| Monday, August 19th, 2013, 00:34 UTC | ||
| [00:34:19] | dgeary2 (dgeary2!~debian@pa49-187-94-47.pa.nsw.optusnet.com.au) has joined #mythtv | |
| [00:43:23] | dgeary2 (dgeary2!~debian@pa49-187-94-47.pa.nsw.optusnet.com.au) has quit (Quit: Ex-Chat) | |
| [00:54:40] | sl1ce (sl1ce!~johnathan@pool-100-0-73-123.bstnma.fios.verizon.net) has joined #mythtv | |
| [01:24:31] | jya_: | wagnerrp: i think his bug report is appropriate… no such thing as "doing it wrong"… should handle the case regardless |
| [02:10:18] | Gibby (Gibby!~Gibby@184.170.249.223) has joined #mythtv | |
| [02:10:32] | knightr_: | jya_, dekarl nope, the French translations are not incomplete, are you using PHP 5.5, people seem to have problem with it and MythWeb... (I can't say, my distro still have 5.4.something...) |
| [02:11:04] | jya_: | knightr_: ubuntu 12.04; ships with php 5.3.10 |
| [02:12:05] | knightr_: | that's weird, what screen was that (or are all screens affected?) |
| [02:13:42] | jya_: | maybe it was just backend status |
| [02:14:02] | jya_: | in recording list, the dates are in english |
| [02:14:53] | jya_: | ah! the show that gave me trouble last week, now opens instantly ! |
| [02:15:00] | jya_: | at leaast on that ipad |
| [02:15:50] | jya_: | go figure… what used to take 2 minutes is now instantaneous… no idea what happened |
| [02:16:20] | jya_: | all I did to the DB is manually remove the index title_subtitle_start |
| [02:17:17] | jya_: | ah other shows are still very slow…. oh well :( |
| [02:25:21] | knightr_: | could be since backend status is served by the backend and is not translated IIRC and should be rewritten to use the new qsp (IIRC) pages... |
| [02:26:11] | knightr_: | (IMHO they shouldn't be served by the main backend process but by another but that's another story...) |
| [02:26:41] | knightr_: | the dates are correctly set in the translation file so something is broken... |
| [02:31:02] | knightr_: | (it looks like it is supposed to get the content of generic_date and generic_time in the translation files, concatenate them together and put them in the session) |
| [02:31:33] | knightr_: | that used to work, I wonder when it got broken (the patch that supposedly fixe the PHP 5.5 problems maybe?) |
| [02:32:24] | knightr_: | lindoro is coming back to haunt us I see... |
| [02:34:42] | nyloc (nyloc!~quassel@p57B4FD31.dip0.t-ipconnect.de) has joined #mythtv | |
| [02:39:16] | _nyloc_ (_nyloc_!~quassel@p3EE2C7D8.dip0.t-ipconnect.de) has quit (Ping timeout: 260 seconds) | |
| [02:47:20] | joki (joki!~joki@p548609EB.dip0.t-ipconnect.de) has quit (Ping timeout: 256 seconds) | |
| [02:52:15] | joki (joki!~joki@p54860BBE.dip0.t-ipconnect.de) has joined #mythtv | |
| [02:52:19] | peper03 (peper03!~peper03@mythtv/developer/peper03) has quit (Ping timeout: 260 seconds) | |
| [02:57:19] | Sharky112065 (Sharky112065!~Sharky112@c-24-19-57-28.hsd1.wa.comcast.net) has joined #mythtv | |
| [03:04:16] | wilmoore-misc (wilmoore-misc!~wilmoore@c-67-190-17-108.hsd1.co.comcast.net) has quit (Remote host closed the connection) | |
| [03:33:39] | knight_ (knight_!~Nicolas@69-165-170-178.dsl.teksavvy.com) has joined #mythtv | |
| [03:35:08] | knight_ (knight_!~Nicolas@69-165-170-178.dsl.teksavvy.com) has quit (Client Quit) | |
| [03:35:19] | knight_ (knight_!~Nicolas@69-165-170-178.dsl.teksavvy.com) has joined #mythtv | |
| [03:36:01] | knight_ (knight_!~Nicolas@69-165-170-178.dsl.teksavvy.com) has quit (Client Quit) | |
| [03:36:03] | knightr (knightr!~Nicolas@69-165-170-178.dsl.teksavvy.com) has joined #mythtv | |
| [03:36:03] | knightr (knightr!~Nicolas@69-165-170-178.dsl.teksavvy.com) has quit (Changing host) | |
| [03:36:03] | knightr (knightr!~Nicolas@mythtv/developer/knightr) has joined #mythtv | |
| [03:36:09] | fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has quit (Ping timeout: 264 seconds) | |
| [03:37:09] | fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has joined #mythtv | |
| [03:38:21] | knightr: | how much of a problem would adding a depency to MythWeb be in general and this late in the release process? |
| [03:44:07] | knightr_ (knightr_!~knight@69-165-170-178.dsl.teksavvy.com) has quit (Quit: Leaving) | |
| [04:15:14] | jya_: | knightr: would that impact the various packagers ? (mythbuntu, mythdora etc...) |
| [04:17:55] | knightr: | jya_, most likely, that's why I am wondering if it would be possible and especially now... (it's possible though that some of them might already be packaging the package I need for other reasons...) |
| [04:19:48] | jya_: | then I would say no |
| [04:20:10] | XDS2010_ (XDS2010_!uid1218@gateway/web/irccloud.com/x-dvtrqefaxdrvhpbz) has quit (Ping timeout: 256 seconds) | |
| [04:20:17] | jya_: | I mean, what positives would you add, as opposed to the inconvenience added |
| [04:20:48] | jya_: | or you solving an actual problem (e.g. is it a bug fix?) |
| [04:20:59] | knightr: | It's a bug fix... |
| [04:21:31] | superm1 (superm1!~uid4318@ubuntu/member/superm1) has quit (Ping timeout: 264 seconds) | |
| [04:22:11] | knightr: | I don't know if all PHP packages are like that but at least in Fedora's case (the distro I currently use), proper multibyte (for character sets such as UTF-8) is available in a separate package... |
| [04:23:26] | wagnerrp: | jya_: you mentioned some python database module error a while back |
| [04:23:38] | wagnerrp: | the bindings look for MySQLdb or Oursql |
| [04:23:54] | wagnerrp: | is it possible you had one installed, but it was installed against a different version of python? |
| [04:24:24] | wagnerrp: | what version of python are you running? |
| [04:27:38] | XDS2010_ (XDS2010_!uid1218@gateway/web/irccloud.com/x-umxqidhizislehkr) has joined #mythtv | |
| [04:29:35] | knightr: | Currently we use functions which do not have UTF-8 support in MythWeb... The reason it works for most of them if even though the files are supposedly in UTF-8 is because most of them don't have diacritics/accents so they are essentially in US-ASCII (UTF-8 being a superset of US-ASCII if the character are only in the US-ASCII range they are single-byte so it doesn't cause problems that we are not using th |
| [04:29:35] | knightr: | e appropriate PHP functions...) |
| [04:29:58] | knightr: | gotta go sleep, goodnight everybody... |
| [04:37:02] | jya_: | wagnerrp: the issue was that I have the libraries installed in a non standard place, but the python binary didn't have the LD_LIBRARY_PATH properly set to find those. So while the mysqldb plugin was installed, it failed to load |
| [04:37:29] | jya_: | knightr: what's the ticket #? no ticket, no bug :) |
| [04:37:41] | wagnerrp: | for some reason, python likes to use its own PYTHON_PATH, rather than the standard LD_LIBRARY_PATH |
| [04:37:55] | jya_: | PYTHON_PATH is for where to find the python plugins |
| [04:38:04] | wagnerrp: | right |
| [04:38:10] | jya_: | LD_LIBRARY_PATH is for where to find the libs |
| [04:38:14] | jya_: | both are required here |
| [04:38:16] | wagnerrp: | oh, it couldn't find the mysql shared library? |
| [04:38:21] | jya_: | that's right |
| [04:38:23] | wagnerrp: | ah |
| [04:38:49] | wagnerrp: | i thought you meant the compiled python module |
| [04:38:51] | jya_: | quite happy now, I have the metadata properly working on a mac… never had that before |
| [04:39:08] | jya_: | yeah, but the compiled python module still looks for libmysql.so |
| [04:40:27] | knightr: | jya_, hmmm, I love your thinking, since MythWeb no longer has a dev maybe we should remove it, then there would no longer be any bug left in it... :) |
| [04:41:22] | jya_: | knightr: hardly relevant… to me, if there's a bug, it should be lodged in trac first, then fixed |
| [04:41:24] | knightr: | (I didn't ask the person who reported the problem to open a ticket that would stay there and never be fixed...) |
| [04:41:30] | jya_: | it's just the proper way to do things |
| [04:42:28] | jya_: | knightr: a project like fisheye/jira has a setting where you can't even commit anything if there's no ticket attached to the commit |
| [04:42:32] | knightr: | I wanted to investigate the problem first, I could open the ticket now... |
| [04:42:49] | jya_: | it's a bit extreme for our use, but IMHO, having some process does help |
| [04:42:58] | knightr: | jya_, we use Jira at work, I know... |
| [04:43:52] | knightr: | (and before that we use an in-house application made in Lotus Notes...) |
| [04:44:02] | wagnerrp: | we use jira... sorta... |
| [04:44:13] | wagnerrp: | it's running, somewhere, but there never seems to be any activity on it |
| [04:44:19] | knightr: | s/use/used |
| [04:44:33] | knightr: | wagnerrp, LOL... |
| [04:45:18] | wagnerrp: | there's also a subversion server, somewhere |
| [04:45:40] | knightr: | wagnerrp, geez, what do you currently use, Clearcase? |
| [04:45:50] | knightr: | (don't tell me CVS...) |
| [04:46:03] | wagnerrp: | heh, nothing |
| [04:46:59] | knightr: | ouch, I hope you have good backups and you comment each modification in the source files... |
| [04:47:00] | wagnerrp: | we just pull a latest revision archive, and then modify locally from there for each installation |
| [04:47:48] | wagnerrp: | there's no real mechanism for our changes to make it back up the chain and into the codebase |
| [04:47:48] | jya_: | wagnerrp: starting to understand things ….. :) |
| [04:48:02] | ** knightr is ROTFL... ** | |
| [04:49:33] | jya_: | knightr: you may laugh, but even this year on a project I worked on, an engineer use of SVN was usually to have a directory where he checks things out. Then copy what he wants to work on somewhere else. Do his changes. When he's done. He does a svn update, then copy/paste his changes back into the svn tree and commit. |
| [04:49:41] | jya_: | it's great that way, there's never any conflicts ! |
| [04:51:44] | knightr: | :) I hope he was the only person work on that project... |
| [04:51:52] | knightr: | s/work/working |
| [04:52:09] | wagnerrp: | well that's the thing, there's typically only ever one or two people working on a project |
| [04:52:40] | jya_: | even working alone on a project I use a repository |
| [04:53:00] | wagnerrp: | i do as well just for convenience |
| [04:53:14] | knightr: | were I work too but we frequently have to work on the same files, we wouldn't be able to work without a repo of some kind.. |
| [04:53:18] | wagnerrp: | but subversion isn't really deigned for local operation |
| [04:53:24] | knightr: | s/were/where |
| [04:53:25] | jya_: | it's just good practice... |
| [04:53:35] | wagnerrp: | and have you ever tried to get git running on qnx? |
| [04:53:51] | jya_: | you develop on qnx? |
| [04:53:53] | wagnerrp: | boy that was a good 6 hours tracking down dependencies and ripping things out of git to make it run |
| [04:53:58] | wagnerrp: | yeah |
| [04:54:01] | jya_: | or you develop "*for*" qnx |
| [04:54:05] | wagnerrp: | on |
| [04:54:14] | wagnerrp: | and an ancient version, 6.3 |
| [04:54:36] | jya_: | why would you want to develop *on*? |
| [04:54:46] | wagnerrp: | real time controller |
| [04:55:07] | jya_: | i've used qnx quite a bit, doesn't have to develop on the board itself… would be mostly waste of time |
| [04:55:30] | wagnerrp: | we're running it on standard PCs |
| [04:55:47] | jya_: | still… |
| [04:57:46] | jya_: | in reference to #11754 |
| [04:57:46] | ** MythLogBot http://code.mythtv.org/trac/ticket/11754 ** | |
| [04:57:51] | jya_: | I have a rip here: Red |
| [04:58:15] | jya_: | doing a metadata search would have allocated 574 MetadataLookup; 62 were freed |
| [05:07:59] | knightr: | owned by is no longer set when a ticket is created? |
| [05:08:14] | jya_: | knightr: only if the component as a default owner |
| [05:08:14] | wagnerrp: | only when the component itself is owned |
| [05:08:59] | knightr: | MythWeb still has one AFAIK and so does translations and it doesn't work for either... |
| [05:11:05] | knightr: | (and I knew that's how it worked but was surprised when I started seeing tickets which didn't have it set... I thought it might be because of the way the translators had opened the ticket but I just opened one and didn't set any owner on purposed to test that...) |
| [05:11:19] | peper03 (peper03!~peper03@mythtv/developer/peper03) has joined #mythtv | |
| [05:11:54] | wagnerrp: | huh, apparently you have to select "default" from the list |
| [05:13:25] | jya_: | well, I get assigned all the mac and audio tickets still |
| [05:13:39] | knightr: | :) that's something most people won't probably do.. |
| [05:14:16] | knightr: | jya_ the tickets you assign to yourself don't count... :) |
| [05:14:30] | jya_: | no no |
| [05:14:41] | knightr: | I was kidding... |
| [05:16:48] | knightr: | I wonder if the way it used to work was an in-house modification or the fact that it only partially work a regression... |
| [05:25:33] | knightr: | I really got to go sleep now, goodnight guys... |
| [05:25:43] | wagnerrp: | ok, fixed |
| [05:28:31] | knightr: | (:) you set it to <default> by default... clever...) |
| [06:19:20] | superm1 (superm1!uid4318@ubuntu/member/superm1) has joined #mythtv | |
| [06:35:24] | FabriceMG (FabriceMG!~Thunderbi@LCaen-156-54-30-212.w80-11.abo.wanadoo.fr) has joined #mythtv | |
| [06:42:01] | dekarl: | jya_: I copy-pasted from the other unit tests. I have to look at the .pro files sometime if I can you tell it to "link everything from ../../library.pro" |
| [06:42:39] | jya_: | dekarl: you can probably do an include of the pro file |
| [06:42:52] | jya_: | but why would you bother ? |
| [06:43:05] | dekarl: | knightr: no Mythbuntu 12.04 LTS with Version: 5.3.10–1ubuntu3.6 |
| [06:43:06] | jya_: | I mean, the unit test is started after the various .o have been compiled |
| [06:44:49] | dekarl: | I haven't really thought about how to build the tests. I was happy that duplicate detection will work again (mostly, they don't signal generic episodes and sometimes have multiple ProgramIDs for the same program and vice versa. But its a big improvement over what I have now) |
| [06:48:09] | dekarl: | Right now I'm trying to get our devs at work to actually run the other 50% of their unit tests. They got removed from the build some years ago and in the meantime some do fail... |
| [06:49:25] | dekarl: | Oh, and we migrated our CVS to SVN some years ago, but still no GIT in sight. |
| [06:49:53] | SteveGoodey (SteveGoodey!~steve@host109-158-211-237.range109-158.btcentralplus.com) has joined #mythtv | |
| [07:26:13] | aloril (aloril!~aloril@dsl-tkubrasgw3-54f97e-153.dhcp.inet.fi) has quit (Ping timeout: 246 seconds) | |
| [07:39:32] | aloril (aloril!~aloril@dsl-tkubrasgw3-54f97e-153.dhcp.inet.fi) has joined #mythtv | |
| [07:56:36] | Merlin83b (Merlin83b!~Daniel@office.34sp.com) has joined #mythtv | |
| [08:55:20] | SteveGoodey (SteveGoodey!~steve@host109-158-211-237.range109-158.btcentralplus.com) has quit (Quit: Konversation terminated!) | |
| [09:02:46] | clever is now known as cleverca22 | |
| [09:32:48] | stuartm: | jya_: 11756 is a backend problem and a frontend log wouldn't show anything |
| [09:33:34] | jya_: | I know that… |
| [09:34:30] | jya_: | just got peeved of on reading a "me too" when there's no indication it's the same whatsoever and the guy doesn't even bother to provide logs |
| [09:54:27] | cleverca22 is now known as clever | |
| [10:09:59] | jya_: | I have a pointer to an object… Is there an easy way to create a new object of the same type? |
| [10:11:04] | jya_: | e.g. template <class T> myClass { myclass() { T = new …. } |
| [10:11:18] | jya_: | where the template type is AnotherClass* |
| [10:19:44] | stuartm: | sphery: as long as there is no viewer for the database logging, is there any good reason for it to be enabled by default? |
| [10:23:09] | stuartm: | jya_: I can't think of anything, in mythui and elsewhere we added a Clone() method to the classes which would return a new object of the same type – we wouldn't need to know that type when calling clone |
| [10:25:07] | stuartm: | this looks relevant and there are a few suggestions there, including the use of a Clone() method (so that seems like a fairly common/standard solution) – http://stackoverflow.com/questions/2032881/cr . . . ic-type-info |
| [10:34:38] | MaverickTech (MaverickTech!~MaverickT@111.86.233.220.static.exetel.com.au) has joined #mythtv | |
| [10:35:54] | FabriceMG (FabriceMG!~Thunderbi@LCaen-156-54-30-212.w80-11.abo.wanadoo.fr) has quit (Ping timeout: 264 seconds) | |
| [10:49:31] | PatrickDickey (PatrickDickey!~quassel@2001:470:1f11:830:a0d0:b436:f564:5ac8) has quit (Remote host closed the connection) | |
| [11:00:39] | FabriceMG (FabriceMG!~Thunderbi@217.112.59.207) has joined #mythtv | |
| [11:14:40] | stuartm: | danielk22: do you mind looking at #11756 ? I think the problem is likely to be the filesystem or a kernel specific issue, and you know more than most about the I/O related gotchas |
| [11:14:40] | ** MythLogBot http://code.mythtv.org/trac/ticket/11756 ** | |
| [11:20:13] | stuartm: | admittedly it's OSX so I'm not sure how relevant that filesystem/kernel knowledge will be, still you might remember what changes went into the writer between 0.25 and 0.26 |
| [11:25:43] | MaverickTech (MaverickTech!~MaverickT@111.86.233.220.static.exetel.com.au) has quit (Ping timeout: 256 seconds) | |
| [11:36:44] | SteveGoodey (SteveGoodey!~steve@host109-158-211-237.range109-158.btcentralplus.com) has joined #mythtv | |
| [11:56:50] | jya_: | stuartm: I've found out why I must do reset for the metadata |
| [11:57:49] | jya_: | for some reasons, the artwork assigned to a particular video or recording is actually using absolute path to the local filesystem. Those are metadata I scanned on the backend itself… So they show fine on that backend machine, but all remote frontend: nope |
| [11:58:07] | jya_: | when I do an update, it sees that there's already an image assigned so it skips it |
| [12:00:46] | sphery: | stuartm: disable it |
| [12:04:46] | sphery: | though we could have a Python bindings log viewer/"pastebin-er" in about an hour of work, but it would rely on the DB logging |
| [12:24:29] | Jordack (Jordack!~Jordack@h69-131-44-221.plmomi.dedicated.static.tds.net) has joined #mythtv | |
| [12:34:21] | jya_: | when you do emit signal(arg) |
| [12:34:51] | wagnerrp: | an hour? |
| [12:34:55] | jya_: | what is arg doing… It it going to be like an event, the object will be deleted when used ? |
| [12:35:08] | wagnerrp: | i admit, i'm a bit groggy having just woken up, but didn't i write one of those several months back? |
| [12:37:56] | danielk221: | stuartm: Too me it looks like horrendously slow writes are blocking reads in #11756. I'd look for evidence to support that. The disk is probably done for. In my experience disks fail at second 5, hour 5, day 5, month 5, year 5, and all times in between. |
| [12:37:56] | ** MythLogBot http://code.mythtv.org/trac/ticket/11756 ** | |
| [12:38:34] | danielk221: | In the case of an external drive it is worth it to try another compatible power brick, these often fail independently of the drive itself. |
| [12:40:48] | danielk221: | In any case not a MythTV bug. We're reporting these stats to wash our hands of the issue, those are "Hey drive no worky, don't blame me if there are problems" error messages. |
| [12:42:02] | jya_: | danielk22: the key thing is that it was working perfectly fine last week with 0.25 |
| [12:42:09] | jya_: | he just upgraded to 0.26 |
| [12:42:14] | jya_: | and is seeing the problem |
| [12:44:22] | jya_: | that bug was lodged at my request, following yet another argument on "is it worth upgrading from 0.25 to 0.26", Craig chipped in with "well, since I upgraded to 0.26 it now doesn't work" type of thing... |
| [12:46:44] | jya_: | danielk22: Qt question… In libmythmetadata, there's a use of signal/slot. One of them is emiting a pointer (MetadataLookup*) and somewhere else, in another thread it gets that pointer and delete it.. |
| [12:47:11] | jya_: | now to me that seems dangerous, say you had two listeners (not that it's the case here) |
| [12:47:46] | jya_: | so I modified the code and encapsulate into a container that handle its deletion automatically once unused |
| [12:48:38] | jya_: | my question is: is this required? are signal/slot working any differently to Event in this case? the argument provided to the slot is deleted once it's all done ? |
| [12:48:51] | jya_: | or I shouldn't bother and leave things as-is |
| [13:01:53] | stichnot (stichnot!~stichnot@ppp-68-126-151-208.dsl.pltn13.pacbell.net) has joined #mythtv | |
| [13:01:53] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv | |
| [13:01:53] | stichnot (stichnot!~stichnot@ppp-68-126-151-208.dsl.pltn13.pacbell.net) has quit (Changing host) | |
| [13:18:03] | SteveGoodey (SteveGoodey!~steve@host109-158-211-237.range109-158.btcentralplus.com) has quit (Quit: Konversation terminated!) | |
| [13:18:55] | IReboot (IReboot!~doug@CPE10bf48e67915-CM78cd8e7e342d.cpe.net.cable.rogers.com) has joined #mythtv | |
| [13:20:31] | rhpot199` (rhpot199`!~rhpot1991@2001:4968:202:3:759d:b77c:affe:a5c2) has joined #mythtv | |
| [13:21:03] | tgm4883_ (tgm4883_!~tgm4883@ubuntu/member/tgm4883) has joined #mythtv | |
| [13:21:15] | `Nothing4You (`Nothing4You!N4Y@Nothing4You.w.tf-w.tf) has joined #mythtv | |
| [13:22:05] | tgm4883 (tgm4883!~tgm4883@ubuntu/member/tgm4883) has quit (*.net *.split) | |
| [13:22:06] | rhpot1991 (rhpot1991!~rhpot1991@ubuntu/member/rhpot1991) has quit (*.net *.split) | |
| [13:22:06] | jwhite (jwhite!~jwhite@75-146-153-89-minnesota.hfc.comcastbusiness.net) has quit (*.net *.split) | |
| [13:22:06] | Nothing4You (Nothing4You!N4Y@Nothing4You.w.tf-w.tf) has quit (*.net *.split) | |
| [13:22:06] | _charly_ (_charly_!kroseneg@sunrise.schmidham.net) has quit (*.net *.split) | |
| [13:23:23] | `Nothing4You is now known as Nothing4You | |
| [13:28:03] | jya_: | wagnerrp: any particular reason the database backup script is in perl rather than python... |
| [13:28:14] | jya_: | would be much easier (for me) if it was all python … |
| [13:29:18] | jwhite (jwhite!~jwhite@75-146-153-89-minnesota.hfc.comcastbusiness.net) has joined #mythtv | |
| [13:29:18] | sphery: | jya_: because it was written long before we had python bindings (or examples) and written by someone who doesn't know python :) |
| [13:29:23] | danielk22: | jya_: For the Qt question, I'd have to see the code in question but it is probably unsafe. With a direct connection you could do this, assuming the lifetime of the sender is fully within the lifetime of the receiver and they are only used within the same thread. |
| [13:29:25] | danielk22: | With an indirect connection you risk leaking resources because there is no guarantee that the signal is received before the receivers lifetime ends. |
| [13:29:48] | danielk22: | +1 for rewriting in python ;) |
| [13:29:55] | sphery: | granted, the backup/restore scripts don't use the bindings so that they work with any version of MythTV |
| [13:30:11] | sphery: | and if someone wants to rewrite it to python, they should probably do the same |
| [13:30:32] | jya_: | danielk22: ok.. so if I encapsulate my pointer into a class instance stored on a stack… that will do the job. |
| [13:30:55] | jya_: | I added a RefCountHandler which is a simple class that can *own* a reference counted object |
| [13:31:32] | jya_: | with that I created a RefCountedList which works just like QList, except it handles ReferenceCounter objects safely |
| [13:32:19] | danielk22: | jya_: For the ownership I'd thought about making our reference counting ownable by the Qt smart pointer classes. I don't thik it would really be hard, we just need to name the ref()/deref() functions whatever Qt whats. |
| [13:32:52] | danielk22: | But I didn't attempt a Proof of Concept. There might be issues I haven't forseen. |
| [13:32:53] | jya_: | danielk22: looking at the Qt code, when you do a emit and it's queued, it's stored in a queue own by the eventloop |
| [13:33:00] | danielk22: | yep |
| [13:33:13] | jya_: | danielk22: well, you see what I've crafted.. you'll let me know what you think. |
| [13:33:44] | sphery: | wagnerrp: I think you wrote something, but never made it available because of the time or something... that said, I don't think you knew about the services api support ( http://www.gossamer-threads.com/lists/mythtv/ . . . 04119#504119 ) when you implemented it. It wouldn't be too hard to do one that uses the services API and interactively allows the user to choose filtering criteria. (And, IIRC, I had a way to get the list of available ... |
| [13:33:45] | jya_: | i'm pretty pleased with that container. |
| [13:33:46] | wagnerrp: | jya_: to be honest, it would probably be easier to keep in perl |
| [13:33:49] | danielk22: | ok. I am about 900 messages behind on the commits mailing list.. so if you hear from me in a while... |
| [13:33:50] | sphery: | ... filter values--and I used it in the backend web server log viewer.) |
| [13:34:20] | wagnerrp: | since a lot of what it does is interface with the mysql command line utilities |
| [13:34:25] | sphery: | really, all I needed to do was put a proxy in place in MythWeb and we'd have had a MythWeb-accessible log viewer |
| [13:34:29] | jya_: | wagnerrp: probably… but making the mac port support the perl binding is looking rather complicated right now (if not outright impossible) |
| [13:34:46] | wagnerrp: | the backup script does not use the mythtv perl bindings |
| [13:34:53] | wagnerrp: | (at least i don't think it does) |
| [13:35:02] | sphery: | correct, it doesn't use the bindings or need them |
| [13:35:06] | sphery: | so it works on any version |
| [13:35:07] | danielk22: | As for why #11756 would pop up with the upgrade.. the only thing I can think of is that some call that is non-blocking in Linux is blocking in OSX. |
| [13:35:07] | ** MythLogBot http://code.mythtv.org/trac/ticket/11756 ** | |
| [13:35:26] | sphery: | just needs valid Perl, config.xml, and (optionally) backuprc |
| [13:35:32] | danielk22: | If O_NONBLOCK isn't respected in OSX I could see that causing problems. |
| [13:35:59] | wagnerrp: | that said, calling command line utilities itself may be made a lot more difficult by using OSX |
| [13:36:31] | jya_: | I need to find out why the call to the perl script is successful |
| [13:36:38] | sphery: | jya_: Nigel tested the perl backup/restore scripts on Mac OS X and they worked fine after I made a couple changes for him |
| [13:36:48] | jya_: | it could just be that as the perl binding configuration fail, it doesn't install any perl script |
| [13:37:10] | danielk22 (danielk22!~danielk@exchange.wgen.net) has left #mythtv () | |
| [13:37:18] | sphery: | it should always be installed--we don't make it depend on bindings in configure or anything |
| [13:37:27] | sphery: | (since it doesn't use them :) |
| [13:37:28] | danielk22 (danielk22!~danielk@exchange.wgen.net) has joined #mythtv | |
| [13:37:39] | jya_: | sphery: I get the error: DBUtil: Error backing up database: /Users/jyavenard/Work/mythtv/.osx-packager/build/bin/share/mythtv/mythconverg_ba ckup.pl /tmp/mythtv_db_backup_conf_Squ7Nf (127) |
| [13:37:48] | jya_: | Script-based database backup failed. Retrying with internal backup. |
| [13:38:13] | jya_: | ohhh… it seems to call mysqldump |
| [13:38:17] | jya_: | I don't have that installed |
| [13:38:19] | sphery: | you'd need to have mythconverg_backup.pl in your PATH somewhere (or whatever Mac OS X uses to find utils) and same for mysqldump |
| [13:38:22] | sphery: | right |
| [13:38:29] | jya_: | I only have the libs |
| [13:38:53] | sphery: | really it does basically the same thing as the internal implementation (which also calls mysqldump) |
| [13:39:04] | sphery: | but it's a bit more robust/configurable |
| [13:39:34] | jya_: | do we have to use mysqldump? the configure I believe only check for sql libraries.. no mysql executable |
| [13:40:10] | wagnerrp: | i'm half surprised mysql lets you install libraries with no binaries |
| [13:41:23] | jya_: | wagnerrp: mysql is compiled from source. the libs are installed. Then the packager checks for the libs linked to the executatble and extra those libs and package them |
| [13:41:46] | sphery: | well, it could just be that the "packagers" for Mac OS X splits them out (like packagers like to do with MythTV) |
| [13:41:57] | jya_: | so the only thing that ships with the OS X package, is mythtv itself and all the dynamic libs it linked against |
| [13:42:59] | sphery: | how much extra does mysqldump require? In theory, it should just be "normal" libs (like C++/crypt/threads, etc.) + libmysqlclient, so maybe you could install it |
| [13:43:41] | jya_: | I like the user questions in regards to adjusting the audio sync. After many years, it still take me multiple goes before I know which way to go to shift the audio… Is it + or is it -, I can never tell |
| [13:44:25] | sphery: | is there a way to tell? I always figure it out by adjusting way too much to make it clear, then going back until it's good |
| [13:44:29] | jya_: | sphery: on the mac here it's 2.5MB |
| [13:44:42] | jya_: | interestingly, it's not linked to any libs but libSystem |
| [13:44:49] | jya_: | seems to have been compiled statically |
| [13:45:10] | jya_: | sphery: that's fine for a few ms.. but when it's several seconds behind |
| [13:45:26] | jya_: | oops, I'm confused… I had the adjustment of the subtitles in mind |
| [13:45:59] | jya_: | it's a similar interface, and I never understood how it works. I certainly never managed to properly adjust subitltes that were way out of sync |
| [13:51:26] | stichnot: | jya_: I have the same problem with audio & subtitle sync. I'm better with subtitle sync than audio sync, maybe because "subtitle delay" is a bit more clear than "audio sync", in terms of the relationship between audio/subtitle and video |
| [13:52:09] | stichnot: | Probably the audio sync OSD timeout should be infinite instead of "medium" |
| [13:52:27] | jya_: | stichnot: last week I was attempting to play a DVD rip… I had a friend here who needed subtitles |
| [13:52:45] | jya_: | I didn't have them.. So I searched opensubtitles.org for subtitles on that particular movie |
| [13:52:56] | jya_: | obviously none of them matched perfectly the rip I had |
| [13:53:31] | jya_: | the best I found appeared to be 4s out… Well, after a good 10 minutes playing with the adjustment |
| [13:53:35] | jya_: | I never managed to align it. |
| [13:54:02] | jya_: | I could see changing the delay did something. but never as much as what the OSD told me… +4s was certainly not +4s |
| [13:54:45] | jya_: | that the OSD gets dismissed is definitely annoying for subtitles, because often you have no subittles for several seconds, by wihich time it's been dismissed already |
| [13:56:30] | sphery: | I'd support making it a no-dismiss for that and audio sync--for the reason that you mention (it takes many seconds of observation to figure out what to do next) |
| [13:56:37] | stichnot: | I've seen srt files with drift, not just a fixed offset, e.g. the srt is based on 30fps and the video is 29.97fps. A linear drift could easily be compensated for, but I have no idea how to design a decent UI for it |
| [13:56:45] | stichnot: | sphery: +1 |
| [13:57:12] | stichnot: | When I did the subtitle delay, I just copied audio sync for consistency |
| [13:57:33] | sphery: | then to dismiss, users can hit ESCAPE or re-hit the keybinding for changing the sync |
| [13:57:51] | stichnot: | (except that pressing ENTER on the subtitle delay dialog doesn't commit/exit like audio sync does, that should be fixed) |
| [13:57:52] | sphery: | yeah, consistency is important, so if we change one, we should change both |
| [13:58:17] | sphery: | and both have everything in place so it should be an easy "just change it to not auto-dismiss" |
| [13:58:21] | jya_: | I don't really see the point of having to validate with enter any sync (audio or subtitle) |
| [13:58:44] | jya_: | changing to get immediate visual feedback is good enough… |
| [13:58:51] | sphery: | yeah, with time stretch, we don't require SELECT to accept |
| [13:59:18] | jya_: | so let's do that… don't dismiss the window and make it immediately effective upon change |
| [13:59:23] | sphery: | so if sync adjustment requires it, and it can be changed to not require it, it would be nice for consistency |
| [13:59:46] | stichnot: | None of these require SELECT to accept, they usually just dismiss the dialog and maybe give one final "commit"-like OSD message |
| [14:00:18] | jya_: | stichnot: with the subtitle: it set/clear the bookmark when I press enter |
| [14:00:40] | stichnot: | jya_: right, that was an oversight and should be fixed |
| [14:00:59] | sphery: | stichnot: does it apply as you're changing the value or does it wait until OSD clear to apply? in timestretch, it does it even while the osd is up |
| [14:01:17] | stichnot: | sphery: all 3 apply it in real time while the dialog is up |
| [14:01:24] | sphery: | ah, ok, great |
| [14:01:45] | sphery: | so SELECT was just supposed to be another "way out" of the OSD (like ESCAPE or re-hitting the change button)? |
| [14:01:50] | stichnot: | otherwise they would be practically useless :) like having to go into Frontend Settings to turn subtitles on/off |
| [14:01:51] | amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has joined #mythtv | |
| [14:01:57] | sphery: | right |
| [14:02:10] | sphery: | or at least very annoying to adjust :) |
| [14:02:20] | stichnot: | sphery: I believe that is true, though I haven't checked the code for any other side effects of SELECT |
| [14:02:53] | stichnot: | one side effect is a "final confirmation" OSD message with a short timeout |
| [14:06:33] | stichnot: | jya_: I'll try to get that SELECT problem on subtitle delay dialog fixed today |
| [14:08:39] | jya_: | thanks |
| [14:09:26] | jya_: | stichnot: I remember reading here about you playing with the ringbuffer to attempt to determine if a file existed or not |
| [14:09:44] | jya_: | regardless of this was a myth:// url or a direct path |
| [14:10:15] | stichnot: | I had to create RemoteFileWrapper class in textsubtitleparser.cpp |
| [14:10:18] | jya_: | any tips ? I'd like to check when searching for metadata to see if current artwork set actually exists (rather than simply checking if the path is set) |
| [14:10:48] | jya_: | there's no generic class to test a path ? |
| [14:11:12] | stichnot: | not sure... I was specifically trying to avoid RingBuffer in this case |
| [14:11:25] | jya_: | seems that's exactly what I need |
| [14:11:51] | jya_: | maybe I should extract that and put it in libmythbase |
| [14:12:46] | jya_: | how would you use RemoteFileWrapper just to check if the file exists or not? |
| [14:14:07] | stichnot: | hmm... I guess I just assume the file exists because the previous RingBuffer call returned the existence of the .srt file that is being opened with RemoteFileWrapper |
| [14:15:20] | jya_: | there RemoteFile::Exiss |
| [14:15:29] | jya_: | Exists |
| [14:15:31] | stichnot: | you probably want QFile::exists(filename) for a local file or RemoteFile::Exists(filename) for remote |
| [14:16:05] | stichnot: | but yeah, if you find a use for the class, please move it somewhere appropriate |
| [14:16:11] | jya_: | i like your wrapper :) makes it easier.. don't have to know about the path type |
| [14:16:38] | stichnot: | yeah, it was a shame that RemoteFile didn't already do that |
| [14:16:55] | jya_: | maybe it will be easier to adjust RemoteFile instead |
| [14:17:02] | stichnot: | anyway, first day of school, gotta get the kids moving :) |
| [14:17:22] | jya_: | do you foresee any issue if RemoteFile is extended to work with local file too? |
| [14:39:36] | IReboot (IReboot!~doug@CPE10bf48e67915-CM78cd8e7e342d.cpe.net.cable.rogers.com) has quit (Remote host closed the connection) | |
| [14:42:21] | IReboot (IReboot!~doug@CPE10bf48e67915-CM78cd8e7e342d.cpe.net.cable.rogers.com) has joined #mythtv | |
| [14:46:13] | jya_: | oh… didn't see stuartm tagged the beta and I pushed my fixes... |
| [14:50:52] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 256 seconds) | |
| [15:01:07] | FabriceMG (FabriceMG!~Thunderbi@217.112.59.207) has quit (Ping timeout: 264 seconds) | |
| [15:04:29] | FabriceMG (FabriceMG!~Thunderbi@217.112.59.207) has joined #mythtv | |
| [15:11:01] | FabriceMG (FabriceMG!~Thunderbi@217.112.59.207) has quit (Quit: FabriceMG) | |
| [15:13:58] | SteveGoodey (SteveGoodey!~steve@host109-158-211-237.range109-158.btcentralplus.com) has joined #mythtv | |
| [15:18:39] | SteveGoodey (SteveGoodey!~steve@host109-158-211-237.range109-158.btcentralplus.com) has quit (Client Quit) | |
| [15:24:53] | stuartm: | stuarta: can we re-sync the osuosl mirror from our end, or do I need to put in a request with them? |
| [15:29:50] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv | |
| [15:35:03] | stichnot: | jya_: Offhand I couldn't see any problems with making RemoteFile deal properly with local files, but for a niche feature (getting .srt updates for an in-progress recording), I wanted to be careful about breaking other things |
| [15:35:28] | jya_: | stichnot: I checked… everywhere remotefile is only ever used for myth:// bit |
| [15:35:40] | jya_: | actually quite thorough the code you have somethine like: |
| [15:36:00] | jya_: | if (url.startsWith("myth://) … RemoteFile code else { QFile code } |
| [15:36:09] | jya_: | seems like a great waste of lines... |
| [15:37:01] | jya_: | stichnot: that's my first attempt at amending remotefile |
| [15:37:06] | jya_: | haven't even compiled yet |
| [15:37:07] | jya_: | http://pastebin.com/TBrEE1QW |
| [15:37:55] | jya_: | need to look on how to simulate GetFileHash |
| [15:38:58] | jya_: | hmmm… does your QDataStream works with the current file position ? (following a seek) ? |
| [15:40:14] | jya_: | ah yes.. doc suggests that yes... |
| [15:46:56] | stichnot: | jya_: looks reasonable (but I didn't study the Seek() changes). I would factor out the multiple is_local calculations. |
| [15:47:17] | jya_: | i was just in the process of doing that... |
| [15:47:27] | jya_: | also want to check that the file is indeed local |
| [15:52:20] | jya_: | stichnot: why check if the file starts with / or that the file exists? |
| [15:52:29] | jya_: | why not just testing if the file exists? |
| [15:53:35] | stichnot: | Not sure... I just copied the logic from FileRingBuffer::OpenFile() without digging into the history/consequences of that decision |
| [15:53:43] | wagnerrp: | the "is_local" calculations are not for a local versus a remote file, but for a file on a local versus remote filesystem |
| [15:53:54] | wagnerrp: | i.e. ext4 vs. nfs |
| [15:54:49] | stichnot: | wagnerrp: I don't see how that particular condition could distinguish between file systems |
| [15:55:09] | wagnerrp: | very crudely |
| [15:55:20] | wagnerrp: | http://code.mythtv.org/cgit/mythtv/tree/mytht . . . nfo.cpp#n243 |
| [15:55:34] | Jordack (Jordack!~Jordack@h69-131-44-221.plmomi.dedicated.static.tds.net) has quit () | |
| [15:56:11] | wagnerrp: | i just lost power... |
| [15:58:26] | jya_: | wagnerrp: I don't see how that code could distinguish between anything other than absolute/relative path and anything else |
| [15:58:58] | wagnerrp: | am i thinking of a different set of code? |
| [15:59:14] | jya_: | we're talking about the pastebin I posted above |
| [16:00:15] | wagnerrp: | ah, nevermind |
| [16:00:29] | wagnerrp: | i was still thinking of the disk scheduler/expirer discussion last night |
| [16:04:36] | jya_: | wagnerrp: in the RemoteFile class, what as the auxfiles list for? |
| [16:05:40] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 256 seconds) | |
| [16:06:08] | wagnerrp: | i'm guessing so you can have one socket that receives a bunch of small files |
| [16:06:11] | dekarl1 (dekarl1!~dekarl@p4FCEF656.dip0.t-ipconnect.de) has joined #mythtv | |
| [16:06:17] | wagnerrp: | rather than opening up a new socket for each |
| [16:06:25] | wagnerrp: | Captain_Murdoch would probably know |
| [16:06:42] | jya_: | ok.. so its not about files having auxilary names |
| [16:07:23] | jya_: | shit.. I forget my wife was working tomorrow early morning.. so got to take the kids to school.. off to bed |
| [16:08:56] | dekarl (dekarl!~dekarl@p4FE84EC6.dip0.t-ipconnect.de) has quit (Ping timeout: 260 seconds) | |
| [16:19:06] | NightMonkey (NightMonkey!~NightrMon@173.164.139.193) has joined #mythtv | |
| [16:19:07] | NightMonkey (NightMonkey!~NightrMon@173.164.139.193) has quit (Changing host) | |
| [16:19:07] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv | |
| [16:21:48] | stichnot (stichnot!~stichnot@216.239.45.77) has joined #mythtv | |
| [16:21:48] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv | |
| [16:21:48] | stichnot (stichnot!~stichnot@216.239.45.77) has quit (Changing host) | |
| [16:26:30] | rsiebert (rsiebert!~quassel@e179133069.adsl.alicedsl.de) has joined #mythtv | |
| [16:29:35] | rsiebert_ (rsiebert_!~quassel@g226060020.adsl.alicedsl.de) has quit (Ping timeout: 245 seconds) | |
| [16:54:12] | stichnot: | jya_: the auxfiles stuff in RemoteFile is so that when you ask to open a RemoteFile, the remote host will also check for the existence of files with the same basename and suffixes in the auxfiles set, and it returns that set, so that you don't have to individually query the existence with separate protocol requests. |
| [16:55:04] | skd5aner (skd5aner!~skd5aner@50-90-5-146.res.bhn.net) has quit (Read error: Connection reset by peer) | |
| [16:55:30] | skd5aner (skd5aner!~skd5aner@50-90-5-146.res.bhn.net) has joined #mythtv | |
| [17:00:06] | Merlin83b (Merlin83b!~Daniel@office.34sp.com) has quit (Quit: Leaving) | |
| [17:03:18] | skd5aner: | My only open ticket from 0.26 at this point is #11234, anyone interested in taking a look at it? |
| [17:03:18] | ** MythLogBot http://code.mythtv.org/trac/ticket/11234 ** | |
| [17:04:11] | skd5aner: | I'm debating if it should just be closed – I've not seen any additional recordings exhibit the issue since I submitted the ticket, so perhaps it's not relavent anymore. I do know I submitted a sample that you guys could look at from a regression test perspective |
| [17:07:12] | tgm4883_ is now known as tgm4883 | |
| [17:11:57] | SteveGoodey (SteveGoodey!~steve@host109-158-211-237.range109-158.btcentralplus.com) has joined #mythtv | |
| [17:41:29] | stichnot: | skd5aner: From looking at the log, it suggest an upstream ffmpeg failure. However, the frontend doesn't need to delay tens of seconds before giving a cryptic message, so that could be improved. |
| [17:55:40] | jpabq: | stichnot: I think jya's new notification stuff will be great for situations like this — the frontend will now be able to tell the user what went wrong, instead of having to dig through the log.... maybe? |
| [18:03:30] | stichnot: | Yeah, that sounds promising. So in such a failure, the user would end up back in the Watch Recordings screen, with a notification popped up waiting for dismissal. |
| [18:05:28] | stichnot: | In this case, I think there is a lack of communication between asynchronous processes, so a little more plumbing is needed to actively get out of the "Please Wait..." screen instead of waiting for a timeout. |
| [18:43:53] | wagnerrp (wagnerrp!~wagnerrp_@mythtv/developer/wagnerrp) has quit (Ping timeout: 240 seconds) | |
| [18:51:01] | stichnot: | danielk22 (and anyone else): What's the best way to yield? sched_yield() , QThread::yieldCurrentThread() , usleep(some_small_value) ? |
| [18:51:22] | danielk22: | myth_yield() :) |
| [18:52:14] | stichnot: | nice, thanks :) |
| [18:59:06] | wagnerrp (wagnerrp!~wagnerrp_@mythtv/developer/wagnerrp) has joined #mythtv | |
| [19:09:53] | _nyloc_ (_nyloc_!~quassel@p5B26F0D9.dip0.t-ipconnect.de) has joined #mythtv | |
| [19:10:47] | nyloc (nyloc!~quassel@p57B4FD31.dip0.t-ipconnect.de) has quit (Ping timeout: 246 seconds) | |
| [19:16:31] | natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has joined #mythtv | |
| [19:40:38] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Ping timeout: 240 seconds) | |
| [21:09:06] | stichnot: | danielk22: MThreadPool::start() takes an optional priority argument, but I don't see any calls with that argument set. Is it suitable for creating a lower-priority thread? (Specifically, a thread that does all the non-MythUI processing of subtitles, that runs at a lower priority than the decoder and UI threads.) |
| [21:11:19] | danielk22: | stichnot: Only if it is a short lived thread. That priority just determines when it will be popped off the work queue not at what niceness level it runs. |
| [21:13:23] | stichnot: | OK. This is a long-lived thread. Is there any way of dropping its priority (and maybe restoring the original priority before returning it to the pool)? |
| [21:14:55] | danielk22: | I don't think so, you can always lower your priority, but I don't think you can then increase it again unless you have special permissions. |
| [21:15:28] | danielk22: | But maybe you can get back up to normal niceness, I haven't tried it. |
| [21:15:34] | stichnot: | OK. I don't know if it's actually necessary/helpful to lower the thread's priority, but I was just wondering about options. |
| [21:17:17] | stichnot: | I'm happy to observe that the non-MythUI preprocessing time is roughly the same as the MythUI time (mostly calling Draw() on the objects), so offloading the preprocessing to a separate thread should be a win. |
| [21:23:00] | stichnot: | I'd *really* like to be able to say that this separate thread should have a timeslice no more than say 25% of a frame interval and that it should always take a back seat to the UI and decoder threads. |
| [21:24:09] | danielk22: | Then just create an MThread for it, then you can give it it's own priority and sleep/yield as you see fit. |
| [21:25:05] | stichnot: | good point, I should just yield frequently |
| [21:26:00] | sraue (sraue!~stephan@xbmc/staff/sraue) has quit (Quit: Ex-Chat) | |
| [21:30:38] | natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has quit (Ping timeout: 240 seconds) | |
| [21:34:23] | SteveGoodey (SteveGoodey!~steve@host109-158-211-237.range109-158.btcentralplus.com) has quit (Quit: Konversation terminated!) | |
| [21:36:25] | sraue (sraue!~stephan@xbmc/staff/sraue) has joined #mythtv | |
| [21:50:54] | sphery: | stichnot: BTW, do you know what TV Playback|SETAUDIOSYNC action is used for? Is that one of those TV-menu-only "key bindings"? |
| [21:52:26] | dekarl1: | Where are we in the release cycle? Can I push fixes to master? |
| [21:52:29] | dekarl1 is now known as dekarl | |
| [21:53:48] | dekarl: | I got a missing piece for #10495 / #10496 in the pipeline. |
| [21:53:48] | ** MythLogBot http://code.mythtv.org/trac/ticket/10495 ** | |
| [21:53:48] | ** MythLogBot http://code.mythtv.org/trac/ticket/10496 ** | |
| [21:58:21] | stichnot: | sphery: I wasn't exactly sure. I'll take a look. |
| [21:59:43] | Merlin83b (Merlin83b!~Daniel@2a00:1ee0:3:1337:dc62:f52a:423:5b6a) has joined #mythtv | |
| [22:00:25] | stuartm: | dekarl: fixes can go in until the moment of release, invasive fixes are best avoided though, especially after the first RC |
| [22:00:40] | stuartm: | we're now in beta (until the weeked) |
| [22:02:06] | stuartm: | tbh, we could have skipped alpha and I think we might do that in future – it serves no particular purpose with open source software |
| [22:02:10] | dekarl: | hmm its just a refactoring moving some stuff from the PESPacket to the PSIPTable class to finally get rid of the 1000+ CRC errors per second... though it's not so bad with the DB Logging turned off :) |
| [22:02:29] | sphery: | stichnot: just wondering based on http://www.gossamer-threads.com/lists/mythtv/users/549638#549638 . Since TV Playback|SETAUDIOSYNC requires a value to be specified, I can't imagine a way to use it as a "user" key binding. Anyway, if it is a menu-only thing, I was wondering if your 0.27 custom/shortcut menu stuff could be used with it (though it would require several entries--probably making TOGGLEAUDIOSYNC (which brings up the OSD to change ... |
| [22:02:36] | sphery: | ... it) a better option, anyway) |
| [22:02:57] | dekarl: | http://pastebin.com/N4qpjJFn http://pastebin.com/TAXrbXhc |
| [22:03:11] | stichnot: | sphery: In one form, it simply "adjusts" the audio sync by 0ms, effectively just bringing up the dialog. However, if you give it a numeric prefix, like '5 0 SETAUDIOSYNC', it would set the audio sync to 50ms. |
| [22:04:34] | stichnot: | btw, this is from inspecting the code, not from actually testing it |
| [22:04:37] | sphery: | ah, so the value is specified with the OSD input--i.e. type using numeric keys |
| [22:04:44] | stichnot: | I believe so |
| [22:04:45] | sphery: | like with channel change |
| [22:04:47] | sphery: | makes sense |
| [22:05:49] | stichnot: | As discussed this morning, I think we'll set audio_sync, timestretch, and subtitle_delay to unlimited timeout for the 0.27 release (though of course not in time for beta) |
| [22:06:13] | stuartm: | unless you have a time machine |
| [22:06:33] | stichnot: | sadly my time machine only enables forward travel |
| [22:06:39] | sphery: | yeah, I was thinking about timestretch... I realize that consistency says it should be unlimited timeout, but as that doesn't require seconds of waiting to decide what value to choose, it's nice for it to time out |
| [22:06:52] | sphery: | the others definitely should be unlimited timeout |
| [22:07:20] | stichnot: | It's a trivial change, but I also need to track down the place to make SELECT close the subtitle_delay dialog |
| [22:07:24] | sphery: | but I'd say go with same for all 3 and when I put in the theme-configured timeouts, I can change mine to timeout for timestretch |
| [22:08:28] | stichnot: | With any of these dialogs where you use the arrow keys to adjust the values, there's always the risk that the dialog goes away and your arrow keys make you jump somewhere pseudo-random in the recording, causing much annoyance |
| [22:09:00] | sphery: | right |
| [22:09:01] | stichnot: | so I think it should just stick until you do EXIT, SELECT, or something else that closes the dialog |
| [22:10:00] | sphery: | it's a good idea (and needed for consistency), so that's the best approach to use |
| [22:10:01] | stichnot: | and it's not an idle risk – it either happens to me, and/or adds an unnecessary level of stress while I'm using the dialog |
| [22:10:14] | stichnot: | so I don't know why I didn't do it long ago :) |
| [22:10:31] | dekarl: | stuartm, I agree. all the phases seem like overkill. Just a feature freeze and release after 7 days without fixes might be enough. (or something in between) |
| [22:11:07] | sphery: | I should look at the schedule, then figure out when I'm going to do that initial DB rollup... guess I need to this week |
| [22:16:51] | Merlin83b (Merlin83b!~Daniel@2a00:1ee0:3:1337:dc62:f52a:423:5b6a) has quit (Quit: Leaving) | |
| [22:18:54] | stuartm: | the RC phase does seem to encourage more people to try out master, while accepting some risk |
| [22:19:29] | Merlin83b (Merlin83b!~Daniel@office.34sp.com) has joined #mythtv | |
| [22:20:06] | stuartm: | of course I wish many more people would get involved in testing so that we don't end up with people reporting "serious breakage" more than eight months _after_ a release |
| [22:21:03] | stuartm: | they don't of course see the irony in telling people not to upgrade (namely that if no-one upgrades, there's no-one to find and report the bug) |
| [22:21:05] | Merlin83b (Merlin83b!~Daniel@office.34sp.com) has quit (Client Quit) | |
| [22:22:51] | dekarl: | according to smolt 2/3 are still on some version of 0.25 :( |
| [22:23:06] | dekarl: | There's no point in porting back fixes from master to fixes if no one upgrades |
| [22:23:20] | sphery: | that's the one users on the list keep saying to stick with (mainly because that's the earliest version you can get on Ubuntu 12.04 with packages) |
| [22:24:02] | stuartm: | tgm4883: superm1: can we 'disappear' the 0.25 packages? :) |
| [22:24:05] | dekarl: | I got mixed success with users testing fixes + a patch by providing PPAs. |
| [22:24:23] | dekarl: | stuartm, sure in april :D |
| [22:25:09] | stuartm: | aww, how hard is it to accidentally type rm -f *0.25* ? |
| [22:29:16] | dekarl: | btw, anybody else seeing "mpeg/pespacket.cpp:161 (VerifyCRC) PESPacket: Failed CRC check 0x00000000 != 0xffffffff for StreamID = 0xXX"? http://pastebin.com/jTRBtwTN |
| [22:29:50] | stuartm: | with 0.26.1 tagged hopefully we'll convince a few more people to move, probably should have tagged it 6 months ago |
| [22:34:25] | rhpot199` is now known as rhpot1991 | |
| [22:34:42] | rhpot1991 (rhpot1991!~rhpot1991@2001:4968:202:3:759d:b77c:affe:a5c2) has quit (Changing host) | |
| [22:34:43] | rhpot1991 (rhpot1991!~rhpot1991@ubuntu/member/rhpot1991) has joined #mythtv | |
| [22:48:38] | Tobbe5178 (Tobbe5178!~asdf@h104n2-sv-a13.ias.bredband.telia.com) has quit (Read error: Connection reset by peer) | |
| [22:50:30] | stuartm: | dekarl: not here |
| [22:51:34] | stuartm: | grepped logs from the last week. not a single instances |
| [22:53:55] | knightr: | stuarm, if we skip the alpha in the future we must not shorten the time before the beta otherwise themers and translators will not have time to work on their themes/translations.. |
| [22:54:36] | stuartm: | knightr: yeah, I'm not suggesting a shorting phase, just combining the current alpha into the 'beta' |
| [22:54:45] | stuartm: | shorter |
| [22:54:49] | knightr: | dekarl ^ |
| [22:56:51] | knightr: | knightr, OK, thank you... |
| [22:57:04] | stuartm: | the 'alpha' designation just doesn't make sense, alpha is just the step before you invite people outside the dev team to test the software and with open source that's already happening, might as well skip to the beta |
| [22:57:31] | stuartm: | especially since 'alpha' is actually more likely to put people off testing |
| [22:57:50] | knightr: | how much time before the RC, do we stick with the original 1 week delay between beta and first RC? |
| [22:58:31] | stuartm: | knightr: we'll stay with one week, I consider that the minimum interval no matter how long we extend the previous phase |
| [23:00:47] | knightr: | yep, I agree... If anything we might have to push the RC back a couple of days depending on what comes up... |
| [23:01:22] | stuartm: | exactly |
| [23:03:09] | stuartm: | and because I'm apparently the only one with the time, and privileges to carry out the release atm, it's going to have to be worked around my schedule ;) |
| [23:04:44] | stuartm: | it's not the best time of year, everything is always quieter in the middle of summer – people have plans for the warm evenings that don't involve being inside sat in front of a computer |
| [23:07:50] | knightr: | stuartm, LOL.. It might be worth looking into what Daniel K had in mind to make the process easier to do and by a greater number of people... |
| [23:08:21] | knightr: | finding somebody to do it was a PITA, I know first hand... :) |
| [23:09:50] | peper03 (peper03!~peper03@mythtv/developer/peper03) has quit (Quit: Quit) | |
| [23:13:48] | peper03 (peper03!~peper03@mythtv/developer/peper03) has joined #mythtv | |
| [23:20:35] | skd5aner (skd5aner!~skd5aner@50-90-5-146.res.bhn.net) has quit (Read error: Connection reset by peer) | |
| [23:21:03] | skd5aner (skd5aner!~skd5aner@50-90-5-146.res.bhn.net) has joined #mythtv | |
| [23:22:00] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 256 seconds) | |
| [23:25:12] | IReboot (IReboot!~doug@CPE10bf48e67915-CM78cd8e7e342d.cpe.net.cable.rogers.com) has quit (Quit: Ex-Chat) | |
IRC Logs collected by
BeirdoBot.
Please use the above link to report any bugs.