Back to Home / #mythtv / 2007 / 01 / Prev Day | Next Day
#mythtv IRC Logs for 2007-01-24

---Logopened Wed Jan 24 00:00:50 2007
00:03|-|eskil [] has joined #mythtv
00:14|-|adante_ [] has joined #mythtv
00:30|-|adante [] has quit [Connection timed out]
00:30|-|adante_ changed nick to adante
00:32|-|cattelan changed nick to cattelan_away
00:38|-|cattelan_away changed nick to cattelan
00:59|-|gnome42 [] has quit [Remote closed the connection]
01:08|-|noddan [] has joined #mythtv
01:22|-|Anduin [] has quit [Read error: 104 (Connection reset by peer)]
01:29|-|Anduin [] has joined #mythtv
01:32|-|noddan [] has quit []
01:44|-|cattelan changed nick to cattelan_away
01:50<xris>gbee: ew
01:51<xris>blech. I really wish that EXPLAIN would tell me WHY mysql chose to ignore the table key.
02:07|-|noddan [] has joined #mythtv
02:10<xris>wtf is this?! SELECT recordid,search,subtitle,description FROM record WHERE search <> 0 AND (recordid = -1 OR -1 = -1);
02:20<Chutt> query = QString("SELECT recordid,search,subtitle,description "
02:20<Chutt> "FROM %1 WHERE search <> %2 AND "
02:20<Chutt> "(recordid = %3 OR %4 = -1) ")
02:20<Chutt> .arg(recordTable).arg(kNoSearch).arg(recordid).arg(recordid);
02:21<xris>ah, ok. as long as it's not hard coded. lol
02:22<xris>shouldn't that be tweaked to do some more in the code, though?
02:22<Chutt>i dunno
02:22<xris>like check for recorded == -1 and alter the query accordingly?
02:23<xris>since if recordid is -1, the OR stuff is redundant, and if it's NOT, you certainly don't eed the OR.
02:23<xris>anyway, not that it's a particularly slow query.. just seemed odd.
02:23<xris>I have a list of indexes to add, but it looks like there's no help for a few of the queries. no matter what indexes I added (and mysql found), it wouldn't use them.
02:24<xris>too many odd joins.
02:24<xris>I'll pastebin them tomorrow. I need to sleep now.
02:26|-|xris [] has quit ["Leaving."]
02:29|-|Cardoe [n=Cardoe@gentoo/developer/Cardoe] has joined #mythtv
02:34|-|eskil [] has quit ["Leaving"]
03:12|-|DrNickRiviera [] has joined #mythtv
03:14|-|stuarta [n=stuart@unaffiliated/stuarta] has joined #mythtv
03:38|-|slaine_ [n=glengray@] has joined #mythtv
04:28|-|adante_ [] has joined #mythtv
04:42|-|DrNickRiviera [] has quit [Remote closed the connection]
04:43|-|DrNickRiviera [] has joined #mythtv
04:43|-|adante [] has quit [Read error: 110 (Connection timed out)]
04:43|-|adante_ changed nick to adante
04:48|-|DrNickRiviera [] has quit ["Leaving."]
04:58|-|aevil [] has joined #mythtv
04:59|-|lucas123 [] has quit [Read error: 110 (Connection timed out)]
04:59|-|Skiingsean changed nick to Skiingsean_sleep
05:36<stuarta>so it's not just me! group filtering's not working #2983, thought it was the mac frontend...
05:38<MrGandalf>i believe someone put in a ticket for that in the last 24 hours..
05:38<MrGandalf>nm, not awake
05:38<stuarta>yeah, testing the new mini mac frontend last night. that's one of the first things i noticed.
05:59|-|MrGandalf [] has quit ["work"]
06:11|-|adante_ [] has joined #mythtv
06:26|-|adante [] has quit [Read error: 110 (Connection timed out)]
06:26|-|adante_ changed nick to adante
06:34|-|adante_ [] has joined #mythtv
06:49|-|adante [] has quit [Read error: 110 (Connection timed out)]
06:49|-|adante_ changed nick to adante
07:04|-|adante_ [] has joined #mythtv
07:11|-|lsobral [n=sobral@] has joined #mythtv
07:20|-|adante [] has quit [Read error: 110 (Connection timed out)]
07:20|-|adante_ changed nick to adante
07:49|-|Cardoe [n=Cardoe@gentoo/developer/Cardoe] has quit ["Leaving"]
07:52|-|nero [n=nero@unaffiliated/nero] has quit ["Leaving"]
08:15|-|Cardoe [n=cardoe@gentoo/developer/Cardoe] has joined #mythtv
08:26|-|MrGandalf [] has joined #mythtv
08:27<MrGandalf>WOw, that OGL ticket looks interesting..
08:30<MrGandalf>just don't know if my card is fast enough for non-deinterlaced 1080i playback.
08:35<MrGandalf>maybe time to upgrade..
08:37<Merlin83b>"non-deinterlaced 1080i" - erm, is there any other sort? ;-)
08:37|-|_Defend [i=defend@] has joined #mythtv
08:37<MrGandalf>you're getting me wrong.. I don't deinterlace the 1080i at playback
08:54|-|Defend [i=defend@] has quit [Read error: 110 (Connection timed out)]
08:55<Merlin83b>So it's 1080i, not 1080p ;-)
09:02<MrGandalf>when I got my TV 1080p wasn't out yet
09:02<MrGandalf>besides, I don't see anything wrong with 1080i, for TV anyway
09:03<MrGandalf>once I tweaked the screen, anyway
09:15|-|cattelan_away changed nick to cattelan
09:15<MrGandalf>interesting.. reading a thread starting out on the ivtv dev list. Does anyone know if it's true that the scheduler huge query locks the recorded table while it's running?
09:16<stuarta>one of the chris's verified it
09:16<stuarta>but it's recordedmarkup
09:16<stuarta>the whole thread is on -dev
09:20<janneg>no, it's ivtv recorder updates recorded.filesize
09:20[~]stuarta rereads
09:22<MrGandalf>all recorders update that field, don't they?
09:22<stuarta>since it's in mpegrecorder, then pretty much yes.
09:23<MrGandalf>may explain my HD playback problem when EIT kicks up..
09:23<stuarta>EIT does do a reschedule() after new data comes in.
09:24|-|noddan [] has quit []
09:24<janneg>that shouldn't affect playback, only recording
09:27<MrGandalf>hmm, true.. I guess I would probably see ringbuffer errors on the backend if that were the case and I'm generally not.
09:29|-|noddan [] has joined #mythtv
09:30<janneg>stuarta: is still valid?
09:31<stuarta>atm yes
09:31<stuarta>i'll have to sync up to the latest ffmpeg and see if it's fixed there
09:31<stuarta>then wait till we sync if it's fixed.
09:35<janneg>stuarta: ~janne/ffmpeg_resync_6351-7671.diff on
09:35<stuarta>i'm updating ffmpeg as we speak.
09:36<MrGandalf>janneg: any real advantages to updating?
09:37|-|gnome42 [] has joined #mythtv
09:38<stuarta>got it, sadly only on command line till the weekend.
09:38<stuarta>will do what testing i can
09:41<janneg>MrGandalf: h264 optimizations which have unfortunately only little effect on playback
09:43<janneg>hah, reduced the load in libmythtv while h264 playback by factor 2 (from 4% to 2%)
09:44<Dibblah>... How?
09:45<janneg>by using bit operations for h264 sync detection
09:49[~]stuarta builds latest ffmpeg
09:51<gnome42>janneg: Is that similar to what I tried for MPEG2 start code detection?
09:54[~]stuarta builds latest x264 libs first
09:54<janneg>gnome42: probably. is that patch committed?
09:54|-|jgarvey [] has joined #mythtv
09:55<janneg>I though someone did it already but I must have mixed up mpeg2 and h264
09:56|-|jmk_ [] has joined #mythtv
09:56<gnome42>janneg: no, it's not, or not that I know of. Also, I did a similar patch for the playback side in avformatdecoder.
09:58<gnome42>janneg: I didn't submit those patches cause I thought the nothl() for endianess, would be unacceptable.
09:59<janneg>why would you need that? bit shifts are portabel
10:00<gnome42>I was using it was a quick/clever(?) way to stash that last bytes from the end of the previous pkt.
10:01<janneg>gnome42: did you send the patches to -dev?
10:02<gnome42>No, I sent the original patch directly to you.
10:03<gnome42>I can clean up and resend?
10:03<janneg>thanks, found it
10:10<janneg>gnome42: please send both patches, I know how the nothl can be avoided
10:12<gnome42>oh yeah? without a memcpy?
10:13<janneg>btw did you profiled the changes?
10:14<gnome42>janneg: no, I didn't. I only micro-benchmarked the various approaches in some test code.
10:15<janneg>that's ok.
10:16<gnome42>should I send the patches to you? or create a ticket?
10:17<janneg>create a ticket and assign it to me
10:19<gnome42>ok will do.
10:21<stuarta>janneg: re #2062: looks like it may have been implemented
10:21<stuarta>now only get lots of "[h264 @ 0x8425ee0]warning: first frame is no keyframe" errors
10:21<stuarta>no idea what the picture looks like though
10:23<janneg>I could check it with synced mythtv if a sample is available
10:27<stuarta>i've a 49Mb sample
10:27<stuarta>tho IIRC correctly it's a test screen with no action...
10:28|-|slaine_ [n=glengray@] has quit []
10:38|-|hooch [n=hooch@] has joined #mythtv
10:41<janneg>stuarta: good enough for testing
10:44|-|hooch [n=hooch@] has left #mythtv []
10:45<stuarta>i'll copy it to in my home dir
11:05<gnome42>janneg: ok, I created #2984, with the two patches. But how do I assign it to you? :)
11:06|-|aevil [] has quit ["Verlassend"]
11:06<stuarta>gnome42: he can take it.
11:08<MrGandalf>stuarta: btw, I have another sample taken from another provider for that issue you were looking at the other day
11:08<MrGandalf>it has slightly different characteristics
11:11<gnome42>stuarta: ok, thanks. I a missing something in trac? Should I be able to assign a ticket to someone?
11:11<stuarta>got disable
11:11<janneg>stuarta: thanks
11:12<gnome42>oh ok
11:16<janneg>gnome42: did you also test single byte only with bit operations? for the h264 case it's equally fast but much simpler code
11:19<gnome42>janneg: nope, I didn't. I would like to, do you have example code?
11:23<janneg>gnome42: I'll send a patch
11:24<gbee>janneg: what revision of ffmpeg did you resync to?
11:29<janneg>gbee: 7671 and I noticed that they applied some of our mpegts changes
11:29|-|stuarta [n=stuart@unaffiliated/stuarta] has quit ["tomorrow people"]
11:32<gbee>janneg: that's what I was wondering :)
11:32|-|briand [] has joined #mythtv
11:33|-|emcnabb [] has quit [Remote closed the connection]
11:34|-|emcnabb [] has joined #mythtv
11:35<jk1joel>so I was asking about running two frontends on the same box in #mythtv-users, and |Torg| pointed out that settings are saved on a per-hostname basis
11:35<jk1joel>what does everyone think of storing settings by hostname:user instead? That way you could have two users each with different settings for their frontends
11:36<Chutt>change the hostname in mysql.txt.
11:36<gbee>hmm, any way to disable or redirect the errors from ffplay? they actually cause greater CPU use than playback with that HD sample
11:37<jk1joel>Chutt: can I have two different mysql.txt files on one box? (that mythfrontend and friends will be able to find?)
11:37<GreyFoxx>jk1joel: jk: yes
11:38<GreyFoxx>jk1joel: ~/.mythtv/mysql.txt overrides any global one
11:38<GreyFoxx>So either have their own mysqk.,txt for each
11:38<jk1joel>ah. perfect. well then I won't have to write a patch. thanks!
11:38<GreyFoxx>or use the environment variable MYTHCONFDIR to specify a location
11:38<GreyFoxx>you can even specifiy a seperate hostname/identifier in it for each
11:39<jk1joel>sweet. so it shouldn't be a big deal at all to run two frontends on one box
11:39<GreyFoxx>No, none at all
11:43<janneg>gbee: set the av_log_level to AV_LOG_FATAL or AV_LOG_QUIET in avformatdecoder.cpp
11:44|-|eskil [] has joined #mythtv
11:49|-|beavis [] has joined #mythtv
11:50<Chutt>janneg, i worked too late last night, didn't have a chance to look over the merge patch
11:50<janneg>Chutt: no problem
11:55|-|xris [] has joined #mythtv
12:13|-|gr8nash [n=andy@] has quit [Read error: 104 (Connection reset by peer)]
12:15<gbee>janneg: was referring to ffplay not mythtv - but at least the solution is still pretty much the same :)
12:25|-|brianc [] has joined #mythtv
12:34<gnome42>janneg: results are kind of surprising
12:37<janneg>gnome42: thanks. the celeron is a netburst chip?
12:40<gnome42>janneg: hmm, not too sure, Netburst is P4 era? This is an early celeron 300A.
12:43<janneg>yes, netburst is p4.
12:43<gnome42>janneg: Gotta love the simplicity of your solution though! :)
12:43<gnome42>janneg: yeah, this is Celeron (Mendocino) very early celeron
12:45<janneg>gnome42: can you test another patch?, just replaced the shift in the if(). shouldn't make a big difference for the core 2 but is maybee on the celeron faster
12:49<janneg>gnome42: are you testing the routines with a complete recording?
12:54<gnome42>janneg: no, not a complete recording only 64MB on a recording. It iterates 64 times over 64MBs of the file.
12:55<gnome42>... 64MB of a recording
12:55<janneg>yeah, figured that already, otherwise it would be io-bound
13:03|-|zigovr3 [] has quit [Read error: 110 (Connection timed out)]
13:03<gnome42>janneg: ok, new results
13:04|-|zigovr3 [] has joined #mythtv
13:04|-|beavis [] has quit ["Verlassend"]
13:04<gnome42>janneg: needed some extra ()'s there
13:05|-|beavis [] has joined #mythtv
13:16<janneg>gnome42: and another one if you don't mind hopefully almost as fast as your 3 byte approach
13:38|-|brianc [] has quit ["."]
13:57|-|Anduin [] has quit [Read error: 110 (Connection timed out)]
14:20|-|zigovr3 [] has quit [Read error: 110 (Connection timed out)]
14:22|-|zigovr3 [] has joined #mythtv
14:55<gnome42>janneg: sorry, I got busy, I'll try and test that last one and get back to you.
14:57|-|someothernick [] has joined #mythtv
14:58|-|someothernick [] has left #mythtv ["Leaving"]
14:59|-|melunko [] has joined #mythtv
15:03|-|spacecoaster [] has joined #mythtv
15:17|-|SeeJay [] has joined #mythtv
15:18|-|SeeJay [] has left #mythtv []
15:37|-|MrGandalf [] has quit ["home"]
15:39<j-rod>just curious... anyone else played with any saa7162-based hardware yet?
15:39<j-rod>just got a boxed dropped on me w/this card in it:
15:45<jams>thats a no from me
15:46<janneg>they aren't supported yet. phillips claims they have a linux driver
15:48<j-rod>I suspected they weren't supported just yet...
15:48[~]j-rod wonders how to get said driver out of philips...
15:49<j-rod>or if it'd just be easier to work on writing one :)
15:51<janneg>according to it's probably faster to write one
15:55|-|gr8nash [n=andy@] has joined #mythtv
15:55<j-rod>I'll ask around internally if we have any contacts at philips to maybe grease the wheels...
15:55<gbee>wasn't it philips who talked about contributing some code (or something else?) to mythtv?
15:56<gbee>it would have been a few months ago, hence why my memory is vague
15:58<j-rod>they should, with the number of mythtv users who buy stuff w/their chips on 'em...
15:58<j-rod>that reminds me, wtf happened to the saa7127 driver in 2.6.19?...
15:59<janneg>i think there is a philips presentation about mythtv and an embedded platform
15:59<j-rod>janneg, elc2006-klaas-de-waal-mythtv-nexperia.ogg ?
16:00<j-rod>that's been sitting on my box here for ages, haven't got around to watching it...
16:00<janneg>no, a pdf iirc
16:01|-|sliverworm [] has joined #mythtv
16:01<janneg>which seems to be the same content
16:02<sliverworm>Hey , wasn't this mythtv program ment for ubuntu?
16:02<gbee>the ogg is the same event
16:02<GreyFoxx>sliverworm: Ummm, what ?
16:03<gbee>sliverworm: #mythtv-users
16:03<sliverworm>alright then.
16:03<gbee>GreyFoxx: don't ask - I don't think we want to know the answer
16:05<j-rod>I should just up and watch the damn video... I think I brezzed through the pdf a while ago, can't really remember it all that well
16:05|-|sliverworm [] has left #mythtv []
16:08<gbee>hmm, might have been a little too hard on him
16:08<GreyFoxx>heh yeah
16:17<janneg>GreyFoxx: I've seen the fast playback with 2.6.20-rc5 yesterday
16:21|-|adante_ [] has joined #mythtv
16:25|-|_Defend changed nick to defend
16:30|-|jgarvey [] has quit ["Leaving"]
16:33|-|adante__ [] has joined #mythtv
16:37|-|adante [] has quit [Connection timed out]
16:38|-|melunko [] has quit ["Leaving"]
16:49|-|adante_ [] has quit [Read error: 110 (Connection timed out)]
16:55|-|hatlevip [] has joined #mythtv
16:56<gnome42>janneg: Hi, I'm having difficulty testing that last version ( did you compile and run that code?
16:59|-|Cardoe [n=cardoe@gentoo/developer/Cardoe] has quit ["Leaving"]
17:00|-|hatlevip [] has quit ["Leaving"]
17:06<janneg>gnome42: only compiled. ah I see the error. the else misses i++;
17:11<gnome42>janneg: can we even hit the 'else'?
17:11|-|beavis [] has quit ["Verlassend"]
17:13<janneg>yes, else is the interesting case
17:16|-|MrGandalf [] has joined #mythtv
17:18|-|jmk_ [] has quit ["Leaving"]
18:03|-|WZ__ [] has quit [Read error: 110 (Connection timed out)]
18:24|-|TTT_Travis [] has joined #mythtv
18:24|-|TTT_Travis [] has left #mythtv ["Leaving"]
18:26|-|fysa [] has joined #mythtv
18:27<fysa>Would a mythvideo patch that enables the ability (with config option) to fall back to displaying a folder.jpg in listing mode (i.e., directory view) if a thumbnail for the selected file isn't available be accepted?
18:45|-|daum [] has joined #mythtv
18:45<daum>hi all
18:50|-|DrNickRiviera [] has joined #mythtv
18:50|-|squidy [n=squidy@unaffiliated/squidy] has joined #mythtv
18:50|-|squidy [n=squidy@unaffiliated/squidy] has left #mythtv ["fuizzz!"]
18:52<GreyFoxx>fysa: It might, since it's what is done in MythGallery (VideoGallery::drawIcon)
19:00<GreyFoxx>janneg: I haven't had a single playback speed issue since I went to
19:00<GreyFoxx>and strangely enough, I haven;t had any of those ivtv "application not reading data fast enough" messages either
19:02|-|daum [] has quit ["Leaving"]
19:04<janneg>GreyFoxx: restarting (and maybe recompiling) fixed it
19:05<GreyFoxx>I wonder if the HZ change from 250 to 1000 had anything to do with the ivtv complaints and dropped frames
19:13|-|jams [] has quit [Read error: 104 (Connection reset by peer)]
19:13|-|jams [] has joined #mythtv
19:17<MrGandalf>the lernel folkes had a huge discussion on that.. it's said that a higher hz is better for multimedia
19:21<GreyFoxx>With ivtv+ I was getting a ton of the complaints from the driver(250HZ), now with and 1000HZ + same ivtv source I haven't had any of them
19:21<GreyFoxx>Iwas getting some dropped frames from it, but so far everything is fine now
19:41|-|gnome42 [] has quit [Read error: 110 (Connection timed out)]
19:41|-|gnome42 [] has joined #mythtv
19:48|-|defend changed nick to Defend
19:54|-|rtsai1111 [] has joined #mythtv
19:56|-|calvarez [] has joined #mythtv
19:56<calvarez>hello, I am thinking of making a MythTV box, and I was wondering if there are any specific PVR cards you would recommend
19:57<GreyFoxx>Try #mythtv-users
19:57<calvarez>oops, sorry, I will
19:57<calvarez>didn't notice it was dev channel
19:58|-|calvarez [] has left #mythtv ["Leaving"]
19:59<fysa>Maybe this weekend I'll have time to play with the myth source. I'd like to debug DVD subtitles also..
19:59<GreyFoxx>What problem are you having with them ?
20:00<fysa>last time I used them on heavy subtitled movies (wife is Indian, so watches losts of Hindi movies w/English subs), the video would freeze for a few frames every subtitle render.
20:00<fysa>.. s/losts/lots
20:00<fysa>unfortunately, it made it unwatchable so I had to switch back to xine.
20:01<fysa>I'll test with latest SVN again this weekend..
20:01<GreyFoxx>I don't use them much here, but so fqr I've not had that problem
20:02<fysa>perhaps it was related to my slow filesystem issues and not being able to pull the subtitles quick enough, causing the video to lock while waiting for them.
20:03<fysa>(i.e., no real clue)
20:03<fysa>we'll see :)
20:05|-|rtsai [] has quit [Read error: 110 (Connection timed out)]
20:11|-|Anduin [] has joined #mythtv
20:22|-|DrNickRiviera [] has quit ["Leaving."]
20:29|-|cattelan changed nick to cattelan_away
20:34|-|cattelan_away changed nick to cattelan
20:45|-|Vyrus [] has joined #mythtv
20:46<Vyrus>im haveing problems with tv playback on a newly installed knoppyth system
20:46<Vyrus>pvr150 atix1300 running in gl mode... evrything works but my frame rate is in the crapper
20:46<Vyrus>idea;s ?
20:50|-|eskil [] has quit [Read error: 110 (Connection timed out)]
20:53|-|eskil [] has joined #mythtv
21:01|-|Vyrus [] has quit ["Leaving"]
21:10<Captain_Murdoch>anyone have any opinions on that asynchronous DB idea that will allow us to update the filesize and recordedseek tables in another thread?
21:11|-|rtsai [] has joined #mythtv
21:11<Captain_Murdoch>looking at the patch and asking the guy to provide a complete cleaned up version. the last version posted has some junk in it and is missing at least one part where they setup the gAsyncDB variable.
21:12<GreyFoxx>I'm curious just how many tuners, channels and schedules some people have that are taking 30 seconds for the scheduler to run
21:12<Captain_Murdoch>well, 19
21:12<GreyFoxx>Mine takes 2
21:13<Captain_Murdoch>I have 5 tuners, 83 recording schedules, and about 700-900 items 'scheduled' according to the scheduler when it runs.
21:13<Captain_Murdoch>probably has a lot to do with hardware and MySQL version.
21:13<GreyFoxx>400 or so channels, 12 days of programs, 4 tuners, 69 schedules
21:13<Captain_Murdoch>and tuning of course.
21:14<Captain_Murdoch>my master backend is only a P3-733, but even on my dev box (a P4 @ 1.8Ghz) it takes ~20 seconds.
21:14<GreyFoxx>though now that I think ab out it, last year I went through and added indexes to stuff that showed up in my slow query log
21:14<Captain_Murdoch>maybe some of xris' stuff will speed that up.
21:15<Captain_Murdoch>used to be about 10 seconds on my system just 6 months ago, maybe I need to optimize the db again. I don't do that regularly.
21:15<GreyFoxx>I never do hehe
21:15<Captain_Murdoch>my DB is a VM on a dual-proc P4 @ 3.0Ghz box.
21:16<GreyFoxx>mine is on my master backend, xp2500, does a lot of stuff around the house including myth
21:16<GreyFoxx>I should compare my table structures to a fresh install
21:16<GreyFoxx>I'm sure I have indexes that are not default
21:17<Captain_Murdoch>looks like it takes about 13-14 on my master backend, so maybe that 20 on my dev box was while I was doing something else. "Scheduled 872 items in 14.3 = 0.00 match + 14.29 place"
21:18<Captain_Murdoch>I regularly truncate the record table on my dev box when I want to test things because it takes so long to reschedule. my production DB is backed up and restored into my dev DB nightly.
21:19<GreyFoxx>Makes me wonder if the "master" query should be rewritten to be several smaller queries and do some of the processing ourselves instead of letting mysql do it
21:20<GreyFoxx>though I suppose mysql in theory should be faster assuming all of the relevant indexes are there
21:20<Captain_Murdoch>oh, I'm still running MySQL v3.23.54a also. I setup a test VM to try switching to 5.x but never made the switch. need to look into that again.
21:21<GreyFoxx>still 4.0.24a here :)
21:21<GreyFoxx>I've had no compelling reason to update it
21:22<Captain_Murdoch>I'd like to take advantage of some of the caching, wondering if that would help me out any. I looked into it a little and that is when I realized how much duplicated settings queries were taking place in Myth so instead of upgrading MySQL, I added the settings cache. :)
21:22<GreyFoxx>Did you ever decide if you were gonna order the Dlink DSM320 ?
21:22<GreyFoxx>I have 20M dedicated to mysqls caching
21:22<GreyFoxx>makes a big difference
21:22<Captain_Murdoch>actually have it setup in the bedroom now. :) that is driving the top 2 things on my TODO list right now.
21:22<GreyFoxx>cool :)
21:23<GreyFoxx>I'm thinking I wanna get2 from newegg and talk someone into turning around and ship them here
21:23<Captain_Murdoch>not sure which order they fall in, but I want to add the ability to have multiple files per 'recording' and also want to be able to have the software encoder (NVR) create mpeg-ps files (or something that can be played on the dsm)
21:23<GreyFoxx>1 for me, and 1 for my dad
21:24<GreyFoxx>Captain_Murdoch: See a new mvpmc out today? Apparently with upnp support
21:24<Captain_Murdoch>yeah, need to send in my rebate form to get my $50 rebate.
21:24<Captain_Murdoch>no, just sold both of my 1000 models on eBay for $60 each. paid $40 at Radio Shack so that wasn't a bad deal.
21:24<Captain_Murdoch>is that the wireless version or a newer model?
21:25<GreyFoxx>If the backend could do on the fly transcoding and tie it into the upnp stuff both mythweb users and mvp users would be in heaven
21:25<Captain_Murdoch>yeah, I've been tossing that idea around also.
21:25<Captain_Murdoch>wanted to take a peek into the upnp code to see how the streaming was handled to know if it could be tied into on-the-fly transcoding or something.
21:25<GreyFoxx>Captain_Murdoch: I dug my mvp out today when I saw the upnp support mentioned
21:26<GreyFoxx>Captain_Murdoch: I'm sure something could be done to tie it in. It would be VERY handy since I could turn around and use that to get xbox360 upnp video support going
21:26<Captain_Murdoch>oh, i read that wrong, I thought you meant new mvp. so they have it in the software now. cool. I never liked the interface much in mvpmc, so I didn't use mine much.
21:26<GreyFoxx>it has to be in acertain codec/res etc
21:27<GreyFoxx>then get mythbackend to allow streaming of mythvideo stuff and it would rock
21:27<Captain_Murdoch>I forgot, did you say you have a dsm
21:27<GreyFoxx>I'm looking to pick up 2, but newegg wont ship here
21:27<GreyFoxx>they hate us canucks :)
21:28|-|rtsai1111 [] has quit [Read error: 110 (Connection timed out)]
21:28<Captain_Murdoch>I couldn't get it to work right with my mythvideo files, but wasn't sure if it was me or what.
21:29<GreyFoxx>If I get a dsm I'll make sure it works, I want to change some stuff related to that anyway to navigate folders
21:29<Captain_Murdoch>problem with it right now is that most shows I watch are HD transcoded to .nuv so I can't watch them on the dsm. most of my wife's shows are off of cable and are recorded using ivtv so they play fine.
21:30<GreyFoxx>With onthe fly transcoding I could use the mvp for everything, and likely get the xbox360 video streaming working
21:30<Captain_Murdoch>yeah, I was also thinking about navigating if I add support for multiple files per recording, so you could navigate to a title then see "1920x1080i" and "640x480" or something on the upnp interface. the Watch Recordings screen would popup and ask you which you wanted to play if there were multiple files.
21:30<CDev>GreyFoxx: I have a few changes I wanted to make to the recorded CDS ... It's not working correctly now, and I wanted to add the ability to delete a recoding.
21:31<Captain_Murdoch>CDev: are there any dangers with that, are there any clients that might walk the tree like search bots have walked people's MythWeb installs and deleted all their recordings?
21:32<CDev>I was planning on having it think it's just another video item. It will walk the tree to get info but shouldn't actually request the video.
21:32<CDev>It would definitely be an option that can be turned on/off in the settings.
21:32<Captain_Murdoch>ah, cool.
21:33<CDev>I've been tied up with the autodiscovery code, so haven't had time to look into it.
21:33<GreyFoxx>CDev: Cool.
21:34<Captain_Murdoch>that will be nice also.
21:34<GreyFoxx>One thing I've noticed is that with the 360 I have to have ":" in my friendly name
21:34<CDev>It should be fairly easy to implemented the transcoding in the upnp code. If you have any questions, just let me know.
21:34<GreyFoxx>So instead of "Mythtv AV MEdia Server" I changed it to "Mythtv: AV Media Server:" and then the 360 sees it as "Mythtv"
21:35<GreyFoxx>CDev: I'm curious about that. Is there a standard of what the requests/parameters passed from client to server should look like ?
21:35<Captain_Murdoch>CDev: would it make sense to change that "Mythtv AV Media Server" (or whatever it is) to something that indicates which myth it is? I'm thinking networks with multiple master backends (like when I bring up my dev backend and can't tell which is which in the upnp browser)
21:36<Captain_Murdoch>s/which myth it is/which backend it is/
21:36<GreyFoxx>add the hostname to it ?
21:36<CDev>GreyFoxx: If the : is a must have, you might want to change the code in upnpdevice.cpp so all xbox360 users would have it working.
21:36<GreyFoxx>CDev: I'll do that
21:36<Captain_Murdoch>GreyFoxx: yeah, I was thinking something like that.
21:36<GreyFoxx>"Mythtv - Host - AV Media Server" and for the xbox "Mythtv - Host: Av MEdia Server:"
21:37<GreyFoxx>that would be handy
21:37<GreyFoxx>especially with multiple master backends on anetwork
21:37<CDev>GreyFoxx: There is a standard for all parameters... however each stack implements things slightly different.
21:38<CDev>Captain_Murdoch: The friendly name is overridable by using a setting per host.
21:38<Captain_Murdoch>still wondering if upnp should be disabled by default with an option to turn it on. especially if you can delete files via it or bypass the recording group passwords.
21:39<Captain_Murdoch>little johnny downloads a upnp client for windows off the net and all the sudden can watch his mommy and daddy's porn channel recordings. :)
21:39<CDev>Right now, you can't delete files... I'm also working on getting the frontend to locate the masterbackend by using the upnp discovery code... so disabling upnp, would make that useless.
21:39<CDev>fyi: the setting to change the device name is: upnpFriendlyName
21:39<GreyFoxx>and with deletion thhey would have to know show ID's and such ahead of time I imahgine
21:39<CDev>just add one per host.
21:40<Captain_Murdoch>cool, thanks.
21:40<GreyFoxx>CDev: That's the setting I'm using now for the xbox :)
21:42[~]CDev just moved all his capture cards to a new server and it's not working :-(
21:48|-|adante [] has joined #mythtv
21:54|-|adante_ [] has joined #mythtv
21:55[~]Captain_Murdoch realizes he already handles the thing he thought might be a bug in the inuseprograms code.
21:57<Captain_Murdoch>janneg: if you see this in scrollback, is the server that is using the file the same one that recorded it? just trying to see where it falls in the logic in ProgramInfo::MarkAsInUse. If it is on the recording host, we should be able to call GetPlaybackURL to find the actual location of the file, but we should also just set the recdir to empty instead of NULL if we only have a filename and can't find the directory.
21:57|-|xris [] has quit [""]
21:58|-|CDev [] has left #mythtv []
21:59|-|CDev [] has joined #mythtv
22:03|-|adante__ [] has quit [Read error: 110 (Connection timed out)]
22:06|-|adante__ [] has joined #mythtv
22:09|-|adante [] has quit [Success]
22:19|-|adante_ [] has quit [Read error: 110 (Connection timed out)]
22:25|-|adante__ [] has quit [Connection timed out]
22:35|-|Nem^1 [] has joined #mythtv
22:39|-|cmug [] has joined #mythtv
22:39<cmug>Hi, does the frontend use perl mysql?
22:46|-|cmug [] has left #mythtv []
22:46|-|jhatch [] has joined #mythtv
22:47|-|jhatch [] has quit [Client Quit]
22:47|-|jhatch [] has joined #mythtv
22:48|-|jhatch [] has quit [Client Quit]
22:49|-|jhatch [] has joined #mythtv
22:53|-|Nem^ [] has quit [Read error: 110 (Connection timed out)]
22:53|-|Nem^1 changed nick to Nem^
22:53|-|rtsai [] has quit ["leaving"]
22:57|-|_lose_thegrim [] has joined #mythtv
22:58|-|_lose_thegrim [] has left #mythtv []
22:59|-|xris [] has joined #mythtv
23:00|-|rtsai [] has joined #mythtv
23:05|-|russellb_ [i=russellb@nat/digium/x-56e401651bc07299] has joined #mythtv
23:07|-|russellb [i=russellb@asterisk/developer-and-stable-maintainer/drumkilla] has quit [Read error: 104 (Connection reset by peer)]
23:10|-|russellb_ [i=russellb@nat/digium/x-56e401651bc07299] has quit [Remote closed the connection]
23:14|-|Skiingsean_sleep changed nick to Skiingsean
23:19|-|cmorgan [] has joined #mythtv
23:23|-|jhatch [] has quit ["Leaving"]
23:28|-|cmorgan [] has quit [Remote closed the connection]
23:37<xris>Chutt: what do you think of a page on the wiki for mythtv devs who would be willing to work on contract stuff related to mythtv? I'm still getting the occasional contact request asking about dev work.
23:39[~]xris grumbles about mysql not indexing NULL fields
---Logclosed Thu Jan 25 00:00:37 2007