| Friday, December 31st, 2010, 00:14 UTC | ||
| [00:14:17] | danielk22: | yes bzero should be replaced with memset. |
| [00:14:31] | danielk22: | which are you saying is depreciated usleep or nanosleep? |
| [00:22:35] | danielk22: | I guess sleeps depreciated in MythTV code when used where a wait condition would be more appropriate. But there are a number of legitimate uses for a *sleep(). AFAIK neither of those calls are depreciated in POSIX. |
| [01:00:43] | abqjp (abqjp!~abqjp@71-38-209-31.albq.qwest.net) has quit (Quit: abqjp) | |
| [01:43:53] | NightMonkey (NightMonkey!debian-tor@pdpc/supporter/professional/nightmonkey) has joined #mythtv | |
| [01:49:25] | JEDIDIAH__ (JEDIDIAH__!~jedi@cpe-76-185-72-21.tx.res.rr.com) has joined #mythtv | |
| [02:01:33] | AriX (AriX!~Ari@c-76-99-118-183.hsd1.pa.comcast.net) has joined #mythtv | |
| [02:06:27] | knightr: | !seen jya |
| [02:06:28] | MythLogBot: | jya was last seen 4 days 2 hours 11 minutes 27 seconds ago |
| [02:17:16] | plut0 (plut0!~cory@cpe-74-76-182-29.nycap.res.rr.com) has joined #mythtv | |
| [02:17:32] | plut0 (plut0!~cory@cpe-74-76-182-29.nycap.res.rr.com) has left #mythtv () | |
| [02:37:44] | pheld (pheld!~heldal@cl-5.osl-01.no.sixxs.net) has quit (Quit: Leaving.) | |
| [02:50:05] | knightr: | !seen stoth |
| [02:50:06] | MythLogBot: | stoth was last seen 49 days 4 hours 12 minutes 6 seconds ago |
| [03:07:08] | xris: | does mythvideo have groups like tv has recgroups? |
| [03:11:51] | danielk22: | knightr: You are more likely to see stoth in linuxtv. |
| [03:12:36] | knightr: | danielk22, I checked there too... Thanks for the suggestion though! |
| [03:13:14] | knightr: | (and Happy Holidays!) |
| [03:16:45] | wagnerrp (wagnerrp!~wagnerrp_@mythtv/developer/wagnerrp) has quit (Remote host closed the connection) | |
| [03:18:33] | wagnerrp (wagnerrp!~wagnerrp_@mythtv/developer/wagnerrp) has joined #mythtv | |
| [03:22:07] | mtrg (mtrg!~mtgr@unaffiliated/mtrg) has quit (Quit: Lost terminal) | |
| [03:25:39] | abqjp (abqjp!~abqjp@174-28-182-216.albq.qwest.net) has joined #mythtv | |
| [03:29:46] | Captain_Murdoch: | xris, not that I know of. |
| [03:30:26] | xris: | Captain_Murdoch: thanks. guess I'll consider "mythvideo" a recgroup in that sense, then. |
| [03:30:29] | Captain_Murdoch: | it has categories, but I don't know if you can group by those |
| [03:30:38] | xris: | well, this would be for mythweb |
| [03:31:02] | Captain_Murdoch: | I had patch way back when that would actually let you view your mythvideo videos under a virtual MythVideo recording group in the Watch Recordings screen. |
| [03:31:23] | Captain_Murdoch: | didn't seem to interest anyone at the time so I dropped it. |
| [03:31:51] | iamlindoro: | Categories are the MythVideo equivalent of recgroup |
| [03:32:07] | iamlindoro: | Videos can be grouped by anything one wishes, see MENU->Browse By |
| [03:32:17] | iamlindoro: | Genre is also a close, but different, concept |
| [03:37:35] | Captain_Murdoch: | danielk22, do you see any issues with http://mythtv.pastebin.ca/2032280 as a fix to #9209 |
| [03:42:29] | xris: | iamlindoro: perfect, thanks |
| [03:43:06] | xris: | yeah, I already have support planned for genre, since that should overlap with TV stuff |
| [03:43:59] | xris: | Captain_Murdoch: take a look at what I'm thinking about for mythweb... it stems from conversations we've had over the years about integrating TV and video into one screen |
| [03:44:13] | xris: | not sure how it would work in a screen UI, but it should work well for web. |
| [04:55:53] | AriX (AriX!~Ari@c-76-99-118-183.hsd1.pa.comcast.net) has quit (Remote host closed the connection) | |
| [05:24:18] | wagnerrp_ (wagnerrp_!~wagnerrp_@NR-FT1-66-42-240-159.fuse.net) has joined #mythtv | |
| [05:24:18] | wagnerrp_ (wagnerrp_!~wagnerrp_@mythtv/developer/wagnerrp) has joined #mythtv | |
| [05:24:18] | wagnerrp_ (wagnerrp_!~wagnerrp_@NR-FT1-66-42-240-159.fuse.net) has quit (Changing host) | |
| [05:26:29] | wagnerrp (wagnerrp!~wagnerrp_@mythtv/developer/wagnerrp) has quit (Ping timeout: 264 seconds) | |
| [08:40:18] | stoffel (stoffel!~quassel@p57B4DB25.dip.t-dialin.net) has joined #mythtv | |
| [09:30:05] | stoffel (stoffel!~quassel@p57B4DB25.dip.t-dialin.net) has quit (Ping timeout: 264 seconds) | |
| [09:45:44] | NightMonkey (NightMonkey!debian-tor@pdpc/supporter/professional/nightmonkey) has quit (Quit: Body blow! Body blow!) | |
| [10:18:51] | ** stuarta wonders what an overnight valgrind will reveal ** | |
| [10:34:36] | stuarta: | a leak here, a leak there |
| [10:43:14] | stoffel (stoffel!~quassel@p57B4DB25.dip.t-dialin.net) has joined #mythtv | |
| [10:46:53] | cocoa117 (cocoa117!~cocoa117@188-223-64-197.zone14.bethere.co.uk) has joined #mythtv | |
| [10:48:58] | balor (balor!~aidan@87.127.55.57) has joined #mythtv | |
| [11:11:35] | stuarta: | stuartm: know much about the recordingprofile code? |
| [11:12:04] | stuarta: | specifically it's use of ConfigurationDialog? |
| [11:13:20] | stuarta: | i ask because there's a comment at recordingprofile.cpp:1169 "// id and name will be deleted by ConfigurationGroup's destructor" however i can't see where ConfigurationGroup's destructor would be cleaning it up, and valgrind has flagged it as a leak |
| [11:15:57] | ** stuarta is playing plug the leaks ** | |
| [11:24:03] | stuartm: | stuarta: one or both are parented to a QObject, when QObject derived classes are deleted they iterate through their list of children and delete them |
| [11:24:49] | stuartm: | that said, although it appears one is not a leak the other I couldn't really be certain about |
| [11:25:24] | stuartm: | I was looking at this just 6 weeks ago ... |
| [11:25:26] | stuarta: | right, however there's a bunch of new's in the constructors without corresponding deletes in the destructor (if a destructor even exists) |
| [11:25:56] | stuarta: | i think i just need to make the destructors delete everything that's new'd in the constructor |
| [11:29:05] | stuarta: | hmmm. ConfigurationWizard, ConfigurationDialog etc. none of them are QObjects |
| [11:32:29] | ** stuarta scratches head ** | |
| [11:35:27] | stuartm: | stuarta: last time you did that you caused crashes and it was reverted |
| [11:36:23] | stuarta: | :-p |
| [11:36:46] | stuartm: | hang on a minute, I'll get that code up in front of me so I can remember what I'm talking about |
| [11:37:44] | stuarta: | https://github.com/MythTV/mythtv/blob/master/ . . . le.cpp#L1169 |
| [11:37:52] | stuarta: | that's where i'm starting from |
| [11:39:53] | stuartm: | "addChild(id);" < RecordingProfile is a QObject, so the QObject destructor will take responsibility for deleting id |
| [11:40:51] | stuarta: | and so too name |
| [11:42:02] | stuarta: | looks like basically everything gets the addChild treatment |
| [11:43:10] | stuartm: | name is added as a child of a ConfigurationGroup (profile) and profile is added as a child of RecordingProfile, so the cascading delete will get both |
| [11:43:42] | ** stuarta tries to work out why the new TranscodeLossless on :1209 is being flagged ** | |
| [11:44:13] | stuartm: | the real mystery is why Valgrind has problem with that code when QObject deletion is used all over the place |
| [11:44:54] | stuarta: | aye |
| [11:45:54] | stuarta: | maybe it just doesn't track it deep enough |
| [11:47:18] | stuartm: | that old configuration UI stuff is a real pain to follow, it's subclassing taken to it's illogical extreme |
| [11:47:50] | stuarta: | well if it's depreciated i won't break any more brain cells over it |
| [11:47:59] | stuarta: | and yes it's a royal pita to follow |
| [11:48:22] | stuartm: | although no-one really thinks that making every settings screen themable is a great idea, I'd love to get rid of that old stuff ASAP |
| [11:48:31] | stuarta: | i'm not too bothered by the subclassing tbh |
| [11:49:02] | stuarta: | it's just finding all the things that are QObjects and remembering the implications of that |
| [11:49:21] | stuarta: | even been using my copy of the class documentation to follow it! |
| [11:51:10] | stuartm: | http://miffteevee.co.uk/documentation/develop . . . onGroup.html |
| [11:51:22] | stuartm: | actually makes it seem reasonable |
| [11:52:12] | stuartm: | http://miffteevee.co.uk/documentation/develop . . . Filters.html |
| [11:52:25] | stuarta: | yeah, i've been hitting the copy on my site |
| [11:52:40] | stuartm: | there are actually better examples, which go much deeper |
| [11:55:14] | stuartm: | anyway, without those diagrams (which I'd forgotten about), tracing it all through is a nightmare :) Whenever fixing a bug in that code I end up with a dozen or more files open |
| [11:55:54] | ** stuarta was up to at least 12 tabs in FF ** | |
| [12:00:39] | stuarta: | is the deletion of a QObjects children an automatic thing? |
| [12:01:05] | stuartm: | yes |
| [12:01:29] | andreax (andreax!~andreaz@p57B95C9F.dip.t-dialin.net) has joined #mythtv | |
| [12:01:34] | ** stuarta gets confused ** | |
| [12:01:35] | stuartm: | http://doc.trolltech.com/4.5/qobject.html#dtor.QObject |
| [12:03:54] | stuarta: | okay, can anyone explain then, why ~ConfigurationGroup() explicitly delete's it's children? |
| [12:04:10] | stuarta: | well, disconnect() & deleteLater() |
| [12:04:40] | andreax (andreax!~andreaz@p57B95C9F.dip.t-dialin.net) has quit (Client Quit) | |
| [12:06:41] | andreax (andreax!~andreaz@p57B95C9F.dip.t-dialin.net) has joined #mythtv | |
| [12:10:08] | stuartm: | I've no idea, I've also no idea why it re-implements DeleteLater() |
| [12:11:40] | stuartm: | probably because it's re-implementing addChild – so it doesn't get to use the QObject stuff |
| [12:11:48] | stuartm: | but again, I've no idea why it would do that |
| [12:12:10] | stuarta: | maybe that's why valgrind can't follow it properly |
| [12:13:05] | stuartm: | seems to me that you could delete deleteLater, addChild and that destructor will no ill-effect, but I'm not going to touch it |
| [12:14:07] | ** stuarta looks for different leaks to attack ** | |
| [12:16:36] | ** stuarta laughs at the inheritance diagram on -> http://www.squashedfrog.net/mythtv/devdocs/cl . . . onGroup.html ** | |
| [12:18:00] | stuarta: | ugh |
| [12:18:06] | stuarta: | i now see what that does |
| [12:18:43] | stuarta: | it's overriding the QObject definition of children with its own |
| [12:18:52] | stuartm: | yup |
| [12:19:03] | stuarta: | why on earth would it do such a crazy thing |
| [12:19:20] | stuarta: | nothing is gained from that apart from use of standard templates rather than Qt |
| [12:21:18] | stuarta: | that said it does do some funky interaction with it's children |
| [12:48:15] | ** stuarta comes to the conclusion that valgrind can't follow qobject based deletes ** | |
| [12:54:27] | jjwin2k (jjwin2k!~jjwin2k@p4FC124D0.dip.t-dialin.net) has joined #mythtv | |
| [12:57:29] | stuarta: | stuartm: i'll give you a gold star when you rip out all that old code |
| [13:03:50] | jjwin2k (jjwin2k!~jjwin2k@p4FC124D0.dip.t-dialin.net) has quit (Quit: Leaving.) | |
| [13:10:51] | highzeth (highzeth!~hz@hoiseth.no) has quit (Ping timeout: 246 seconds) | |
| [13:12:05] | stoffel (stoffel!~quassel@p57B4DB25.dip.t-dialin.net) has quit (Ping timeout: 264 seconds) | |
| [13:22:58] | stuartm: | danielk22: usleep is deprecated in favour of nanosleep in POSIX 2001 |
| [13:23:12] | stuarta: | is it portable tho? |
| [13:23:43] | Kunalagon (Kunalagon!~Kunalagon@212.200.240.68) has joined #mythtv | |
| [13:24:58] | stuartm: | yeah, it's still part of the spec, it's use it just discouraged |
| [13:25:01] | stuartm: | is |
| [13:26:52] | stuartm: | nanosleep supposedly has some advantages |
| [13:29:37] | danielk22: | stuartm: They have different signal handling behavior. nanosleep can in theory sleep to a greater accuracy, but only if you've configured your kernel to have more than 1000 tics per second (which requires hand editing the Linux kernel config). & the API is more complex. |
| [13:30:26] | danielk22: | Neither one is really good at handling signals, there is an even more complicated clock_nanosleep() call for that. |
| [13:33:56] | stuartm: | hence 'supposedly', it was the deprecation which caused me to raise the issue not any benefits, tangible or otherwise of nanosleep over usleep |
| [13:34:31] | danielk22: | Heh, just reading up on it, usleep() is gone in POSIX.1–2008 ... |
| [13:34:37] | stuartm: | and I don't really mind keeping usleep, I just want to know which deprecated stuff we're aiming to replace |
| [13:36:08] | stuarta: | pretty much all of the supposed leaks are triggered from the recordingprofile constructor |
| [13:37:07] | stuartm: | stuarta: that was my experience when I spent time looking at it |
| [13:37:09] | danielk22: | stuartm: We should probably just write a usleep() replacement, since it's not available on windows anyway. |
| [13:37:42] | stuartm: | I thought Windows had some posix support? |
| [13:38:00] | danielk22: | stuarta: I haven't caught up on the backlog yet, but valgrind can not be trusted with QObject derived stuff. It doesn't know about deleteLater(). |
| [13:38:48] | stuarta: | no, i've been going throught it with a fine tooth comb and come to the conclusion |
| [13:39:04] | stuartm: | danielk22: actually it does seem to handle a lot of QObject stuff, but ConfigurationGroup overrides the QObject deleteLater, addChild and deletes it's own children for some reason |
| [13:39:05] | danielk22: | stuartm: It is an extra cost feature. You need either the Ultimate or Enterprise version. |
| [13:39:09] | stuarta: | there's a lot of stuff that should be handled by the qobject auto child deletion |
| [13:39:15] | stuartm: | danielk22: ahh |
| [13:39:44] | stuarta: | of valgrind? |
| [13:40:13] | stuartm: | stuarta: of Windows |
| [13:40:14] | danielk22: | stuarta: it needs to do that because it's sometimes added to the UI where it will auto delete, and sometimes not (i.e. in the backend). |
| [13:40:18] | stoffel (stoffel!~quassel@p57B4DB25.dip.t-dialin.net) has joined #mythtv | |
| [13:40:44] | stuartm: | any gotchas I should be aware of if I prepare a patch to replace every instance of bzero with memset? |
| [13:40:47] | stuarta: | ugh, nasty |
| [13:41:01] | stuartm: | since I _know_ bzero isn't available on windows |
| [13:41:21] | stuartm: | stuarta: well VS at least |
| [13:41:34] | danielk22: | stuarta: POSIX compliance in Windows. It was a free add on in 2000 and earlier releases, a $100 add on in XP and only available in those two editions in later releases. |
| [13:42:01] | ** stuarta adds that to the long list of why i hate windows ** | |
| [13:42:11] | stuartm: | they really don't like POSIX I guess |
| [13:42:21] | stuarta: | of course not, it's a standard |
| [13:42:29] | danielk22: | stuarta: bzero is only used in non-windows code at this point, so it's more of a cleanup. |
| [13:42:31] | stuartm: | doh, of course |
| [13:42:43] | danielk22: | stuartm: no it should be perfectly safe. |
| [13:42:59] | stuarta: | MS deal with standards by doing "Embrace, Extend, Extinguish" |
| [13:43:19] | stuarta: | Extend make their implementation incompatible with everyone elses |
| [13:43:36] | okolsi: | segfault in MNews, looks like MythUI related: http://mythtv.pastebin.ca/2035066 |
| [13:43:55] | stuarta: | read something yesterday that when they tried that with java, sun took back control of java on the windows platform |
| [13:44:17] | stuartm: | danielk22: I see it used all over the place including areas that I don't believe are linux only e.g. videooutputbase, the CC decoders, the file writer |
| [13:44:34] | stuarta: | look in the compat header |
| [13:44:51] | stuartm: | those probably compile because we have a bzero > memset wrapper in a windows only header |
| [13:45:45] | stuartm: | stuarta: aye, it's going to be there, but ultimately it's deprecated on linux too so we might as well replace the lot |
| [13:45:58] | stuarta: | cunning plan |
| [13:52:01] | stuarta: | since i'm about to wander off for the evening. Have a Good New Years everyone, and i'll see you all next year |
| [14:00:57] | stuartm: | you too |
| [14:16:53] | stuartm: | danielk22, jannau: I'm thinking that it would be nice to write some guidance on which standards should be followed into the coding standards, we've got this far without them so it's not especially important but it might help reduce the frequency of broken OSX/Windows builds etc |
| [14:19:21] | stuartm: | and fixing those issues gives me some no-thought-required busy work which is so much more productive than playing games of Solitaire |
| [14:29:10] | cocoa117 (cocoa117!~cocoa117@188-223-64-197.zone14.bethere.co.uk) has quit (Ping timeout: 250 seconds) | |
| [14:29:57] | cocoa117 (cocoa117!~cocoa117@188-223-64-197.zone14.bethere.co.uk) has joined #mythtv | |
| [15:59:51] | mrand (mrand!~mrand@ubuntu/member/mrand) has quit (Read error: Connection reset by peer) | |
| [16:16:35] | Lunar_Lamp (Lunar_Lamp!~Lunar_Lam@unaffiliated/lunarlamp/x-038437) has joined #mythtv | |
| [16:16:49] | Lunar_Lamp (Lunar_Lamp!~Lunar_Lam@unaffiliated/lunarlamp/x-038437) has left #mythtv () | |
| [16:25:50] | superm1: | xris, right, but that's not part of the standard working model, it's a bandaid in case of a fsck up. so if there's a break again, then just mark it to do a fresh start |
| [16:40:43] | stoffel (stoffel!~quassel@p57B4DB25.dip.t-dialin.net) has quit (Ping timeout: 260 seconds) | |
| [16:41:25] | stuartm: | Beirdo: any progress on the windows build slave(s)? |
| [16:44:17] | Steve_Goodey (Steve_Goodey!~steve@host109-158-208-60.range109-158.btcentralplus.com) has joined #mythtv | |
| [17:18:07] | Steve_Goodey (Steve_Goodey!~steve@host109-158-208-60.range109-158.btcentralplus.com) has quit (Remote host closed the connection) | |
| [17:27:58] | Steve_Goodey (Steve_Goodey!~steve@host109-158-208-60.range109-158.btcentralplus.com) has joined #mythtv | |
| [17:31:18] | mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv | |
| [17:52:09] | balor (balor!~aidan@87.127.55.57) has quit (Quit: Ex-Chat) | |
| [17:57:25] | okolsi: | stuartm: you have time to check bt, looks like mythui ^^^^ |
| [17:59:32] | stuartm: | looks suspiciously like the the ticket I just assigned to Captain_Murdoch |
| [17:59:55] | stuartm: | although without the second thread modifying QRegion at the same time |
| [18:00:36] | okolsi: | okay, thought to check before ticket cause I suspected the same |
| [18:00:59] | stuartm: | which theme? Shouldn't matter but .. |
| [18:01:51] | okolsi: | Blue Abstract, thanks to the on-line theme chooser :) |
| [18:01:55] | Steve_Goodey (Steve_Goodey!~steve@host109-158-208-60.range109-158.btcentralplus.com) has quit (Remote host closed the connection) | |
| [18:02:24] | okolsi: | just happened once, not a frequent thing |
| [18:32:02] | skd5aner (skd5aner!~skd5aner@cpe-069-132-082-180.carolina.res.rr.com) has quit () | |
| [18:49:10] | Kunalagon (Kunalagon!~Kunalagon@212.200.240.68) has quit (Quit: Leaving.) | |
| [18:49:28] | danielk22: | Captain_Murdoch: AFACIT It looks ok. (#9209 patch) |
| [19:10:17] | danielk22: | stuartm: https://github.com/MythTV/mythtv/commit/dd9e1ac08 <--- I tried this before and it caused segfaults. The problem is that every QString use involving MythEvent needs to be threadsafe before doing this is safe. |
| [19:10:58] | danielk22: | stuartm: There are a number of places still where multiple threads update the same QString instance without any locking. |
| [19:11:12] | stuartm: | hmm |
| [19:11:34] | Captain_Murdoch: | xris, I skimmed the doc, but haven't digested it yet. right now my brain is in fried, end of year mode taking a break. |
| [19:11:57] | danielk22 (danielk22!~danielk@96.57.9.142) has quit (Quit: Leaving.) | |
| [19:12:32] | stuartm: | danielk22: ok, I'll revert and add comments to prevent it being done again, seemed safe since message is immediately assigned and const |
| [19:16:28] | danielk22 (danielk22!~danielk@96.57.9.142) has joined #mythtv | |
| [19:31:18] | Captain_Murdoch: | stuartm, I see what's going on with that ticket you assigned and will work on a fix. |
| [19:32:07] | stuartm: | Captain_Murdoch: great, I was hoping that you might have some idea :) |
| [19:32:53] | stuartm: | I didn't really fancy working on it myself, so I'm sorry about that ;) |
| [19:33:55] | Captain_Murdoch: | yeah, when I did that animated image code, I had in the back of my mind that there was probably something I was missing, and sure enough there was. we do locking inside mythuiimage, but it can call back up to it's parent which is what's happening outside the UI thread. I need to do the same for these images as I did for the original animated image work, so that this part of the code gets called in the UI thread when we receive |
| [19:33:55] | Captain_Murdoch: | an image loaded event. |
| [19:41:26] | Dave123-road (Dave123-road!~dave@cpe-74-74-222-96.rochester.res.rr.com) has quit (Ping timeout: 265 seconds) | |
| [20:39:56] | highzeth (highzeth!~hz@hoiseth.no) has joined #mythtv | |
| [20:40:41] | danielk22: | Captain_Murdoch: I have another stuck mythcommflag and the file IS a zero byte file as confirmed on the command line. |
| [20:42:05] | danielk22: | backtrace: http://www.pastebin.ca/2035266 |
| [20:42:41] | danielk22: | This is using the mythtv-rec branch so line numbers may be a little off from trunk. |
| [20:44:33] | danielk22: | Looks like it may be a ringbuffer problem |
| [20:51:41] | rkulagow (rkulagow!~rkulagow@c-67-176-129-251.hsd1.il.comcast.net) has joined #mythtv | |
| [20:55:45] | rkulagow: | any git-masters around? |
| [20:58:14] | danielk22: | rkulagow: don't think so. but I may be of some minimal help. |
| [20:58:34] | iamlindoro: | One might instead ask whether git masters even exist :) |
| [20:58:48] | rkulagow: | :) |
| [20:59:16] | rkulagow: | maybe i should just send an email to linus. he might know it better than i do. |
| [20:59:53] | rkulagow: | i've read through the git users manual and i think i'm getting myself deeper into trouble. |
| [21:00:05] | rkulagow: | i've cloned master and been making trivial edits |
| [21:00:19] | rkulagow: | i've created a local branch |
| [21:00:41] | rkulagow: | merged in the changes from danielk's 8262 ticket |
| [21:00:44] | danielk22: | Captain_Murdoch: I think I see the problem. We're never releasing the rwlock in the RingBuffer::run() loop under a complicated set of conditons. |
| [21:01:05] | rkulagow: | but then noticed that people keep spelling "separated" incorrectly. |
| [21:01:48] | rkulagow: | so i did git checkout master |
| [21:02:15] | rkulagow: | and git status showed that i was back in master |
| [21:02:19] | rkulagow: | i made my fixes |
| [21:02:53] | rkulagow: | but when i go to do a commit, it looks like it's going to also push the changes that i don't want to make. |
| [21:03:02] | rkulagow: | and right now it will break master |
| [21:03:50] | rkulagow: | because mythbackend can't bootstrap itself into running yet without running mythtv-setup first |
| [21:04:05] | danielk22: | rkulagow: I believe what you want to do is create another clone of master, cherry pick this one change and then push it. |
| [21:08:57] | rkulagow: | back to TFM |
| [21:09:08] | xris: | ok, services.mythtv.org moved to new server |
| [21:11:22] | NightMonkey (NightMonkey!debian-tor@pdpc/supporter/professional/nightmonkey) has joined #mythtv | |
| [21:13:13] | rkulagow: | danielk: so maybe git branch temp HEAD |
| [21:13:25] | rkulagow: | git checkout temp |
| [21:14:56] | danielk22: | rkulagow: that might work, but I'm perplexed as to the first checkout didn't work. Did you have uncommitted changes in the tree? |
| [21:15:35] | rkulagow: | in the mythtvsetup-www branch that i have local? |
| [21:19:32] | rkulagow: | here's what happens when i do git checkout master |
| [21:19:36] | rkulagow: | Switched to branch 'master' |
| [21:19:38] | rkulagow: | Your branch is behind 'origin/master' by 5 commits, and can be fast-forwarded. |
| [21:19:40] | rkulagow: | rkulagow@mythtv-dev:~/mythtv$ git status |
| [21:19:42] | rkulagow: | # On branch master |
| [21:19:44] | rkulagow: | # Your branch is behind 'origin/master' by 5 commits, and can be fast-forwarded. |
| [21:19:46] | rkulagow: | # |
| [21:19:48] | rkulagow: | # Changes to be committed: |
| [21:19:50] | rkulagow: | # (use "git reset HEAD <file>..." to unstage) |
| [21:19:52] | rkulagow: | # |
| [21:19:54] | rkulagow: | # new file: mythtv/programs/mythbackend/config_backend_database.xml |
| [21:19:56] | rkulagow: | # new file: mythtv/programs/mythbackend/config_backend_general.xml |
| [21:19:58] | rkulagow: | # new file: mythtv/programs/mythbackend/httpconfig.cpp |
| [21:20:00] | rkulagow: | # new file: mythtv/programs/mythbackend/httpconfig.h |
| [21:20:02] | rkulagow: | # new file: mythtv/programs/mythbackend/mythsettings.cpp |
| [21:20:04] | rkulagow: | # new file: mythtv/programs/mythbackend/mythsettings.h |
| [21:20:06] | rkulagow: | # Changed but not updated: |
| [21:20:08] | rkulagow: | # (use "git add <file>..." to update what will be committed) |
| [21:20:10] | rkulagow: | # (use "git checkout — <file>..." to discard changes in working directory) |
| [21:20:12] | rkulagow: | # |
| [21:20:14] | rkulagow: | # modified: mythplugins/mytharchive/mytharchivehelper/main.cpp |
| [21:20:16] | rkulagow: | # modified: mythplugins/mythgame/mythgame/gamesettings.cpp |
| [21:20:18] | rkulagow: | # modified: mythplugins/mythvideo/mythvideo/scripts/jamu-example.conf |
| [21:20:20] | rkulagow: | # modified: mythtv/bindings/python/MythTV/tmdb/tmdb_api.py |
| [21:20:22] | rkulagow: | # modified: mythtv/contrib/development/MythXMLTest/CDS.html |
| [21:20:24] | rkulagow: | # modified: mythtv/libs/libmyth/mythcommandlineparser.cpp |
| [21:20:26] | rkulagow: | # modified: mythtv/libs/libmyth/mythcontext.cpp |
| [21:20:28] | rkulagow: | # modified: mythtv/libs/libmythdb/mythdb.cpp |
| [21:20:30] | rkulagow: | # modified: mythtv/libs/libmythdb/mythdb.h |
| [21:20:32] | rkulagow: | # modified: mythtv/libs/libmythtv/eitfixup.cpp |
| [21:20:34] | rkulagow: | # modified: mythtv/libs/libmythupnp/httprequest.cpp |
| [21:20:36] | rkulagow: | # modified: mythtv/programs/mythbackend/backendcontext.h |
| [21:20:38] | rkulagow: | # modified: mythtv/programs/mythbackend/main.cpp |
| [21:20:40] | rkulagow: | # modified: mythtv/programs/mythbackend/main_helpers.cpp |
| [21:20:42] | rkulagow: | # modified: mythtv/programs/mythbackend/main_helpers.h |
| [21:20:44] | rkulagow: | # modified: mythtv/programs/mythbackend/mainserver.cpp |
| [21:20:46] | rkulagow: | # modified: mythtv/programs/mythbackend/mediaserver.cpp |
| [21:20:48] | rkulagow: | # modified: mythtv/programs/mythbackend/mediaserver.h |
| [21:20:50] | rkulagow: | # modified: mythtv/programs/mythbackend/mythbackend.pro |
| [21:20:52] | rkulagow: | # modified: mythtv/programs/mythcommflag/README.txt |
| [21:21:22] | rkulagow: | so the issue is that my master seems to be a mix of changes that i want pushed back (like the mythplugins stuff) |
| [21:21:28] | danielk22: | rkulagow: please use pastebin.ca The mistake you made was to switch trees while you had uncommitted changes. |
| [21:21:41] | rkulagow: | ok |
| [21:21:55] | danielk22: | rkulagow: Are the changes you want to commit in a certain limited set of files? |
| [21:22:07] | rkulagow: | yes |
| [21:23:59] | danielk22: | I believe you can run "git reset HEAD <file>" to remove a files from your next commit then use "git add <file>" to add just the files you want committed. If you only want part of the changes in a file "git add -p <file>" will allow you to interactively decide what will be in the next commit. |
| [21:24:28] | rkulagow: | in a pinch i guess i could use git diff to capture the changes that i've made |
| [21:25:17] | rkulagow: | so what i _should_ have done: |
| [21:25:25] | danielk22: | Then you can run a "git status" and make sure only the files you want committed are in the top list of files. Then run "git commit" to commit those changes. Now when you push only those changes will be pushed. |
| [21:25:31] | rkulagow: | synced master using git pull |
| [21:25:47] | rkulagow: | created a branch locally |
| [21:26:00] | rkulagow: | made edits there for the 8262 patch |
| [21:26:08] | rkulagow: | realized that there were spelling errors in master |
| [21:26:21] | rkulagow: | did a git commit within my local branch |
| [21:26:35] | rkulagow: | then did a git checkout master to switch back to the top level |
| [21:26:42] | rkulagow: | made the spelling corrections there |
| [21:26:46] | rkulagow: | git commit |
| [21:26:50] | rkulagow: | git push |
| [21:27:31] | danielk22: | yup, and then switch back to the local branch, do a git pull and keep working... |
| [21:45:19] | Dave123-road (Dave123-road!~dave@74-37-180-238.dsl2-pixley.roch.ny.frontiernet.net) has joined #mythtv | |
| [22:01:29] | rkulagow: | ok, i think i'm better now |
| [22:02:45] | rkulagow: | so in my local branch, if i want to keep track of my changes and the ones done in master, i would use git merge? |
| [22:10:26] | danielk22: | rkulagow: git pull will merge in the master by default, if you use "git pull --rebase" it functions like "svn update" without the --rebase it applies upstream changes on top of your changes. |
| [22:11:23] | danielk22: | "on top of" vs. "temporarily unapplies your changes, then applies changes from master, then reapplies your changes" |
| [22:11:52] | bindi (bindi!~bindi@dsl-vsabrasgw1-fe03dc00-157.dhcp.inet.fi) has joined #mythtv | |
| [22:12:12] | bindi: | hey! is MythTV able to use multiple digituners? |
| [22:12:13] | danielk22: | Typically I don't think you call "git merge" or "git rebase" explicitly. |
| [22:12:28] | bindi: | say I have 2 tuners, and I set it to record 2 different programs, from 2 different channels, at the same time. |
| [22:12:36] | bindi: | Would it use card 1 for program 1 and card 2 for program 2? |
| [22:12:38] | danielk22: | bindi: wrong channel, see topic. |
| [22:12:41] | bindi: | oh? D: |
| [22:12:44] | bindi: | damn. |
| [22:13:25] | bindi: | excuse me then :< |
| [22:13:26] | bindi (bindi!~bindi@dsl-vsabrasgw1-fe03dc00-157.dhcp.inet.fi) has left #mythtv () | |
| [22:26:30] | Beirdo: | stuartm: no progress on the windows side yet. I guess I could take a look at that today. I think it was kenni that was planning on running one for us? |
| [22:26:42] | balor (balor!~aidan@87.127.55.57) has joined #mythtv | |
| [22:27:26] | balor: | In the master backend configuration, should you set mysql to listen to the public IP or on 127.0.0.1? |
| [22:28:44] | wagnerrp: | balor: this is the development channel |
| [22:30:27] | balor (balor!~aidan@87.127.55.57) has quit (Read error: Connection reset by peer) | |
| [22:30:44] | balor (balor!~aidan@87.127.55.57) has joined #mythtv | |
| [22:40:37] | balor (balor!~aidan@87.127.55.57) has quit (Read error: Connection reset by peer) | |
| [22:40:39] | balor_ (balor_!~aidan@87.127.55.57) has joined #mythtv | |
| [22:42:32] | kormoc_afk is now known as kormoc | |
| [22:43:56] | danielk22: | stuartm: Nokia makes no promise that QString is thread-safe. The difference between Qt3 and Qt4 is that after an "QString a("A"), b=a;" it is now safe to use a and b in different threads. |
| [22:46:49] | danielk22: | But "a" still can't be used in more than one thread at a time without locking, and "b" still can't be used in more than one thread at a time without locking. |
| [23:02:33] | balor_ is now known as balor | |
| [23:07:52] | rkulagow (rkulagow!~rkulagow@c-67-176-129-251.hsd1.il.comcast.net) has quit (Ping timeout: 250 seconds) | |
| [23:19:53] | cocoa117 (cocoa117!~cocoa117@188-223-64-197.zone14.bethere.co.uk) has quit (Ping timeout: 264 seconds) | |
| [23:20:41] | cocoa117 (cocoa117!~cocoa117@188-223-64-197.zone14.bethere.co.uk) has joined #mythtv | |
| [23:23:11] | andreax1 (andreax1!~andreaz@p57B946E9.dip.t-dialin.net) has joined #mythtv | |
| [23:24:07] | balor (balor!~aidan@87.127.55.57) has quit (Read error: Connection reset by peer) | |
| [23:24:15] | andreax (andreax!~andreaz@p57B95C9F.dip.t-dialin.net) has quit (Ping timeout: 265 seconds) | |
| [23:24:16] | balor (balor!~aidan@87.127.55.57) has joined #mythtv | |
| [23:25:07] | kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has left #mythtv () | |
| [23:27:22] | balor (balor!~aidan@87.127.55.57) has quit (Read error: Connection reset by peer) | |
| [23:27:38] | balor (balor!~aidan@87.127.55.57) has joined #mythtv | |
| [23:49:43] | cocoa117 (cocoa117!~cocoa117@188-223-64-197.zone14.bethere.co.uk) has quit (Read error: Connection reset by peer) | |
IRC Logs collected by
BeirdoBot.
Please use the above link to report any bugs.