| Thursday, August 29th, 2013, 00:02 UTC | ||
| [00:02:55] | stuartm: | oh ffs, I know what happened, someone pushed it to github mirror (or clicked on the github pull link) instead of committing it to the main tree |
| [00:03:12] | stuartm: | that's why it's there in github but not in our trees |
| [00:04:18] | stuartm: | #7486 |
| [00:04:18] | ** MythLogBot http://code.mythtv.org/trac/ticket/7486 ** | |
| [00:05:16] | stuartm: | it's a very short list of people who still have perms to action pull requests or push to github |
| [00:06:26] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 256 seconds) | |
| [00:11:43] | stuartm: | stuarta: that would be you, or Beirdo |
| [00:15:52] | stuartm: | for those who have asked in the past why they don't have perms to handle pull requests – _this_ is the reason |
| [00:27:39] | dmfrey (dmfrey!~dmfrey@65-78-98-83.c3-0.smt-ubr1.atw-smt.pa.cable.rcn.com) has joined #mythtv | |
| [00:28:17] | knightr: | jpabq, stuarta, as soon as the first RC is cut we are in hard string freeze. Before the first RC we are in soft string freeze, as long as you don't add too many strings it's OK. If possible ping me to let me know you added new strings. |
| [00:30:17] | knightr: | hard string freeze means no new strings... |
| [00:41:51] | jya (jya!~jya@1.138.30.224) has joined #mythtv | |
| [00:43:35] | jya: | jpabq: The IPTV stuff is jut udp/rtp and if http it's assumed it's http live streaming. It you link to a mpegts file instead, those assumptions would have to be tweaked.. |
| [00:43:55] | ** jya in the plane, catching up on the last bugs. ** | |
| [00:47:46] | knightr: | stuarta, if you really have to get a new string in and can't get it in before the RC do it before I contact the translators. Since Stuart M doesn't seem to have cut the RC and he's most likely asleep (like you most likely are) that means that the earliest I could contact them (assuming the RC was cut Thursday) would be in 24 hours... |
| [00:49:37] | knightr: | stuartm, maybe nobody should have perms to that repo with their regular users... |
| [00:49:40] | knightr (knightr!~Nicolas@mythtv/developer/knightr) has left #mythtv ("Leaving") | |
| [00:49:57] | knightr (knightr!~Nicolas@mythtv/developer/knightr) has joined #mythtv | |
| [00:50:46] | jya: | gigem: I'll have a look at db logging... Don't see why it would work with mythlogserver but not without. It's the same code... I admit.. I didn't test it.. The --disable-mythlogserver primary intent was for the mac build as mythlogserver doesn't work there at all, that and I can't compile zeromq in universal mode: |
| [00:50:56] | jya: | Always linking errors |
| [00:51:18] | knightr: | when I do admin stuff at work I use another account (created for me) that has admin accesses my regular account doesn't have... |
| [00:53:19] | jya: | gigem: Vaapi works well. I've fixed a couple of bugs with it in 0.27 (wasn't working for live TV) de interlacing is okay-ish. Not as good a vdpau. But it works surprisingly we |
| [00:53:31] | jya: | Otherwise.. Certainly a usable options on my hd3000 |
| [00:54:35] | jya: | stuartm: You need a HD intel adapter: so i3 with hd3000 and up. Won't work with the previous intel ISP GMAXXX |
| [00:59:01] | jya: | Tobbe5178: I've reworked all that metadata thingy.. It shouldn't crash anymore... If some metadata were running in the background and was updating elements: it coUld have crashed if the parent changed... Now it's all checked up properly. |
| [00:59:42] | knightr: | stuartm, if one of us wanted to work on BCM70012 support I think I have one lying somewhere... |
| [01:00:13] | jya: | stuartm: I have a whole myth development setup with me.. Even took slow USB stick so I can reproduce the bugs Paul-h found in the writer/reader... |
| [01:01:41] | jya: | While at it; danielk22 you changed the TFW write buffer from 2MB to 128MB... That's an insanely huge value.. And worse, the TFW code now only works if the call to write doesn't overflow the buffer. It used to simply wait until there was more room available. |
| [01:03:06] | jya: | Have you changed the behaviour in the recorder, and it's expected to always have space in the buffer and never block? I'm keen otherwise to implement a wait for space and reduce the buffer... There's always several writer being instantiated at any one time.. 128MB for each is massive. 2–4MB is more reasonable for a write buffer. |
| [01:03:22] | jya: | All right, being asked to turn off my iPad... See ya |
| [01:08:18] | jya (jya!~jya@1.138.30.224) has quit (Ping timeout: 264 seconds) | |
| [01:30:04] | jheizer: | I am workign on updating MobileMyth to account for service api changes in .27. Has there been any thought to adding a Mythtv version number to the services API for a client to automatically know which version of the backend it is accessing? |
| [01:31:22] | jheizer: | By that I mean a call that returns it. |
| [01:31:37] | stichnot (stichnot!~stichnot@adsl-68-127-103-225.dsl.pltn13.pacbell.net) has joined #mythtv | |
| [01:31:37] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv | |
| [01:31:37] | stichnot (stichnot!~stichnot@adsl-68-127-103-225.dsl.pltn13.pacbell.net) has quit (Changing host) | |
| [01:35:56] | amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has quit (Quit: Konversation terminated!) | |
| [01:38:31] | dblain: | jheizer: I've thought about it, but was going to wait until .28 since it's not a bug fix. |
| [01:40:09] | jheizer: | Ok, no problem. |
| [01:40:18] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Quit: Body blow! Body blow!) | |
| [01:40:40] | jheizer: | Not like I have updated it in forever. Went to try soemthing on it last night and saw it was throwing errors. |
| [01:42:16] | jheizer: | I believe only 2 are effected for me. Also good call on the Titles one. I ended up caching the entire recordings list so I can use it to generate the titles list quicker. |
| [01:42:34] | jheizer: | Can use the wsdl version. |
| [01:45:54] | gigem: | jya: The db logging segfault might be a red herring. I did a complete clean, configure and build and it didn't happen anymore. |
| [01:50:29] | gigem: | jya: I know about the deinterlacing deficiencies with VAAPI. I'll probably wait until the motion adaptive support for Haswell is ready before taking a serious look at it again. In my latest dabbling, I was quickly turned off by some decode corruption that didn't occur with VDPAU. I need to give software decoding with OpenGL rendering a try since my desktop is more than powerful enough to do it. I'd like |
| [01:50:31] | gigem: | to move the GT430 in it to my main frontend since it's on-board GT9300 can't do Advanced deinterlacing. |
| [02:03:11] | jheizer: | dblain: I believe when we spoke in the past you said that you had used .Net as a testing platform. Do you happen to any tips on casting/converted a .27 wsdl return object to a .26 one? I don't see an easy way to modify the generated reference to return the previously made type. |
| [02:13:28] | dmfrey (dmfrey!~dmfrey@65-78-98-83.c3-0.smt-ubr1.atw-smt.pa.cable.rcn.com) has quit (Quit: Ex-Chat) | |
| [02:24:28] | jheizer: | nm, used reflectino to duplicate the objects |
| [02:24:32] | jheizer: | reflection |
| [02:36:31] | _nyloc_ (_nyloc_!~quassel@p57B4FFC8.dip0.t-ipconnect.de) has joined #mythtv | |
| [02:40:54] | nyloc (nyloc!~quassel@p3EE2D1CA.dip0.t-ipconnect.de) has quit (Ping timeout: 276 seconds) | |
| [02:52:05] | peper03 (peper03!~peper03@mythtv/developer/peper03) has quit (Ping timeout: 240 seconds) | |
| [02:57:22] | peper03 (peper03!~peper03@mythtv/developer/peper03) has joined #mythtv | |
| [03:19:53] | superm1: | dekarl: FYI i had meant to add this for a while. https://github.com/MythTV/packaging/commit/18 . . . 7238e0e5ff1d thta's why I wanted to prefix all stuff commited to ubuntu packaging with deb: |
| [03:21:06] | joki (joki!~joki@p54863F1A.dip0.t-ipconnect.de) has quit (Ping timeout: 264 seconds) | |
| [03:27:01] | joki (joki!~joki@p5486380C.dip0.t-ipconnect.de) has joined #mythtv | |
| [03:33:46] | fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has quit (Ping timeout: 245 seconds) | |
| [03:35:18] | fetzerch (fetzerch!~quassel@unaffiliated/fetzerch) has joined #mythtv | |
| [04:19:04] | robink (robink!~quassel@unaffilated/robink) has quit (Quit: No Ping reply in 180 seconds.) | |
| [04:19:32] | robink (robink!~quassel@unaffilated/robink) has joined #mythtv | |
| [05:17:56] | dekarl: | stuarta re #11701 this "uses a ONID thats not assigned to them" has been brought up in the discussion but didn't make it to the ticket :( |
| [05:17:56] | ** MythLogBot http://code.mythtv.org/trac/ticket/11701 ** | |
| [05:23:28] | dekarl: | stuarta doh, it actually got commited? but is hanging in the void? (the parent made it to master and 0.27-beta) |
| [05:25:52] | dekarl: | superm1: sorry that I slipped some in without :/ |
| [05:27:50] | dekarl: | stuarta: the iptv recorder not caring about service_id has never been an issue because you only use MPTS around broadcasters (and some hacks). But the easy fix to just use "whatever the correct service_id is" assumes that there will be only one entry in the PAT instead of one service entry in the PAT. |
| [05:28:33] | dekarl: | Instead of complicating things for everyone I'd prefer to see the mess of "it has a service_id but it is no service" handled in the PAT class. |
| [05:28:47] | dekarl: | And fwiw the patch makes the M3U unsexy :D |
| [05:31:20] | dekarl: | lets stay with one kind of #EXTMYTHTV: followed by key=value stuff instead of adding another #EXTMYTHTV2: with another key |
| [05:36:54] | dekarl: | Tobbe5178: from going back to out discussion you need the service_id because the dreambox has not PAT rewrite (its broken) and outputs the full PAT instead of only the PAT that belongs to the channel its tuned to. |
| [05:38:42] | dekarl: | stuarta: wrt http://code.mythtv.org/trac/ticket/11791#comment:3 see above. autodetection is broken on these streams... but it is an easy fix |
| [05:39:37] | dekarl: | MuMuDVBs does it right with the PAT/PMT rewrite function |
| [05:41:37] | dekarl: | to sum it up, I'm happy to see the M3U enhanced to support setting the program_number/service_id but would prefer to do it nice and unbreak the two use cases that have been brought up where its a workaround bugs instead of a neccesary part of the design. (that would be MPTS at tv playout centers and similar) |
| [06:04:03] | FabriceMG (FabriceMG!~Thunderbi@217.112.59.207) has joined #mythtv | |
| [06:11:32] | FabriceMG (FabriceMG!~Thunderbi@217.112.59.207) has quit (Ping timeout: 268 seconds) | |
| [06:12:20] | Tobbe5178: | dekarl: yes, the dreambox most likely just filters the pids and then outputs them as is over http |
| [06:12:45] | Tobbe5178: | i would be surprised if it does any rewriting since the streaming service is very small and simple |
| [06:14:26] | dekarl: | Tobbe5178: I don't remember if we looked at the actual PAT to see how many entries it has... the proper fix is to copy PAT filtering from MuMuDVB to the Dreambox's streaming component, but I'm happy with the workaround of specifying the service_id in the M3U. But I'd like to avoid polluting the namespace even more when we already have lots of parameters in #EXTMYTHTV:key=value |
| [06:15:14] | dekarl: | can you dump the first few KB of such a stream to a file and let dvbsnoop or mythutil print out the PAT? should be a straight forward test |
| [06:15:15] | Tobbe5178: | i have no objection to putting it in some other tag in the m3u file, thats fine by me |
| [06:15:26] | Tobbe5178: | lets see... |
| [06:16:30] | Tobbe5178: | what pid is the pat on? |
| [06:16:32] | Tobbe5178: | 0x10 ? |
| [06:16:37] | Tobbe5178: | or was it something with 31 |
| [06:18:10] | SteveGoodey (SteveGoodey!~steve@host217-42-221-180.range217-42.btcentralplus.com) has joined #mythtv | |
| [06:18:49] | dekarl: | The PAT goes to PID 0. |
| [06:19:28] | dekarl: | The NIT goes to PID 0x10 by default unless you specify another PID in the PAT with the pseudo-program number 0. Which was mentioned in your ticket, too, and is part of the confusion :) |
| [06:19:49] | ** dekarl wanders off to work ** | |
| [06:23:32] | ** Tobbe5178 too ** | |
| [06:23:59] | Tobbe5178: | i'll see if i can create a dump, had some problems now, dvbsnoop output was just a mess |
| [07:11:33] | FabriceMG (FabriceMG!~Thunderbi@217.112.59.207) has joined #mythtv | |
| [07:35:10] | stuarta: | knightr: there would only be one string |
| [07:37:35] | stuarta: | knightr: it would be the one in #7486, though it looks like i (and maybe dekarl) need to test it some more |
| [07:37:35] | ** MythLogBot http://code.mythtv.org/trac/ticket/7486 ** | |
| [07:42:11] | stuarta: | stuartm: i'll see if i can work out how that commit made it into github |
| [07:45:38] | stuarta: | i may be able to cherry pick it back, but i'm not sure if that'll break github |
| [08:05:05] | Merlin83b (Merlin83b!~Daniel@office.34sp.com) has joined #mythtv | |
| [08:13:18] | stuarta: | Beirdo: care to comment on what would break, if a commit that exists only in github were cherry picked back to our git master? |
| [08:47:07] | amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has joined #mythtv | |
| [08:53:45] | jams (jams!~jams@cpe-24-92-85-1.wi.res.rr.com) has quit (Ping timeout: 245 seconds) | |
| [08:57:09] | dgeary2 (dgeary2!~debian@pa49-187-94-200.pa.nsw.optusnet.com.au) has joined #mythtv | |
| [09:01:39] | dgeary2_ (dgeary2_!~debian@pa49-187-86-189.pa.nsw.optusnet.com.au) has joined #mythtv | |
| [09:03:33] | dgeary2 (dgeary2!~debian@pa49-187-94-200.pa.nsw.optusnet.com.au) has quit (Ping timeout: 248 seconds) | |
| [09:13:23] | dgeary2_ is now known as dgeary2 | |
| [09:32:41] | jams (jams!~jams@cpe-24-92-85-1.wi.res.rr.com) has joined #mythtv | |
| [09:33:27] | dgeary2 (dgeary2!~debian@pa49-187-86-189.pa.nsw.optusnet.com.au) has quit (Ping timeout: 260 seconds) | |
| [09:57:40] | stuartm: | sphery: since you could reproduce it, maybe you'll test 0.27? http://code.mythtv.org/trac/ticket/9954 |
| [10:00:08] | stuartm: | anyone have an HDHomeRun? |
| [10:22:53] | IReboot (IReboot!~doug@CPE10bf48e67915-CM78cd8e7e342d.cpe.net.cable.rogers.com) has quit (Remote host closed the connection) | |
| [10:22:55] | jarryd (jarryd!jarryd@im.jarryd.net) has quit (Ping timeout: 245 seconds) | |
| [10:25:45] | IReboot (IReboot!~doug@CPE10bf48e67915-CM78cd8e7e342d.cpe.net.cable.rogers.com) has joined #mythtv | |
| [10:33:39] | jarryd (jarryd!~jarryd@im.jarryd.net) has joined #mythtv | |
| [10:39:42] | peper03: | Anyone tried MythArchive on master? When I try to create an ISO image, mythffmpeg sits in a loop doing *something* but not converting. According to top, CPU usage is around 115% but nothing gets written. Same if I run mythffmpeg run from the command line. mythffmpeg from 0.26-fixes works fine. |
| [10:40:04] | FabriceMG (FabriceMG!~Thunderbi@217.112.59.207) has quit (Ping timeout: 264 seconds) | |
| [10:42:40] | peper03: | paul-h mentioned something about a problem with ffmpeg that had been fixed but I'm getting this even after pulling the latest changes and a dist-clean. |
| [10:43:19] | stuarta: | hmmm, might have to try later tonight |
| [10:45:16] | SteveGoodey (SteveGoodey!~steve@host217-42-221-180.range217-42.btcentralplus.com) has quit (Quit: Konversation terminated!) | |
| [11:15:00] | svalcken (svalcken!~svalcken@dslb-084-059-107-242.pools.arcor-ip.net) has quit (Read error: Connection reset by peer) | |
| [11:27:49] | FabriceMG (FabriceMG!~Thunderbi@217.112.59.207) has joined #mythtv | |
| [11:29:14] | stuartm: | sphery: note the UTF-8 codeset warning in this log http://code.mythtv.org/trac/attachment/ticket/10631/mheg.log |
| [11:31:11] | knightr: | stuarta, you could get those two strings in right now if you wish. Just add some dummy code that invoke QObject::tr() with the text of those two strings and "comment it out" using "#if 0". |
| [11:33:13] | knightr: | The Qt tools don't understand preprocessor directives so they will extract the strings just as if they were in code that wasn't "commented out" using "#if 0"... |
| [11:34:41] | knightr: | Normally I wouldn't recommended using QObject::tr() but I haven't changed the one in that file yet which I will do post 0.27... |
| [11:36:48] | amessina (amessina!~amessina@2001:470:c1dc:7779:d6be:d9ff:fe8d:7c1e) has quit (Quit: Konversation terminated!) | |
| [11:39:14] | knightr: | Since it's in QObject it doesn't even have to be in the same class, it just has to be in the libraries or in MythFrontend (they use the same translation file). The only restriction is that you must not create a new directory for the source file you put those in (if you add a new directory you must tell me as I must change the config files used by the Qt translation tools...) |
| [11:42:29] | stuarta: | knightr: okay, that's interesting |
| [11:45:09] | stuarta: | i might just do that, as it solves the translation issue, whilst still allowing the code to be worked on |
| [11:46:27] | knightr: | yep, that would be your best bet in this situation I believe... |
| [12:07:40] | jarryd (jarryd!~jarryd@im.jarryd.net) has quit (Ping timeout: 264 seconds) | |
| [12:37:54] | FabriceMG (FabriceMG!~Thunderbi@217.112.59.207) has quit (Quit: FabriceMG) | |
| [12:38:14] | FabriceMG (FabriceMG!~Thunderbi@LCaen-156-54-30-212.w80-11.abo.wanadoo.fr) has joined #mythtv | |
| [12:56:37] | jarryd (jarryd!jarryd@im.jarryd.net) has joined #mythtv | |
| [13:13:16] | Jordack (Jordack!~Jordack@h69-131-44-221.plmomi.dedicated.static.tds.net) has joined #mythtv | |
| [13:31:25] | FabriceMG (FabriceMG!~Thunderbi@LCaen-156-54-30-212.w80-11.abo.wanadoo.fr) has quit (Ping timeout: 256 seconds) | |
| [13:40:51] | knightr: | stuarta, the idea of commenting out with "#if 0" is obviously especially useful if you decide to put the real code in but comment it out. As long as it's "commented out" with "#if 0" the strings will string be extracted. The reason why I was suggesting putting the strings someplace else (than the real source file they should be in) is to avoid possible problems when you resolve the conflict between the r |
| [13:40:51] | knightr: | epo (cherry picking most likely work like applying a diff file, the code around the area affected by the patch must likely be identical or at least trivially mergeable). |
| [13:41:32] | stuarta: | knightr: oh, i'll just split the patch up. no drama at all |
| [13:42:32] | knightr: | I thought the github repo "knew" it was a copy/slave of the other repo and simply modifying the file in our repo would simply force our repo version of the file to the github repo but I guess it's not as simple as that... |
| [13:43:35] | superm1: | dekarl: hmm maybe i should just have it do git log on the deb directory itself and then no need to put that prefix in the future |
| [13:44:43] | knightr: | stuarta, perfect. gotta go, ttyl |
| [13:56:20] | stuartm: | knightr: github doesn't offer such functionality, it assumes it is the master repo, not merely a read-only mirror |
| [14:16:28] | FabriceMG (FabriceMG!~Thunderbi@LCaen-156-54-30-212.w80-11.abo.wanadoo.fr) has joined #mythtv | |
| [14:51:54] | sphery: | stuartm: I can't say I can reproduce the problem in #9954 , but the drivers for my HD-3000's can get into a corrupt state where once they try to tune a no-longer-existent channel (i.e. from channels moving around since my scan), the card is unusable until a reboot (possibly could recover with module unload/reload, but never tried). |
| [14:51:54] | ** MythLogBot http://code.mythtv.org/trac/ticket/9954 ** | |
| [14:52:12] | sphery: | stuartm: If it's that problem, I'd say it's an upstream bug in the drivers |
| [15:01:40] | sphery: | stuartm: as for #10631 , that user has a bad environment where they've specified en_IE.utf8 rather than en_IE.utf-8 . "utf8" isn't a valid charmap on any system I've seen. You can check your system with: localedef --help (which should end with something like, "System's directory for character maps : /usr/share/i18n/charmaps"), then use that value to: find /usr/share/i18n/charmaps -iname '*utf8*' -print |
| [15:01:40] | ** MythLogBot http://code.mythtv.org/trac/ticket/10631 ** | |
| [15:03:24] | sphery: | (note, though, that you do get a result with: find /usr/share/i18n/charmaps -iname '*utf-8*' -print ) |
| [15:05:17] | sphery: | if distros are creating a (non-standard) utf8.gz link or something, we can allow it, but I wrote the code to warn if /anything/ is wrong that may confuse Qt, since we're not in control of the code that decides what charmap to use (especially since the Qt code that chooses is a huge mess, because of the lack of standardization and code creep that's occurred over the decades of *nix usage) |
| [15:06:01] | sphery: | so I'd prefer leaving the warning and being pedantic about their environment variable specification... if they really feel that they must use utf8 , instead of utf-8 , they can and they can live with the warning |
| [15:10:01] | Chutt_ is now known as Chutt | |
| [15:45:42] | FabriceMG (FabriceMG!~Thunderbi@LCaen-156-54-30-212.w80-11.abo.wanadoo.fr) has quit (Quit: FabriceMG) | |
| [15:49:20] | dekarl-work (dekarl-work!51c8c614@gateway/web/freenode/ip.81.200.198.20) has joined #mythtv | |
| [15:50:22] | dekarl-work: | sphery: smolt says we have en_US.utf8 on Top 11 of the language charts |
| [15:52:36] | stichnot (stichnot!~stichnot@207.198.105.23) has joined #mythtv | |
| [15:52:36] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv | |
| [15:52:36] | stichnot (stichnot!~stichnot@207.198.105.23) has quit (Changing host) | |
| [15:56:32] | stichnot: | stuartm: For drawing text outlines efficiently, do you know if anyone ever considered using QPainterPath::drawText() and then QPainter::drawPath() with a wide pen? I'm trying that now and so far it looks promising. |
| [15:57:00] | stichnot: | It only requires drawing the text once, instead of 8*outlinesize times. |
| [15:58:17] | danielk22: | stichnot: I don't think that was ever considered. |
| [16:00:32] | stichnot: | Cool. It gives what I think is a nicer, more rounded outline, rather than the squarish outline. |
| [16:01:00] | stuarta: | that's got to help performance |
| [16:01:09] | stichnot: | yep |
| [16:03:27] | stuartm: | stichnot: yes, it was first considered 2–3 days ago when I proposed it to you as a solution :) |
| [16:05:02] | dekarl-work (dekarl-work!51c8c614@gateway/web/freenode/ip.81.200.198.20) has quit () | |
| [16:05:46] | stuartm: | http://irc.mythtv.org/ircLog/channel/4/2013-08-26:20:47:57 |
| [16:06:00] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 268 seconds) | |
| [16:06:12] | stuarta: | haha |
| [16:06:53] | stuarta: | but how rude of him to ping out just as you pointed that out |
| [16:08:00] | dekarl (dekarl!~dekarl@p4FCEEB64.dip0.t-ipconnect.de) has quit (Read error: Operation timed out) | |
| [16:08:18] | dekarl (dekarl!~dekarl@p4FCEEBD2.dip0.t-ipconnect.de) has joined #mythtv | |
| [16:14:38] | superm1: | thanks for squeezing that in stuartm |
| [16:17:56] | stichnot (stichnot!~stichnot@216.239.45.84) has joined #mythtv | |
| [16:17:56] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv | |
| [16:17:56] | stichnot (stichnot!~stichnot@216.239.45.84) has quit (Changing host) | |
| [16:18:25] | stichnot: | sorry stuart[am] my bus "suddenly" arrived |
| [16:23:07] | stichnot: | so it seems that if you use QPainterPath::addText() and then draw the path with Qt::NoBrush, performance is good, but it's bad if you use e.g. Qt::SolidPattern, which would match my expectations |
| [16:24:32] | jya (jya!~jyavenard@mythtv/developer/jya) has joined #mythtv | |
| [16:25:43] | jya: | danielk22: are you there? |
| [16:26:36] | jya: | in was reviewing ThreadedFileWriter during the flight... and there are things that greatly concerns me in there.. |
| [16:26:56] | rsiebert_ (rsiebert_!~quassel@g225052117.adsl.alicedsl.de) has joined #mythtv | |
| [16:27:09] | jya: | 1- you can never write in a single time an amount of data greater than the buffer. it would fail and prevent all future write |
| [16:28:49] | jya: | 2- a pointer to the data passed on by write is stored in buffer and used later in the writing thread... If the data is being provided to TFW::Write faster than it can be written to disk; and should the caller re-used the buffer. the data will be overwritten by the time it's processed in the writing thread... |
| [16:29:46] | jya: | (we really should copy that data in another buffer an use that one instead... we can't rely on the pointer provided to TFW::Write to be still valid by the time it's processed later |
| [16:29:51] | stichnot: | the only annoyance with the QPainterPath::addText approach is that QPainter::drawText() has these nice flags that handle alignment for you, whereas QPainterPath lacks them so alignment has to be figured out manually |
| [16:30:19] | rsiebert (rsiebert!~quassel@g231186239.adsl.alicedsl.de) has quit (Ping timeout: 260 seconds) | |
| [16:31:01] | jya: | 3- Why bother with wait at all in the writing thread. the test if ((tot < sz) && !in_dtor) |
| [16:31:02] | jya: | bufferHasData.wait(locker.mutex(), 50); |
| [16:31:03] | jya: | tot will be == to sz in 99% of the cases... and why would you wait 50ms when we're trying to write as fast as possible |
| [16:34:09] | sphery: | dekarl: I'd still want to see a distro that actually adds that non-standard/not-part-of-glibc-or-eglibc charmap name before we got rid of the warning, and even then (since it's still non-standard and a standard utf-8 should /always/ work), I still think the warning is appropriate. (In other words, common misconfiguration doesn't make it a valid configuration--and if there is no utf8.gz file on the system, Qt is likely to pick up the wrong ... |
| [16:34:16] | sphery: | ... charmap.) |
| [16:34:32] | sphery: | I'd guess a ton of people are either using the same broken mythtv start script or copied/pasted from the same broken web page |
| [16:35:04] | sphery: | FWIW: glibc: http://sourceware.org/git/?p=glibc.git;a=tree . . . ata/charmaps and eglibc: http://www.eglibc.org/cgi-bin/viewvc.cgi/trun . . . ta/charmaps/ |
| [16:37:41] | sphery: | and, again, even if we wanted to, we couldn't do anything to make it work for them since it's all handled by Qt (in about 8 different Qt classes, implemented separately and independently--not all supporting one approach... it's a huge mess) |
| [16:38:52] | sphery: | wonder if their using utf8 explains any of the character set problems users have reported |
| [16:42:25] | SteveGoodey (SteveGoodey!~steve@host217-42-221-180.range217-42.btcentralplus.com) has joined #mythtv | |
| [16:43:41] | jya: | stuartm: I had a crash in flight (no network connectivity) when running mythtv-setup: in main.cpp, there's a call to menu->Close(). But menu was null there... The test for a menu failure is: if (!RunMenu(themedir, themename) && !resetTheme(themedir, themename)) |
| [16:43:48] | jya: | shouldn't that be a || ? |
| [16:45:01] | jya: | sphery: for some reasons, because I had no connectivity, my hostname on my mac changed from jyamba to jyamba.local ; this totally screwed up my config... In the end I had to to a database dump, edit the dump and change all instances of jyamba to jyamba.local |
| [16:45:22] | jya: | surely, there got to be a better way to handle a temporary hostname change rather than loosing it all ! |
| [16:45:36] | sphery: | you don't lose it all--you just create a "new" host (with default configurations) |
| [16:45:42] | sphery: | i.e. a new profile |
| [16:46:17] | sphery: | if you want "constant" profile regardless of host name (i.e. you don't want to use the host name as your profile identifier), you have to use the <LocalHostName> override in config.xml to tell it what profile identifier to use |
| [16:46:27] | jya: | as far as I'm concerned, when i start the backend and it tells me that no storage group is set, and when you start the fronted, there's no more recordings: it does look like I lost it all |
| [16:47:02] | jya: | I have "<LocalHostName>my-unique-identifier-goes-here</LocalHostName>& quot; |
| [16:47:09] | jya: | so it's already set to a unique value :) |
| [16:47:42] | sphery: | my-unique-identifier-goes-here is a sentinel that says "no value specified" |
| [16:47:46] | jya: | surely, there got to be a more intuitive way to handle change of hostname... somewhere else than config.xml |
| [16:47:47] | sphery: | so set it to a real value :) |
| [16:48:40] | jya: | could be that the backend stores a hash a unique id, and we can check that the ID is carried across so we know that despite the different hostname its one and the same |
| [16:48:49] | sphery: | the right approach is to always specify a profile identifier that's /not/ a host name and then host name becomes 100% irrelevent |
| [16:49:31] | jya: | IMHO: anything even remotely connected to a file stored in a hidden directory isn't the right approach |
| [16:49:32] | sphery: | the fact that someone once said, "rather than forcing the user to configure a profile identifier for every system, I'll just pull the host name as use that as the identifier if they haven't already configured one" is the only reason host name is even a concern |
| [16:50:31] | sphery: | so, basically, we just need to rewrite all the code that does initial configuration and force a profile identifier (and allow choosing from available identifiers and creating new ones and ...) and then change the name <LocalHostName> to <ProfileID> or something so that users don't think hostname actually means anything (because it doesn't--it's just a profile identifier) |
| [16:50:39] | sphery: | not something I think we should try before 0.27 |
| [16:51:04] | sphery: | and until then I think anyone who wants to be able to handle host name changes should just specify <LocalHostName> (once--for always) in config.xml |
| [16:51:10] | jya: | what I'm surprised is that the definition of storage group got removed |
| [16:51:20] | sphery: | I think you can even specify it in the gui prompt that comes up when you don't have a config.xml |
| [16:51:36] | sphery: | no, the definition wasn't removed... it only applied to a different host |
| [16:51:44] | sphery: | one that didn't exist/wasn't running |
| [16:51:55] | jya: | that's just as bad as editing the config.xml :) |
| [16:52:21] | sphery: | again, specify it once, and it will always work... you don't have to edit it if you "do it right" the first time |
| [16:52:22] | jya: | I would guess that many start mythfrontend or mythtv-setup from a menu rather than a command line.. |
| [16:52:34] | sphery: | the only reason you're using host name is because you didn't specify a profile ID for us to use |
| [16:52:37] | jya: | and as such have no ability to pass different arguments |
| [16:53:09] | Merlin83b (Merlin83b!~Daniel@office.34sp.com) has quit (Quit: Leaving) | |
| [16:53:10] | sphery: | no arguments necessary--just don't have a config.xml and it will prompt for the information required to create one |
| [16:53:33] | sphery: | i.e. first-time setup it works, so if you specify the profile id there, it will be valid forever |
| [16:53:38] | sphery: | regardless of your host name |
| [16:53:55] | jya: | and why would I want to setup a hostname, when at no time I got prompted to do so? ~/.mythtv/config.xml is created automatically, weather I like it or not... or are you saying it is expected that the user after starting myth the first time, must kick a command line, and edit a file in a hidden directory? |
| [16:54:10] | jya: | that's a champion in user conviviality for sure :) |
| [16:55:37] | sphery: | no, the user who has never run any mythtv program on a system (and, therefore, has no config.xml) will be prompted for the information (iff they run a GUI app, like mythtv-setup, which should be run first on any mythtv host, anyway) and it will create the config.xml for them |
| [16:55:52] | jya: | well, I just deleted ~/.mythtv/config.xml ; restarted the fronted... and I certainly didn't get prompted for a hostname... mythfrontend started like it always does, except that it is now saying can't connect to the backend, is it running is mythtv-setup correct |
| [16:56:45] | stichnot: | stuartm: I remember you mentioning QPainterPath a few days ago. I thought you were referring to the approach of drawing the path once and then reusing it for all 8N outline steps. I didn't realize the opportunity to draw once with the wide pen. |
| [16:56:51] | jya: | well, sorry but that sucks |
| [16:57:35] | jya: | this needs to be improved... i don't ever recall having been asked what my hostname is, or what identifier myth should be using |
| [16:57:54] | sphery: | well, there's the "try localhost/mythtv/mythtv connection to the db", which seems to be preventing it from asking you for the information |
| [16:58:24] | sphery: | and/or the "if we find one (and only one) backend running on the network and the pin is set to allow anyone to connect, autoconfig from that" which may prevent the prompt |
| [16:59:05] | jya: | i got things going by editing /etc/hosts so jyamba.local would resolve to something. and editing the database |
| [16:59:52] | sphery: | anyway, I'm the first to admit that we have a huge mess when it comes to use of the host name and it needs to be removed, but I haven't had time to rewrite all of the UI and such |
| [17:00:17] | stichnot: | I'm not sure if the QPainterPath can be reused for drawing the actual text since afaik the path stuff doesn't do antialiasing. Lack of antialiasing should be fine for outlines and shadows, but not for the main text. |
| [17:00:27] | jya: | no worries.. just wanted to raise the issue... i had no access to any documentation, so wasted a good 2hours to get going |
| [17:01:12] | sphery: | and since you can work around its use of host name as a default identifier without too much trouble (just editing an xml file or using -p on a gui program), it's not the highest priority for me |
| [17:01:37] | jya: | you can do it in setup -> general -> page 2: check the box |
| [17:01:57] | jya: | the core issue was identifying what the problem was... jyamba vs jyamba.local |
| [17:01:58] | sphery: | only thing making it even remotely closer to the top is the fact that people think host name is actually important to mythtv/mythtv can only be run on a system with a static host name and IP address (which is annoying) |
| [17:02:09] | sphery: | yeah, definitely not intuitive or easy to find |
| [17:02:24] | jya: | what's surprising is that it wasn't just the fronted that stopped working, but the backend too |
| [17:02:24] | sphery: | and, yeah, forgot to mention that you can get there in mythtv-setup general settings |
| [17:03:19] | stichnot: | jya: this is reminiscent of a somewhat recent Udo semicentithread – http://www.gossamer-threads.com/lists/mythtv/users/534447#534447 |
| [17:03:25] | sphery: | right, because you would have started up the backend and it wouldn't have had any configuration (meaning no tuners or anything and just default settings), so it wouldn't be able to do anything |
| [17:04:01] | jya: | stichnot: sorry, i don't follow |
| [17:05:26] | sphery: | jya: at the bottom of that thread was a discussion of the use of host name (his host name changed) and use of my-unique-identifier-goes-here as a sentinel value |
| [17:05:29] | jya: | sphery: if the backend also relies on a hostname value (or identifier) then we should be able to set it up in mythtv-setup |
| [17:05:31] | stichnot: | jya: Udo complained greatly about "<LocalHostName>my-unique-identifier-goes-here</LocalHostName>& quot; |
| [17:05:44] | jya: | they are calling me at front desk |
| [17:05:49] | jya: | be back later |
| [17:05:52] | jya (jya!~jyavenard@mythtv/developer/jya) has quit (Quit: jya) | |
| [17:06:00] | danielk22: | jya: @work so I can't answer the questions where I need to refer to the code, but the 128MB is the maximum buffer size. When the buffer size was 2MB it was a pre-allocated buffer. 128MB is huge, but it eliminates complaints about broken recordings when people use the CFQ disk scheduler and their hard disks refuse new writes for minutes at a time. |
| [17:10:01] | stuartm: | jya: I've been working on a fix for the hostname issue that will be one of the first things I commit to 0.28 |
| [17:11:01] | sphery: | does it include removing the name "hostname" from all UI code |
| [17:11:20] | stuartm: | what will happen is that the very first run, the system hostname will be inserted into config.xml <LocalHostName> then that will be used for all future runs, we won't ever depend on the system hostname remaining static |
| [17:12:07] | sphery: | that's the most important thing--just don't ever say host name, and say profile name or component name or whatever? |
| [17:12:53] | sphery: | so would be nice if you changed <LocalHostName>, too (but otherwise your approach is fine--as long as you can still specify your own value in backend settings/-p prompt) |
| [17:13:07] | stuartm: | secondly all current places that use the hostname will be replaced with a hostid – in future changing your 'machine identifier' won't break everything, the 'hostname' or whatever you call it will just be a user friendly string associated with a static host id |
| [17:13:14] | sphery: | and if you do that, I can change the db schema to rename the hostname fields |
| [17:13:45] | stuartm: | that will have a small performance boost, since all those string comparisons are replaced with int comparisons |
| [17:13:46] | sphery: | I'd prefer leaving "host" out of the name, though--just because it implies that things are locked to one machine |
| [17:14:02] | stuartm: | bbl, got to eat |
| [17:14:39] | sphery: | you can easily run a frontend on myhost1 and then (after shutting down the frontend on that host) run "the same" frontend on myhost2 by using <LocalHostName>myhost1</LocalHostName> |
| [17:15:07] | sphery: | so something that says it's just a user-specified name for a configuration |
| [17:15:31] | sphery: | whether it's profile id or profile name or ? |
| [17:25:33] | dekarl: | sphery: I was actually thinking the opposite... now that we have nice notifications => show the warning more prominently :) |
| [17:26:05] | jya (jya!~jyavenard@mythtv/developer/jya) has joined #mythtv | |
| [17:27:46] | jya: | danielk22: my bad, the buffer are properly copied... so ignore point 2 above |
| [17:28:45] | jya: | paul-h: please try this patch: http://pastebin.com/a4DUGeNq |
| [17:30:10] | jya: | this should fix all your issues in regards to TFW... it fixes two things in TFW: handle the issue when we write at a much slower rate than we read from (e.g. in your case copying to a USB stick). and it solves the issue should you attempt to write a big buffer on disk at once |
| [17:30:59] | jya: | I reduced the size of the maximum write buffer to 8MB, and we now write in blocks of 2MB max.. could reduce those two by half I believe (Which would put us back to the values prior 0.25) |
| [17:31:43] | sphery: | dekarl: that would work (oh, and btw, thanks for the fix on the mythweb templates stuff--not sure why it worked on my system, maybe a change in a newer php version?) |
| [17:36:51] | dekarl: | now we only need to update RPM Fusion to actually distribute the fix ;) |
| [17:37:46] | jya (jya!~jyavenard@mythtv/developer/jya) has quit (Quit: jya) | |
| [18:14:04] | CiscOH (CiscOH!~CiscOH@cpe-174-101-188-36.cinci.res.rr.com) has joined #mythtv | |
| [18:15:53] | NightMonkey (NightMonkey!~NightrMon@64.124.185.45) has joined #mythtv | |
| [18:15:58] | NightMonkey (NightMonkey!~NightrMon@64.124.185.45) has quit (Changing host) | |
| [18:15:59] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has joined #mythtv | |
| [18:20:01] | dmfrey (dmfrey!~dmfrey@65-78-98-83.c3-0.smt-ubr1.atw-smt.pa.cable.rcn.com) has joined #mythtv | |
| [18:22:12] | gigem (gigem!~david@mythtv/developer/gigem) has quit (Quit: WeeChat 0.4.1) | |
| [18:22:56] | IReboot (IReboot!~doug@CPE10bf48e67915-CM78cd8e7e342d.cpe.net.cable.rogers.com) has quit (Quit: Ex-Chat) | |
| [18:23:09] | gigem (gigem!~david@pool-96-226-13-10.dllstx.fios.verizon.net) has joined #mythtv | |
| [18:23:10] | gigem (gigem!~david@pool-96-226-13-10.dllstx.fios.verizon.net) has quit (Changing host) | |
| [18:23:10] | gigem (gigem!~david@mythtv/developer/gigem) has joined #mythtv | |
| [18:24:29] | CiscOH is now known as CiscOH|Away | |
| [18:37:00] | CiscOH|Away is now known as CiscOH | |
| [18:40:32] | stuartm: | sphery: I was looking at #10789 – it seems reasonable to ignore a keypress when the screensaver or dpms is in effect, simply have it wake up the screen, the problem is that for X11 (I've not checked OSX), there's currently no _reliable_ way of checking that the screensaver is running |
| [18:40:32] | ** MythLogBot http://code.mythtv.org/trac/ticket/10789 ** | |
| [18:41:20] | stuartm: | the best solution seems the use of dbus, org.freedesktop.ScreenSaver.GetActive but I seem to remember there was a reason we're not using dbus for the screensaver? |
| [18:43:59] | sphery: | a) kde and gnome screensavers are using different dbus names for the GetActive |
| [18:44:48] | sphery: | and b) on both of them GetActive tells you whether the screensaver /program/ is "active", not whether the screen is displaying a screensaver |
| [18:45:07] | stuartm: | oh, I knew it was something that makes me want to put the kde/gnome devs in front of a firing squad |
| [18:45:31] | sphery: | yeah |
| [18:46:28] | CiscOH is now known as CiscOH|Away | |
| [18:46:39] | stuartm: | so there's really no way of handling this properly (for screensavers) |
| [18:47:09] | sphery: | I actually have a patch for xdg-screensaver that changes it to make it work for kde (they also thought GetActive indicated whether the screen was displaying a screensaver – https://bugs.freedesktop.org/show_bug.cgi?id=29859 ), but the project seems dead |
| [18:47:36] | stuartm: | we can't rely on the internal state var, because the screensaver may have been disabled/enabled outside of the frontend |
| [18:47:43] | sphery: | note their confidence-inspiring web site for xdg-utils, too: http://portland.freedesktop.org/wiki/ |
| [18:47:57] | sphery: | right... we don't have a good way of knowing anything other than the X dpms state |
| [18:48:24] | sphery: | and even that requires some ugly X code |
| [18:50:26] | sphery: | hmmm, I may have misremembered which way GetActive works based on that bug |
| [18:51:02] | sphery: | note, though, that xscreensaver does not--and will never--support the dbus api (at least I'm pretty sure JWZ will never accept any patch for it) |
| [18:51:27] | sphery: | so, GetActive would tell you whether gnome or kde screensaver is running, but not whether xscreensaver is |
| [18:51:50] | sphery: | and wouldn't help with systems using dpms only (no screensaver program) |
| [18:52:25] | sphery: | and note that we don't currently even support gnome and kde screensavers (since we have no means of poking either since they both broke and refuse to fix their --poke type functionality) |
| [18:53:32] | sphery: | I've considered doing a screensaver handler for gnome and kde, but haven't yet because I philosophically despise their "every single app in existence must have support for multiple different screensaver programs" approach |
| [18:54:05] | stuartm: | yeah, it's all coming flooding back now ... and I was it wasn't |
| [18:54:08] | sphery: | it's just too bad that freedesktop decided that xdg-utils wasn't necessary after the magic of d-bus |
| [18:54:09] | stuartm: | wish |
| [18:54:37] | sphery: | hehe, yeah, I know the feeling (and sorry for my ranting here, but every time it comes up, so does my blood pressure :) |
| [18:54:41] | stuartm: | dbus is great, but only if everyone sticks to a single implementation |
| [18:55:07] | stuartm: | ok, that bug is off the table for 0.27 |
| [18:55:09] | sphery: | right, and that's the challenge (though jwz dislikes it because it's mainly a Linux thing) |
| [18:55:33] | stuartm: | well screw him anyway, X11 is dead |
| [18:56:59] | stuartm: | hasn't been fit for purpose for 15+ years now, unbelievable that it's still being used today |
| [18:57:29] | sphery: | yeah, that will be interesting, too--Wayland and Mir. Why not force devs to rewrite everything to support 3 different (or 2 if you throw out X support) platforms |
| [18:58:26] | sphery: | seems both are working on Qt support, but who knows how much change it will require from us |
| [18:59:13] | stuartm: | Canonical are out of hand |
| [19:16:35] | Chutt_ (Chutt_!~ijr@cpe-76-190-199-73.neo.res.rr.com) has joined #mythtv | |
| [19:17:05] | Chutt (Chutt!~ijr@cpe-76-190-199-73.neo.res.rr.com) has quit (Ping timeout: 245 seconds) | |
| [19:22:55] | stuartm: | https://github.com/MythTV/mythtv/commit/c60f3314 |
| [19:23:55] | CiscOH|Away is now known as CiscOH | |
| [19:26:22] | CiscOH is now known as CiscOH|Away | |
| [19:41:24] | dekarl: | stuartm: ? |
| [19:45:36] | stuartm: | dekarl: wrong window :) |
| [19:47:33] | stuartm: | I was looking into http://code.mythtv.org/trac/ticket/10543 |
| [19:49:16] | sphery: | stuartm: having read that ticket, I think there's an invalid assumption in there. The user seems to think we're telling X to "wake up" the monitor, but we're not. X's built in DPMS support is enabling the monitor in response to a keyboard event, not mythfrontend. All we ever do with DPMS is disable DPMS blanking support in X when we start playback and enable it on when we stop playback. |
| [19:50:02] | stuartm: | sphery: actually ... no, at least for LIRC events, we call screensaver->Reset() |
| [19:50:13] | stuartm: | we are waking it up |
| [19:50:40] | stuartm: | but that dates back to a time before lirc events went through the kernel, and it's probably not needed with recent kernels |
| [19:52:01] | sphery: | ah, didn't know that got put in... |
| [19:52:17] | stuartm: | since remote input is now treated the same as keyboard input (need to test that theory where lirc is still being used), a keyboard and a remote should both wake up dpms at a lower level than mythfrontend |
| [19:53:12] | sphery: | I'd be all for taking it out and those users who use lirc properly (as opposed to using a remote like a keyboard) can decide if they want a button on the remote to wake mythfrontend or not themselves |
| [19:53:16] | sphery: | rather than our forcing it on them |
| [19:53:54] | sphery: | if you have the kernel driver configured with protocol=lirc only, I"m sure it will work like lirc always has (where the dpms won't wake up) |
| [19:53:57] | stuartm: | where lirc is used with a modern kernel, the driver is told not to forward remote input through to applications as keyboard events, I don't know if that affects dpms wakeups |
| [19:54:56] | sphery: | but, again, it's not hard for the user to configure a button to wake the screen with lirc if they want, and doing it that way allows them to decide which button(s) (i.e. nice "safe" ones) should be used to wake the screen |
| [19:55:06] | dekarl: | stuartm: ahh. I see. |
| [19:55:45] | stuartm: | sphery: well I for one WANT it to wake up and that's where I think the reporter had a valid point, an event received when the screen is blank should _only_ unblank it, not do anything else (like delete a recording) |
| [19:56:43] | sphery: | well, if I have a highlight in LibreOffice and I hit a key on my keyboard when the screen is dpms blanked (but without a screensaver eating the keys), it will delete the highlight and replace it with the key I typed |
| [19:56:54] | sphery: | (which is why I wake computers with safe keys like shift :) |
| [19:57:10] | stuartm: | I haven't got a free button I can dedicate just to exiting from dpms, and requiring such a thing doesn't seem user friendly, after all you don't need a special button on the keyboard or mouse to do the same thing |
| [19:57:20] | sphery: | it doesn't have to be a free button |
| [19:57:39] | sphery: | just choose any button you want and have a stanza for prog = mythtv and another for prog = irexec |
| [19:57:50] | sphery: | both will handle the button press |
| [19:58:24] | sphery: | or have prog = irexec stanzas for every button |
| [19:58:30] | stuartm: | sphery: I think we can implement it as I suggest, at least for dpms and personally I only use dpms these days |
| [19:59:04] | stuartm: | a screensaver is pointless with LCDs – doesn't save power or prevent screen burn in |
| [19:59:43] | sphery: | agreed |
| [20:00:37] | sphery: | just seems that checking dpms state (when X is dpms aware) on every key or button press is kind of heavy handed... won't bother me, though, because I disable dpms support in X, anyway |
| [20:01:24] | sphery: | (assuming we're only doing anything for dpms aware X--which is true in screensaver-x11, at least, so I hope it's true for resetscreensaver) |
| [20:02:00] | stuartm: | we don't have to check it on every keypress, dpms only kicks in after a period of inactivity – do we only need to check for it on the first keypress after such a period |
| [20:02:28] | stuartm: | s/do/so/ |
| [20:03:36] | stuartm: | if we can query the dpms timeout even better, but that's not crucial, we just pick a reasonable figure – i.e. 10 minutes |
| [20:04:29] | stuartm: | 10 minutes is I think the minimum that most GUI tools allow you to configure for dpms |
| [20:08:25] | sphery: | so, seems we're not doing anything with the dpms--only running xscreensaver-command -deactivate --in ResetScreenSaver() |
| [20:09:50] | stuartm: | it's ScreenSaverX11::Reset() which gets called |
| [20:10:04] | stuartm: | that calls DPMSForceLevel(dsp, DPMSModeOn); |
| [20:10:04] | sphery: | but looks like you could use Asleep() to check DPMS state if m_dpmsaware |
| [20:11:26] | sphery: | ah, yeah, too many similarly-named functions there... got that confused when find in page took me to ResetScreenSaver() first |
| [20:13:04] | sphery: | anyway, seems it's just a matter of checking asleep, which only does something if dpms is enabled |
| [20:21:33] | danielk22: | I don't know if this is the same thing, but ever since I last upgraded the OS on one of my boxes from ubuntu 1004 to 1204 the remote is no longer enough to completely wake up the screen. It will turn on the backlight, but the screen will still be blanked. |
| [20:22:14] | danielk22: | I assumed it was really a problem at the KDE level and not with MythTV so I never mentioned it or filed a ticket. |
| [20:22:44] | sphery: | right, we don't have support for waking (or inhibiting) kde screensaver |
| [20:23:04] | danielk22: | I don't have any screensaver installed. |
| [20:23:39] | danielk22: | I believe it's that old pre-dpms screen blanking feature of X11 somehow getting activated. |
| [20:24:01] | sphery: | I remember seeing (about a month after you posted it, which is why I never replied) a mention of a running kscreenlocker process on your system--that's the KDE screensaver |
| [20:24:28] | danielk22: | different system |
| [20:24:41] | sphery: | ah, ok... not sure what that would be then |
| [20:24:43] | danielk22: | I run kscreenlocker on my laptop |
| [20:28:16] | sphery: | danielk22: have you tried running (at X startup): xset s off |
| [20:28:55] | sphery: | I think that's the control for the pre-dpms screensaver |
| [20:29:21] | danielk22: | sphery: I'll give it a try. I'll try the blanking option too: http://www.shallowsky.com/linux/x-screen-blanking.html |
| [20:30:01] | danielk22: | It was just one of those little annoyances I haven't gotten around to looking at, but I thought it might have some relevance to the discussion |
| [20:30:07] | sphery: | yeah, fixing it with the X config would mean you wouldn't have to run xset (as the defaults would be right) |
| [20:30:55] | danielk22: | If it really is that I'll file a bug against Ubuntu too, really monitor produced in the last decade lacks DPMS support... |
| [20:31:50] | sphery: | hehe, yeah--maybe that's an artifact of their using config-less X and X's default for BlankTime being 10min |
| [20:32:15] | sphery: | (I'm too OCD to use config-less X, so don't know how it works. :) |
| [20:32:26] | danielk22: | probably... in which case I guess the report would need to go upstream of Ubuntu :) |
| [20:33:02] | sphery: | yeah... I'd agree the default makes no sense, now--especially since it's not saving power since backlight stays on and such |
| [20:33:43] | danielk22: | right, it's just to prevent CRT burnin.. as a default.. in 2013 |
| [20:35:44] | stuarta: | back to the future eh? |
| [20:36:03] | danielk22: | (We were pondering a -43 years elapsed bug @ work yesterday.. and someone realized it has been 43 years since the epoch, which made some of us feel old.) |
| [20:46:58] | stuartm: | danielk22: were you the one recently looking at mythsystem? |
| [20:48:26] | stuartm: | or more to the point, before I go ahead and commit it, does anyone have thoughts on https://github.com/garybuhrmaster/mythtv/comm . . . e24b19a991a2 that they want to share? |
| [20:48:32] | danielk22: | Depends what you mean by recently.. several months ago I started adding tests and doing some cleanup. |
| [20:50:26] | stuartm: | danielk22: was it several months? Time really does fly |
| [20:51:11] | danielk22: | Pretty sure it was a few weeks before Henning was born and he's 12 or 13 weeks now. |
| [20:53:08] | NightMonkey (NightMonkey!~NightrMon@pdpc/supporter/professional/nightmonkey) has quit (Quit: Body blow! Body blow!) | |
| [20:56:34] | stuartm: | hey, congrats! (If I've already said that before, then just put it down to my terrible memory) |
| [21:03:47] | skd5aner (skd5aner!~skd5aner@50-90-5-146.res.bhn.net) has quit (Read error: Connection reset by peer) | |
| [21:04:12] | skd5aner (skd5aner!~skd5aner@50-90-5-146.res.bhn.net) has joined #mythtv | |
| [21:04:20] | danielk22: | thx :) |
| [21:08:26] | dekarl: | stuarta: http://code.mythtv.org/trac/attachment/ticket . . . s-but-.patch just needs someone to test it |
| [21:08:52] | stuartm: | actually, going to defer that commit until after 0.27, not entirely sure I understand what would happen if you close a dup'd file descriptor |
| [21:10:35] | robink (robink!~quassel@unaffilated/robink) has quit (Remote host closed the connection) | |
| [21:10:42] | robink (robink!~quassel@unaffilated/robink) has joined #mythtv | |
| [21:23:03] | dmfrey (dmfrey!~dmfrey@65-78-98-83.c3-0.smt-ubr1.atw-smt.pa.cable.rcn.com) has quit (Quit: Ex-Chat) | |
| [21:25:13] | stuarta: | dekarl: rather than (pat->ProgramNumber(0) == 0) || (pat->ProgramNumber(1) == 0)) maybe use a FindPID(0) ?? |
| [21:25:18] | stuarta: | er FindProgram |
| [21:28:43] | danielk22 (danielk22!~danielk@exchange.wgen.net) has left #mythtv () | |
| [21:30:30] | stuartm: | so it seems the XBox 360 takes advantage of the fact that the DLNA spec only deals in what must _minimally_ be supported – which is why although MythTV is DLNA compliant, at least as far as upnp:class operators are concerned, it still doesn't work with the XBox |
| [21:31:23] | stuartm: | WTF is the point of a spec if it doesn't guarantee that devices/software can interoperate? |
| [21:33:09] | sphery: | yeah, since xbox 360 doesn't support mpeg systems streams (or NuppelVideo) or mpeg-2 in UPnP, it's pretty much useless for upnp with MythTV recordings |
| [21:33:54] | stuarta: | dekarl: am i making any sense? i'm not sure i am |
| [21:33:56] | sphery: | as far as the spec, ask the html5 video tag implementers... or Netflix/Google/Microsoft with their HTML5 Encrypted Media Extensions spec |
| [21:37:56] | stuartm: | I had this idea that I'd ensure mythtv was DLNA compliant, but it seems that wouldn't actually be worth a great deal, since two DLNA compliant devices/apps won't necessarily be able to talk to each other |
| [21:40:16] | jya (jya!~jyavenard@mythtv/developer/jya) has joined #mythtv | |
| [21:40:22] | stuarta: | dekarl: looking a bit deeper, what you've done will work just fine. FindPID(0) can't differentiate between the 1st program being "0" and not finding the program |
| [21:49:52] | jya: | pretty nice gogo in the air... |
| [21:59:00] | SteveGoodey (SteveGoodey!~steve@host217-42-221-180.range217-42.btcentralplus.com) has quit (Quit: Konversation terminated!) | |
| [22:26:24] | Jordack (Jordack!~Jordack@h69-131-44-221.plmomi.dedicated.static.tds.net) has quit () | |
| [22:33:57] | stuartm: | wagnerrp: now would seem like a good time to sort out packaging scripts? http://code.mythtv.org/trac/ticket/11165 |
| [22:34:18] | jya: | gigem: I'm surprised you're not seeing a crash when mythlogserver is running... there's a bug in the reference counter set... |
| [22:42:31] | jya (jya!~jyavenard@mythtv/developer/jya) has quit (Ping timeout: 245 seconds) | |
| [23:12:39] | stuarta: | well i've now fixed and smoke tested #7486 changes. now it's time for bed |
| [23:12:39] | ** MythLogBot http://code.mythtv.org/trac/ticket/7486 ** | |
| [23:15:28] | stuartm: | wagnerrp: re 10339, can't we just wrap titles/subtitles in quotes? |
| [23:16:24] | stuartm: | How are we currently handling other characters that must be escaped, i.e. exclamation marks? |
| [23:26:55] | sphery: | stuartm: I'm pretty sure it's not passing the command to a shell, but passing it as an array of args (so no quoting required to "group" stuff) |
| [23:27:35] | sphery: | and most commands, even if args are passed that way or in quotes, will have problems if an argument starts with a dash |
| [23:31:35] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has quit (Ping timeout: 260 seconds) | |
| [23:41:35] | stichnot (stichnot!~stichnot@mythtv/developer/stichnot) has joined #mythtv | |
IRC Logs collected by
BeirdoBot.
Please use the above link to report any bugs.