Back to Home / #mythtv / 2008 / 02 / Prev Day | Next Day
#mythtv IRC Logs for 2008-02-21

---Logopened Thu Feb 21 00:00:29 2008
00:02-!-leprechau [] has joined #mythtv
00:02-!-greg__ [] has quit ["Ex-Chat"]
00:18-!-sc00p_ [] has joined #mythtv
00:20-!-sc00p [] has quit [Read error: 110 (Connection timed out)]
00:24-!-ahbritto [] has joined #mythtv
00:33-!-jmblack [] has quit []
01:08-!-grokky [] has joined #mythtv
01:26-!-JoeBorn [] has joined #mythtv
01:31-!-leprechau [] has quit [Remote closed the connection]
01:31-!-Netsplit <-> quits: Puhi
01:32-!-grokky [] has quit [Read error: 110 (Connection timed out)]
01:32-!-Netsplit over, joins: Puhi
01:38-!-leprechau [] has joined #mythtv
01:58-!-davilla [] has quit ["Leaving"]
01:59-!-davilla [] has joined #mythtv
02:00-!-davilla [] has quit [Client Quit]
02:08-!-xris [] has quit []
02:11-!-SlimG [n=SlimG@] has joined #mythtv
02:36-!-SlimG [n=SlimG@] has quit []
02:40-!-MGisbers [n=mgisbers@] has joined #mythtv
02:47-!-SlimG [n=SlimG@] has joined #mythtv
03:16-!-SlimG [n=SlimG@] has quit []
03:16-!-beavis [] has joined #mythtv
03:17-!-gnome42 [] has quit [Remote closed the connection]
03:18-!-loops [] has quit [Read error: 104 (Connection reset by peer)]
03:19-!-MGisbers [n=mgisbers@] has quit [Remote closed the connection]
03:19-!-MGisbers [n=mgisbers@] has joined #mythtv
03:20-!-SlimG [n=SlimG@] has joined #mythtv
04:08-!-JoeyBorn [] has joined #mythtv
04:08-!-waini [] has joined #mythtv
04:08<waini>is it possible to start mythbackend with debg mode (i want to see the frequency, sid, .. send out to the dvb-module)?
04:10<stuarta>--verbose channel,record,siparser
04:12-!-grokky [] has joined #mythtv
04:22-!-JoeBorn [] has quit [Read error: 110 (Connection timed out)]
04:22-!-SlimG [n=SlimG@] has quit []
04:31-!-SlimG [n=SlimG@] has joined #mythtv
04:45-!-SlimG [n=SlimG@] has quit []
04:48-!-SlimG [n=SlimG@] has joined #mythtv
05:05-!-grokky [] has quit []
05:40-!-onixian [] has joined #mythtv
06:50-!-sunrop__ [] has quit [Read error: 110 (Connection timed out)]
06:54-!-sunrop [] has joined #mythtv
06:57-!-MaverickTech [] has joined #mythtv
06:59-!-MrGandalf [] has quit ["work"]
07:01-!-waini [] has quit ["This computer has gone to sleep"]
07:01-!-mzb [] has quit [Read error: 113 (No route to host)]
07:02-!-mzb_d800 [] has quit [Read error: 113 (No route to host)]
07:05-!-MavT [] has quit [Read error: 148 (No route to host)]
07:06-!-TelnetManta [] has quit ["Ex-Chat"]
07:54-!-TelnetManta [n=benwilli@] has joined #mythtv
07:59-!-mattwire [] has joined #mythtv
08:03<stuarta>server still fubar then?
08:21-!-ben1092 [n=benben@] has joined #mythtv
08:23-!-waini [] has joined #mythtv
08:24<ben1092>hi guys, I want to know something about mythuitest, It has a dialog that says "Well, how about moving?". The point is that i can't see anything moving in this dialog. I checked MythUIType::HandleMovementPulse the curXY changes each time the function is called. So what do i need to do to see movement effects?
08:37-!-dekar1 [] has joined #mythtv
08:48<gbee>thought the images were missing last time I looked, but I can't check since mythuitest doesn't exist anymore
08:52-!-dekarl [] has quit [Read error: 110 (Connection timed out)]
09:04-!-SlimG [n=SlimG@] has quit []
09:14-!-mykeul [n=mykeul@] has joined #mythtv
09:25-!-Cardoe [n=Cardoe@gentoo/developer/Cardoe] has joined #mythtv
09:26-!-skamithi [] has joined #mythtv
09:35-!-Ojg [] has joined #mythtv
09:36-!-foxhunt [] has joined #mythtv
09:42-!-onixian [] has quit ["This computer has gone to sleep"]
09:42<gbee>baylink on an extended holiday? Wiki still has a notice about SD taking up half the front page
09:45-!-ben1092 [n=benben@] has left #mythtv []
09:54-!-foxhunt [] has quit [Remote closed the connection]
09:58-!-SlimG [n=SlimG@] has joined #mythtv
10:10<GreyFoxx>Is there a config option to turn off the database backup stuff ?
10:18-!-hads [] has quit [Read error: 104 (Connection reset by peer)]
10:18-!-hads [] has joined #mythtv
10:18<jams>i dont think so
10:21<gbee>f'king bastards broke into my shed again >:(
10:25-!-mntmst [] has quit [Read error: 110 (Connection timed out)]
10:25-!-mntmst [] has joined #mythtv
10:26-!-waini [] has quit ["This computer has gone to sleep"]
10:35-!-mick_laptop [n=chatzill@clamwin/admin/mickhome] has joined #mythtv
10:35<mick_laptop>is there a "history of mythtv" somewhere? i'd just kind of be interested in how things started
10:40<GreyFoxx>The background section
10:40<GreyFoxx>that's about it I think :)
10:43-!-MrGandalf [] has joined #mythtv
10:47<mick_laptop>GreyFoxx: thanks
10:59<MrGandalf>danielk22: there?
11:00-!-gnome42 [] has joined #mythtv
11:04*stuarta reads -commits to find out why there was a schema change in -fixes
11:07<gbee>stuarta: janneg committed a fix for the #cpu setting not being saved, required that the value be prepopulated in the database because it used an update statement
11:08<stuarta>fair enuf
11:10-!-onixian [] has joined #mythtv
11:10<GreyFoxx>gbee: Think I've got the "starting at the start of the video" thing fixed properly now
11:11<gbee>GreyFoxx: yeah?
11:11<GreyFoxx>AFD::OpenFile, we are doing av_read_frame_flush before doing the scan for streams, when it should be after
11:11<GreyFoxx>moving it to just after the ScanStreams seems to fix it up
11:12<gbee>cool, makes sense - wish I'd spotted that earlier, drove myself mad looking for the cause :)
11:14<GreyFoxx>You can thank libav's more verbose logging for matroka parsing :)
11:14<GreyFoxx>I started a matroska file this time and it spitt out a bunch of logs just before the ScanStreams output
11:14<GreyFoxx>which led me right too it
11:19-!-loops [] has joined #mythtv
11:21<gbee>just shopping for stuff I can use to secure my garden, unfortunately they don't stock landmines at the DIY
11:21<Hannibal->motion activated sentry guns
11:21-!-waini [] has joined #mythtv
11:21<gbee>Hannibal-: they've got a couple of those, but they are outside my price range ;)
11:22<stuarta>a vicious dog?
11:22<GreyFoxx>so, are we commiting fixes to both trunk and fixes now ?
11:22<gbee>stuarta: cat might eat it
11:23<stuarta>put the cat in the shed
11:23<stuarta>for the robbers
11:23<stuarta>GreyFoxx: depends if you want to leave it up the auto sync or not
11:23<gbee>GreyFoxx: personal preference, anything committed to -fixes will be backported to trunk by Isaac, but not everyone wants the hassle of switching their working copy to -fixes for the duration
11:23<stuarta>personally i like doing both
11:24<GreyFoxx>a goes looking for tickets relating to this bug
11:24<gbee>not everyone knows the 'commit to fixes' plan since they aren't on the mailing list
11:25<stuarta>which they should be, but that's not my call
11:25-!-wain [] has joined #mythtv
11:26<Hannibal->i upgraded to the latest fixes, and pip started working - so thanks to whomever committed that.
11:27<gbee> << For the price, reckon I'll get one of these + PVR-150, will stop me having to run out into the garden everytime I suspect someone is out there
11:27<laga_>gbee: i think justinh has some cctv kit
11:27<gbee>heh, yeah
11:28-!-siXy [i=siXy@] has joined #mythtv
11:28<mick_laptop>who is developing the netflix plugin? wanted to know if adding the username/password into the UI was planned
11:28<mick_laptop>if not, i'd write it
11:30<gbee>think Kevin is the only one who has done any netflix work in the last year
11:30-!-rod_ [] has joined #mythtv
11:32<laga_>oops, wrong window
11:32-!-wain [] has left #mythtv ["Apple forever!"]
11:36-!-johnp__ [] has joined #mythtv
11:39-!-JoeyBorn [] has quit [Read error: 104 (Connection reset by peer)]
11:42-!-waini [] has quit [Read error: 110 (Connection timed out)]
11:48-!-rod__ [] has quit [Read error: 113 (No route to host)]
11:49-!-xris [] has joined #mythtv
11:51-!-MGisbers [n=mgisbers@] has quit [Read error: 110 (Connection timed out)]
11:51-!-t0ny-p40 [n=t0ny-p40@] has quit [Read error: 110 (Connection timed out)]
11:54-!-t0ny-p40 [n=t0ny-p40@] has joined #mythtv
12:01-!-xris [] has quit []
12:04<mick_laptop>gbee: what is kevin's nick?
12:04<gbee>mick_laptop: don't know that I've ever seen him in IRC, sorry
12:05<mick_laptop>k - thnaks
12:05-!-xris [] has joined #mythtv
12:05-!-MGisbers [] has joined #mythtv
12:10<Anduin>Under the "last to touch owns" rule, it is Captain_Murdoch's plugin.
12:10*stuarta chuckles
12:16-!-t0ny-p40 [n=t0ny-p40@] has quit [Connection timed out]
12:24<danielk22>why are people other than Isaac doing syncs from fixes to trunk? Did I miss an e-mail?
12:28<MrGandalf>danielk22: I know your busy, was wondering if you had a few seconds for a question regarding that ticket you helped me with (stuck tuning)
12:29<MrGandalf>danielk22: was wondering if something simular could effect IVTV recording. I'm seeing a case of stuck tuning when going from IVTV -> DVB waiting for the ringbuffer to close.
12:30<danielk22>it would probably be something entirely different..
12:30<Chutt>danielk22, individual people can, if they want to
12:30<Chutt>i'll do it every day or two as a convenience, though
12:30<MrGandalf>k, I have a bt but I didn't step through any threads but one so I'll have to investigate further next time it happens.
12:30<danielk22>chutt: doesn't that complicate your merge?
12:31<Chutt>not as long as the change is the same in both places
12:31<Chutt>svn merge isn't _entirely_ stupid
12:31<danielk22>ok, I won't worry about it then..
12:31<Chutt>some people are going trunk->branch, too, not the other way 'round
12:32<danielk22>mrg: look at dmesg too...
12:32<Chutt>danielk22, any idea why video playback takes so long to start these days?
12:33<danielk22>there might be some problem with closing the ivtv device.
12:33<Chutt>seemed to get much longer after the multi-rec merge
12:33<Chutt>or, maybe the -vid merge, i can't remember =)
12:33<GreyFoxx>Chutt: One thing that adds time, and more so if you are nfs rooting is the way we preload CC fonts
12:33<danielk22>chutt: don't know. but I think it predated multi-rec merge.
12:33<Chutt>GreyFoxx, local root, nfs videos
12:33<danielk22>vid merge might be a more likely target.
12:34<danielk22>I haven't tried to trace it yet.
12:34<GreyFoxx>Even if you don't use CC we preload, and prescale fonts adds seconds on a nfsrooted setup
12:35<danielk22>greyfoxx: couldn't we do that once and cache?
12:35<GreyFoxx>yeah. I had planned on change it to load them on demand, and if we had to scale cache that for the next load
12:35-!-JoeBorn [] has joined #mythtv
12:36<danielk22>For the 708 fonts, 99% of the time captioners only use one font too...
12:36<GreyFoxx>I think we are even doing the rescaling and such for each channel change
12:36<GreyFoxx>for livetvf users
12:37<GreyFoxx>but it was a while ago I really looked at it
12:38<Chutt>those need to be on demand :p
12:40<danielk22>I'd be willing to look at that after #3995 & #4631.. (I'm waiting for a better fix for #4513 from Mark Buechler)
12:40*stuarta wonders if that's MrGandalf
12:40<GreyFoxx>it is
12:40<stuarta>thought it was
12:41<danielk22>sorry not Mark Buechler, the other Mark, Mr Kendall :)
12:42<danielk22>he understands the problem Markus was seeing...
12:42<MrGandalf>first too many Stuarts, now too many Marks.. oh my.
12:43<danielk22>heh, I just took a quick look at my inbox and saw the wrong mark e-mail for the name..
12:44<danielk22>and we had too many chris' before that :)
12:44<danielk22>too many chrisi ? I know the apostrophe is wrong, but what is the plurals of Chris?
12:45-!-onixian [] has quit ["Leaving"]
12:52<laga_>chrises? hum
12:52<laga_>yeah, chrises seems to be right
12:53-!-mykeul [n=mykeul@] has left #mythtv []
12:54<danielk22>heh. Yes the Joneses argument swayed me :)
12:56<laga_>although that guy's question is messed up because he missed the difference between possessive case and plural
12:56*laga_ blames web 2.0
12:58<danielk22>Hmm, with local playback using an NFS filesystem playback starts in a few hundred milliseconds here, without a CPU spike.
12:59<danielk22>Chutt do you know if you're perhaps using MythTV streaming?
13:01<Chutt>it sure better not be
13:01<Chutt>the files are mounted locally
13:01<Chutt>it's a single backend+frontend box
13:02<Chutt>i don't have logs, though
13:02<MrGandalf>select * from settings where value = 'AlwaysStreamFiles';
13:03<Chutt>0, of course
13:04<GreyFoxx>danielk22: Wow, that's fast
13:04<Chutt>danielk22, hd feels like it takes at least 20 seconds to start
13:04<Chutt>i haven't measured it
13:05<Chutt>i don't remember if it takes as long for SD or not
13:07-!-GreyFoxx [] has quit [Remote closed the connection]
13:07-!-GreyFoxx [] has joined #Mythtv
13:09-!-t0ny-p40 [n=t0ny-p40@] has joined #mythtv
13:11*Dibblah thinks someone on the mailing list doesn't quite realise that the mythtv executable is not really the ideal thing to be running :)
13:15<gnome42>20 seconds! Holy sheep sugar! :)
13:15<gnome42>GreyFoxx: you see similar startup times?
13:15<gnome42>mine are in line with danielk22's times.
13:16<gnome42>nfs mounted as well
13:18<GreyFoxx>gnome42: my nfsrooted celeron733 take 9-10 seconds, BUT if I commend out the CC font loading it drops to 4
13:18<GreyFoxx>my regular frontend (via nfs) takes 2 seconds, 3 or one files
13:18<Daviey>that is floody fast
13:19<GreyFoxx>Daviey: It use to be 1 :/ gotten slower
13:19*Daviey misunderstood
13:19<Daviey>thought it was nfsroot boot time
13:19<GreyFoxx>my nfsroot boottime is 12 seconds :)
13:20<danielk22>are these slow start times with slow CPUs?
13:20<Daviey>inc. bios post?
13:20<GreyFoxx>Daviey: Yup
13:21<Daviey>thats great!
13:21<Daviey>I was thinking of suspending to disk rather than shutdown for quicker boots.
13:22<GreyFoxx>danielk22: my "regular" frontend, showing 2-3 seconds is an AMD sempron 2800
13:22<GreyFoxx>nvidia board
13:22<Daviey>ati :(
13:22<danielk22>hmm, that's a pretty reasonable processor...
13:22<GreyFoxx>Daviey: I have that on my main frontend, the difference is crazy :)
13:22*Daviey needs more freetime to play
13:23<Chutt>danielk22, model name : AMD Athlon(tm) X2 Dual Core Processor BE-2350
13:23<Chutt>so, not really a slow cpu :p
13:23<danielk22>chutt: yours is taking seconds too?
13:23-!-TelnetManta [n=benwilli@] has quit ["Ex-Chat"]
13:23<Chutt>it takes a very long time to start HD playback
13:23<Daviey>last week my main FE was an old laptop due to borked xorg on real main fe, and not enough free time
13:24<gnome42>GreyFoxx: hmm, does the DB do a longish CPU spike at startup?
13:25<GreyFoxx>gnome42: Not sure actually
13:25<Chutt>2008-02-21 13:24:16.192
13:25<Chutt>2008-02-21 13:24:23.937
13:25*GreyFoxx goes to grab a log entry
13:25<danielk22>gnome: you think it's downloading the keyframe map?
13:25<Chutt>mythtv 'filename'
13:25<gbee>IVTV -> DVB switch takes an hour here (well upto 10 seconds)
13:25<Chutt>livetv startup takes a really long time, too.
13:26<danielk22>heh, I feel like bjm being told the scheduler takes forever.. It's really fast HERE :)
13:26<gbee>not really related to the current discussion, just throwing it out there in case someone wants to take look at speeding things up, this was before multirec as well
13:27<gnome42>danielk22: I suspect something to do with the DB. But it should be obvious (via top or whatever) to see if its the DB spiking at startup.
13:27<Chutt>2008-02-21 13:24:18.658 Resyncing position map. posmapStarted = 0 livetv(0) watchingRec(0)
13:27<Chutt>2008-02-21 13:24:21.251 Position map filled from DB to: 125174
13:27<gbee>one day I'll look at what is taking so long myself - seems like it's the closing down of IVTV which takes most of the time
13:27<Chutt>appears to be taking way too long
13:27<danielk22>I'm not seeing mysql pop up at all in top
13:28<gnome42>danielk22: My DB suffered some silent borkage after updating to mysql 5, which caused similar issues.
13:28<danielk22>chutt: yeah that is a long time..
13:28<GreyFoxx>just now kicking off a mp4 file I encoded ... 9 seconds
13:28<gnome42>danielk22: I get cranky if I have to wait more than a second :)
13:29<danielk22>yeah, my cranky starts at 250 ms
13:30<Chutt>480658 rows in recordedseek;
13:30-!-jgarvey [] has joined #mythtv
13:31<danielk22>927539 here
13:31<GreyFoxx>That's from a local database, local backend, playing a mp4 file I generated with no seektable entries
13:31<gnome42>Chutt: I have 2.8M rows
13:32<GreyFoxx>the upnpo auto discover seens to be causing 3 seconds of my delay
13:32<GreyFoxx>on this particular system
13:33<Chutt>doing the resync select manually takes .01 seconds in mysql.
13:33<gnome42>oh, I use the -d switch and pnp disabled on the backend
13:34<danielk22>chutt: moving the data may take most of the time... still 3 seconds is excessive
13:34-!-_gunni_ [] has joined #mythtv
13:34<Chutt>probably inserting those values one at a time into the map takes forever.
13:34<Chutt>if it's resizing the map..
13:36<gnome42>13:34:43.068 ProgramInfo: GetPositionMap(chanid: 3342 recstartts: 2007-12-11T03:00:00 type: 9)
13:36<gnome42>13:34:43.165 Position map filled from DB to: 204896
13:36<gnome42>That's an old rec. so it wasn't in the sql cache
13:38<danielk22> ?
13:39<Chutt>2008-02-21 13:39:09.391 Resyncing position map. posmapStarted = 0 livetv(0) watchingRec(0)
13:39<Chutt>2008-02-21 13:39:09.578 Position map filled from DB to: 125173
13:39<Chutt>after optimize, on a new file.
13:40-!-SlimG [n=SlimG@] has quit [Read error: 113 (No route to host)]
13:40-!-SlimG [n=SlimG@] has joined #mythtv
13:40<Chutt>total time from start to video (end of debug spew) with mythtv: 13:39:06.219 -> 13:39:10.771
13:40<Chutt>so, better, but still 4 seconds
13:41<gnome42>Chutt: you have Gigs of ram in the DB I assume?
13:42<gbee>we ought to run the optimise/repair stuff in the housekeeper on at least a weekly, if not daily basis
13:42<Chutt>in the db?
13:42<danielk22>hmmm, I started a new recording 10 minutes ago, and I got this just now:
13:42<danielk22>2008-02-21 13:40:27.913 TV: Attempting to change from None to WatchingPreRecorded
13:42<danielk22>2008-02-21 13:40:27.915 RingBuf(/video/testing/1019021_20080221132449.mpg): OpenFile(/video/testing/1019021_20080221132449.mpg, 12)
13:42<danielk22>2008-02-21 13:40:27.915 RingBuf(/video/testing/1019021_20080221132449.mpg): CalcReadAheadThresh(10966 KB)
13:42<danielk22> -> threshhold(64 KB) min read(0 KB) blk size(32 KB)
13:42<danielk22>2008-02-21 13:40:29.042 RingBuf(/video/testing/1019021_20080221132449.mpg): Waited 1.0 seconds for data to become available...
13:42<danielk22>2008-02-21 13:40:30.050 RingBuf(/video/testing/1019021_20080221132449.mpg): Waited 2.0 seconds for data to become available...
13:42<danielk22>2008-02-21 13:40:30.437 AFD: Stream #0, has id 0x49 codec id MPEG2VIDEO, type Video, bitrate 18500000 at 0x0x951b10
13:42<gbee>danielk22: might be easier to paste it to
13:42<danielk22>Which doesn't make much sense.. The file was definately there..
13:43<danielk22>thx gbee
13:43<Dibblah>danielk22: Live TV?
13:43<Dibblah>Oh - No.
13:43<danielk22>well it is still recording, but not livetv
13:43<gnome42>Chutt: never mind. I have a slower DB server with only 512 MB so your DB times are still slow I think.
13:43<Dibblah>Nah - If it was live, it could have been NFS attribute caching.
13:43<Chutt>i just have a GB of ram on this box
13:44<Chutt>that better be enough :p
13:44<gnome42>Chutt: yeh, agreed!
13:45-!-TelnetManta [n=benwilli@] has joined #mythtv
13:45-!-johnp__ [] has left #mythtv []
13:46<gnome42>hmm, not using the indexes kinda fits the scenario
13:46-!-Ojg [] has quit ["( :: NoNameScript 4.2 :: )"]
13:46<danielk22>dibblah: that's possible.. i'll turn that off..
13:46<Dibblah>Not if it's a recording, it's not.
13:46<gbee>GreyFoxx: start fix didn't work here on the h.264 file I've been using for debugging
13:47<MrGandalf>danielk22: the file may be there but not >= CHUNK, as defined in RingBuffer.cpp
13:47-!-ahbritto [] has quit [Client Quit]
13:47<GreyFoxx>gbee: Yeah I've noticd that it works for my mkv file, but NOT for my mp4
13:47<GreyFoxx>so I'm going to put in the DFF anyway
13:47<danielk22>mrg: 1.2 GB..
13:48<gnome42>Chutt: Hmm, maybe should check if you have the same DB borkage I came across. optimize/repair and stuff will pass OK without fixing it I think
13:48<MrGandalf>yea, that's bigger
13:48-!-Cheavera [] has joined #mythtv
13:48<Chutt>gnome42, after running optimize, it's fast at loading the posmap.
13:48*MrGandalf should read more carefully
13:49<gbee>yeah, should help for now - though I don't think it's going to work for files with a posmap since those are created with a bad initial offset (I think) - might be worth double checking that, what's the value for the first offset in the database?
13:49<Chutt>that posmap insert thing should be reworked, though
13:49<Chutt>map[key] = value
13:49<Chutt>is slower than map.insert(value)
13:49<Chutt>(first does a find, then an insert if it fails)
13:49-!-Cheavera [] has left #mythtv []
13:50<Chutt>inserting all new guaranteed unique items, so..
13:50<gbee>GreyFoxx: whoah, weird - that file played with 'mythtv' show the start problem, but when played through mythvideo it doesn't
13:50<danielk22>chutt are you sure it's slower? an insert needs to do a find for insert on a balanced tree anyway
13:51<gbee>there is no good reason for that to happen, they use exactly the same code path, the only difference is the logging
13:51<Chutt>danielk22, check qt's code
13:52<Chutt> detach();
13:52<Chutt> QMapNode<Key,T>* p = sh->find( k ).node;
13:52<Chutt> if ( p != sh->end().node )
13:52<Chutt> return p->data;
13:52<Chutt> return insert( k, T() ).data();
13:52<Chutt>that's operator[]
13:52<danielk22>heh, ok then..
13:52<Chutt>so even if insert does the fine..
13:52<gbee>ahh, I was wrong, your fix _is_ working for the positionmap ... I'll take a look at it in a while, I think the fix I had earlier might now work with your fix in place
13:52<danielk22>i guess i was spoiled by stl
13:52<Chutt>err, find. =)
13:56<danielk22>dibblah: I turned off NFS attribute caching and I'm still seeing the RingBuffer slow start..
13:57<gnome42>Chutt: But still too slow! :) Your posmap load took twice as long as mine to load roughly half as many entries.
13:57<Dibblah>Well, for non-live I wouldn't expect anything eles.
13:57<Dibblah>else, even.
13:58<gnome42>Chutt: The speedup could have come from priming the fs cache by running optimize. Just an idea :)
14:03*gnome42 steps away for a bit
14:07-!-mykeul [] has joined #mythtv
14:13<danielk22>I've opened a ticket for the RingBuffer problem I'm seeing #4727.
14:15-!-RaYmAn-Bx [] has quit [Read error: 110 (Connection timed out)]
14:21<gnome42>danielk22: is that with Greg's av_read_frame_flush fix in place? [16187]
14:24-!-beavis_ [] has joined #mythtv
14:41-!-beavis [] has quit [Read error: 113 (No route to host)]
14:49-!-melunko_ [n=hmelo@] has joined #mythtv
14:50-!-melunko_ [n=hmelo@] has quit [Read error: 104 (Connection reset by peer)]
14:53-!-reynaldo [] has joined #mythtv
15:25-!-skamithi [] has quit ["WeeChat 0.2.3"]
15:27-!-|gunni| [] has joined #mythtv
15:31-!-hume [] has joined #mythtv
15:32<hume>hello.... i've got a problem: suddenly myth turned quiet - probably my kid did something with the keyboard that I dont know now, while playin - is there a "mute" function or something?
15:33*stuarta points at the topic
15:33<hume>sorry.... wrong channel...
15:33<Hannibal->20 lashes.
15:34<stuarta>electrocution followed by a good ole fashioned tar n feathering
15:43-!-_gunni_ [] has quit [Read error: 110 (Connection timed out)]
15:46-!-hume [] has left #mythtv []
15:47-!-RaYmAn-Bx [] has joined #mythtv
15:49-!-foxhunt [] has joined #mythtv
15:51-!-grokky [] has joined #mythtv
15:52-!-mick_laptop [n=chatzill@clamwin/admin/mickhome] has quit [Read error: 110 (Connection timed out)]
15:56<Daviey>win 1
16:03<ma9mwah>does SVN diff create a file?
16:05-!-kormoc [n=kormoc@unaffiliated/kormoc] has joined #mythtv
16:06<hads>ma9mwah: svn diff > ~/mythtv_mybugfix.diff
16:06<gnome42>ma9mwah: svn diff > the_new_patch.diff will create a diff file
16:08-!-richards [] has joined #mythtv
16:11-!-JoeBorn [] has quit [Read error: 104 (Connection reset by peer)]
16:11<ma9mwah>and i just attach that file to the ticket
16:12<gnome42>ma9mwah: yep, and provide a description of what it hopes to fix. :)
16:27-!-MrGandalf [] has quit ["home"]
16:33<Hannibal->i might love you if it's the crashing on changing tuners, or the mythtv channel scanner crash ;-)
16:36<ma9mwah>sadly not :( my programming skills/knowledge aren't up to something like that. mine was just a simple division by zero error patch for the mythweb stats page
16:39-!-TelnetManta [n=benwilli@] has quit ["Ex-Chat"]
16:40-!-MavT [] has joined #mythtv
16:40*xris is slightly disturbed:
16:40-!-absalom [n=simon@] has joined #mythtv
16:47<gnome42>xris: that ohloh is neat, haven't seen it before
16:49<Chutt>they used to have better graphs
16:49<xris>gnome42: was recently posted to /.
16:49<xris>surprised to see that I have a bunch of kudos points. heh
16:49<xris>not sure whether or not I should create an acct
16:51<gnome42>does it look accurate to you guys?
16:55<xris>ohloh? looks like a weird way to rank people.
16:55<xris>I'm apparently ranked 3rd for mythtv devs by number of commits.
16:55<xris>which is a silly ranking system
16:57<janneg>number of committed lines is totally of for me due to ffmpeg merges
16:57-!-djc__ [n=djc@] has quit [Read error: 110 (Connection timed out)]
16:57-!-MaverickTech [] has quit [Read error: 113 (No route to host)]
16:58-!-djc__ [n=djc@] has joined #mythtv
16:59<gnome42>Yeah, I wasn't thinking ranking person against person, just a look at the version control history data :)
17:00<gnome42>pretty graphs and what not
17:05-!-absalom [n=simon@] has quit [Remote closed the connection]
17:05<Cardoe>someone just sent me a comment on my blog about the SA 4250 HDCs
17:06<Cardoe>might be interesting if I can get in touch with this guy and get some info
17:06<Cardoe>but it basically looks like SA has an option for cable companies to "obfuscate" so that it doesn't follow the standard properly.
17:07<Cardoe>but it might be possible to make these suckers work with myth
17:08-!-foxhunt [] has quit [Remote closed the connection]
17:13-!-reynaldo [] has quit [Read error: 113 (No route to host)]
17:22-!-jwhite [] has quit [Read error: 104 (Connection reset by peer)]
17:23-!-jwhite [] has joined #mythtv
17:27-!-MrGandalf [] has joined #mythtv
17:32-!-mykeul [] has left #mythtv []
17:39-!-jmk_ [] has joined #mythtv
17:39-!-jmk [] has quit [Read error: 113 (No route to host)]
17:46<danielk22>ohloh is way off on actual written lines of code, since much of what committers do is review and commit other people's code..
17:46-!-noisymime [] has joined #mythtv
17:47<danielk22>i like the graphs though..
17:47-!-jgarvey [] has quit ["Leaving"]
17:48<danielk22>gnome42: "danielk22: is that with Greg's av_read_frame_flush fix in place? [16187]" I believe so, I did an svn up + compile and install, just before doing the test.
17:50<danielk22>cardoe: It would be great to get these 4250's working.. we can detect them, but AFAIK we just print out a warning message that they don't work...
17:59<Cardoe>well I'll grab some more info when I get it
17:59-!-Cardoe [n=Cardoe@gentoo/developer/Cardoe] has quit ["Leaving"]
18:00<janneg>danielk22: I've checked the qt source and they are using boyer moore if you don't use a regexp
18:00<danielk22>great boyer moore is fast..
18:01<danielk22>I wonder if some of the eit fixups might not be able to use QString.. rather than regular expressions..
18:01<danielk22>a 0.22 consideration, of course
18:01<MrGandalf>hmm, always wondered what would happen if a rungbuffer change and a channel change happened at the same time.. not good :(
18:02<danielk22>mrg: those are supposed to be blocked by tv_play's keyboard/network handling..
18:02<gbee>xris: found ohloh a while back, they weren't tracking plugins so I registered an account just to change it, didn't see fair to ignore plugin committers :)
18:02<danielk22>most input is blocked during a ring buffer change...
18:03<MrGandalf>danielk22: scenerio, channel change in effect (rotor still moving), ringbuffer changes due to a new program on the target channel
18:03<xris>gbee: heh
18:03<danielk22>mrg: we should probably delay the ringbuffer change in that case...
18:03<MrGandalf>it's pretty rare
18:04<gbee>it's not really an accurate picture on a per-person basis and I'm not sure I like the idea of ranking contributors - it's not a competition
18:04-!-t0ny2 [n=t0ny-p40@] has joined #mythtv
18:05-!-Netsplit <-> quits: jwhite, renatofilho^_, jd86, onyxsoft, sunrop, richards, gnome42, _crichardson, beavis_, nordenm_, (+5 more, use /NETSPLIT to show all of them)
18:05<janneg>danielk22: probably. I'm still wondering why we see the pthread regexp issue only now
18:05-!-t0ny2 is now known as t0ny-p40
18:06<danielk22>janne, I'm a bit sceptical of his analysis.
18:07-!-crichardson [n=crichard@] has joined #mythtv
18:07<danielk22>but if the QThread stuff fixes it for now.... we can find out why later.
18:08<janneg>I'm not sure, the qt documentation is pretty clear on it though
18:10<janneg>but QThread seems to be the way to go if we want to have a native windows port
18:10<Chutt>might as well convert the other threads
18:10<danielk22>But we can't use it to set special priorities to the threads can we?
18:10<janneg>not sure if it's possible due ffmpeg depending on cygwin or mingw
18:11<danielk22>I'm assuming it supports joinable and detached threads..
18:11-!-nordenm [] has joined #mythtv
18:11-!-SlimG [n=SlimG@] has joined #mythtv
18:11<danielk22>We also sometimes need to use pthreads on OSX because we need to attach special cocoa things...
18:12<danielk22>I think if we use qt threads we'll need to put em in a myth thread wrapper which uses pthreads when needed...
18:13<danielk22>not that this wouldn't be a bad idea anyway...
18:14-!-yfaykya [] has joined #mythtv
18:15<gbee>QThread supports priorities
18:15<janneg>qt 3.3 doesn't seem to support detached threads
18:15<danielk22>FYI Qt docs say QRegExp says that it is re-entrant, not thread-safe...
18:15<danielk22>janne, yeah I just looked, that's a bit surprising.
18:16-!-noisymime [] has joined #mythtv
18:16-!-jwhite [] has joined #mythtv
18:16-!-richards [] has joined #mythtv
18:16-!-beavis_ [] has joined #mythtv
18:16-!-gnome42 [] has joined #mythtv
18:16-!-sunrop [] has joined #mythtv
18:16-!-onyxsoft [] has joined #mythtv
18:16-!-renatofilho^_ [n=renato@] has joined #mythtv
18:16-!-lsobral [n=sobral@] has joined #mythtv
18:16-!-Tanthrix [] has joined #mythtv
18:16-!-nordenm_ [] has joined #mythtv
18:16-!-_crichardson [n=crichard@] has joined #mythtv
18:16-!-jd86 [n=jdio@pdpc/supporter/student/jd86] has joined #mythtv
18:18<danielk22>hmm, Qt 4.0 QThread seems pretty braindead too :(
18:18<janneg>yes, it can't be detached either but emits a signal when it's finished
18:19<danielk22>maybe we just need to create QThreadStorage() ?
18:19-!-Netsplit <-> quits: onyxsoft, _crichardson, sunrop, gnome42, renatofilho^_, richards, jwhite, noisymime, nordenm_, Tanthrix, (+3 more, use /NETSPLIT to show all of them)
18:19-!-Tanthrix_AFK [] has joined #mythtv
18:19-!-Netsplit over, joins: gnome42
18:19-!-Netsplit over, joins: onyxsoft, jd86, jwhite
18:19-!-Netsplit over, joins: renatofilho^_
18:19-!-Netsplit over, joins: sunrop, lsobral
18:20-!-Netsplit over, joins: beavis_
18:20<danielk22>That's basically the type of thing we need to create for the OSX pthreads to use the firewire device..
18:20<janneg>danielk22: qt docs claim we can't
18:20<danielk22>janne, are signals/slots threadsafe in Qt4.0?
18:20-!-beavis_ [] has quit [Client Quit]
18:20<danielk22>janne: they may claim that, but they implement it somehow...
18:22<danielk22>hmm, maybe we can just replace QRegExp... maybe that's simpler...
18:23<janneg>iirc they are thread safe, would be braindead if the weren't to use them to signal thread status
18:24<janneg>danielk22: I doubt it would be simpler for eitfixups. might be possible for the channel seperators
18:24<danielk22>janne, I'm looking at
18:24-!-hads [] has quit [Remote closed the connection]
18:24-!-hads [] has joined #mythtv
18:25<danielk22>by default signals and slots are still not thread-safe, but there is a new "queued" signal/slot mechanism, which is threadsafe...
18:27<janneg>danielk22: By default, QObject::connect() establishes ..., and a queued connection if they live in different threads.
18:28<danielk22>Hmm, it sounds like Qt containers are still not thread-safe in Qt 4.0, but they are mostly reentrant now (I know "fully" is an exageration in that doc, there are still unsafe QString functions like ::ascii().)
18:29<danielk22>janne, how does it know which thread the pointer lives in? Or is it based on who first created the object?
18:32-!-czth_ [n=dbrobins@nat/microsoft/x-966ff26775d6cccb] has joined #mythtv
18:34<janneg>I would guess that qt records which thread created the object
18:36<danielk22>Makes sense, instead of all deletes needing to be made in the one Qt event thread, you can delete an object within the thread that created it; without worrying about disconnecting all the slots first.
18:37<danielk22>It's gonna be a PITA to port MythTV to Qt 4.0 :[
18:46-!-ahbritto [] has joined #mythtv
18:49-!-czth [n=dbrobins@nat/microsoft/x-4d5225f129bc8786] has quit [Read error: 110 (Connection timed out)]
18:52-!-mzb_d800 [] has joined #mythtv
18:52-!-mzb [] has joined #mythtv
18:58-!-JoeBorn [] has joined #mythtv
19:01-!-CDev [] has left #mythtv []
19:02-!-CDev [] has joined #MYTHTV
19:04-!-CDev [] has left #MYTHTV []
19:04-!-CDev [] has joined #mythtv
19:07-!-siXy [i=siXy@] has quit ["bye!"]
19:11-!-yfaykya [] has left #mythtv []
19:14-!-skwashd [n=skwashd@phpgroupware/skwashd] has joined #mythtv
19:15-!-skwashd [n=skwashd@phpgroupware/skwashd] has left #mythtv ["tia"]
19:19-!-TelnetManta [] has joined #mythtv
19:31-!-|gunni| [] has quit ["KVIrc 3.2.4 Anomalies"]
19:35-!-SlimG [n=SlimG@] has quit []
19:36-!-eskil [] has joined #mythtv
19:36-!-JoeBorn [] has quit ["Leaving"]
19:36-!-SlimG [n=SlimG@] has joined #mythtv
19:37-!-eskil [] has quit [Client Quit]
19:38-!-mattwire [] has quit ["Leaving"]
19:41-!-eskil [] has joined #mythtv
19:50-!-jamesd [] has joined #mythtv
19:58<Chutt>danielk22, we need to move away from using qobject for stuff, i think
19:58<Chutt>even though it's handy :/
20:05<Chutt>now if only there weren't things like 'customers' and 'software qa' teams, i could write mythtv code..
20:07*jamesd doesn't like using code that isn't QA'd... its pretty painful
20:09<Chutt>conflict in mythnews
20:10<Chutt>wasn't at tot
20:11<Chutt>damn, more conflicts
20:22-!-xris [] has quit []
20:25-!-jamesd [] has quit [Read error: 104 (Connection reset by peer)]
20:34<Hannibal->any suggestions on how to capture debug for this? my wife is upstairs watching tv. i come downstairs; flip on the tv - starts up on a diff tuner, i switch channels to a satellite input - and it makes her tv upstairs change channels... tho it still thinks it's tuned to the current one. i believe it's a channel on the same transport, i'm guessing it COULD have something to do with multirec?
20:36<hads>I saw something vaguely similar once after multirec was merged. Wife changed channel in the bedroom and the channel info poped up on the lounge frontend. Only happened once though.
20:37<Hannibal->i'm going to try setting the number of recordings down to 1
20:37<Hannibal->it's at 2 right now.
20:37<Hannibal->it doesn't happen all the time, but it happens enough.
20:45-!-rmeden [] has joined #mythtv
21:01-!-rmeden [] has left #mythtv []
21:02-!-MGisbers_ [] has joined #mythtv
21:03-!-xris [] has joined #mythtv
21:19-!-MGisbers [] has quit [Read error: 110 (Connection timed out)]
21:34-!-danielk22 [] has left #mythtv []
21:38-!-superm1 is now known as tgm4883_
21:38-!-kormoc [n=kormoc@unaffiliated/kormoc] has quit []
21:39-!-tgm4883_ is now known as superm1
21:53-!-davez0r2222 [] has joined #mythtv
21:54-!-davez0r2222 [] has left #mythtv []
21:55-!-davilla [] has joined #mythtv
22:07-!-Cardoe [n=Cardoe@gentoo/developer/Cardoe] has joined #mythtv
22:31-!-eskil is now known as eskil_afk
22:58-!-xris [] has quit []
23:05-!-Cardoe [n=Cardoe@gentoo/developer/Cardoe] has quit ["Leaving"]
23:06-!-danielk22 [] has joined #mythtv
23:10<Chutt>dumb vfd :/
23:11-!-xris [] has joined #mythtv
23:32<Chutt>man, ubuntu really is crappy at times
23:32<Chutt>all i want to do is recompile lirc
23:33<Chutt>well, the modules
23:38<superm1>Chutt, hardy or gutsy?
23:38<superm1>lirc-modules-source should do it by its own
23:38<Chutt>m-a insists that lirc-modules-source isn't installed
23:38<superm1>via dkms
23:38<Chutt>i need to recompile it, local change
23:38<Chutt>to try to fix my vfd
23:38<superm1>add the local change to /usr/src/lirc
23:39<superm1>and then use dkms to recompile it
23:39<superm1>let me check , sec
23:40<superm1>it should be something like this: dkms -m lirc -v 0.8.3~pre1 build
23:40-!-grokky [] has quit [Read error: 110 (Connection timed out)]
23:41<superm1>the exact command form the postinst is this
23:41<superm1> dkms build -m lirc -v $CVERSION > /dev/null
23:41<superm1> dkms install -m lirc -v $CVERSION --force > /dev/null
23:41<superm1>where CVERSION is the version installed
23:41<superm1>queried something like this
23:41<Chutt>had to do a remove, first
23:41<superm1>CVERSION=`dpkg-query -W -f='${Version}' lirc-modules-source | awk -F "-" '{print $1}'`
23:42<Chutt>Error! DKMS tree does not contain: lirc-0.8.3~pre1
23:42<Chutt>Build cannot continue without the proper tree.
23:42<Chutt>after doing a 'remove' like it told me to in the error
23:42<superm1>you might have to readd it to the tree if you removed it
23:43<superm1>(dkms add -m lirc -v $CVERSION)
23:45-!-Cardoe [n=Cardoe@gentoo/developer/Cardoe] has joined #mythtv
23:45<superm1>no prob. the nicer thing about this versus module assistant is that it does it for u when you switch kernels
23:45<superm1>so if they decide to roll a 2.6.24-9-generic, you just upgrade your system like normal, reboot and its done
23:48-!-Cardoe [n=Cardoe@gentoo/developer/Cardoe] has quit [Client Quit]
23:49<Chutt>ah, nice
23:49<Chutt>it won't keep my local change, will it?
23:53<superm1>it will
23:53<superm1>your local change would only go away if you installed a newer lirc-modules-source
23:57<Chutt>still garbage on the vfd
---Logclosed Fri Feb 22 00:00:14 2008