Ping timeout: 240 seconds) | |
| [03:28:42] | jpabq: | Beirdo, do you allow commflag jobs to run as soon as a recording starts? If not, that may be why you don't see the problem. |
| [03:29:38] | Beirdo: | no, I wait until the recording's done. That could well be a distinction |
| [03:37:57] | kc (kc!~Casper@unaffiliated/kc) has quit (Ping timeout: 264 seconds) | |
| [03:49:48] | kc (kc!~Casper@unaffiliated/kc) has joined #mythtv | |
| [03:52:54] | jpabq: | Beirdo, even with --daemon, mythbackend spits "Error in my_thread_global_end(): 1 threads didn't exit" out to the console. |
| [03:54:07] | Beirdo: | yeah |
| [03:54:16] | Beirdo: | that's not under our control directly |
| [03:54:32] | Beirdo: | that's the mysql driver spewing... |
| [03:54:54] | jpabq: | Ah. Okay. |
| [03:55:13] | jpabq: | Why was it not a problem before? |
| [03:55:21] | Beirdo: | the core issue is how we handle our mysql connections. It has to be disconnected by the same thread that connects, and we have a couple spots where it doesn't |
| [03:55:26] | Beirdo: | it was at times |
| [03:55:56] | Beirdo: | but it's more predictably like that now, part of which was the db logging, but I have that shutting down properly AFAIK now |
| [03:56:09] | Beirdo: | try with --nodblog to see if that's still the issue |
| [03:56:43] | Beirdo: | the other one that sometimes catches us is the ProgramInfoUpdater |
| [03:56:48] | jpabq: | Yeah, I am actually using --nodblog |
| [03:57:01] | Beirdo: | K, so it won't be the db logging |
| [03:57:09] | jpabq: | mythmetadatalookup is running, so that is probably it |
| [03:58:10] | iamlindoro: | jpabq: again? |
| [03:58:10] | Beirdo: | the ProgramInfoUpdater sometimes is that one thread that has a db connection and doesn't shut it down. So when you quit, the connection gets shut down by the core thread in the dtor for the context |
| [03:58:35] | jpabq: | iamlindoro, going through the 3 steps on the wiki. |
| [03:58:43] | iamlindoro: | jpabq: oh, ok, gotcha |
| [03:58:52] | iamlindoro: | jpabq: was worried the housekeeper had somehow gone nuts |
| [03:58:57] | jpabq: | :) |
| [03:59:17] | Beirdo: | I think sphery was planning to look into the idea of redoing the db connection handling to make QSQLDatabase and mysql driver happier |
| [04:39:54] | tris (tris!tristan@2001:1868:a00a::4) has quit (Excess Flood) | |
| [04:43:10] | noisymouse (noisymouse!~david@pool-71-106-10-205.lsanca.dsl-w.verizon.net) has quit (Quit: Leaving) | |
| [04:45:31] | tris (tris!tristan@2001:1868:a00a::4) has joined #mythtv | |
| [04:57:41] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has joined #mythtv | |
| [05:09:14] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has quit (Ping timeout: 252 seconds) | |
| [05:09:39] | taylorr (taylorr!~taylorr@unaffiliated/elmojo) has quit (Ping timeout: 258 seconds) | |
| [05:10:06] | taylorr (taylorr!~taylorr@cpe-173-095-144-220.nc.res.rr.com) has joined #mythtv | |
| [05:10:06] | taylorr (taylorr!~taylorr@unaffiliated/elmojo) has joined #mythtv | |
| [05:10:06] | taylorr (taylorr!~taylorr@cpe-173-095-144-220.nc.res.rr.com) has quit (Changing host) | |
| [05:14:06] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has joined #mythtv | |
| [05:16:48] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has quit (Client Quit) | |
| [05:16:48] | dekarl (dekarl!~dekarl@dslb-084-058-100-069.pools.arcor-ip.net) has joined #mythtv | |
| [05:20:07] | davide_ (davide_!~david@mythtv/developer/gigem) has quit (Remote host closed the connection) | |
| [05:22:33] | davide_ (davide_!~david@mythtv/developer/gigem) has joined #mythtv | |
| [05:24:36] | wagnerrp (wagnerrp!~wagnerrp_@mythtv/developer/wagnerrp) has quit (Ping timeout: 260 seconds) | |
| [05:24:48] | wagnerrp (wagnerrp!~Wagner@nr-ft1-66-42-241-137.fuse.net) has joined #mythtv | |
| [05:37:15] | eharris (eharris!~eharris@99-179-7-82.lightspeed.austtx.sbcglobal.net) has joined #mythtv | |
| [05:40:57] | quentusrex_ (quentusrex_!~quentusre@freeswitch/developer/quentusrex) has quit (Ping timeout: 252 seconds) | |
| [05:42:26] | andreax (andreax!~andreaz@p57B94355.dip.t-dialin.net) has joined #mythtv | |
| [05:42:47] | MythBuild: | build #480 of master-freebsd-64bit is complete: Failure [failed git] Build details are at http://code.mythtv.org/buildbot/builders/mast . . . t/builds/480 blamelist: Mark Kendall <mkendall@mythtv.org > |
| [05:48:27] | iamlindoro: | *git update* failed? |
| [05:50:36] | wagnerrp: | Beirdo: do you know if github has AAAA records? |
| [05:51:43] | Beirdo: | not sure |
| [05:51:52] | andreax (andreax!~andreaz@p57B94355.dip.t-dialin.net) has quit (Quit: Leaving.) | |
| [05:52:02] | wagnerrp: | thats my guess |
| [05:52:05] | Beirdo: | !dig github.com aaaa |
| [05:52:05] | MythLogBot: | github.com AAAA: No results |
| [05:52:15] | Beirdo: | !dig github.com a |
| [05:52:15] | MythLogBot: | github.com A: 207.97.227.239 |
| [05:52:16] | wagnerrp: | i got disconnected a while back because my tunnel went down |
| [05:52:28] | dlblog (dlblog!~dlblog@c-76-127-227-175.hsd1.ma.comcast.net) has quit (Ping timeout: 250 seconds) | |
| [05:52:31] | Beirdo: | looks like no AAAA to me |
| [05:53:54] | dlblog (dlblog!~dlblog@c-76-127-227-175.hsd1.ma.comcast.net) has joined #mythtv | |
| [05:57:16] | MythBuild: | build #481 of master-freebsd-64bit is complete: Exception [exception interrupted] Build details are at http://code.mythtv.org/buildbot/builders/mast . . . t/builds/481 blamelist: Robert McNamara <rmcnamara@mythtv.org > |
| [05:57:29] | wagnerrp: | how could you robert |
| [05:58:43] | Beirdo: | heh, restarted it? |
| [05:58:54] | wagnerrp: | yeah |
| [05:59:20] | iamlindoro: | Compilus Interruptus |
| [05:59:39] | iamlindoro: | Two back to back WTF buildbot messages |
| [06:00:04] | wagnerrp: | how do you force a build? |
| [06:03:54] | Beirdo: | MythBuild: force build master-freebsd-64bit now |
| [06:03:56] | MythBuild: | build forced [ETA 8m31s] |
| [06:03:56] | MythBuild: | I'll give a shout when the build finishes |
| [06:03:59] | Beirdo: | like that :) |
| [06:09:00] | jmartens (jmartens!~jmartens@109.232.42.33) has joined #mythtv | |
| [06:12:32] | jmartens (jmartens!~jmartens@109.232.42.33) has quit (Read error: Connection reset by peer) | |
| [06:12:56] | jmartens (jmartens!~jmartens@109.232.42.33) has joined #mythtv | |
| [06:14:13] | jmartens (jmartens!~jmartens@109.232.42.33) has quit (Client Quit) | |
| [06:19:27] | xris: | Beirdo: your logging changes (or something) play weird stuff with init script usage. |
| [06:19:43] | xris: | I restrart the backend and suddenly get a bunch of logs on the console from the backgrounded process |
| [06:19:50] | MythBuild: | Hey! build master-freebsd-64bit #482 is complete: Success [build successful] |
| [06:19:50] | MythBuild: | Build details are at http://code.mythtv.org/buildbot/builders/mast . . . t/builds/482 |
| [06:24:13] | Beirdo: | xris: you'd have to be more specific as to which stuff is showing up |
| [06:24:59] | Beirdo: | we can work on that though. It could be something I had fixed since, not sure how recent your build is |
| [06:35:24] | Beirdo: | and also, I'd need to know what your command line is :) |
| [06:59:27] | Captain_Murdoch (Captain_Murdoch!~cpinkham@mythtv/developer/CaptainMurdoch) has joined #mythtv | |
| [07:09:41] | dlblog (dlblog!~dlblog@c-76-127-227-175.hsd1.ma.comcast.net) has quit (Read error: Connection reset by peer) | |
| [07:10:39] | dlblog (dlblog!~dlblog@c-76-127-227-175.hsd1.ma.comcast.net) has joined #mythtv | |
| [07:10:42] | Unhelpful (Unhelpful!~quassel@rockbox/developer/Unhelpful) has quit (Ping timeout: 255 seconds) | |
| [07:10:50] | Unhelpful (Unhelpful!~quassel@rockbox/developer/Unhelpful) has joined #mythtv | |
| [07:12:00] | xris: | wagnerrp: how do you make mythcommflag behave like it did before when run without args? |
| [07:14:56] | xris: | I think this preview gen issue is killing commflagging. running mythcommflag is my usual way to see what's stuck in the queue. |
| [07:15:12] | dlblog (dlblog!~dlblog@c-76-127-227-175.hsd1.ma.comcast.net) has quit (Ping timeout: 255 seconds) | |
| [07:57:20] | Giannirs4 (Giannirs4!~Giannirs4@dslb-084-058-137-214.pools.arcor-ip.net) has joined #mythtv | |
| [07:57:39] | Giannirs4 (Giannirs4!~Giannirs4@dslb-084-058-137-214.pools.arcor-ip.net) has left #mythtv () | |
| [08:39:04] | kth (kth!~kth@unaffiliated/kth) has joined #mythtv | |
| [08:39:38] | Giannirs4 (Giannirs4!~Giannirs4@dslb-084-058-137-214.pools.arcor-ip.net) has joined #mythtv | |
| [09:01:11] | gandalfcome (gandalfcome!~gandalfco@mithrandir.anu.edu.au) has joined #mythtv | |
| [09:15:04] | Giannirs4 (Giannirs4!~Giannirs4@dslb-084-058-137-214.pools.arcor-ip.net) has quit (Ping timeout: 252 seconds) | |
| [09:28:57] | mrand (mrand!~mrand@ubuntu/member/mrand) has quit (Ping timeout: 264 seconds) | |
| [09:47:48] | kurre_ (kurre_!~tomimo@xdsl-83-150-88-111.nebulazone.fi) has quit (Ping timeout: 255 seconds) | |
| [09:49:10] | mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv | |
| [09:53:20] | jcarlos_ (jcarlos_!~quassel@85.137.99.76.dyn.user.ono.com) has joined #mythtv | |
| [10:05:01] | mike|2 (mike|2!~mike@c-24-21-63-118.hsd1.or.comcast.net) has quit (Remote host closed the connection) | |
| [10:05:31] | jcarlos (jcarlos!~quassel@85.137.99.76.dyn.user.ono.com) has quit (Ping timeout: 255 seconds) | |
| [10:05:33] | kurre_ (kurre_!~tomimo@xdsl-83-150-88-111.nebulazone.fi) has joined #mythtv | |
| [10:05:56] | mike|2 (mike|2!~mike@c-24-21-63-118.hsd1.or.comcast.net) has joined #mythtv | |
| [10:36:03] | andreax (andreax!~andreaz@p57B94355.dip.t-dialin.net) has joined #mythtv | |
| [11:41:32] | davide_ (davide_!~david@mythtv/developer/gigem) has quit (Remote host closed the connection) | |
| [11:41:42] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [11:41:58] | davide_ (davide_!~david@mythtv/developer/gigem) has joined #mythtv | |
| [11:47:24] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection) | |
| [11:50:23] | paul-h (paul-h!~Paul@5adce259.bb.sky.com) has joined #mythtv | |
| [11:50:23] | paul-h (paul-h!~Paul@5adce259.bb.sky.com) has quit (Changing host) | |
| [11:50:23] | paul-h (paul-h!~Paul@mythtv/developer/paul-h) has joined #mythtv | |
| [12:04:03] | andreax (andreax!~andreaz@p57B94355.dip.t-dialin.net) has quit (Read error: Connection reset by peer) | |
| [12:10:22] | MaverickTech (MaverickTech!~MaverickT@111.86.233.220.static.exetel.com.au) has joined #mythtv | |
| [12:35:12] | gandalfcome (gandalfcome!~gandalfco@mithrandir.anu.edu.au) has quit (Ping timeout: 276 seconds) | |
| [12:54:20] | PointyPumper (PointyPumper!~pintlezz@190.244.73.13) has quit (Ping timeout: 252 seconds) | |
| [13:00:51] | PointyPumper (PointyPumper!~pintlezz@190.244.73.13) has joined #mythtv | |
| [13:19:13] | Jordack (Jordack!~jordack@h69-131-44-221.mdsnwi.tisp.static.tds.net) has joined #mythtv | |
| [13:24:22] | abqjp (abqjp!~abqjp@97-119-166-191.albq.qwest.net) has joined #mythtv | |
| [13:56:36] | j-rod|afk is now known as j-rod | |
| [14:10:39] | zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [14:25:21] | JamesHarrison (JamesHarrison!~jharrison@cpc3-addl4-2-0-cust97.6-3.cable.virginmedia.com) has quit (Ping timeout: 260 seconds) | |
| [14:26:24] | JamesHarrison (JamesHarrison!~jharrison@cpc3-addl4-2-0-cust97.6-3.cable.virginmedia.com) has joined #mythtv | |
| [14:41:35] | davide_: | jya: bumping alsa prealloc to the max of 32768 and using vdpaubuffersize=50 gets rid of almost all of the audio underruns at 2x timestretch. they are virtually all gone at 1.95x. do you have any more suggestions to get that last 0.05x? or is 1.95x right at the limit of the decoding support on the gt430? |
| [14:42:06] | davide_: | jya: interestingly, the above changes all but fix the repeated audio problem. i wonder if the underlying problem for that is faulty buffer handling on partial writes. anyway, just trying to give you another data point in case you're working on that problem. |
| [14:48:11] | wagnerrp: | xris: you cant run it in its previous manner, where with no args, it would run through all recordings in order |
| [14:48:38] | wagnerrp: | however with --queue, it will add a job for each recording before terminating |
| [15:05:13] | danielk22 (danielk22!~danielk@96.57.9.142) has joined #mythtv | |
| [15:09:07] | danielk22: | OMG Did someone actually add over a dozen globals to mythcommflag in the last week?? Is it 1941 again? The advances of Cobol are still in the future?? |
| [15:10:12] | iamlindoro: | Careful, you're liable to get punchcarded |
| [15:16:50] | danielk22: | iamlindoro: it really looks quite broken.. i can't see anyone defending that code, more likely an "oops! i didn't mean to push that crap! it was just for testing.. SORRY EVERYBODY!!!!" kinda thing. ;] |
| [15:17:41] | danielk22: | Everyone eventually has a their much Scotch to commit moment. ;] |
| [15:18:27] | danielk22: | s/much/Too much/ |
| [15:34:29] | Jordack (Jordack!~jordack@h69-131-44-221.mdsnwi.tisp.static.tds.net) has quit () | |
| [15:40:50] | dlblog (dlblog!~dlblog@c-76-127-227-175.hsd1.ma.comcast.net) has joined #mythtv | |
| [15:53:08] | Mousey (Mousey!~wtfisme@ross154.net) has joined #mythtv | |
| [16:05:34] | dblain (dblain!~dblain@c-76-127-227-175.hsd1.ma.comcast.net) has joined #mythtv | |
| [16:05:34] | dblain (dblain!~dblain@mythtv/developer/dblain) has joined #mythtv | |
| [16:05:34] | dblain (dblain!~dblain@c-76-127-227-175.hsd1.ma.comcast.net) has quit (Changing host) | |
| [16:28:40] | kormoc is now known as kormoc_afk | |
| [16:40:37] | davide_ (davide_!~david@mythtv/developer/gigem) has quit (Remote host closed the connection) | |
| [16:41:05] | davide_ (davide_!~david@host103.16.intrusion.com) has joined #mythtv | |
| [16:41:05] | davide_ (davide_!~david@host103.16.intrusion.com) has quit (Changing host) | |
| [16:41:05] | davide_ (davide_!~david@mythtv/developer/gigem) has joined #mythtv | |
| [16:46:44] | kth (kth!~kth@unaffiliated/kth) has quit (Quit: Leaving.) | |
| [17:12:38] | Goga777 (Goga777!~Goga777@shpd-95-53-223-88.vologda.ru) has joined #mythtv | |
| [17:18:20] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has joined #mythtv | |
| [17:21:45] | kormoc_afk is now known as kormoc | |
| [17:31:39] | danielk22: | is the commflagger working from the command line at all anymore? It doesn't appear to provide any progress indication, it just exits after a few seconds. |
| [17:33:40] | Beirdo: | it will queue it for flagging |
| [17:33:56] | Beirdo: | if you want to flag a single file, it should still work with --chanid --starttime |
| [17:34:10] | danielk22: | I don't want to run it on a recording in the DB. |
| [17:34:19] | Beirdo: | I dunno, wagnerrp changed it around a few days back, and I'm still trying to figure out the changes |
| [17:34:33] | Beirdo: | --file I think is what you are looking for |
| [17:35:06] | danielk22: | Beirdo: it just doesn't work.. mythcommflag --skipdb --file /frogger1/video/4050_20110302233100.mpg ... nada |
| [17:35:23] | Beirdo: | oh. wagnerrp ^^^ |
| [17:35:32] | Beirdo: | not sure what all changed in that |
| [17:36:26] | Beirdo: | https://github.com/MythTV/mythtv/commit/baf6105331 |
| [17:36:33] | danielk22: | It looks like that commit reverted it to it's 1941 level of functionality :) (Assuming it's the same commit that added all the globals.) |
| [17:36:34] | Beirdo: | that was the change that likely changed it |
| [17:37:12] | wagnerrp: | danielk22, i tested the commflagger prior to commit, it worked fine |
| [17:37:30] | wagnerrp: | although it was a bit odd how long it took to terminate after the jobqueue said it hit 100% |
| [17:37:49] | danielk22: | wagnerrp: did you try it with --skipdb ? |
| [17:38:03] | wagnerrp: | --skipdb and --file, yes |
| [17:38:44] | wagnerrp: | right now, --skipdb only works when used with --file |
| [17:38:57] | danielk22: | Also.. is that the commit that introduced all those globals that alias the names of local variables? |
| [17:39:07] | wagnerrp: | no, globals were added during the logging stuff |
| [17:39:25] | danielk22: | Beirdo: ^^^^ |
| [17:39:39] | wagnerrp: | the only global i added was 'MythCommflagCommandLineParser cmdline' |
| [17:40:13] | danielk22: | wagnerrp: ok, maybe there is something specific to my files or setup, I'll dig into it. |
| [17:40:14] | Beirdo: | that file has had a lot of globals for quite some time |
| [17:40:33] | wagnerrp: | now something beirdo did mention, there is no longer support for using --file without --skipdb |
| [17:40:46] | wagnerrp: | where formerly, it used to look up the recording attached to that, and operate as normal |
| [17:40:49] | xris: | wagnerrp: so mythcommflag *should* work as before, processing everything and displaying stuff back to the console? |
| [17:40:51] | wagnerrp: | i intend to add that behavior back |
| [17:41:06] | wagnerrp: | xris: when running mythcommflag as part of the jobqueue, it displays nothing |
| [17:41:17] | Beirdo: | danielk22: look at a version from last August... just as many globals |
| [17:41:21] | mrand (mrand!~mrand@ubuntu/member/mrand) has quit (Ping timeout: 276 seconds) | |
| [17:41:36] | xris: | wagnerrp: when I type `mythcommflag` into bash, it prints out its help message and exits |
| [17:41:42] | Beirdo: | wagnerrp: I put an error in there for the time being so people don't get too confused |
| [17:41:51] | wagnerrp: | yes, mythcommflag no longer supports no options |
| [17:42:05] | xris: | how do I get it to process what's in the queue? |
| [17:42:14] | wagnerrp: | the queue? |
| [17:42:19] | xris: | of stuff to flag. |
| [17:42:30] | wagnerrp: | the jobqueue? |
| [17:42:31] | xris: | I occasionally get bad recordings that hold up the queue. |
| [17:42:43] | xris: | jobqueue, sure. but only talking about commercials |
| [17:42:51] | danielk22: | Beirdo: ok, I guess it just looked like they were added from the git blame. |
| [17:42:52] | Beirdo: | danielk22: see: https://github.com/MythTV/mythtv/blob/59f10b3 . . . lag/main.cpp |
| [17:42:59] | Beirdo: | that's from over a year ago |
| [17:43:10] | xris: | no idea which recording is the problem, so watching it process things via a shell is the easiest way to identify the issue |
| [17:43:26] | wagnerrp: | xris: the old behavior of mythcommflag with no options was to go through and run commercial detection on every recording with no skiplist |
| [17:43:28] | Beirdo: | yeah, gotcha |
| [17:43:34] | wagnerrp: | that behavior is no longer available |
| [17:43:46] | danielk22: | wagnerrp: without the --skipdb it appears to be doing something.. |
| [17:43:57] | wagnerrp: | however, you can give it a --queue option, where all of those recordings it would have run through, it instead queues to the jobqueue |
| [17:44:03] | xris: | which is what I'm experiencing now... none of my commercials are getting flagged, I think because of whatever is causing the preview pixmap problem. |
| [17:44:05] | wagnerrp: | and then the jobqueue slowly churns through them |
| [17:44:22] | xris: | wagnerrp: so how do you get the old behavior? |
| [17:44:25] | wagnerrp: | i also added --queue --rebuild, so the jobqueue will rebuild the seektables if needed |
| [17:44:44] | xris: | I don't know which files are causing the problem. this is my way to find out. |
| [17:44:46] | danielk22: | xris: previews have their own queue they shouldn't hold up flagging. |
| [17:44:50] | wagnerrp: | looping through all unflagged recordings in that specific process? |
| [17:44:55] | wagnerrp: | that behavior no longer exists |
| [17:44:58] | wagnerrp: | i didnt see a purpose to it |
| [17:45:09] | xris: | wagnerrp: I just gave you a purpose. :) |
| [17:45:26] | wagnerrp: | if you want to flag one specific recording, the --chanid/--starttime inputs work |
| [17:45:42] | xris: | wagnerrp: I don't know which recording to flag.. that's the problem. |
| [17:46:15] | xris: | if you remove the option to process, then could you at least add a way to list out what hasn't been done? maybe something like a queue/dryrun option? |
| [17:46:41] | xris: | danielk22: that was just my first step. I just happened to notice that the recordings without previews are also missing commercial flags |
| [17:46:43] | wagnerrp: | dryrun i could do |
| [17:47:41] | xris: | wagnerrp: something to restore the old behavior would be better (even an option). it's been a useful debugging tool for me |
| [17:48:13] | wagnerrp: | running through all the programs, a task that could take several days? |
| [17:48:22] | xris: | run through all that haven't yet been flagged |
| [17:49:04] | xris: | there was a force-all type thing before, too.. which does take days. I run it in screen. have used it before when my db got corrupt, though. |
| [17:49:10] | xris: | though that's more rare. |
| [17:49:21] | wagnerrp: | id rather add a new flag to the jobqueue to turn logging back on |
| [17:49:29] | wagnerrp: | or change the logging so it dumps to a new file |
| [17:49:51] | xris: | is there a way to run the job queue in a shell to see the output? |
| [17:51:08] | wagnerrp: | Beirdo: any way to bypass the lack of logging in the jobqueue? |
| [17:51:34] | xris: | let me rephrase it as a user story: commercials are no longer getting flagged, and I need a way to find out why. |
| [17:51:53] | xris: | digging through large log files after the fact would be VERY tedious |
| [17:52:17] | xris: | previously, I could run mythcommflag on the shell and wait for it to segfault. |
| [17:52:26] | xris: | and then would get an error I could use to dig deeper |
| [17:52:41] | wagnerrp: | segfaults should show up as such in the queue |
| [17:53:16] | Beirdo: | wagnerrp: huh? lack of logging in the jobqueue? It should log as it is |
| [17:53:36] | Beirdo: | assuming you start it with the logging turned on. I'd have to take a look |
| [17:53:57] | wagnerrp: | nevermind, xris is talking about something separate than the logging |
| [17:54:02] | Beirdo: | OK :) |
| [17:54:23] | Beirdo: | yeah, mythjobqueue defaults to general as a mask. it should work |
| [17:54:46] | xris: | wagnerrp: "in the queue"? I guess I can check the table.... |
| [17:54:48] | xris: | exit status 140, job status was "Running" |
| [17:55:17] | wagnerrp: | 140 means killed, manually terminated |
| [17:55:27] | danielk22: | wagnerrp: if I add "--method blank" to the command line I get the progress bar.. so it appears to just be a defaults issue. |
| [17:55:39] | sphery: | xris: why not just look at the frontend MythTV Status page, look at the job queue, and see what's stuck? |
| [17:55:48] | xris: | danielk22: gotcha. |
| [17:56:07] | xris: | sphery: because it's usually busy. |
| [17:56:09] | wagnerrp: | hmm... i dont recall what segfault show show up as |
| [17:56:16] | xris: | (in use by someone else) |
| [17:56:27] | sphery: | segfault would be 139 (128 + 11) |
| [17:56:50] | xris: | yeah, this might be something else. mythweb shows that one as a transcode job. |
| [17:56:54] | Beirdo: | #define GENERIC_EXIT_KILLED 140 ///< Process killed or stopped |
| [17:56:55] | sphery: | sig 12 = sigusr2, which doesn't make sense |
| [17:57:03] | Beirdo: | it's an exit code |
| [17:57:09] | sphery: | so, yeah, likely that was our exit code |
| [17:57:13] | wagnerrp: | segfault does show up as killed? |
| [17:57:19] | wagnerrp: | we should probably break out that specific |
| [17:57:32] | Beirdo: | why? |
| [17:57:37] | xris: | whatever this particular problem is, it's seemingly not a segfault |
| [17:57:42] | wagnerrp: | killed would mean manually termianted |
| [17:57:44] | Beirdo: | it's logged why it got killed |
| [17:57:50] | wagnerrp: | as opposed to faulted |
| [17:58:17] | Beirdo: | we don't really need to know the difference, do we? |
| [17:58:24] | xris: | danielk22: --method thing apparently doesn't work without providing file info |
| [17:58:43] | Beirdo: | I guess we could add back an ABORTED which would be manual abort/quit |
| [17:59:19] | xris: | wagnerrp: the more important question is why is stuff not getting added to the queue? |
| [17:59:47] | xris: | I just ran --queue and it added 6 things, from over the last 3–4 days |
| [17:59:51] | danielk22: | xris: sorry, that wasn't for your problem :P |
| [18:01:04] | Beirdo: | anyways, I got a meeting. YAY |
| [18:01:11] | wagnerrp: | xris: theres a big query at line 710, that checks for any commercial markup in a file |
| [18:01:28] | wagnerrp: | by default with queue, if there is any markup, it skips that recording, assuming it is already flagged |
| [18:01:51] | wagnerrp: | even if that previous flagging was bad or incomplete |
| [18:02:33] | RDV_Linux (RDV_Linux!~doug@CPE1caff7df6774-CM00252eac6f40.cpe.net.cable.rogers.com) has joined #mythtv | |
| [18:02:39] | wagnerrp: | other than that, if your channels are marked commercial free, it will skip those as well |
| [18:02:47] | wagnerrp: | that is new behavior |
| [18:02:58] | wagnerrp: | the old behavior was to ignore the channel property |
| [18:03:18] | xris: | wagnerrp: I ran `mythcommflag --queue` and it added 6 things. question is why there were any that needed to be added at all. |
| [18:03:29] | xris: | shouldn't the normal process have added them? |
| [18:03:51] | xris: | ok, did get a flag job fail with 140, too |
| [18:04:15] | wagnerrp: | were these things that have since been transcoded, and the flagging removed? |
| [18:04:21] | xris: | no |
| [18:04:24] | xris: | new recordings |
| [18:04:32] | wagnerrp: | yeah, you said last few days so i figured not |
| [18:05:39] | xris: | quick glance confirms that they're ones that don't have previews |
| [18:06:42] | danielk22: | wagnerrp: ok, I just committed something that gets us to the point where the flagging occurs with with --skipdb.. But I'm not seeing any output.. it's possible that it didn't find any ads though, need to test some bigger files. |
| [18:07:30] | ** xris fixes silly href="#" on the backend status page ** | |
| [18:07:34] | wagnerrp: | danielk22: what exactly line did you run? |
| [18:07:43] | wagnerrp: | *exact |
| [18:09:03] | danielk22: | after the fix: mythcommflag --skipdb --file ~/1001_1.1.0.mpg |
| [18:09:28] | wagnerrp: | hmm... i see what youre saying, it falls through to no flagging method, and drops out |
| [18:12:00] | danielk22: | wagnerrp: usually we run on the command line to check how a specific flagger deals with a file, so this limitation may have been in there a long time. i just wasn't thinking of the method since i was testing the command itself. |
| [18:12:22] | wagnerrp: | no, i rewrote the code that chose what method to use |
| [18:12:57] | wagnerrp: | it should be defaulting to COMM_DETECT_ALL |
| [18:13:00] | wagnerrp: | not sure why its not |
| [18:15:45] | danielk22: | wagnerrp: me neither.. but it looks like the output I'm looking for is commented out with "is there a purpose to this not fulfilled by --getskiplist?" |
| [18:16:21] | danielk22: | nah, maybe not.. not sure why i'm not seeing output. |
| [18:16:50] | wagnerrp: | yeah, onlyDumpDBCommercialBreakList sounds like its just grabbing the list out of the database and printing to screen |
| [18:16:56] | wagnerrp: | which is what --getskiplist already does |
| [18:17:11] | danielk22: | yep |
| [18:17:18] | wagnerrp: | since it returns before it actually runs any commflagging |
| [18:17:25] | mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv | |
| [18:19:46] | danielk22: | I debugging code always goes to shot when there are these big re-orgs.. |
| [18:19:48] | wagnerrp: | why would "blank" be the default for --skipdb? |
| [18:21:15] | danielk22: | wagnerrp: no real reason — just runs a little quicker, all is just as good... |
| [18:21:49] | wagnerrp: | do you mind if i move that down to the bottom as another else if? |
| [18:22:15] | danielk22: | wagnerrp: nah, just so long as it still works :) |
| [18:22:30] | wagnerrp: | since if --method is defined, default is ignored anyway, and it faults if it is not processed properly |
| [18:23:24] | Mousey (Mousey!~wtfisme@ross154.net) has quit (Remote host closed the connection) | |
| [18:24:15] | wagnerrp: | OH... blank doesnt mean no method, that means blank screen detection |
| [18:24:17] | wagnerrp: | right... |
| [18:24:30] | wagnerrp: | was confused by you were setting it to what i thought was no detection method |
| [18:24:56] | danielk22: | heh, no that wouldn't had the desired effect. |
| [18:28:20] | wagnerrp: | now something i dont understand, and i saw this happening with normal commercial flagging too |
| [18:28:28] | wagnerrp: | it hits 100%, and just sits there |
| [18:28:34] | xris: | lol. I broke the backend status page.. too many things in the queue. |
| [18:28:39] | wagnerrp: | i dont recall it previously doing that |
| [18:32:07] | xris: | ah, just mythweb's version of it. |
| [18:32:14] | xris: | stupid php |
| [18:33:49] | xris: | whoever is working on setup stuff, please use href="javascript:void(0)" instead of href="#" |
| [18:34:05] | xris: | or just remove the href parameter altogether. |
| [18:34:58] | wagnerrp: | i noticed that clicking the job dumped you back to the front page, is that what youre talking about? |
| [18:35:31] | xris: | wagnerrp: yeah, that's one of the effects of doing that |
| [18:35:39] | xris: | # is actually a valid link |
| [18:35:55] | xris: | and if you have a base href on the page, it'll send you there instead of taking you to a nonexistent anchr on the current page |
| [18:36:54] | xris: | used to be that A tags needed href, but you don't anymore. and things look fine as long as you are using css to set the style on them (without the href, the default link styles don't show, and they just look like normal text) |
| [18:37:37] | xris: | so, um.. how can I change my web setup password if it requires a password to get in? |
| [18:38:06] | wagnerrp: | i dont think theres a UI option for that yet |
| [18:38:55] | Captain_Murdoch: | you use the default password specified in the commit log. :) |
| [18:39:06] | Captain_Murdoch: | once in, you can change the password. |
| [18:39:21] | xris: | Captain_Murdoch: yeah. finding that specific commit is difficult. was hoping it'd be in the wiki. |
| [18:40:20] | xris: | wagnerrp: fwiw I just fixed the href="#" stuff on the status page.. but don't know where to look for the stuff driving the menus in web setup |
| [18:40:26] | xris: | git-grep didn't help much |
| [18:40:57] | Captain_Murdoch: | xris, I sent login/pw via /msg |
| [18:41:03] | xris: | thx |
| [18:41:15] | davide_ is now known as davide | |
| [18:42:23] | Goga777 (Goga777!~Goga777@shpd-95-53-223-88.vologda.ru) has quit (Remote host closed the connection) | |
| [18:42:24] | Captain_Murdoch: | default is buried in libmyth* somewhere since its' part of the webserver itself. |
| [18:43:11] | xris: | wagnerrp: nevermind, found them. |
| [18:43:53] | jmartens1 (jmartens1!~jmartens@s5597ca60.adsl.wanadoo.nl) has joined #mythtv | |
| [18:44:35] | jmartens2 (jmartens2!~jmartens@s5597ca60.adsl.wanadoo.nl) has joined #mythtv | |
| [18:46:53] | wagnerrp: | ugh... i hate how git pushes everything, and not just the working branch |
| [18:47:09] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has quit (Ping timeout: 240 seconds) | |
| [18:47:48] | Beirdo: | wagnerrp: there is a note on the UsingGit page on how to change that |
| [18:48:10] | jmartens1 (jmartens1!~jmartens@s5597ca60.adsl.wanadoo.nl) has quit (Ping timeout: 252 seconds) | |
| [18:48:11] | iamlindoro: | wagnerrp: will it take git push -b branchname? |
| [18:48:22] | Beirdo: | git config --global push.default tracking |
| [18:49:09] | wagnerrp: | i knew of the -b option, but not of the global directive |
| [18:50:34] | Beirdo: | it's been on the page for months :) I don't recall who added it |
| [18:52:50] | xris: | I have a default=current setting in my ~/.gitconfig |
| [18:53:08] | xris: | https://github.com/ex-nerd/scripts/blob/master/.gitconfig |
| [18:53:11] | xris: | highly recommended |
| [18:54:13] | xris: | well, except for the my name part. :) |
| [18:55:03] | Beirdo: | current is OK if the upstream branch name is always identical to the local branch name |
| [18:55:30] | Beirdo: | tracking will cover the possibility of the upstream branch having a different name |
| [18:55:43] | Beirdo: | for most people in most cases, they will act identically |
| [18:57:25] | xris: | ah, yeah. tracking is probably a better default. |
| [19:13:51] | jmartens2 (jmartens2!~jmartens@s5597ca60.adsl.wanadoo.nl) has quit (Ping timeout: 255 seconds) | |
| [19:14:57] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has joined #mythtv | |
| [19:20:08] | Giannirs4 (Giannirs4!~Giannirs4@dslb-084-058-137-214.pools.arcor-ip.net) has joined #mythtv | |
| [19:23:20] | andreax (andreax!~andreaz@p57B94355.dip.t-dialin.net) has joined #mythtv | |
| [19:23:58] | danielk22: | wagnerrp: the --skipdb setting is now working as expected here.. I added a small commit that just made sure it output the commflag map and didn't try to save it to the DB. The problem with output was just that the outputfile default wasn't set to "-" when --skipdb was used. |
| [19:25:33] | danielk22: | Any third part users should be using --outputfile, and for debugging purposes "-" is the sane default.. I never want to run a full commflag and then send the result to /dev/null :) |
| [19:27:17] | wagnerrp: | yeah, i was actually about to add a check that would fault out if skipdb was set and outputfile was not |
| [19:27:29] | wagnerrp: | but changing the default works too |
| [19:28:48] | danielk22: | wagnerrp: Yeah, faulting out would be better than running and then producing no output, but it's always good when you can come up with sane defaults instead. |
| [19:30:06] | wagnerrp: | actually, since we dont actually check that the value is defined |
| [19:30:14] | wagnerrp: | im going to move the default to the commandlineparser.cpp |
| [19:34:41] | Mousey (Mousey!~wtfisme@ross154.net) has joined #mythtv | |
| [19:38:22] | danielk22: | wagnerrp: PS I'm sorry for thinking you might have introduced all those globals. |
| [19:38:54] | wagnerrp: | no worries, i was simply the last one to hit the file |
| [19:39:42] | wagnerrp: | for what its worth, part of the re-ordering was to get right of some of those |
| [19:40:57] | wagnerrp: | well at least that was the original intend, it seems i didnt remove anyu |
| [19:41:00] | Beirdo: | yeah, I think you got rid of many of the variables scoped to main() |
| [19:41:16] | Beirdo: | not quite globals, but still :) |
| [19:41:33] | wagnerrp: | yeah, i did git rid of a bunch of the main() variables in favor of pulling them straight out of the parser as needed |
| [19:41:57] | wagnerrp: | consequence of the old parser that variables had to be used as they were looped past |
| [19:42:11] | Beirdo: | yeah |
| [19:42:25] | wagnerrp: | and the whole reason for the rewrite of the parser |
| [19:42:44] | Beirdo: | this way is a touch slower (several map lookups instead of a local variable), but makes it cleaner in that way |
| [19:43:06] | Beirdo: | so minorly slower that it really doesn't matter, it's not like it's in a tight loop :) |
| [19:43:23] | Beirdo: | ooh, it takes 100us instead of 50us! |
| [19:43:32] | Beirdo: | or whatever |
| [19:43:35] | wagnerrp: | well the only time speed might be a problem would be things like the preview generator |
| [19:43:43] | wagnerrp: | which could conceivably be fired in rapid succession |
| [19:44:36] | Beirdo: | yeah, once we re-fix the previews (likely a reoccurring seek table issue), we might want to look at how to streamline it more if needed |
| [19:44:54] | Beirdo: | right now, it's borked in some cases, and times out and kills mythpreviewgen |
| [19:45:44] | Beirdo: | be nice if a successful preview took like 500ms |
| [19:46:23] | Beirdo: | or even less |
| [19:46:27] | Giannirs4 (Giannirs4!~Giannirs4@dslb-084-058-137-214.pools.arcor-ip.net) has left #mythtv () | |
| [19:47:10] | danielk22: | Beirdo: most of the preview time is setup and teardown of the sockets and whatnot.. what we've talked about in the past is a long lived process that handles many previews and can be auto-restarted by the backend.. the apache approach.. |
| [19:47:24] | Beirdo: | yeah, that would be nice too |
| [19:47:32] | Beirdo: | a very nice way to handle it |
| [19:48:20] | Beirdo: | but while seeking seems to be borked... that probably should get fixed first and increase the % of successful previews back to nearly 100% |
| [19:48:23] | wagnerrp: | would that be run along side each mythbackend then? |
| [19:49:02] | Beirdo: | I would think so |
| [19:49:13] | Beirdo: | yeah, a lot of the startup time is also loading all the libraries |
| [19:49:59] | Beirdo: | doing so once (and db connections, and sockets, blah blah) would significantly speed up the preview process on the whole |
| [19:50:16] | wagnerrp: | wasnt the preview generator configured as a separate process each in the first place in case it crashed/ |
| [19:50:23] | Beirdo: | yup |
| [19:50:30] | Beirdo: | so it wouldn't take down the backend |
| [19:50:33] | wagnerrp: | is that no longer an issue? |
| [19:50:40] | Beirdo: | sure it is. |
| [19:50:42] | wagnerrp: | or would we just restart the process if needed |
| [19:50:52] | Beirdo: | but we can restart the process, yeah |
| [19:51:05] | Beirdo: | and it would just take itself out, not the ongoing recordings |
| [19:51:39] | Beirdo: | make it queue-driven, and voila |
| [19:53:23] | danielk22: | wagnerrp: it used to be a long running thread, but that was a stability problem. the idea is to make it a long lived process instead; so if it crashes, we just blacklist the crash inducing recording and continue processing the rest of the queue. |
| [19:53:45] | Beirdo: | hmmm, food. |
| [19:54:18] | simonckenyon (simonckenyon!~simoncken@195.7.61.12) has joined #mythtv | |
| [19:55:30] | wagnerrp: | well you can manually add a new connection into libmythprotoserver |
| [19:55:46] | wagnerrp: | and get fed new queries coming out of it, and a callback when it closes |
| [19:56:09] | wagnerrp: | you dont have to actually have it listen on something |
| [19:56:34] | kth (kth!~kth@unaffiliated/kth) has joined #mythtv | |
| [19:56:54] | wagnerrp: | that way the backend would not have to send out global events when it wants a single preview generator to run |
| [19:57:09] | sailerboy (sailerboy!sailerboy@ipv61.sailerboy.net) has quit (Ping timeout: 264 seconds) | |
| [19:59:19] | danielk22: | wagnerrp: I think the default outputfile should be "" not "-", unless --skipdb is set.. by default we shouldn't be printing out anything extra... |
| [19:59:26] | sailerboy (sailerboy!sailerboy@ipv61.sailerboy.net) has joined #mythtv | |
| [20:00:00] | jmartens1 (jmartens1!~jmartens@s5597ca60.adsl.wanadoo.nl) has joined #mythtv | |
| [20:02:29] | wagnerrp: | done |
| [20:02:42] | wagnerrp: | the other option... |
| [20:02:58] | wagnerrp: | right now, i was going to continue to use the jobqueue as a single process per host |
| [20:03:21] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has quit (Ping timeout: 252 seconds) | |
| [20:03:21] | wagnerrp: | i could change it so that the jobs allowed are handled per-process rather than per-host |
| [20:04:05] | wagnerrp: | so on a host, you would run a mythjobqueue that would handle normal external jobqueue tasks |
| [20:04:37] | wagnerrp: | but you could also run a mythpreviewgenerator, that would hook into the jobqueue interface, but would specifically only run preview generation tasks |
| [20:04:50] | wagnerrp: | and run then internally upon command from the jobqueue |
| [20:05:05] | wagnerrp: | using the jobqueue to balance out as needed |
| [20:07:56] | xris: | wagnerrp: how many jobs can take advantage of multiproc? |
| [20:09:04] | wagnerrp: | multiproc? |
| [20:09:18] | wagnerrp: | oh, how many are SMP |
| [20:09:31] | wagnerrp: | only mythcommflag, as far as i know |
| [20:09:52] | danielk22: | wagnerrp: the preview generator has it's own queue which is importantly FILO not FIFO.. this means that previews that are more likely still on screen get priority. |
| [20:10:43] | xris: | wagnerrp: just wondering if the "one job processor per host" is the most efficient, if running more than one on a multiproc box (most of them these days) would get more work done |
| [20:12:10] | xris: | though I suppose it's probably not important enough to matter |
| [20:20:06] | kth (kth!~kth@unaffiliated/kth) has quit (Quit: Leaving.) | |
| [20:25:48] | mrand: | Beirdo: friendly reminder to grab the file attached to Bug #9581 before it disappears from the file sharing site |
| [20:39:07] | _charly_ (_charly_!kroseneg@sunrise.schmidham.net) has quit (Ping timeout: 246 seconds) | |
| [20:39:16] | Beirdo: | ahh, will do, thanks for the reminder |
| [20:44:20] | wagnerrp: | xris: when i say 'one job processor', i mean one connected jobqueue |
| [20:44:31] | wagnerrp: | that jobqueue process might spawn off multiple tasks simultaneously |
| [20:44:33] | xris: | ah, gotcha |
| [20:44:36] | xris: | cool |
| [20:45:05] | wagnerrp: | and if the preview generator were added, it might run multiple threads of preview generation simultaneously |
| [20:45:17] | wagnerrp: | but if its FILO, then that would not work well with the jobqueue |
| [20:46:02] | wagnerrp: | and theyre not long lived enough to bother trying to make the two schedulers interact |
| [20:46:18] | wagnerrp: | i.e. to pause one queue when trying to generate |
| [20:46:43] | xris: | yeah |
| [20:51:09] | mrand (mrand!~mrand@ubuntu/member/mrand) has quit (Ping timeout: 264 seconds) | |
| [20:57:56] | mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv | |
| [21:10:25] | taylorr: | just saw this at a bookstore last night -> nice review of MythTV (showing off ArcLight) -> http://www.ubuntu-user.com/Magazine/Archive/2011/9/DVR-meets-web |
| [21:15:33] | jmartens1 (jmartens1!~jmartens@s5597ca60.adsl.wanadoo.nl) has quit (Quit: Leaving.) | |
| [21:16:32] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has joined #mythtv | |
| [21:16:46] | jmartens (jmartens!~jmartens@s5597ca60.adsl.wanadoo.nl) has quit (Client Quit) | |
| [21:23:07] | simonckenyon (simonckenyon!~simoncken@195.7.61.12) has quit (Ping timeout: 240 seconds) | |
| [21:34:44] | iamlindoro: | taylorr: Heh, nice |
| [21:37:15] | wagnerrp: | what, no gant? |
| [21:56:37] | zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Read error: Connection reset by peer) | |
| [21:56:44] | zombor_ (zombor_!~zombor_@kohana/developer/zombor) has joined #mythtv | |
| [21:58:27] | ** xris needs to update his arclight download ** | |
| [22:03:55] | andreax (andreax!~andreaz@p57B94355.dip.t-dialin.net) has quit (Ping timeout: 240 seconds) | |
| [22:04:13] | andreax (andreax!~andreaz@p57B94355.dip.t-dialin.net) has joined #mythtv | |
| [22:09:30] | branden__ (branden__!~branden@70.36.67.211) has joined #mythtv | |
| [22:09:48] | branden__: | Is it true they are dropping support for mpeg2? |
| [22:10:49] | wagnerrp: | they who? |
| [22:11:27] | wagnerrp: | branden__? |
| [22:12:24] | branden__: | mythtv |
| [22:12:50] | wagnerrp: | dropping support in what fashion? |
| [22:12:54] | branden__: | 2010-12–02: XvMC and libmpeg2 to be dropped in 0.25 |
| [22:13:07] | wagnerrp: | nearly all the content we record, either from MPEG encoder cards, or digital tuners, comes in MPEG2 |
| [22:13:18] | wagnerrp: | dropping MPEG2 support would be shooting ourselves in the foot |
| [22:13:40] | wagnerrp: | libmpeg2 is an alternate, and outdated, mpeg decoder, from the standard ones available through ffmpeg |
| [22:13:55] | wagnerrp: | it only exists as it offered some limited optimizations for very old hardware |
| [22:14:05] | wagnerrp: | and these days causes more problems than it is worth |
| [22:14:25] | wagnerrp: | XvMC is a very limited, partial offload of MPEG2 decoding |
| [22:14:41] | Beirdo: | the one exception may be in the mpeg2-mpeg2 "lossless" decoder |
| [22:14:53] | kormoc: | branden__, libmpeg2 is a mpeg2 decoder that was specifically better for older amd cpus, it's not overly useful anymore |
| [22:14:57] | Beirdo: | wher we *may* still have libmpeg2 |
| [22:15:11] | wagnerrp: | that has been abandoned by nVidia since the 8-series cards, was replaced by VAAPI on intel stuff, and was never even supported by ATI hardware |
| [22:15:11] | kormoc: | Beirdo, vs libavcodec's mpeg2? |
| [22:15:26] | Beirdo: | I don't recall if it still does or not |
| [22:15:29] | branden__: | ah snap |
| [22:15:31] | wahrhaft_ (wahrhaft_!~quassel@cpe-24-210-71-26.columbus.res.rr.com) has quit (Ping timeout: 260 seconds) | |
| [22:15:31] | Beirdo: | it needs a lobotomy anyways |
| [22:15:37] | branden__: | I have nvidia card |
| [22:15:38] | branden__: | s |
| [22:15:43] | kormoc: | I was fairly certain libmpeg2 was only for frontend display |
| [22:16:20] | Beirdo: | K, I had thought some of it was used for parsing in there too, certainly nothing that can't be overhauled if it's still thre |
| [22:16:22] | kormoc: | and it was targeted at the sub 600 mhz amd k processors |
| [22:16:26] | Beirdo: | yeah |
| [22:16:35] | wagnerrp: | branden__: this is the development channel, and if you wish to continue this discussion, it should be in #mythtv-users |
| [22:16:57] | Beirdo: | it's a pretty good mpeg2 decoder but is not of much use to us anymore |
| [22:17:12] | Beirdo: | s/much/any real/ |
| [22:17:13] | Beirdo: | hehe |
| [22:17:17] | branden__: | Well I thought it was appropriate for this channel, since it had to do with development. |
| [22:18:02] | wahrhaft (wahrhaft!~quassel@cpe-24-210-71-26.columbus.res.rr.com) has joined #mythtv | |
| [22:18:16] | Beirdo: | kormoc: it's still linked into mythtranscode as libmythmpeg2 |
| [22:18:30] | kormoc: | huh |
| [22:19:08] | Beirdo: | I say we find out why and fix it :) |
| [22:19:30] | Beirdo: | I also want to remove the Qt3 support required in there too |
| [22:19:43] | Beirdo: | I had one partial run at it and found the roadblocks. |
| [22:23:01] | branden__: | Is there gtk3 support or just qt? |
| [22:23:15] | wagnerrp: | iamlindoro, taylorr: its a shame they refer to our insistence of adherence to ToS like its a bad thing |
| [22:24:58] | iamlindoro: | branden__: This isn't meet the devs, it's "talk about development you personally are doing" |
| [22:25:08] | branden__: | ok |
| [22:25:56] | iamlindoro: | wagnerrp: If anything I'm proud we've gotten that reputation... all in all, as accurate a Myth article as I've seen in a bit |
| [22:29:35] | j-rod is now known as j-rod|afk | |
| [22:35:54] | wahrhaft (wahrhaft!~quassel@cpe-24-210-71-26.columbus.res.rr.com) has quit (Ping timeout: 255 seconds) | |
| [22:36:13] | paul-h (paul-h!~Paul@mythtv/developer/paul-h) has quit (Remote host closed the connection) | |
| [22:52:29] | danielk22 (danielk22!~danielk@96.57.9.142) has quit (Quit: Leaving.) | |
| [22:54:42] | zombor_ is now known as zombor | |
| [23:05:14] | branden__ (branden__!~branden@70.36.67.211) has quit (Quit: Leaving) | |
| [23:06:50] | abqjp (abqjp!~abqjp@97-119-166-191.albq.qwest.net) has quit (Quit: abqjp) | |
| [23:09:31] | wahrhaft (wahrhaft!~quassel@cpe-24-210-71-26.columbus.res.rr.com) has joined #mythtv | |
IRC Logs collected by
BeirdoBot.
Please use the above link to report any bugs.