Beirdobot, irc.freenode.net

Tuesday, December 14th, 2010, 00:34 UTC

[00:34:04]

Mudboy (Mudboy!~zod@cpe-75-83-190-215.socal.res.rr.com) has joined #mythtv

[00:38:36]

castlec (castlec!~christoph@cpe-76-181-105-97.insight.res.rr.com) has joined #mythtv

[00:40:39]

Beirdo:

so. anyone have a Windows box that they leave on mostly 24/7 that has the capacity to run as a build slave once I iron out the Windows build sequence?

[00:41:06]

Beirdo:

likewise for OSX.

[00:41:19]

iamlindoro:

I can probably manage OS X from my work workstation

[00:41:41]

kormoc:

OS X builds are kinda broken right now

[00:41:47]

iamlindoro:

It'd be better for me if we can keep it on a nighttime duty cycle, but could probably manage one or two during the day too

[00:41:50]

kormoc:

least with the packager script they are

[00:42:07]

Beirdo:

cool. I still need to figure out how I want to make it build in those two cases, but if we have willing/able targets, all the better

[00:42:11]

iamlindoro:

kormoc, Where are they dying?

[00:42:18]

iamlindoro:

I haven't done it for a few weeks

[00:42:57]

kormoc:

iamlindoro, the packager script hasn't been updated for GIT and it's been proving kinda a pain

[00:43:03]

iamlindoro:

ah, right, of course

[00:43:04]

iamlindoro:

duh

[00:43:26]

kormoc:

I haven't gotten it to handle things cleanly yet at all and it's been a pretty invasive change so far

[00:59:32]

superm1:

Beirdo, is it not going to be possible to cross compile for OSX or Windows?

[00:59:39]

the_alien (the_alien!~till@xbmc/staff/thealien) has quit (Quit: the_alien)

[00:59:51]

Beirdo:

for windows, yes

[00:59:57]

Beirdo:

no idea for OSX

[01:00:19]

Beirdo:

but if we are also supporting a native mingw build, I'd like to have it in the buildbot too

[01:01:08]

Beirdo:

the cross-compile one (once I figure out the buildbot steps), I don't mind putting on my backend box along with the linux-64bit

[01:01:38]

Beirdo:

but mingw-native and OSX will need slaves that I don't currently have

[01:01:56]

Beirdo:

which is why I'm looking to see if others do :)

[01:02:17]

superm1:

what format are the linux-64bit ones going to spit out in?

[01:02:32]

Beirdo:

they do a test compile, that is all

[01:02:35]

superm1:

ah

[01:03:12]

Beirdo:

and tonight, I will (hopefully) be able to add the plugin compile into there

[01:03:57]

superm1:

are you going to do it via fake make install into a chroot that is axed at the end or somehow setting link paths for plugins to look in the build dir for standard?

[01:04:18]

Beirdo:

thinking of the former at the moment

[01:04:49]

Beirdo:

but I'm not quite sure :)

[01:04:59]

superm1:

ubuntu's sbuild does something really neat with an LVM snapshot process

[01:05:11]

Beirdo:

I'd like to do it with as little customization as possible

[01:05:31]

superm1:

so the LVM snapshot is the starting with same chroot each time and is generated within seconds

[01:05:55]

wylie (wylie!~wylie@ip24-251-22-58.ph.ph.cox.net) has quit (Ping timeout: 240 seconds)

[01:11:24]

kormoc is now known as kormoc_afk

[01:11:56]

Tanthrix (Tanthrix!~tanthrix@c-98-246-133-36.hsd1.or.comcast.net) has joined #mythtv

[01:12:18]

Beirdo:

I don't really need a chroot though, just an install dir

[01:12:40]

Beirdo:

I won't be running anything, just linking against it

[01:12:52]

superm1:

oh right

[01:13:19]

Beirdo:

it's just a matter of making buildslave OK with how I lay it out

[01:13:28]

Beirdo:

so it can auto-clean for each build

[01:26:22]

jya:

i backported all the new audio code to 0.24 so I can easily test it at home (and make the wife an unsuspected tester)... TrueHD passthrough works perfectly in trunk, but on 0.24 I get plenty of Waited 100ms for video buffers AAAAAAAAAAAAAAAAAAAAAAAAAAAAaLLL

[01:27:39]

jya:

the audio code is identical between the two (trunk and the 0.24 backport), I can't explain the difference in behaviour from a sound point of view.. something has changed in the player.

[01:28:45]

jya:

markk_: that would be your area of expertise, would you have any pointers?

[01:37:57]

markk_:

jya: MythPlayer::PreBufferEnoughFrames is waiting on decoded video frames – hence all of the As. presumably if you look at the running threads, the decoder thread is stuck somewhere in AvFormatDecoder::GetFrame

[01:38:53]

jya:

I actually have some doubts it's uniquely related to the video...

[01:39:12]

jya:

for example, f I experience lots of alsa buffer underruns , I hit the issue

[01:40:02]

jya:

like if the alsa buffer size is 64kB, I'll get buffer underruns -> plenty of 100ms wait

[01:40:31]

jya:

same is the audio timestamp calculations are wrong..

[01:41:01]

gigem_ (gigem_!~david@host137.12.intrusion.com) has joined #mythtv

[01:41:18]

jya:

I have the most bizarre thing here. A particular video, plays fine with audio upmixed under all circumstances, except if I output 5.1 LPCM audio, if I re-encode it to AC3 it plays fine

[01:44:50]

gigem (gigem!~david@host137.12.intrusion.com) has quit (Ping timeout: 245 seconds)

[01:53:03]

jya:

I guess I will look into the GetFrame as you suggest

[01:54:07]

kormoc_afk is now known as kormoc

[02:16:50]

elmojo:

jya: can you update and see if you still get the segfault?

[02:17:08]

jya:

elmojo: sorry, I forgot about that one..

[02:17:14]

jya:

let me try

[02:17:22]

elmojo:

I think I may have been passing an uninitialized stream index

[02:18:26]

elmojo:

jya: re: GetFrame has conditions where it will get skipped completely... like if the audio buffer is almost full

[02:19:02]

jya:

what's "almost full" ?

[02:19:08]

jya:

that could well be the problem

[02:19:20]

jya:

cause I'm playing 32 bits audio, 192kHz, 8 channels

[02:19:25]

jya:

they are pretty big frames

[02:20:15]

elmojo:

I think it's 3/4

[02:20:23]

elmojo:

of the buffer size

[02:20:31]

elmojo:

let me check the code and I'll point you to it

[02:21:31]

elmojo:

in avformatdecoder around line 3874 – look at othresh and allowedquit

[02:21:36]

jya:

why not let the audio class manage its buffer entirely ?and only fails when it reports the buffer as being full?

[02:22:16]

jya:

hum.. hit another bug in the alsa driver I think ... have to reboot

[02:22:30]

elmojo:

dunno why it was done that way

[02:23:18]

elmojo:

it can cause deadlocks with hardware decoders that buffer frames which is why the privatedec immediately follows it

[02:24:23]

jya:

could always increase the audio buffer size

[02:24:27]

jya:

but it's 3MB already

[02:24:35]

jya:

it's been increased heaps

[02:24:54]

jya:

it would just delay the problem from occurring not prevent it

[02:25:11]

jya:

but for some reason, getframe is skipped

[02:25:17]

elmojo:

to be honest it's a situation we shouldn't hit anyways

[02:25:27]

jya:

so audio playback stop, because it relies on the timestamp to be updated before it plays

[02:25:44]

jya:

as timestamp aren't updated ; it stops playing

[02:25:49]

jya:

it's a chicken&egg problem

[02:33:42]

jya:

of well, can reproduce the alsa driver bug each time now... rebooting, then I look at that avformatdecoder

[02:34:17]

jya:

elmojo: the weird thing, is that if I run mythavtest in gdb, it doesn't crash

[02:34:47]

jya:

http://mythtv.pastebin.com/sTsZ2jhk

[02:37:04]

cdev_:

stuartm: Thanks for making the change to the #include. I was just surprised that SSDPCache was used outside the lib. I thought I designed that to be an internal class only (Its been a while so I could be mistaken)

[02:38:41]

bgood (bgood!~chatzilla@gate-isdn.admin.theplanet.net) has joined #mythtv

[02:39:04]

bgood (bgood!~chatzilla@gate-isdn.admin.theplanet.net) has left #mythtv ()

[02:41:49]

jya:

elmojo: started playback a few time; no crash..

[02:42:22]

jya:

but it does get stuck a long time there TV: Changing from None to WatchingVideo

[02:43:03]

jya:

about 4s on average

[02:45:26]

pronewbie (pronewbie!~sdfa@203-206-128-240.perm.iinet.net.au) has joined #mythtv

[02:46:46]

jya:

elmojo: you're my hero !

[02:47:25]

jya:

changed line 3878 from othresh = ((ototal>>1) + (ototal>>2))

[02:47:42]

jya:

to othresh = ototal – 100

[02:47:46]

jya:

plays fine now

[02:47:59]

jya:

if you knew how many hours I have spent on that particular video...

[02:48:11]

pronewbie:

Hello everyone, i was hoping someone could assist me in the hardware specs?

[02:48:26]

wagnerrp:

no, but the people in #mythtv-users can

[02:48:27]

jya:

the audio packet decoded at a time are huge, it fills the buffer very easly

[02:48:47]

pronewbie:

thanks wagnerrp

[02:49:02]

jya:

can't believe I've spent so much time on this bug...

[02:49:02]

stuartm (stuartm!~gbee@cpc4-derb9-0-0-cust534.8-3.cable.virginmedia.com) has quit (Read error: Connection reset by peer)

[02:49:06]

pronewbie (pronewbie!~sdfa@203-206-128-240.perm.iinet.net.au) has left #mythtv ()

[02:49:17]

stuartm (stuartm!~gbee@cpc4-derb9-0-0-cust534.8-3.cable.virginmedia.com) has joined #mythtv

[02:50:09]

ybot (ybot!~quassel@61.14.141.36) has quit (Ping timeout: 250 seconds)

[02:51:14]

aputerboy (aputerboy!~chatzilla@pool-96-233-32-137.bstnma.fios.verizon.net) has joined #mythtv

[02:51:38]

jya:

elmojo: so that solution isn't good anyway with checking if the audio buffer is almost full. At it stops getting video frame, the player asks the audio to pause, so the audio buffer never empties , so if this condifition occur, you'll never recover

[02:52:24]

jya:

markk_: I think we should handle the Pause call differently.. or have a different Pause mode, where it will pause and stop feeding the audio card

[02:52:28]

danielk22:

elmojo: jya: The GetFrame() stuff comes out of a demo player. It's not the proper way to handle demuxing with ffmpeg.

[02:52:35]

aputerboy:

How can I find the recording resolution from the recorded table? If it's not there, what is the best way to find out if it was recorded in HD, SD and the resolution

[02:53:13]

elmojo:

jya: yes me and markk_ hit the deadlock you describe with the crystalhd decoder

[02:53:17]

ybot (ybot!~quassel@61.14.141.36) has joined #mythtv

[02:53:49]

jya:

danielk22: that could be the case, but the deadlock occurring is our fault

[02:54:16]

danielk22:

You should have a loop that reads from libav until the audio, video queues reach a sufficient fill level, then block until there is work to do

[02:54:21]

jya:

elmojo: I can add a different pause state, where instead of pause the audio playback (in fact it plays silence) it completely stops

[02:55:06]

jya:

danielk22: yeah, I just re-implemented the iec958 encoder using ffmpeg muxer.. it's quite an easy architecture to use

[02:55:33]

elmojo:

danielk22: didn't you have that on your list at one point?

[02:55:55]

danielk22:

jya: prolly, but our decoder design is whack and countless hours have been spent adding packet queues and such as workarounds.. The code would be much simpler and less likely to contain hidden deadlocks if we fixed this.

[02:55:57]

jya:

simplified the code by several hundred lines of code compare to the previous code where we did all ourselves

[02:56:18]

danielk22:

elmojo: yes..

[02:56:51]

jya:

elmojo: for the time being, there are a few ways I can handle this ; the quick and easy: just increase the size of the audio ringbuffer.

[02:57:01]

jya:

but that will only delay the problem from ever occurring

[02:57:04]

danielk22:

elmojo: right now the backend setup and backend stuff in general has been pushed ahead of that for me.

[02:57:15]

aputerboy:

Also, for some reason Data Direct in my area seems to get a lot of the "Descriptions" and "Subtitles" wrong for shows though it is often right in one of the showings. So, I wrote a perl-based CLI script that allows you to enter a recording and then it looks through the record, recorded, oldrecorded, and recordedprogram tables to find a program match and then lets you choose which one seems...

[02:57:16]

aputerboy:

...most valid to automatically transfer over the 'title' 'subtitle' 'description' fields as appropriate. You can also add your own text — If anyone is interested I am happy to share the code. But it sure as heck beats manually going into mysql and trying to fix bad titles/subtitles/descriptions.

[02:57:33]

jya:

aputerboy: #mythtv-users

[02:57:56]

elmojo:

danielk22: that's fine... it still works remarkably well given it's not the best approach :)

[02:57:57]

aputerboy:

jya – ahh sorry – didn't realize i was in the wron node

[02:57:59]

jya:

even though it's slightly mythtv code related

[02:58:19]

aputerboy:

well my second comment is 'code' that i am happy to share

[02:58:47]

aputerboy:

and the first question about 'recorded' table is not usually the level of users.. since I am writing perl/sql code

[02:59:03]

danielk22:

elmojo: true that :) It really works perfectly well if the audio and video are well muxed, but that is sometimes not the case in real world cases.

[02:59:03]

aputerboy:

but i'm happy to go there if you think it's better

[02:59:36]

elmojo:

aputerboy: many devs hang out in -users

[03:00:04]

aputerboy:

kkkk

[03:01:38]

aputerboy:

Also, for some reason Data Direct in my area seems to get a lot of the "Descriptions" and "Subtitles" wrong for shows though it is often right in one of the showings. So, I wrote a perl-based CLI script that allows you to enter a recording and then it looks through the record, recorded, oldrecorded, and recordedprogram tables to find a program match and then lets you choose which one seems

[03:01:57]

Beirdo:

oO

[03:02:13]

iamlindoro:

Was it anotehr black cat, or was it the SAME black cat?

[03:02:22]

iamlindoro:

This is important

[03:02:44]

jya:

i thought it was black..

[03:04:14]

elmojo:

jya: so did you say if you still get the segfault?

[03:04:23]

jya:

elmojo: I don't

[03:04:32]

jya:

I do get extra warning at the start of playback

[03:04:50]

jya:

like about a hundred of pmt_cb() got a TS packet that doesn't have PMT TID

[03:05:12]

jya:

followed by [NULL @ 0x7fb758057120]start time is not set in av_estimate_timings_from_pts

[03:05:20]

jya:

but no segfault

[03:06:02]

elmojo:

my code that I added should not cause that.... if you want to eliminate my code as suspect then comment out av_update_stream_timings_video in avformatdecoder.cpp

[03:06:17]

jya:

what line ?

[03:06:37]

elmojo:

923

[03:06:37]

ybot (ybot!~quassel@61.14.141.36) has quit (Ping timeout: 260 seconds)

[03:08:27]

jya:

elmojo: what about if I had a quick test, like we've been waiting for video packet for too long, then I just reset the audio buffer entirely ?

[03:09:02]

jya:

problem is in the player, you can't tell why GetFrame stopped, if it was because the audio was too full or something else

[03:09:59]

ybot (ybot!~quassel@61.14.141.36) has joined #mythtv

[03:10:40]

elmojo:

jya: I dunno what is the best method to handle this.... why are we hitting the audio full in the first place... is this a startup problem?

[03:11:18]

jya:

in this particular video I'm playing

[03:11:30]

jya:

the audio frame decoded by ffmpeg are pretty big

[03:11:45]

jya:

I mean it decodes a lot of audio frame at once

[03:11:52]

jya:

they are > 100kb

[03:12:10]

jya:

for DTS-HD MA passthrough, an IEC958 packet is 61kB

[03:12:26]

jya:

when the playback starts and it buffers

[03:12:56]

jya:

the audio plays at 192kHz

[03:13:09]

jya:

so 61kB doesn't last long..

[03:13:18]

elmojo:

I think Mark Speith has a ticket that addresses the startup delay in a better manner

[03:13:38]

elmojo:

it waits till the audio and video timestamps are synced before playing

[03:14:00]

jya:

as you have the "Changing from None -> VideoPlayback" for quite a while

[03:14:06]

jya:

the audio buffer is full by then

[03:15:04]

kenni:

Beirdo: what is the capacity requirements of this buildslave? And what is it for, just for testing if the current trunk is buildable, or? I have a Core i5–670 with 12GB RAM running various 95% idle KVM guests 24/7. If 1 core @ 3.46GHz and 2GB RAM is sufficient, then I can probably help out with a dedicated VM.

[03:15:08]

ybot (ybot!~quassel@61.14.141.36) has quit (Ping timeout: 240 seconds)

[03:15:14]

elmojo:

jya: that patch I think will fix this problem

[03:15:17]

jya:

I will see what are the side effect of instead calling Pause (which shouldn't be called under those circumstances)

[03:15:30]

Beirdo:

well, all it does is compiles on demand

[03:15:33]

elmojo:

jya: baby crying... I'll be back in a few

[03:15:40]

jya:

calling the Reset, which wil lempty the audio buffer

[03:16:01]

Beirdo:

right now, I have my backend acting as the 64bit linux slave, but I believe we can have more than one of a type

[03:16:07]

jya:

or add a different pause state where we stop feeding the audio silence

[03:16:34]

Beirdo:

all it's doing at this time is git pull, ./configure, make

[03:16:48]

kenni:

Beirdo: ok, how often?

[03:16:57]

Beirdo:

once I get home from work, I intend to add in the plugins building

[03:17:03]

Beirdo:

after every commit

[03:17:27]

Beirdo:

it gives a 2min (currently) quiet period to allow more commits to come in, then compiles

[03:18:04]

Beirdo:

on my backend box (i7–860), the compiles are taking ~2 min each at this time

[03:18:09]

Beirdo:

as I use ccache :)

[03:18:42]

Beirdo:

I'll have a 32-bit linux one up soon too

[03:19:19]

kenni:

ok...hmm, perhaps I can figure out something else, I have access to some (mostly idle) core i7 systems in a hosting center.

[03:20:07]

Beirdo:

I'll double-check what it does when it has more than one of the same slave type reporting in.

[03:20:49]

Beirdo:

I also (at some point) would like to have it doing doxygen runs, and the cppcheck, etc runs too

[03:22:50]

kenni:

do you need direct access to the server or will it just pull jobs from your server?

[03:23:36]

Beirdo:

it connects to the new mythtv server over a tcp socket

[03:23:44]

kenni:

perfect

[03:23:45]

Beirdo:

and the server tells it what to do

[03:24:17]

Beirdo:

Oooh, we could run automatically the translation stuff if it makes sense too

[03:24:23]

ybot (ybot!~quassel@61.14.141.36) has joined #mythtv

[03:26:24]

kenni:

I don't think we would benefit much from it...but on a related note, the binary translation files should really be generated on build time and should not be included in git/svn

[03:26:57]

kenni:

but we have discussed that before, we just need someone to do the scripts

[03:26:57]

knightr:

kenni, they would take time to generate, are you sure?

[03:27:09]

knightr:

kenni, not with me though...

[03:30:15]

ybot (ybot!~quassel@61.14.141.36) has quit (Ping timeout: 245 seconds)

[03:31:04]

kenni:

we really shouldn't keep those binary files in git, but let's take that discussion some other day. AFAIR, several of the devs looked into it generating them at compile time, but I think some QT cross-platform build issues, delayed a solution

[03:31:35]

kenni:

something about some tools were missing in QT on Windows or similar

[03:32:18]

knightr:

interesting, I did not find any info on previous attempts on this...

[03:33:07]

xris:

how long is "take time"?

[03:33:08]

danielk22:

knightr: It was brought up before. I thought it was OSX, not Windows which was the problem though.

[03:33:14]

iamlindoro:

Was never a valid argument, though, as the last time it was made I went and verified that all necessary translation tools are built on both windows and OS X

[03:33:40]

knightr:

ok, interesting...

[03:34:11]

iamlindoro:

xris, not long at all

[03:34:19]

iamlindoro:

think seconds, not minutes

[03:34:20]

knightr:

I guess another solution could be to download them from somewhere already generated but without the overhead of being versions in git/svn/whatever...

[03:34:57]

kenni:

but why? How long time is it? 2min on a average system?

[03:35:26]

iamlindoro:

I would be shocked if building the binary translation files took even a single minute

[03:35:32]

xris:

if it's insignificant compared to myth build time (15–20 minutes on a below average system), it's not really a big deal

[03:35:34]

iamlindoro:

I've done it a number of times, it's never taken remotely near that

[03:35:58]

kenni:

I do it every time before I commit a translation update

[03:36:01]

xris:

storing in the cloud becomes a pain for people using the unstable version, too

[03:36:07]

knightr:

hmm, I guess your right I was thinking more about lupdate, sorry guys, was sick for the last couple of days and still not working on all trusters...

[03:36:57]

iamlindoro:

Last time it was discussed, before kenni and knightr came onboard, reynaldo was going to work on making the modification to the build... just think he never got around to it

[03:37:14]

iamlindoro:

but FWIW lrelease and lupdate are both available in the Qt produced/downloaded on both windows and OS X

[03:38:01]

knightr:

it's sometimes called lupdate-qt4 on some platforms though I think...

[03:38:14]

iamlindoro:

it's just plain lupdate and lrelease in windows/OS X

[03:38:20]

knightr:

s/lupdate-qt4/lrelease-qt4

[03:38:30]

iamlindoro:

it's the same as qmake

[03:38:44]

iamlindoro:

lupdate is a link to the version of Qt you are using

[03:38:49]

knightr:

iamlindoro, yep, you're right...

[03:38:59]

knightr:

iamlindoro, not on my distro...

[03:39:12]

xris:

hmm.. just occurred to me that our auto documentation generators for the website are not pulling from git

[03:39:15]

xris:

I should fix that

[03:39:17]

iamlindoro:

knightr, what distro is that?

[03:39:29]

iamlindoro:

(as your distro is doing it wrong)

[03:39:48]

knightr:

Fedora 14... I only have it under thje name lupdate-qt4 and lrelease-qt4...

[03:39:52]

kenni:

anyway, I've gotta go. Beirdo, if the load isn't too extensive (fx. 10min per commit with some ccache optimizations or similar), then count me in with a dedicated Windows VM.

[03:40:35]

Beirdo:

Cool. I expect that it will come out to that. We can try it, and if it is too much, we can drop it easily enough

[03:41:05]

** xris hopes to finally fix the build scripts for fedora tonight. **

[03:41:12]

xris:

and maybe then I can upgrade to f14

[03:41:17]

knightr:

iamlindoro, not very difficult to make a symbolic link but it's not there by default so if we look for lrelease and not lrelease-qt4 it wouldn't work on F14...

[03:41:42]

Beirdo:

OK, going home

[03:41:44]

kenni:

Beirdo: good, does the Windows version matter or will a plain old XP do?

[03:41:55]

Beirdo:

I think XP will be fine

[03:42:02]

ybot (ybot!~quassel@61.14.141.36) has joined #mythtv

[03:42:19]

Beirdo:

iamlindoro would likely know better on that part as he's messed with the Windows builds more than I have

[03:42:20]

stuartm (stuartm!~gbee@cpc4-derb9-0-0-cust534.8-3.cable.virginmedia.com) has quit (Ping timeout: 245 seconds)

[03:42:29]

Beirdo:

anyways, I gotta head out.

[03:42:31]

iamlindoro:

yeah, XP's fine

[03:42:41]

Beirdo:

Be back in a while when I'm home

[03:44:09]

stuartm (stuartm!~gbee@cpc4-derb9-0-0-cust534.8-3.cable.virginmedia.com) has joined #mythtv

[03:46:42]

jya:

elmojo , markk : http://mythtv.pastebin.com/N02R5xKy

[03:46:55]

jya:

I've been able to play all the files giving me issues before

[03:47:15]

jya:

basically it's very simple: just issue a reset the first time we send pause

[03:48:01]

knightr:

is the test for the ssh key on github (using the command line ssh) supposed to work, it complains about my key every time I tried...

[03:49:28]

ybot (ybot!~quassel@61.14.141.36) has quit (Ping timeout: 276 seconds)

[03:50:43]

ybot (ybot!~quassel@61.14.141.36) has joined #mythtv

[03:57:40]

Snow-Man (Snow-Man!~sfrost@tamriel.snowman.net) has joined #mythtv

[03:58:27]

elmojo:

jya: would it be better to only do the reset if the audio is partially full?

[03:58:27]

stuartm (stuartm!~gbee@cpc4-derb9-0-0-cust534.8-3.cable.virginmedia.com) has quit (Read error: Connection reset by peer)

[03:58:32]

stuartm (stuartm!~gbee@cpc4-derb9-0-0-cust534.8-3.cable.virginmedia.com) has joined #mythtv

[03:59:09]

jya:

elmojo: I don't know in mythplayer if that's why GetFrame got interrupted

[03:59:27]

jya:

unless I redo the same kind of test (checking at 2/3)

[03:59:58]

jya:

only AVFD knows why the decoding got interrupted

[04:01:18]

jya:

I guess I could make the same kind of calculation as done in AVFD

[04:02:18]

elmojo:

jya: just suggesting so that it at least confines the work around to only be applied to a known deadlock condition

[04:02:35]

jya:

yeah.. I can add a method to the audioplayer class

[04:02:48]

jya:

returning true of buffer is almost full

[04:03:05]

elmojo:

that would be a nice change... it will simplify avformatdecoder as well

[04:03:08]

jya:

and use that method in avfd as well as mythplayer

[04:04:03]

jya:

I think the underlying problem, is that somehow calling the Pause method of the AudioOutput class will make the audio buffer smaller

[04:04:05]

jya:

it doesn't

[04:04:25]

jya:

it only make AudioOutput plays silence instead of playing the buffer

[04:09:44]

elmojo:

jya: I'm ok with that work around but a comment or two in the code about why it's needed is always appreciated

[04:09:56]

jya:

i am redoing it now

[04:11:37]

ybot (ybot!~quassel@61.14.141.36) has quit (Ping timeout: 250 seconds)

[04:14:25]

jya:

there is code in there, I have no idea what for

[04:14:45]

jya:

not only it checks that the audio buffer is 2/3 full

[04:15:21]

jya:

but it checks that size_audio_to_be_added * 6 won't fill the audio buffer more than 2/3rd

[04:16:00]

jya:

that gives us only a leeway of 3 audio packets from the stream

[04:19:49]

Tanthrix (Tanthrix!~tanthrix@c-98-246-133-36.hsd1.or.comcast.net) has quit ()

[04:22:20]

knightr:

iamlindoro, most qmake files will be converted to real makefiles in the (not so distant future?) right?

[04:22:33]

iamlindoro:

knightr, not planned as far as Iknow

[04:23:12]

jya:

elmogo: http://mythtv.pastebin.com/j4WCtqLd

[04:23:36]

jya:

now check if the buffer is 2/3rd full, if it is, and waited more than 100ms, then we reset it

[04:24:11]

jya:

removed that check to see if the buffer could contain 6 times ; it was a very wild assumption

[04:24:26]

knightr:

iamlindoro, ok so I guess it partially wishful thinking on the part of the person who had mentionned this... If so I'll try to see if I can dig up the info to automatize the running of lrelease...

[04:24:40]

iamlindoro:

knightr, The two that have been transitioned were because their individual maintainers chose to do so-- but no move to remove them from the project as a whole

[04:24:48]

jya:

and IMHO, an external class should never have guesses on what another class is doing internally

[04:25:09]

knightr:

(I'm not very familiar with qmake but I can probably find the info somewhere...)

[04:25:21]

knightr:

iamlindoro, ah, thanks...

[04:25:25]

iamlindoro:

knightr, running a command after a build is trivial

[04:25:45]

jya:

elmogo: seems to work for all my troubling videos now

[04:25:52]

knightr:

iamlindoro, that's what I figured, I just have to dig up the info...

[04:26:06]

knightr:

thanks!

[04:28:26]

jya:

elmojo: now I see "AudioPlayer: AddAudioData():p1: Audio buffer overflow, audio data lost!" which was hidden before.. that's the true cause of all those deadlocks..

[04:29:21]

jya:

wonder if I should double the audio ringbuffer from 3MB to 6MB.

[04:29:51]

jya:

with the extent of big audio frames like the one seen with hdmi HD audio

[04:30:02]

jya:

it's a problem that will occur more and more

[04:30:42]

ybot (ybot!~quassel@61.14.141.36) has joined #mythtv

[04:30:48]

jya:

when with a bigger buffer, the more nasty a reset will sound

[04:34:04]

elmojo:

jya: it seems natural that we will need the bigger buffer for larger frames

[04:34:26]

elmojo:

in your latest patch why did you delete all that code related to audio only mode?

[04:34:45]

jya:

because it will never be applicable.

[04:34:57]

jya:

in only override the detection found earlier

[04:35:03]

jya:

that the buffer is almost full

[04:35:05]

elmojo:

ok, I don't know that code at all

[04:35:18]

jya:

with an even stricter test

[04:35:37]

jya:

it came with Mark Spieth audiostamp accuracy patch

[04:35:57]

jya:

i committed it.. should have looked closer.

[04:36:22]

jya:

under most circumstances, we wouldn't get there, and with the HD audio, we get there too quickly

[04:37:17]

jya:

but all of this wouldn't be required if instead of doing all the getframe ourselves, with used ffmpeg muxer

[04:37:18]

elmojo:

ah, so it's fairly new addition

[04:37:40]

jya:

it is, it's post HDaudio branch merge

[04:38:10]

jya:

in any case, i think all it does is trying to recover from an unrecoverable situation

[04:38:24]

jya:

we are going to loose data because of buffer overflow

[04:38:45]

jya:

not much we can do there, because we only know that info when the data has already been decoded.

[04:39:06]

jya:

what are we going to do with it? place it in a 2nd buffer to buffer access to the buffer ? :P

[04:40:04]

jya:

plus i think most of the issues would only occur when playback starts

[04:40:13]

eharris_ (eharris_!~eharris@99-179-7-82.lightspeed.austtx.sbcglobal.net) has joined #mythtv

[04:40:24]

elmojo:

and then you got to sync it to the video... it's a balancing act

[04:40:27]

jya:

i'd be surprised if it happened once everything has started, and all threads are going smoothly.

[04:41:03]

jya:

exactly.. to be honnest, I've always been amazed that our audio is mostly in sync with the video :)

[04:41:05]

boshhead_ (boshhead_!~boshhead@adsl-71-138-128-90.dsl.pltn13.pacbell.net) has joined #mythtv

[04:41:11]

jya:

it's vooodoo stuff

[04:41:18]

boshhead (boshhead!~boshhead@adsl-71-138-128-90.dsl.pltn13.pacbell.net) has quit (Disconnected by services)

[04:41:21]

boshhead_ is now known as boshhead

[04:42:36]

elmojo:

the audio timestamp code is a bit scary

[04:43:44]

rhpot199` (rhpot199`!~rhpot1991@204.8.45.13) has joined #mythtv

[04:52:00]

ybot (ybot!~quassel@61.14.141.36) has quit (*.net *.split)

[04:52:00]

sutula (sutula!sutula@nat/hp/x-zshqvxkdufneuhwb) has quit (*.net *.split)

[04:52:00]

jya (jya!~avenardj@60-242-40-141.static.tpgi.com.au) has quit (*.net *.split)

[04:52:01]

Kevin` (Kevin`!~kevin@router.kwzs.be) has quit (*.net *.split)

[04:52:02]

okolsi (okolsi!~mythtv@a88-112-74-247.elisa-laajakaista.fi) has quit (*.net *.split)

[04:52:03]

high-rez (high-rez!~gus@2001:470:e935::2) has quit (*.net *.split)

[04:52:03]

tgm4883 (tgm4883!~tgm4883@ubuntu/member/tgm4883) has quit (*.net *.split)

[04:52:03]

jwhite (jwhite!~jwhite@75-146-153-89-minnesota.hfc.comcastbusiness.net) has quit (*.net *.split)

[04:52:03]

dagar (dagar!~dagar@agar.ca) has quit (*.net *.split)

[04:52:03]

XChatMav (XChatMav!~MaverickT@111.86.233.220.static.exetel.com.au) has quit (*.net *.split)

[04:52:03]

gregL (gregL!~greg@cpe-74-76-125-87.nycap.res.rr.com) has quit (*.net *.split)

[04:52:03]

kurre_ (kurre_!~tomimo@xdsl-83-150-88-111.nebulazone.fi) has quit (*.net *.split)

[04:52:03]

Mudboy (Mudboy!~zod@cpe-75-83-190-215.socal.res.rr.com) has quit (*.net *.split)

[04:52:03]

pheld (pheld!~heldal@cl-5.osl-01.no.sixxs.net) has quit (*.net *.split)

[04:52:06]

knightr (knightr!~knightr@69-165-170-178.dsl.teksavvy.com) has quit (*.net *.split)

[04:52:06]

mrand (mrand!~mrand@ubuntu/member/mrand) has quit (*.net *.split)

[04:52:06]

foxbuntu (foxbuntu!~foxbuntu@ubuntu/member/foxbuntu) has quit (*.net *.split)

[04:52:06]

ozatomic (ozatomic!~ozatomic@bb.rapidspeeds.com) has quit (*.net *.split)

[04:52:06]

Computer_Czar (Computer_Czar!~dcooper@69.4.155.83) has quit (*.net *.split)

[04:52:06]

caelor (caelor!~caelor@93-97-184-87.zone5.bethere.co.uk) has quit (*.net *.split)

[04:52:06]

dlblog (dlblog!~dlblog@c-76-127-227-175.hsd1.ma.comcast.net) has quit (*.net *.split)

[04:52:06]

skd5aner (skd5aner!~skd5aner@cpe-069-132-082-180.carolina.res.rr.com) has quit (*.net *.split)

[04:52:06]

castlec (castlec!~christoph@cpe-76-181-105-97.insight.res.rr.com) has quit (*.net *.split)

[04:52:06]

kenni (kenni!~kenni@pfsense.dhcp.pop.k-net.dk) has quit (*.net *.split)

[04:52:06]

simonckenyon (simonckenyon!~simoncken@195.7.61.12) has quit (*.net *.split)

[04:52:06]

tomimo (tomimo!~kurre@xdsl-83-150-88-111.nebulazone.fi) has quit (*.net *.split)

[04:52:06]

xris (xris!~xris@xris.forevermore.net) has quit (*.net *.split)

[04:52:07]

stuarta (stuarta!~stuarta@unaffiliated/stuarta) has quit (*.net *.split)

[04:52:07]

poptix (poptix!poptix@poptix.net) has quit (*.net *.split)

[04:52:07]

dekarl (dekarl!~deKarl@e180155217.adsl.alicedsl.de) has quit (*.net *.split)

[04:52:08]

eharris (eharris!~eharris@99-179-7-82.lightspeed.austtx.sbcglobal.net) has quit (*.net *.split)

[04:52:10]

rhpot1991 (rhpot1991!~rhpot1991@ubuntu/member/rhpot1991) has quit (*.net *.split)

[04:52:11]

gigem_ (gigem_!~david@host137.12.intrusion.com) has quit (*.net *.split)

[04:52:11]

gilesw (gilesw!~chatzilla@91.208.163.140) has quit (*.net *.split)

[04:52:11]

Dave123 (Dave123!~dave@cpe-74-74-222-96.rochester.res.rr.com) has quit (*.net *.split)

[04:52:11]

ikonia (ikonia!~mattd@unaffiliated/ikonia) has quit (*.net *.split)

[04:52:11]

jstenback (jstenback!~jstenback@64.151.104.172) has quit (*.net *.split)

[04:52:11]

Snow-Man (Snow-Man!~sfrost@tamriel.snowman.net) has quit (*.net *.split)

[04:52:11]

ThisNewGuy (ThisNewGuy!~doug@pool-96-242-160-27.nwrknj.fios.verizon.net) has quit (*.net *.split)

[04:52:11]

kormoc (kormoc!~kormoc@unaffiliated/kormoc) has quit (*.net *.split)

[04:52:11]

davide (davide!~gigem@cpe-76-187-29-95.tx.res.rr.com) has quit (*.net *.split)

[04:52:11]

jpabq- (jpabq-!~jpabq@174-28-182-216.albq.qwest.net) has quit (*.net *.split)

[04:52:11]

jpabq (jpabq!~jpabq@174-28-182-216.albq.qwest.net) has quit (*.net *.split)

[04:52:11]

_charly_ (_charly_!kroseneg@sunrise.schmidham.net) has quit (*.net *.split)

[04:52:11]

jams (jams!~jams@cpe-184-58-217-97.wi.res.rr.com) has quit (*.net *.split)

[04:52:11]

elvum_ (elvum_!~elvum@zenit.dh.bytemark.co.uk) has quit (*.net *.split)

[04:52:11]

Cougar (Cougar!~cougar@lost.data.ee) has quit (*.net *.split)

[04:52:11]

mag0o (mag0o!20001@slackhost.lynchmv.com) has quit (*.net *.split)

[04:53:37]

poptix (poptix!poptix@poptix.net) has joined #mythtv

[04:53:37]

pheld (pheld!~heldal@cl-5.osl-01.no.sixxs.net) has joined #mythtv

[04:53:37]

kurre_ (kurre_!~tomimo@xdsl-83-150-88-111.nebulazone.fi) has joined #mythtv

[04:53:37]

gregL (gregL!~greg@cpe-74-76-125-87.nycap.res.rr.com) has joined #mythtv

[04:53:37]

XChatMav (XChatMav!~MaverickT@111.86.233.220.static.exetel.com.au) has joined #mythtv

[04:53:37]

dagar (dagar!~dagar@agar.ca) has joined #mythtv

[04:53:37]

jwhite (jwhite!~jwhite@75-146-153-89-minnesota.hfc.comcastbusiness.net) has joined #mythtv

[04:53:37]

Mudboy (Mudboy!~zod@cpe-75-83-190-215.socal.res.rr.com) has joined #mythtv

[04:54:35]

dekarl (dekarl!~deKarl@e180155217.adsl.alicedsl.de) has joined #mythtv

[04:54:35]

gigem_ (gigem_!~david@host137.12.intrusion.com) has joined #mythtv

[04:54:35]

gilesw (gilesw!~chatzilla@91.208.163.140) has joined #mythtv

[04:54:35]

Dave123 (Dave123!~dave@cpe-74-74-222-96.rochester.res.rr.com) has joined #mythtv

[04:54:35]

ikonia (ikonia!~mattd@unaffiliated/ikonia) has joined #mythtv

[04:54:35]

jstenback (jstenback!~jstenback@64.151.104.172) has joined #mythtv

[04:54:47]

ybot (ybot!~quassel@61.14.141.36) has joined #mythtv

[04:54:47]

sutula (sutula!sutula@nat/hp/x-zshqvxkdufneuhwb) has joined #mythtv

[04:54:47]

jya (jya!~avenardj@60-242-40-141.static.tpgi.com.au) has joined #mythtv

[04:54:47]

Kevin` (Kevin`!~kevin@router.kwzs.be) has joined #mythtv

[04:55:09]

castlec (castlec!~christoph@cpe-76-181-105-97.insight.res.rr.com) has joined #mythtv

[04:55:10]

kenni (kenni!~kenni@pfsense.dhcp.pop.k-net.dk) has joined #mythtv

[04:55:10]

simonckenyon (simonckenyon!~simoncken@195.7.61.12) has joined #mythtv

[04:55:10]

tomimo (tomimo!~kurre@xdsl-83-150-88-111.nebulazone.fi) has joined #mythtv

[04:55:10]

xris (xris!~xris@xris.forevermore.net) has joined #mythtv

[04:55:10]

stuarta (stuarta!~stuarta@unaffiliated/stuarta) has joined #mythtv

[04:55:22]

knightr (knightr!~knightr@69-165-170-178.dsl.teksavvy.com) has joined #mythtv

[04:55:22]

mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv

[04:55:22]

foxbuntu (foxbuntu!~foxbuntu@ubuntu/member/foxbuntu) has joined #mythtv

[04:55:22]

ozatomic (ozatomic!~ozatomic@bb.rapidspeeds.com) has joined #mythtv

[04:55:22]

Computer_Czar (Computer_Czar!~dcooper@69.4.155.83) has joined #mythtv

[04:55:22]

caelor (caelor!~caelor@93-97-184-87.zone5.bethere.co.uk) has joined #mythtv

[04:55:22]

dlblog (dlblog!~dlblog@c-76-127-227-175.hsd1.ma.comcast.net) has joined #mythtv

[04:55:22]

skd5aner (skd5aner!~skd5aner@cpe-069-132-082-180.carolina.res.rr.com) has joined #mythtv

[04:55:47]

Snow-Man (Snow-Man!~sfrost@tamriel.snowman.net) has joined #mythtv

[04:55:47]

ThisNewGuy (ThisNewGuy!~doug@pool-96-242-160-27.nwrknj.fios.verizon.net) has joined #mythtv

[04:55:47]

kormoc (kormoc!~kormoc@unaffiliated/kormoc) has joined #mythtv

[04:55:47]

davide (davide!~gigem@cpe-76-187-29-95.tx.res.rr.com) has joined #mythtv

[04:55:47]

jpabq- (jpabq-!~jpabq@174-28-182-216.albq.qwest.net) has joined #mythtv

[04:55:47]

jpabq (jpabq!~jpabq@174-28-182-216.albq.qwest.net) has joined #mythtv

[04:55:47]

_charly_ (_charly_!kroseneg@sunrise.schmidham.net) has joined #mythtv

[04:55:47]

jams (jams!~jams@cpe-184-58-217-97.wi.res.rr.com) has joined #mythtv

[04:55:47]

elvum_ (elvum_!~elvum@zenit.dh.bytemark.co.uk) has joined #mythtv

[04:55:47]

Cougar (Cougar!~cougar@lost.data.ee) has joined #mythtv

[04:55:47]

mag0o (mag0o!20001@slackhost.lynchmv.com) has joined #mythtv

[04:56:13]

okolsi (okolsi!~mythtv@a88-112-74-247.elisa-laajakaista.fi) has joined #mythtv

[04:56:13]

high-rez (high-rez!~gus@2001:470:e935::2) has joined #mythtv

[04:56:13]

tgm4883 (tgm4883!~tgm4883@ubuntu/member/tgm4883) has joined #mythtv

[04:56:19]

ikonia (ikonia!~mattd@unaffiliated/ikonia) has quit (Max SendQ exceeded)

[04:56:37]

Snow-Man (Snow-Man!~sfrost@tamriel.snowman.net) has quit (Remote host closed the connection)

[04:56:38]

ikonia (ikonia!~mattd@unaffiliated/ikonia) has joined #mythtv

[04:57:07]

Anduin_ (Anduin_!~awithers@pdpc/supporter/professional/anduin) has joined #mythtv

[04:57:21]

okolsi (okolsi!~mythtv@a88-112-74-247.elisa-laajakaista.fi) has quit (*.net *.split)

[04:57:22]

high-rez (high-rez!~gus@2001:470:e935::2) has quit (*.net *.split)

[04:57:22]

tgm4883 (tgm4883!~tgm4883@ubuntu/member/tgm4883) has quit (*.net *.split)

[04:57:22]

ThisNewGuy (ThisNewGuy!~doug@pool-96-242-160-27.nwrknj.fios.verizon.net) has quit (*.net *.split)

[04:57:22]

kormoc (kormoc!~kormoc@unaffiliated/kormoc) has quit (*.net *.split)

[04:57:22]

davide (davide!~gigem@cpe-76-187-29-95.tx.res.rr.com) has quit (*.net *.split)

[04:57:22]

jpabq- (jpabq-!~jpabq@174-28-182-216.albq.qwest.net) has quit (*.net *.split)

[04:57:23]

jpabq (jpabq!~jpabq@174-28-182-216.albq.qwest.net) has quit (*.net *.split)

[04:57:23]

_charly_ (_charly_!kroseneg@sunrise.schmidham.net) has quit (*.net *.split)

[04:57:23]

jams (jams!~jams@cpe-184-58-217-97.wi.res.rr.com) has quit (*.net *.split)

[04:57:23]

elvum_ (elvum_!~elvum@zenit.dh.bytemark.co.uk) has quit (*.net *.split)

[04:57:24]

Cougar (Cougar!~cougar@lost.data.ee) has quit (*.net *.split)

[04:57:24]

mag0o (mag0o!20001@slackhost.lynchmv.com) has quit (*.net *.split)

[04:57:24]

knightr (knightr!~knightr@69-165-170-178.dsl.teksavvy.com) has quit (*.net *.split)

[04:57:24]

mrand (mrand!~mrand@ubuntu/member/mrand) has quit (*.net *.split)

[04:57:24]

foxbuntu (foxbuntu!~foxbuntu@ubuntu/member/foxbuntu) has quit (*.net *.split)

[04:57:24]

ozatomic (ozatomic!~ozatomic@bb.rapidspeeds.com) has quit (*.net *.split)

[04:57:24]

Computer_Czar (Computer_Czar!~dcooper@69.4.155.83) has quit (*.net *.split)

[04:57:24]

caelor (caelor!~caelor@93-97-184-87.zone5.bethere.co.uk) has quit (*.net *.split)

[04:57:25]

dlblog (dlblog!~dlblog@c-76-127-227-175.hsd1.ma.comcast.net) has quit (*.net *.split)

[04:57:25]

skd5aner (skd5aner!~skd5aner@cpe-069-132-082-180.carolina.res.rr.com) has quit (*.net *.split)

[04:57:25]

castlec (castlec!~christoph@cpe-76-181-105-97.insight.res.rr.com) has quit (*.net *.split)

[04:57:25]

kenni (kenni!~kenni@pfsense.dhcp.pop.k-net.dk) has quit (*.net *.split)

[04:57:26]

simonckenyon (simonckenyon!~simoncken@195.7.61.12) has quit (*.net *.split)

[04:57:27]

tomimo (tomimo!~kurre@xdsl-83-150-88-111.nebulazone.fi) has quit (*.net *.split)

[04:57:27]

xris (xris!~xris@xris.forevermore.net) has quit (*.net *.split)

[04:57:27]

stuarta (stuarta!~stuarta@unaffiliated/stuarta) has quit (*.net *.split)

[04:57:27]

ybot (ybot!~quassel@61.14.141.36) has quit (*.net *.split)

[04:57:27]

sutula (sutula!sutula@nat/hp/x-zshqvxkdufneuhwb) has quit (*.net *.split)

[04:57:27]

jya (jya!~avenardj@60-242-40-141.static.tpgi.com.au) has quit (*.net *.split)

[04:57:27]

Kevin` (Kevin`!~kevin@router.kwzs.be) has quit (*.net *.split)

[04:57:27]

dekarl (dekarl!~deKarl@e180155217.adsl.alicedsl.de) has quit (*.net *.split)

[04:57:27]

ikonia (ikonia!~mattd@unaffiliated/ikonia) has quit (*.net *.split)

[04:57:27]

gigem_ (gigem_!~david@host137.12.intrusion.com) has quit (*.net *.split)

[04:57:27]

gilesw (gilesw!~chatzilla@91.208.163.140) has quit (*.net *.split)

[04:57:28]

Dave123 (Dave123!~dave@cpe-74-74-222-96.rochester.res.rr.com) has quit (*.net *.split)

[04:57:28]

jstenback (jstenback!~jstenback@64.151.104.172) has quit (*.net *.split)

[04:57:30]

poptix (poptix!poptix@poptix.net) has quit (*.net *.split)

[04:57:30]

jwhite (jwhite!~jwhite@75-146-153-89-minnesota.hfc.comcastbusiness.net) has quit (*.net *.split)

[04:57:30]

dagar (dagar!~dagar@agar.ca) has quit (*.net *.split)

[04:57:30]

XChatMav (XChatMav!~MaverickT@111.86.233.220.static.exetel.com.au) has quit (*.net *.split)

[04:57:30]

gregL (gregL!~greg@cpe-74-76-125-87.nycap.res.rr.com) has quit (*.net *.split)

[04:57:30]

kurre_ (kurre_!~tomimo@xdsl-83-150-88-111.nebulazone.fi) has quit (*.net *.split)

[04:57:30]

Mudboy (Mudboy!~zod@cpe-75-83-190-215.socal.res.rr.com) has quit (*.net *.split)

[04:57:30]

pheld (pheld!~heldal@cl-5.osl-01.no.sixxs.net) has quit (*.net *.split)

[04:57:37]

anykey__ (anykey__!~guedel@46-126-244-251.dclient.hispeed.ch) has joined #mythtv

[04:57:56]

anykey_ (anykey_!~guedel@46-126-244-251.dclient.hispeed.ch) has quit (Write error: Broken pipe)

[04:57:57]

Anduin (Anduin!~awithers@pdpc/supporter/professional/anduin) has quit (Write error: Broken pipe)

[04:57:58]

paul-h (paul-h!~paulh@5ad84320.bb.sky.com) has quit (Write error: Broken pipe)

[04:58:31]

ikonia (ikonia!~mattd@unaffiliated/ikonia) has joined #mythtv

[04:58:31]

tgm4883 (tgm4883!~tgm4883@ubuntu/member/tgm4883) has joined #mythtv

[04:58:31]

high-rez (high-rez!~gus@2001:470:e935::2) has joined #mythtv

[04:58:31]

okolsi (okolsi!~mythtv@a88-112-74-247.elisa-laajakaista.fi) has joined #mythtv

[04:58:31]

mag0o (mag0o!20001@slackhost.lynchmv.com) has joined #mythtv

[04:58:31]

Cougar (Cougar!~cougar@lost.data.ee) has joined #mythtv

[04:58:31]

elvum_ (elvum_!~elvum@zenit.dh.bytemark.co.uk) has joined #mythtv

[04:58:31]

jams (jams!~jams@cpe-184-58-217-97.wi.res.rr.com) has joined #mythtv

[04:58:32]

_charly_ (_charly_!kroseneg@sunrise.schmidham.net) has joined #mythtv

[04:58:32]

jpabq (jpabq!~jpabq@174-28-182-216.albq.qwest.net) has joined #mythtv

[04:58:32]

jpabq- (jpabq-!~jpabq@174-28-182-216.albq.qwest.net) has joined #mythtv

[04:58:32]

davide (davide!~gigem@cpe-76-187-29-95.tx.res.rr.com) has joined #mythtv

[04:58:32]

kormoc (kormoc!~kormoc@unaffiliated/kormoc) has joined #mythtv

[04:58:32]

ThisNewGuy (ThisNewGuy!~doug@pool-96-242-160-27.nwrknj.fios.verizon.net) has joined #mythtv

[04:58:32]

skd5aner (skd5aner!~skd5aner@cpe-069-132-082-180.carolina.res.rr.com) has joined #mythtv

[04:58:32]

dlblog (dlblog!~dlblog@c-76-127-227-175.hsd1.ma.comcast.net) has joined #mythtv

[04:58:32]

caelor (caelor!~caelor@93-97-184-87.zone5.bethere.co.uk) has joined #mythtv

[04:58:32]

Computer_Czar (Computer_Czar!~dcooper@69.4.155.83) has joined #mythtv

[04:58:32]

ozatomic (ozatomic!~ozatomic@bb.rapidspeeds.com) has joined #mythtv

[04:58:32]

foxbuntu (foxbuntu!~foxbuntu@ubuntu/member/foxbuntu) has joined #mythtv

[04:58:32]

mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv

[04:58:32]

knightr (knightr!~knightr@69-165-170-178.dsl.teksavvy.com) has joined #mythtv

[04:58:32]

stuarta (stuarta!~stuarta@unaffiliated/stuarta) has joined #mythtv

[04:58:32]

xris (xris!~xris@xris.forevermore.net) has joined #mythtv

[04:58:32]

tomimo (tomimo!~kurre@xdsl-83-150-88-111.nebulazone.fi) has joined #mythtv

[04:58:32]

simonckenyon (simonckenyon!~simoncken@195.7.61.12) has joined #mythtv

[04:58:32]

kenni (kenni!~kenni@pfsense.dhcp.pop.k-net.dk) has joined #mythtv

[04:58:32]

castlec (castlec!~christoph@cpe-76-181-105-97.insight.res.rr.com) has joined #mythtv

[04:58:32]

Kevin` (Kevin`!~kevin@router.kwzs.be) has joined #mythtv

[04:58:32]

jya (jya!~avenardj@60-242-40-141.static.tpgi.com.au) has joined #mythtv

[04:58:32]

sutula (sutula!sutula@nat/hp/x-zshqvxkdufneuhwb) has joined #mythtv

[04:58:32]

ybot (ybot!~quassel@61.14.141.36) has joined #mythtv

[04:58:32]

jstenback (jstenback!~jstenback@64.151.104.172) has joined #mythtv

[04:58:32]

Dave123 (Dave123!~dave@cpe-74-74-222-96.rochester.res.rr.com) has joined #mythtv

[04:58:32]

gilesw (gilesw!~chatzilla@91.208.163.140) has joined #mythtv

[04:58:32]

gigem_ (gigem_!~david@host137.12.intrusion.com) has joined #mythtv

[04:58:32]

dekarl (dekarl!~deKarl@e180155217.adsl.alicedsl.de) has joined #mythtv

[04:58:32]

poptix (poptix!poptix@poptix.net) has joined #mythtv

[04:58:32]

pheld (pheld!~heldal@cl-5.osl-01.no.sixxs.net) has joined #mythtv

[04:58:32]

kurre_ (kurre_!~tomimo@xdsl-83-150-88-111.nebulazone.fi) has joined #mythtv

[04:58:32]

gregL (gregL!~greg@cpe-74-76-125-87.nycap.res.rr.com) has joined #mythtv

[04:58:32]

XChatMav (XChatMav!~MaverickT@111.86.233.220.static.exetel.com.au) has joined #mythtv

[04:58:32]

dagar (dagar!~dagar@agar.ca) has joined #mythtv

[04:58:32]

jwhite (jwhite!~jwhite@75-146-153-89-minnesota.hfc.comcastbusiness.net) has joined #mythtv

[04:58:32]

Mudboy (Mudboy!~zod@cpe-75-83-190-215.socal.res.rr.com) has joined #mythtv

[04:59:15]

rhpot199` is now known as rhpot1991

[04:59:24]

rhpot1991 (rhpot1991!~rhpot1991@204.8.45.13) has quit (Changing host)

[04:59:25]

rhpot1991 (rhpot1991!~rhpot1991@ubuntu/member/rhpot1991) has joined #mythtv

[05:01:19]

rooaus (rooaus!~cameron@ppp59-167-126-229.static.internode.on.net) has quit (Ping timeout: 333 seconds)

[05:01:29]

paul-h (paul-h!~paulh@5ad84320.bb.sky.com) has joined #mythtv

[05:01:50]

Snow-Man (Snow-Man!~sfrost@tamriel.snowman.net) has joined #mythtv

[05:02:03]

Snow-Man_ (Snow-Man_!~sfrost@tamriel.snowman.net) has joined #mythtv

[05:02:05]

rooaus (rooaus!~cameron@ppp59-167-126-229.static.internode.on.net) has joined #mythtv

[05:02:27]

Snow-Man_ (Snow-Man_!~sfrost@tamriel.snowman.net) has quit (Client Quit)

[05:03:20]

Snow-Man (Snow-Man!~sfrost@tamriel.snowman.net) has quit (Client Quit)

[05:03:27]

Snow-Man (Snow-Man!~sfrost@tamriel.snowman.net) has joined #mythtv

[05:05:56]

jya:

elmojo: what's your mythtv email address?

[05:08:55]

knightr:

jya, it should probably be tralph@mythtv.org ... BTW, just got my git key working should fix your translation problem this week...

[05:09:16]

jya:

cool, was wondering what was happening to that one

[05:10:55]

knightr:

sorry, was sick for most of last week (and I'm under medications so I'm not fully back in working order yet... :)

[05:11:55]

momelod (momelod!~smelo@dsl-173-248-230-1.acanac.net) has joined #mythtv

[05:12:02]

momelod:

greetings

[05:12:35]

knightr:

momelod, you are probably looking for #mythtv-users...

[05:12:56]

knightr:

(but Hi!)

[05:13:23]

momelod:

hello, and thanks

[05:13:30]

knightr:

np

[05:24:16]

elmojo:

jya: haven't tested that address yet or not... it should forward to my private email account

[05:34:47]

jya:

elmojo: it was for the commit log..

[05:36:16]

aputerboy (aputerboy!~chatzilla@pool-96-233-32-137.bstnma.fios.verizon.net) has quit (Quit: ChatZilla 0.9.86 [Firefox 3.6.12/20101026210630])

[05:46:12]

knightr:

kenni, sorry about earlier... I guess I should have waited until I was no longer sick and under medications... For some weird reason I was thinking more about lupdate but this makes no sense in the context of building the app... :)

[06:10:18]

ybot (ybot!~quassel@61.14.141.36) has quit (*.net *.split)

[06:10:18]

sutula (sutula!sutula@nat/hp/x-zshqvxkdufneuhwb) has quit (*.net *.split)

[06:10:18]

jya (jya!~avenardj@60-242-40-141.static.tpgi.com.au) has quit (*.net *.split)

[06:10:19]

Kevin` (Kevin`!~kevin@router.kwzs.be) has quit (*.net *.split)

[06:10:54]

ybot (ybot!~quassel@61.14.141.36) has joined #mythtv

[06:10:54]

sutula (sutula!sutula@nat/hp/x-zshqvxkdufneuhwb) has joined #mythtv

[06:10:54]

jya (jya!~avenardj@60-242-40-141.static.tpgi.com.au) has joined #mythtv

[06:10:54]

Kevin` (Kevin`!~kevin@router.kwzs.be) has joined #mythtv

[06:11:18]

ybot (ybot!~quassel@61.14.141.36) has quit (Quit: No Ping reply in 180 seconds.)

[06:11:29]

jya (jya!~avenardj@60-242-40-141.static.tpgi.com.au) has quit (Ping timeout: 240 seconds)

[06:11:36]

ybot (ybot!~quassel@61.14.141.36) has joined #mythtv

[06:12:57]

jya (jya!~avenardj@60-242-40-141.static.tpgi.com.au) has joined #mythtv

[06:32:27]

jya:

been wondering why suddenly stuff wouldn't built

[06:32:37]

jya:

I see it's related to visibility now being the default

[06:32:39]

jya:

damn

[06:33:14]

markk_:

jya: just been doing a little testing with the last of your patches above. seems to help with various things – some dvd issues are , I think, resolved. certain files now start up much more cleanly (to the extent that you get an extra second or two of video at startup). also seems to fix some troublesome mkv's

[06:33:38]

jya:

markk_: that's great ..

[06:37:50]

jya:

[27419]

[06:37:50]

MythLogBot:

SVN 27419: (branch fixes/0.24) https://github.com/MythTV/mythtv/commit/316f85f7

[06:38:31]

okolsi (okolsi!~mythtv@a88-112-74-247.elisa-laajakaista.fi) has quit (Read error: Operation timed out)

[06:39:54]

dekarl1 (dekarl1!~deKarl@e180155217.adsl.alicedsl.de) has joined #mythtv

[06:41:26]

okolsi (okolsi!~mythtv@a88-112-74-247.elisa-laajakaista.fi) has joined #mythtv

[06:42:22]

dekarl (dekarl!~deKarl@e180155217.adsl.alicedsl.de) has quit (Ping timeout: 276 seconds)

[06:47:06]

jya:

which header do I need to include to get MPUBLIC ?

[06:47:46]

markk_:

mythexp.h ?

[06:49:16]

jya:

"output.h" seems to do the trick

[06:50:28]

jya:

marrk_: re the patch, as I mentioned earlier, that patch break the deadlock but I think there could be more elegant way to not ever entering this condition

[06:54:27]

markk_:

jya: there are always more elegant solutions – the trick is knowing what they are

[06:54:43]

jya:

sure

[06:55:21]

jya:

but here we pause the audio because the buffer is getting full. pausing the audio will not make the audio buffer start to empty

[06:55:45]

jya:

it will stay at the same level. Pause just means that instead of playing audio from the buffer, it plays silence instead (0s)

[06:56:22]

jya:

so once the buffer is seen as almost full, it never comes out of that mode

[06:58:08]

dekarl1 is now known as dekarl

[07:19:00]

MavT (MavT!~MaverickT@111.86.233.220.static.exetel.com.au) has joined #mythtv

[07:20:21]

natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has joined #mythtv

[07:22:26]

natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has quit (Read error: Operation timed out)

[07:22:37]

XChatMav (XChatMav!~MaverickT@111.86.233.220.static.exetel.com.au) has quit (Ping timeout: 272 seconds)

[07:22:46]

xris:

Captain_Murdoch: poke

[07:38:57]

Beirdo:

wow, the linux 32bit initial builds were slooooow

[07:39:38]

Beirdo (Beirdo!~gjhurlbu@unaffiliated/beirdo) has quit (Remote host closed the connection)

[07:39:45]

Beirdo (Beirdo!~gjhurlbu@unaffiliated/beirdo) has joined #mythtv

[07:39:58]

Beirdo:

of course, it was doing 3 at the same time

[07:41:35]

jya:

Beirdo: when I compile mythtv for 4 ubuntu distrib, 32 and 64 bits mode

[07:41:41]

jya:

it takes about 2 hours

[07:41:55]

jya:

and that's a quad-xeon 2.8GHz with 8GB of ram

[07:42:39]

Beirdo:

heh. I'm really spoiled with my i7

[07:43:01]

jya:

the xeon is based on the same architecture as the i7..

[07:43:08]

jya:

so it won't be that much slower :)

[07:43:12]

Beirdo:

with make -j9, it takes roughly 10min for a distclean and ccache flush

[07:43:22]

jya:

not using ccache though

[07:43:32]

Beirdo:

including plugins

[07:43:38]

jya:

this includes creating the new distribution from scratch in a jail

[07:43:42]

Beirdo:

cool

[07:52:47]

markk_:

jya: we're pausing the audio because we don't have any video to display – and we have no idea how long that might be for. so maybe you just need something less aggresive than audio.Reset()

[07:53:52]

jya:

but it's still trying to fill the audio buffer

[07:53:54]

markk_:

I did have a patch somewhere that didn't pause the audio on the first pass – only after a certain time. that won't really help here though

[07:54:09]

markk_:

?

[07:54:31]

boshhead (boshhead!~boshhead@adsl-71-138-128-90.dsl.pltn13.pacbell.net) has quit (Quit: brb)

[07:55:12]

jya:

I mean: the test to see if the audio buffer is 2/3rd full (Why that test too btw?) is right before adding more audio to the buffer

[07:55:56]

markk_:

jya: you've lost me

[07:56:07]

jya:

ok.. I start again..

[07:56:11]

jya:

right now

[07:56:43]

jya:

avfd checks if the audio buffer is 2/3rd full in two places

[07:57:25]

jya:

the place where it made the most difference is in ProcessAudioFrame

[07:57:59]

jya:

so we've decoded the audio, check that the packet is 2/3 full, and as such exit

[07:58:29]

jya:

audio is paused

[07:59:44]

jya:

could always try to remove the first audio frames from the ring buffer instead once in pause mode

[08:00:07]

jya:

that means readjusting the timestamp of the buffer etc.. not going to be trivial

[08:01:14]

jya:

do you see what I mean ?

[08:12:22]

markk_:

jya: you could just try not pausing the audio – but I doubt it would fix the problem as you're relying on the audio device draining the buffers fast enough. whichever way you look at it, you need a method of removing audio data from the buffer if that buffer is almost full and the player is waiting for video frames. in your patch you used reset() which is obviously a sledgehammer approach. otherwise you need a method to remove the oldest data

[08:12:23]

markk_:

in the buffer (which is what I think you suggested)

[08:12:50]

jya:

yeah... problem is removing audio from the beginning of the buffer

[08:13:01]

jya:

while the audio playback thread is about to play it

[08:13:06]

jya:

nasty race here

[08:14:37]

markk_:

jya: the patch seems to work well enough for now – and I suspect it would help with a couple of issues in 0.24

[08:15:25]

jya:

i think so too

[08:15:32]

jya:

can always try to refine it later

[08:16:01]

jya:

i think the problem is only at startup anyway

[08:16:13]

jya:

why does it take so long for playback to start?

[08:16:18]

jya:

I'm seeing 4–5s here

[08:16:27]

jya:

during which it read the audio and fill 3MB of buffer

[08:16:43]

dekarl1 (dekarl1!~deKarl@e180155217.adsl.alicedsl.de) has joined #mythtv

[08:19:13]

dekarl (dekarl!~deKarl@e180155217.adsl.alicedsl.de) has quit (Ping timeout: 276 seconds)

[08:33:32]

clever (clever!~clever@142.167.134.163) has quit (Ping timeout: 255 seconds)

[08:40:18]

clever (clever!~clever@142.167.201.71) has joined #mythtv

[08:48:10]

NightMonkey (NightMonkey!debian-tor@pdpc/supporter/professional/nightmonkey) has quit (Ping timeout: 245 seconds)

[08:54:47]

Mudboy (Mudboy!~zod@cpe-75-83-190-215.socal.res.rr.com) has quit (Quit: No Ping reply in 180 seconds.)

[08:54:54]

Mudboy (Mudboy!~zod@cpe-75-83-190-215.socal.res.rr.com) has joined #mythtv

[08:56:21]

jwhite (jwhite!~jwhite@75-146-153-89-minnesota.hfc.comcastbusiness.net) has quit (Ping timeout: 272 seconds)

[08:56:35]

jwhite (jwhite!~jwhite@75-146-153-89-minnesota.hfc.comcastbusiness.net) has joined #mythtv

[08:56:56]

kurre_ (kurre_!~tomimo@xdsl-83-150-88-111.nebulazone.fi) has quit (Remote host closed the connection)

[08:56:59]

gregL (gregL!~greg@cpe-74-76-125-87.nycap.res.rr.com) has quit (Ping timeout: 272 seconds)

[08:57:03]

kurre (kurre!~tomimo@xdsl-83-150-88-111.nebulazone.fi) has joined #mythtv

[08:59:41]

gregL (gregL!~greg@cpe-74-76-125-87.nycap.res.rr.com) has joined #mythtv

[09:17:39]

Beirdo:

WOOHOO

[09:18:07]

Beirdo:

got the steps (a lot of em, granted) to compile with plugins for linux

[09:20:43]

stuarta:

cross build?

[09:21:43]

Beirdo:

no, just compiling core mythtv, installing locally, then compiling mythplugins

[09:22:01]

stuarta:

oh...

[09:22:12]

stuarta:

not quite the achievement i thought...

[09:22:21]

Beirdo:

now to put it in place for ALL linux slave instances

[09:22:43]

Beirdo:

well, it was tricky to get it to let me do the install prefix locally

[09:31:12]

stuarta:

ah, outside of /

[09:34:45]

Beirdo:

yeah, I wanted it to be relative to the build dir

[09:35:11]

Beirdo:

which took some farting around, and a wee bit of help from the buildbot authors :)

[09:35:30]

stuarta:

ooo farting around... great pastime that is

[09:37:59]

Beirdo:

ooops

[09:39:42]

Beirdo:

hehe

[09:39:49]

Beirdo:

colorful waterfall

[09:40:04]

Beirdo:

the failures were a buildbot config oops

[09:42:12]

Beirdo:

the exception was me restarting buildbot master

[09:42:18]

Beirdo:

after fixing said oops

[09:43:31]

jannau:

Beirdo: make INSTALL_ROOT=../temp install

[09:44:44]

Beirdo:

doubt that will work right

[09:45:00]

stuarta:

it's supposed to

[09:45:10]

Beirdo:

as it recurses down the directories, does it add more ../ to the beginning?

[09:45:58]

Beirdo:

pretty sure it will need an absolute path for that to work right

[09:46:19]

Beirdo:

unless our makefiles are smart about INSTALL_ROOT and not about PREFIX

[09:46:36]

Beirdo:

as on PREFIX is certainly doesn't track directory level

[09:48:24]

Beirdo:

the trick was to hand it an absolute path, but I have it right now

[09:49:12]

Beirdo:

all that mess because we have to make install on the core before building plugins :)

[09:50:46]

XChatMav (XChatMav!~MaverickT@111.86.233.220.static.exetel.com.au) has joined #mythtv

[09:52:19]

MavT (MavT!~MaverickT@111.86.233.220.static.exetel.com.au) has quit (Ping timeout: 240 seconds)

[10:14:41]

stuarta:

stuartm: things build with whatever you submitted

[10:19:11]

deaman (deaman!~deaman@nat/trolltech/x-gpgwosgjnesmurwd) has joined #mythtv

[10:20:07]

elvum_ is now known as elvum

[10:20:22]

jya:

Beirdo: you may want to look at one of the patch used by the mythbuntu. they can build mythplugins by using the main source tree (no need to install mythtv first)

[10:24:17]

jya:

I wanted to commit similar once I properly review it

[10:24:17]

jya:

http://mythtv.pastebin.com/FedNcvyB

[10:24:49]

jya:

you can then provide --mythroot to where the mythtv source code and compile

[10:24:56]

jya:

no need to run make install first

[10:25:31]

jannau:

Beirdo: you're right, it will be relative to the current dir

[10:25:46]

** jannau curses recursive make **

[10:28:22]

jamesba (jamesba!~jamesba@cpc2-stre1-0-0-cust374.1-1.cable.virginmedia.com) has joined #mythtv

[10:30:08]

** jamesba tries to catch up with emails **

[10:30:28]

jamesba:

was working a second ago, but no longer seems to be

[10:33:02]

stuarta:

select all, delete

[10:34:52]

jamesba:

close thunderbird, reopen it

[10:35:21]

jamesba:

the issue was that it seemed to have lost its secure connection to the forge mail server

[10:35:27]

jamesba:

but didn't realise it had

[10:35:43]

jamesba:

so I just needed to give it a kick and make it reconnect from the beginning again

[10:37:10]

dekarl (dekarl!~deKarl@e180155146.adsl.alicedsl.de) has joined #mythtv

[10:38:58]

dekarl1 (dekarl1!~deKarl@e180155217.adsl.alicedsl.de) has quit (Ping timeout: 276 seconds)

[10:49:32]

stuartm:

cdev_: SSDPCache::Find() is used in libmyth/mythcontext.cpp and libmyth/backendselect.cpp

[10:51:44]

stuartm:

stuarta: thanks for testing

[10:56:03]

markk_:

is anyone actually working on the channel scanner?

[10:59:33]

stuarta:

not that i know of atm

[10:59:50]

stuarta:

i need to get an RFC for design out for the browser based setup

[11:01:38]

stuartm:

stuarta: on the -developers list, yes?

[11:03:00]

stuarta:

well that's where i would send it yes.

[11:06:15]

stuartm:

ok, I was just checking :) If it went to -dev I think we could say that it would kill this task stone dead

[11:07:44]

wagnerrp (wagnerrp!~wagnerrp_@NR-FT1-66-42-240-159.fuse.net) has quit (Ping timeout: 265 seconds)

[11:17:07]

gigem_ (gigem_!~david@host137.12.intrusion.com) has quit (Ping timeout: 240 seconds)

[11:17:18]

gigem_ (gigem_!~david@host137.12.intrusion.com) has joined #mythtv

[11:18:20]

wagnerrp (wagnerrp!~wagnerrp_@NR-FT1-66-42-240-159.fuse.net) has joined #mythtv

[11:18:56]

markk_:

danielk22: re safe_read returning and error – I suspect this is exactly the sort of thing that is needed.

[11:20:15]

markk_:

what would be extremely useful in the short term would be some fairly verbose logging around both the ringbuffers in both the frontend and backend. is -v file sufficient to give that sort of information?

[11:31:26]

markk_:

danielk22: I spotted a couple of other things today. when I first start livetv here, before the first transition to the next program, I get a number of 'waited for 0.2 seconds for data to become available' followed by I check for the next program. when the transition arrives, there is a significant discontinuity as we wait for the decoder to pause – and the decoder is (I think) stuck waiting on data, usually for around 1.5 seconds.

[11:33:40]

stuarta:

that's where that guy proposing we don't pause and unpause the decoder all the time was right, we should just mark the transistion in the decoder and make it internally reset it's state

[11:34:08]

markk_:

danielk22: if I pause a little after starting livetv I don't get the 0.2second wait messages, the tv chain doesn't think there is a discontinuity and the transition is nearly seamless – there is no wait for the decoder to pause, presumably because there is still data available.

[11:34:39]

markk_:

so my question is – can we tweak something somewhere to ensure we always hit the second case?

[11:37:29]

markk_:

stuarta: the ideal would be seemlessly transition the data handed to the decoder. i.e. somehow merge the output of the 2 different 'recordings'

[11:37:43]

stuarta:

magic markers :)

[11:38:03]

stuarta:

isn't that what the livetv chain was meant to achieve tho....

[11:42:19]

Kevin` (Kevin`!~kevin@router.kwzs.be) has quit (Read error: Connection reset by peer)

[11:46:55]

Kevin` (Kevin`!~kevin@router.kwzs.be) has joined #mythtv

[11:50:18]

nutron (nutron!~dserban@S010600195b2f0f96.ok.shawcable.net) has quit (Remote host closed the connection)

[12:38:44]

simonckenyon (simonckenyon!~simoncken@195.7.61.12) has quit (Remote host closed the connection)

[12:41:01]

gigem (gigem!~david@host137.12.intrusion.com) has joined #mythtv

[12:44:45]

gigem_ (gigem_!~david@host137.12.intrusion.com) has quit (Ping timeout: 240 seconds)

[13:20:56]

deaman (deaman!~deaman@nat/trolltech/x-gpgwosgjnesmurwd) has quit (Quit: deaman)

[13:33:52]

okolsi (okolsi!~mythtv@a88-112-74-247.elisa-laajakaista.fi) has quit (Read error: Operation timed out)

[13:36:30]

okolsi (okolsi!~mythtv@a88-112-74-247.elisa-laajakaista.fi) has joined #mythtv

[13:49:53]

jya:

the new audio playback feature have now been tested and working quite reliably, so I am going to merge back to master. I will not merge the ffmpeg changes, so HD audio passthrough will not work until ffmpeg has been resync with upstream

[14:09:46]

j-rod|afk is now known as j-rod

[14:22:16]

Feico (Feico!~Feico@sd5116475.adsl.wanadoo.nl) has joined #mythtv

[14:39:36]

jya:

In order to prevent having the ffmpeg related commit hitting master

[14:39:41]

jya:

I've done the following:

[14:39:53]

jya:

merge temp/newaudiosetup back to merge

[14:40:34]

jya:

then did a git rebase from the new point where the ffmpeg commit were done in the branch, removed all of those (git rebase --interactive)

[14:40:47]

jya:

is it ok to push it back to github ?

[14:41:23]

jya:

it seems to have changed some of the SHA for more recent commit ...

[14:41:28]

jya:

that can't be a good thing

[14:43:32]

jya:

SHA: c67b3b29becd7f4fac607db90d8c538f617b91d1

[14:54:05]

jarle (jarle!~jarle@70.84-234-133.customer.lyse.net) has joined #mythtv

[15:12:43]

AriX (AriX!~Ari@207.245.119.4) has joined #mythtv

[15:42:26]

Kunalagon (Kunalagon!~Kunalagon@212.200.241.156) has joined #mythtv

[15:43:11]

superm1:

hum, i think the data/ directory in mythweb might have been lost to git because it was empty

[15:51:39]

AriX (AriX!~Ari@207.245.119.4) has quit (Quit: Leaving...)

[15:54:31]

stuartm:

superm1: didn't xris just fix that?

[15:55:16]

stuartm:

nope, different directory

[15:55:18]

superm1:

maybe? my checkout was a few days old

[15:57:39]

iamlindoro:

sweet zombie jesus, what happened to commits now

[15:57:55]

** stuarta points up to a rebase **

[15:58:14]

stuarta:

that is pure speculation btw

[15:58:39]

iamlindoro:

sounds likely

[15:58:59]

stuarta:

i haven't even got to the bottom of my inbox

[15:59:16]

stuartm:

ick

[15:59:51]

iamlindoro:

Does that mean that any changes to the release notes/notes on tickets for those commits point to nothing now?

[16:01:15]

stuartm:

when will it end?

[16:01:22]

** stuartm drowns in commit spam **

[16:02:07]

** stuarta shrugs. **

[16:02:17]

stuarta:

i'm used to it. rarely up to date anyway

[16:03:26]

stuartm:

there has got to be a better way to do that, having the last 100 commits 'recommitted' to the repo is chronologically confusing in addition to being noisy

[16:03:37]

stuarta:

agreed

[16:03:43]

iamlindoro:

I'm much more concerned about static references to those commits

[16:04:00]

iamlindoro:

Does anyone know? are the release notes/ticket references for those commits now "wrong"?

[16:04:11]

stuartm:

jya: broke the build – http://code.mythtv.org/buildbot/waterfall

[16:04:42]

jya:

that's a cool page..

[16:04:59]

jya:

stuartm: it is very noisy :(

[16:05:12]

jya:

could have made it far less noisy if I had used rebase

[16:05:27]

jya:

but as rebase is a no-no for the time being

[16:05:52]

jya:

iamlindoro: I didn't push my original merge I discussed earlier

[16:05:57]

iamlindoro:

ah, cool

[16:05:58]

jya:

so nothing got rebased

[16:06:15]

jya:

for that precise reason ; didn't want references to suddenly be broken

[16:06:29]

jya:

it's already hard enough with references breaking between svn and git

[16:06:41]

iamlindoro:

yeah, cool. There must be a happy medium somehow

[16:07:32]

jya:

stuartm: what's that cool page you point too.. what did I break now?

[16:07:43]

stuarta:

i might have to get my OSX box up and running permanently to be a buildbot

[16:08:13]

iamlindoro:

stuarta: My work workstation is OS X, having a buildbot running wouldn't be a problem-- I've already offered but nobody has spoken to me yet

[16:08:22]

stuarta:

Beirdo: ^^^

[16:08:26]

jya:

ah, I saw that error with mythnetvision

[16:08:39]

jannau:

jya and everyone: please try to avoid merging master in feature branches

[16:08:55]

jya:

jannau: what do we do then ?

[16:09:21]

stuarta:

which way is that? master -> branch?

[16:09:37]

JEDIDIAH__ (JEDIDIAH__!~jedi@cpe-76-185-72-21.tx.res.rr.com) has joined #mythtv

[16:09:43]

jya:

iamlindoro: mythnetvision needs to be updated to link to libavxxx

[16:10:05]

jya:

i didn't fix it, because mythnetvision didn't build on an unpatched tree either

[16:10:08]

jannau:

either don't push and rebase or develop from a static point and do a merge whenn finished

[16:10:09]

Mudboy:

anyone tried running a revo has a myth backend, using a hdhomerun?

[16:10:16]

iamlindoro:

MNV was building just fine as of last night

[16:10:32]

jannau:

Mudboy: please read /topic

[16:10:38]

jya:

jannau: i didn't rebase...

[16:10:43]

Mudboy:

ahh, my bad

[16:11:13]

jya:

not master anyway

[16:11:38]

Mudboy (Mudboy!~zod@cpe-75-83-190-215.socal.res.rr.com) has left #mythtv ("bbiab")

[16:11:39]

iamlindoro:

jya: It appears (according to the waterfall anyway) that your merge broke the compile-- it was building fine beforehand

[16:11:41]

jya:

git is pretty good at merging i have to admit

[16:12:01]

jya:

iamlindoro: this is in relation to a question I asked two days ago

[16:12:10]

iamlindoro:

FWIW MNV doesn't use any libav* functionality whatsoever

[16:12:12]

jya:

when I made references to libavformat in libmyth

[16:12:23]

jannau:

jya: yes, but you merged master over 10 times into the newaudiosetup branch

[16:12:51]

jya:

iamlindoro: it does ; because it plays audio... and audio is linked to libavformat

[16:13:02]

jya:

jannau: sure.. what's wrong with that?

[16:13:03]

iamlindoro:

jya: It does not play audio

[16:13:11]

jannau:

this will result in ugly history and duplicate commits. don't doo that

[16:13:15]

jya:

it doesn't play video?

[16:13:18]

iamlindoro:

jya: MNV is a metadata browser-- it has no playback code whatsoever

[16:13:20]

iamlindoro:

no

[16:13:28]

jya:

well, it's using libmyth

[16:13:29]

iamlindoro:

MNV merely invokes mythbrowser

[16:13:36]

jya:

and libmyth is linked to libavformat

[16:14:07]

iamlindoro:

So what has changed in your merge that it no longer works?

[16:14:19]

jya:

jannau: great, so the whole idea behind git is how much easier it is to merge and use branches.. and now we're told not too.

[16:14:51]

jya:

and honnestly, the most noisy thing is the commit maillist

[16:15:12]

jya:

probably could start by fixing this instead of creating even more noise

[16:15:31]

jya:

iamlindoro: i use ffmpeg muxer

[16:15:49]

jya:

to encapsulate the audio stream in an iec958 frame

[16:15:56]

iamlindoro:

jya: So any plugin that links libmyth from now on will also need to link libav*? That seems unfortunate

[16:16:20]

jya:

before there was spdif encoder in libmythtv as well as libmyth

[16:16:23]

jya:

two of them...

[16:16:30]

jya:

duplicated code, and lots of line

[16:16:53]

jya:

iamlindoro: I agree. Which is why I asked where I should put the new spdif encoder: in libmythtv or libmyth

[16:17:02]

stuarta:

libmyth is meant to be the core and have no other lib dependencies

[16:17:10]

jya:

and asked if it was okay to create a dependency to libav...

[16:17:14]

stuarta:

if that is no longer true, then the code is in the wrong place

[16:17:17]

jya:

danielk22 said that it was fine

[16:17:22]

jannau:

jya: it's not only the commits mailing list but also git log

[16:17:34]

jannau:

and I don't care about the mailing list

[16:18:20]

jannau:

commit history becomes pretty much useless if every 5th commit is "merge ..."

[16:18:36]

jya:

jannau: I understand that it's noisy.. but heck... now if we can't use one of the only cool feature of git: branching ... why did we move from svn

[16:18:54]

jya:

jannau: we wouldn't have all those "merge..." if we weren't told at the beginning not to use rebase

[16:18:55]

jannau:

jya: you can use it

[16:18:59]

stuarta:

jannau: which way do you prefer it's done then? the svn way was to merge all head changes to your branch, and then merge back the result.

[16:19:20]

jya:

stuarta: that's what I've done...

[16:19:26]

stuarta:

i know

[16:19:40]

stuarta:

and i would do exactly the same thing if i were working on a big feature

[16:19:49]

stuarta:

so i want to know why this is now the wrong thing todo

[16:20:00]

jannau:

branch, develop, merge to master

[16:20:14]

stuarta:

so how do you deal with merge conflicts?

[16:20:16]

jya:

stuarta: and I was fixing a lot of issues on master at the same time... fixes that I relied on in my branch

[16:20:44]

jya:

I honnestly don't see how I could have done differently without using rebase

[16:20:51]

stuarta:

jannau: what about long running branches then?

[16:21:00]

jannau:

jya: using rebase is perfectly fine

[16:21:17]

jya:

jannau: we were explicitely told not to use rebase when we started git

[16:21:40]

** jannau wouldn't call 7 days long running **

[16:21:46]

jya:

which lead to all the noise with all the merge

[16:21:49]

stuarta:

so what is rebase actually for?

[16:22:02]

stuarta:

rebasing your branch post a merge to trunk?

[16:22:06]

jya:

rebase move your commit as if you were doing them now

[16:22:20]

** stuarta scratches head **

[16:22:22]

jya:

on a recent version.. so you don't create a branch when you push

[16:22:36]

jannau:

I always said don't use rebase on public branches

[16:23:11]

stuarta:

so if you started coding against A, and head went A->B->C

[16:23:13]

jya:

say you make change 1, I come I make change 2, I push. now you push, you have to merge with my change first, so you effectively created a branch which you merged

[16:23:17]

sutula (sutula!sutula@nat/hp/x-zshqvxkdufneuhwb) has quit (*.net *.split)

[16:23:21]

jya:

that makes a lot of noise in the log

[16:23:33]

jya:

if you rebase, it's as if change 1 was done after my change 2

[16:23:45]

jya:

so there's no branching occurring and no merge

[16:23:52]

** stuarta hmmms **

[16:24:13]

jya:

so you do A -> B -> C

[16:24:19]

jya:

it's like you did B -> C -> A

[16:24:41]

stuarta:

my diff would have been against A

[16:24:43]

jya:

and you don't have that line leaveing A, and going back to C

[16:24:48]

jya:

it is

[16:24:57]

jya:

but if there's no conflict with B and C

[16:25:05]

jya:

then there's no problem

[16:25:06]

stuarta:

what happens when B & C create a conflicting merge?

[16:25:13]

jya:

then you don't rebase :)

[16:25:20]

stuarta:

wtf do you do?

[16:25:27]

jya:

well

[16:25:33]

jya:

when doing the rebase

[16:25:38]

jya:

it tells you there's a conflict

[16:25:43]

jya:

so you have to edit the files

[16:25:44]

stuarta:

run around until you find your copy of hitchikers guide to the galaxy with "Don't Panic!" on the front?

[16:25:52]

jya:

when done you git add the new file

[16:25:53]

jannau:

resolving the conflicts during a rebase is no problem

[16:25:55]

jya:

and rebase --continue

[16:26:07]

** stuarta reads up on git-rebase **

[16:26:28]

Kunalagon (Kunalagon!~Kunalagon@212.200.241.156) has quit (Quit: Leaving.)

[16:26:45]

jya:

jannau: one of the issue at hand, is that there were bugs I felt were important to be backported ASAP to fixes

[16:26:57]

jya:

so I made all of those fix in master , backported them to 0.24

[16:27:06]

jya:

then I merged those changes back to my branch

[16:27:38]

** iamlindoro hands stuarta a towel **

[16:27:50]

** stuarta demands a babelfish **

[16:28:06]

jya:

jannau: the noise appears to be only in github commit list, and in our mailing list

[16:28:13]

jya:

when I do git log

[16:28:19]

iamlindoro:

In fairness, MythTV *is* powered by an infinite improbability drive

[16:28:20]

jya:

I see no problem there

[16:28:42]

jya:

the merge appears as one changeset only: b2a86a2ad29cef7b7d8c19cc5929510be30c94fa

[16:29:06]

jya:

it's github that's crap and can't manage his post-commit script properly

[16:29:24]

jya:

they have the same issues as our post-commit script used for the mailing list

[16:30:15]

** jya I'll move the spdif encoder back to libmythtv then **

[16:30:27]

jya:

sigh ...

[16:30:42]

sutula (sutula!sutula@nat/hp/x-biuyifjicxaykase) has joined #mythtv

[16:30:51]

stuarta:

sure there isn't a better way?

[16:31:19]

iamlindoro:

jya: Wouldn't that mean that mythmusic couldn't use the new encoder?

[16:31:27]

jannau:

jya: was there need to port them to your branch?

[16:31:32]

jannau:

jya: git log origin/master --pretty=raw | grep "configure: disable compiler warnings for external/FFmpeg"

[16:32:01]

jya:

hum.. can libmyth rely on libmythtv then ?

[16:32:14]

jannau:

jya: no

[16:32:17]

jya:

iamlindoro: I don't know anymore

[16:32:28]

jya:

so where am I going to put the spdif encoder ?

[16:32:41]

jya:

am I supposed to duplicate the code wherever it's being used?

[16:32:59]

jya:

that sounds awfully innefective

[16:33:27]

iamlindoro:

jya: I take it back, we recently actually started installing some headers from libmythtv, so just as long as it gets added to installs, I guess it'd be okay

[16:33:39]

iamlindoro:

(and obviously mythmusic will need to link libmythtv)

[16:33:55]

iamlindoro:

So sounds like libmythtv will work

[16:34:26]

jya:

I can move the related audio stuff in libmythtv/audio , like I did in libmyth/audio

[16:35:07]

jya:

it's not very elegant to have code used by the audio class in libmyth, be in libmythtv

[16:36:35]

jya:

for the time being, if someone could write some A-B-C instructions on how to handle branches and merge them back, while keeping the branch up to date

[16:37:04]

jya:

jannau: to your question "was there need to port them to your branch?" yes.. because I found the bug as I encountered the issue

[16:37:20]

jya:

those issues turned out to be relevant to many of the playback bugs seen in 0.24

[16:37:48]

jya:

now I could have made them in my branch only, then merge back to master, then cherry pick them to fixes

[16:38:02]

jannau:

cherry pick from your branch might have been nicer

[16:38:39]

jya:

but it would have ended in the same result: whatever I do, someone tells me I shouldn't have done it... it's getting very frustrating. this project starts to require more and more goodwill to keep forward

[16:39:23]

laga:

jya: your work is much appreciated; sorry to hear it's frustrating you so much

[16:39:51]

jya:

jannau: ok, I'll cherry pick next time... but you have to admit, merging is so easy compare to svn, that one could be forgiven for actually using it

[16:40:05]

jya:

laga: I don't mind dong the coding, it's publishing it that is frustrating

[16:40:13]

jya:

and thanks for the kind words

[16:40:29]

iamlindoro:

jya: I know how that feels, believe me. We'll figure it out, I think the change to git is confusing all of us.

[16:40:37]

laga:

jya: i realize that, but i'm sure i'll get easier in no time

[16:41:10]

jya:

actually, as I was doing all the merge... when I would have had to do almost everything manually in svn

[16:41:16]

jya:

like backporting to fixes

[16:41:22]

jya:

the directory structure is different

[16:41:33]

jya:

yet it has absolutely 0 problem doing the cherry pick

[16:41:37]

jya:

it's really amazing

[16:41:38]

jannau:

jya: just because it's easy it doesn't need to be done all the time

[16:41:57]

jya:

isn't git there to make our life easier ? :P

[16:42:05]

jannau:

and now bad feelings, we just switched to git and all need to learn how to handle it

[16:42:16]

jannau:

=w

[16:42:17]

jya:

for the first time, I wasn't cursing at how much more complex the whole thing was

[16:42:17]

jannau:

-w

[16:42:43]

jya:

jannau: in reference to ffmpeg, is there a way you could make it use ccache ?

[16:42:57]

jya:

when I recompile 2/3 of the time is spent in ffmepg compiling

[16:43:07]

jya:

all the c++ part is almost instant

[16:43:07]

jannau:

--cc="ccache gcc"

[16:43:23]

jya:

can't it be done by default? it's done for everything else

[16:43:48]

jannau:

you mean in mythtv build? it uses ccache here?

[16:44:13]

jya:

all libmyth*

[16:44:18]

jya:

compiles with ccache

[16:44:25]

jannau:

make -C external/FFmpeg -j3 4.01s user 1.90s system 105% cpu 5.593 total

[16:44:25]

jya:

I certainly didn't tell configure to do so

[16:44:39]

jya:

not after a reconfigure or make clean

[16:45:03]

jannau:

that was after a make clean

[16:45:15]

jya:

well, you certainly have a faster machine than me then

[16:45:24]

jya:

when the compilation restart

[16:45:33]

jya:

I can't read the c++ line for all libmyth

[16:45:37]

jya:

it scrolls too fast

[16:45:44]

jya:

I certainly can with ffmpeg

[16:45:49]

jya:

anyhow, it was a suggestion

[16:46:04]

jannau:

a ./configure with different options will cause many cache misses due to config.mak|h changes

[16:46:39]

jya:

jannau: I use this a lot swapping branch

[16:46:47]

jya:

like checkout fixes/0.24

[16:46:55]

jya:

git cherry-pick blash

[16:47:04]

jya:

make distclean ; ./configure

[16:47:05]

jya:

make

[16:47:07]

jannau:

increase your cache size then

[16:47:09]

jya:

make install

[16:47:12]

jya:

I test

[16:47:43]

jya:

i give up.... forget me ever asking...

[16:51:55]

iamlindoro:

stuartm: Thanks for the symbol visibliity stuff for libmythmetadata

[16:58:28]

jya:

jannau: if I add "--cc="ccache gcc" to configure, then none of the c++ code use ccache anymore

[16:59:02]

jannau:

jya: that was for a ffmpeg repo

[16:59:35]

jya:

I'm not using a ffmpeg repo.. I'm only interested in mythtv code.

[17:00:22]

jya:

hum... libmyth has currently no dependencies whatsoever to libmythtv

[17:00:38]

jya:

I don't like the idea of being the first to add one

[17:01:14]

jya:

when would you ever have libmyth installed but not libmythavblah ?

[17:01:56]

jya:

i guess we could use weak-links instead

[17:02:01]

jya:

that would be the cleanest

[17:06:43]

jya:

anyone knows if gcc on windows supports weak links?

[17:09:49]

jannau:

jya: libmythaudio

[17:10:14]

jya:

move everything in mythaudio ? I had thought about it for a while this one

[17:11:00]

jya:

I thought the code wasn't significant enough in size to warrant its own library

[17:13:15]

jannau:

dependency issues are good supporting argument though

[17:13:57]

jya:

they are... and something as tiny as nvctrl having its own library...

[17:15:02]

jamesba (jamesba!~jamesba@cpc2-stre1-0-0-cust374.1-1.cable.virginmedia.com) has quit (Remote host closed the connection)

[17:15:24]

jya:

10318 lines of code... not that much

[17:15:44]

jya:

it's like what 0.1% of ffmpeg ? :)

[17:16:48]

jya:

from the gcc manual: Weak symbols are supported for ELF targets, and also for a.out targets when using the GNU assembler and linker.

[17:17:01]

jya:

so I'm guessing it's not available in windows

[17:17:10]

jya:

though it works on a mac, and that's not elf either

[17:17:44]

jarle (jarle!~jarle@70.84-234-133.customer.lyse.net) has quit (Remote host closed the connection)

[17:21:48]

iamlindoro:

We should do up a patch submission howto for git

[17:22:08]

jya:

telling people to use pull request is pretty neat

[17:22:14]

iamlindoro:

So that folks can properly git format patches and we can give them proper attribution/use their commit log when relevant/etc.

[17:22:35]

iamlindoro:

VLC's guide for doing so is decent: http://wiki.videolan.org/Git#Submitting_patch . . . r_x264-devel

[17:23:23]

jya:

http://help.github.com/pull-requests/

[17:24:14]

iamlindoro:

I think pull requests are good too, but what about someone who just wants to send a simple patch?

[17:24:34]

iamlindoro:

Versus having to set up a github account and follow a half dozen+ steps

[17:25:45]

jannau:

I prefer git format-patch / send-mail

[17:26:32]

jya:

sure, but wouldn't he had to git clone and all just to create the patch?

[17:27:43]

jannau:

wouldn't the same be the case for github pull request?

[17:27:51]

jya:

exactly.

[17:28:13]

jya:

this is in regards to the arguments that sending an email/patch is easier

[17:28:25]

jannau:

or does github has a web based patch editor

[17:28:27]

jannau:

?

[17:28:29]

jya:

what i like with the pull request, is that you go and get it when you feel like it

[17:28:51]

jya:

whole pull vs push debate i gues

[17:28:51]

jya:

s

[17:28:57]

Steve_Goodey (Steve_Goodey!~steve@host86-145-237-137.range86-145.btcentralplus.com) has joined #mythtv

[17:29:30]

jannau:

how's that different from a patch by mail?

[17:29:50]

jya:

push vs pull..

[17:30:34]

SteveGoodey (SteveGoodey!~steve@host86-145-237-137.range86-145.btcentralplus.com) has joined #mythtv

[17:30:37]

jya:

and a pull request, can be changed easily. once you've sent your patch, it's sent... have to deal with version , updates etc

[17:30:38]

Beirdo:

the warnings count should appear in buildbot again now. I forgot to check that last night

[17:30:46]

Beirdo:

and... off to work.

[17:32:05]

jya:

Beirdo

[17:32:18]

jya:

I wrote earlier about having a --mythroot option

[17:32:23]

jya:

to compile mythplugins in place

[17:32:31]

jya:

and not requiring to do mythtv install first

[17:37:49]

stuartm:

stuarta: libmyth is no longer the core lib, that's now libmythdb (mis-named, should be called libmythbase), libmyth is being dismantled and ultimately will cease to exist

[17:38:12]

stuartm:

but libmyth does link other libs such as libmythui and libmythupnp

[17:38:27]

dekarl:

fwiw the main hurdle of contributing patches is not the generation of the patch (in whichever form) but catching the interest of a developer to spend time on looking at it and pushing it to the master (anybody mind looking at my patch from this morning, it's really simple and in pretty git format-patch format for your testing pleasures ;)

[17:38:33]

jarle (jarle!~jarle@70.84-234-133.customer.lyse.net) has joined #mythtv

[17:40:00]

gigem_ (gigem_!~david@host137.12.intrusion.com) has joined #mythtv

[17:40:13]

iamlindoro:

Depends entirely on what part of myth is affected

[17:40:28]

iamlindoro:

There are those tickets that I can close in minutes because they are obvious no brainer fixes

[17:41:06]

iamlindoro:

Then there are those patches that require specialized knowledge that few of us have/code that we don't/can't use and thus needs to wait on a much smaller group of potential reviewers, most of whom have dozens or more tickets in the queue ahead of yours

[17:41:59]

jya:

is it possible to set a __attribute__ to a whole bunch of functons in a block rather than doing it for each ?

[17:42:16]

kormoc:

of course, if the users pooled money together and hired the devs so they didn't need day jobs anymore, patches would get handled faster I'm sure

[17:42:19]

dekarl:

imalindoro: PESPacket CRC for TDT is affected, so it's quite central ( http://mythtv.org/pipermail/mythtv-dev/2010-December/070118.html I can reopen the old ticket if you like)

[17:42:19]

dekarl:

I don't know if it's simple enough for someone to try the patch submission process

[17:42:46]

sphery:

dekarl: you mean your patch in thread, "Ticket #5013: Constant PESPacket CRC check failure" on -dev list? If so, your best bet is to post it to a new ticket in trac for consideration. Then the EIT/Player/... whoever guys can decide if it's appropriate to just ignore that message in that case or more appropriate to make it VERBOSE(VB_SIPARSER|VB_EXTRA, ...)

[17:43:20]

iamlindoro:

dekarl: That would be code that most of us don't use/maintain, thus personally I wouldn't feel comfortable reviewing or committing it-- it *sounds* good to me, but really on stuarta or janne would be able to comfortably review and apply it

[17:43:41]

jya:

kormoc: not sure about that, I haven't been to work for days thanks to myth getting my attention :P

[17:43:47]

iamlindoro:

er really only

[17:43:53]

gigem (gigem!~david@host137.12.intrusion.com) has quit (Ping timeout: 264 seconds)

[17:44:23]

dekarl:

sphery: ok, i'll open a fresh ticket

[17:45:25]

sphery:

dekarl: thanks. patches on the lists can go overlooked, plus even for simple things, we like the history provided by trac

[17:50:22]

mrand (mrand!~mrand@ubuntu/member/mrand) has quit (Ping timeout: 265 seconds)

[17:51:35]

mrand (mrand!~mrand@ubuntu/member/mrand) has joined #mythtv

[17:57:08]

jannau:

dekarl: don't, I'll handle it

[17:58:21]

dekarl:

jannau: meh, just hit "open"

[18:01:27]

jya:

how often is this autobuild started?

[18:01:37]

iamlindoro:

I think it's triggered by a commit

[18:01:51]

jya:

does it stop the previous one?

[18:01:53]

iamlindoro:

er.. psuh

[18:01:56]

iamlindoro:

No, don't think so

[18:01:57]

jya:

this is a pretty cool system

[18:02:15]

jya:

now if only it would also generate packages :)

[18:02:35]

iamlindoro:

If it did, we couldn't distribute them

[18:03:01]

iamlindoro:

(ie, I'm sure it could, but we'd just have to rm them anyway)

[18:03:24]

jya:

why?

[18:03:40]

jya:

there are so many packages anyway... would be impossible to please everyone

[18:03:44]

iamlindoro:

Because we are afforded a legal protection by not distributing binaries

[18:03:49]

jya:

ahhh

[18:04:06]

iamlindoro:

(it's one of the same protections ffmpeg uses)

[18:04:46]

jya:

ok, I have a version that works with weak linking

[18:04:57]

jya:

won't complain the build of mythplugins

[18:05:05]

jya:

but there's another one (not mind)

[18:05:17]

jya:

thumbfinder.cpp:572: error: â€˜av_estimate_timingsâ€™ was not declared in this scope

[18:05:50]

jya:

(not mine that is)

[18:06:23]

jya:

it's the same deal.. sub-dependencies to libavsomething

[18:07:10]

jya:

it's used by mythmusic and mytharchive

[18:09:26]

SteveGoodey (SteveGoodey!~steve@host86-145-237-137.range86-145.btcentralplus.com) has quit (Ping timeout: 240 seconds)

[18:09:39]

jya:

how can this fail now when it was added in 2007 ?

[18:09:54]

jya:

probably the visibility thing..

[18:13:41]

hashbang (hashbang!~alex@213-152-35-50.dsl.eclipse.net.uk) has joined #mythtv

[18:20:42]

kth (kth!~kth@dyndsl-080-228-190-123.ewe-ip-backbone.de) has joined #mythtv

[18:22:04]

gigem_ is now known as gigem

[18:28:20]

jya:

iamlindoro: the push I'm about to make , make you change to mythnetvision unecessary

[18:28:29]

jya:

and you also need libavutil and libavcodec

[18:28:30]

iamlindoro:

ok, feel free to undo it

[18:28:49]

iamlindoro:

well, the change I pushed fixed the waterfall, so I presumed it was adequate

[18:28:52]

jya:

nah: you're thing :)

[18:29:08]

jya:

we need to add something similar to libmytharchive and libmythmusic

[18:29:19]

iamlindoro:

sigh

[18:29:27]

jya:

not my fauld :)

[18:29:33]

iamlindoro:

of course not

[18:29:39]

jya:

actually, the problem was there since it was added in 2007

[18:29:53]

jya:

as now there's a visibility by default

[18:29:56]

jya:

it now complains

[18:30:07]

jya:

it's a different call this time

[18:30:14]

jya:

to av_estimate_timings

[18:33:10]

Beirdo:

ah, much better with the warnings listed :)

[18:33:14]

xris:

superm1 / wagnerrp: what package tag format are you using for your trunk/master packages?

[18:34:37]

natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has joined #mythtv

[18:35:42]

Beirdo:

wonder why 64bit and 32bit have a different warning count in plugins.

[18:35:54]

Beirdo:

must be a pointer->int conversion somewhere, I guess

[18:37:53]

Beirdo:

or a gcc version difference?

[18:47:07]

jya:

iamlindoro: can you check that it works for you with SHA: 83e43f3cee161e996f41ff8413896b8a1080c54c

[18:47:29]

jya:

and see if mythnetvision properly starts ? (I can only check that it compiles)

[18:48:05]

iamlindoro:

I can't check operation for another 8 hours or so, but can test compile

[18:49:06]

stuartm:

Beirdo: I'll do a diff of the warnings in a little while

[18:49:26]

Beirdo:

Oooh, I see why it's not failing in 32bit

[18:49:50]

Beirdo:

it's missing pycurl, so mythnetvision is disabled. Let me fix that

[18:50:57]

Beirdo:

There, added pycurl

[18:55:16]

iamlindoro:

jya: compiling fine, can't test operation until tonight

[18:55:24]

jya:

ok

[18:55:32]

jya:

should be fine i believe

[18:55:42]

iamlindoro:

yeah, should be okay

[18:55:54]

jya:

i wonder why mytharchive compile though

[19:00:03]

Beirdo:

there. warning counts now match. That means the extra warning was in mythnetvision :)

[19:03:53]

wagnerrp:

xris: tag... format...?

[19:07:14]

kth (kth!~kth@dyndsl-080-228-190-123.ewe-ip-backbone.de) has quit (Quit: Leaving.)

[19:09:20]

stuartm:

avformatdecoder.cpp:3640: warning: unused variable â€˜ofillâ€™

[19:09:33]

stuartm:

jya: ^^ Yours?

[19:09:53]

jya:

ah yeah, I removed the use of that variable

[19:10:19]

stuartm:

three warnings for three different variables

[19:10:23]

stuartm:

audiogeneralsettings.cpp:1124: warning: â€˜settings6â€™ may be used uninitialized in this function

[19:11:34]

** jya always amazed that stuartm finds time to pick on the warnings of others :) **

[19:12:40]

stuartm:

jya: it's in the buildbot output, the count rose from 129 to 136 and I happen to be working on reducing it so that stood out :)

[19:13:02]

jya:

yeah sure ....

[19:13:22]

jya:

I can imagine the feeling: how look someone committed, what mistake did they make :)

[19:14:36]

stuartm:

heh, ask Beirdo, we're trying to get the count down to zero so that it becomes a useful metric – http://code.mythtv.org/buildbot/waterfall << Warning count in brackets in the compile section for each build, links to the list

[19:15:01]

** jya wonders why av_estimate_timings prevents to compile when it's been there for the past 4 years **

[19:15:02]

stuartm:

iamlindoro: thanks for that, I trust that it worked as expected

[19:15:20]

iamlindoro:

stuartm: Yep, like a charm-- figure it'll knock down a few more warnings too

[19:15:31]

stuartm:

aye, it should

[19:15:32]

jya:

this is awesome this buildbot

[19:17:03]

stuartm:

Beirdo: I assume that it would be good to have multiple slaves for each build, as a backup or to share load?

[19:19:33]

jya:

if I understand right, that build bot says everything compile properly, including mytharchive

[19:19:47]

jya:

i get thumbfinder.cpp:572: error: â€˜av_estimate_timingsâ€™ was not declared in this scope

[19:21:07]

stuartm:

jya: this is where having multiple build slaves for each build would be good, differences in gcc version, linkers and other factors can cause build failures on one machine and not others

[19:22:02]

jya:

pproblem is that I can't for the life of me explain why my version does it

[19:22:32]

jya:

av_estimate_timings is defined in libavformat/avformat.h

[19:22:44]

jya:

which is included in thumbfinder.h

[19:25:39]

jya:

oh well, I'm sure someone will find out why

[19:27:55]

stuartm:

it's odd ... kdevelop seems to think that it's undeclared too, but av_find_stream_info() and av_close_input_file() are fine

[19:30:05]

stuartm:

av_estimate_timings() is also used in mythmusic, is mythmusic building for you?

[19:30:13]

jya:

nope

[19:30:56]

Beirdo:

stuartm: yeah, I think so. I still need to read up on how in particular that's configured

[19:33:56]

stuartm:

Beirdo: wish I had some spare hardware I could offer but I scaled back and threw out what I couldn't justify keeping

[19:34:26]

iamlindoro:

stuartm: I'll work on getting the 2 button dialogs converted over in the next week or two-- I am not sure if I'll find the time to work on the three-view MythGame before Christmas but it's still on my list

[19:34:54]

stuartm:

iamlindoro: ok thanks, no hurry

[19:35:19]

iamlindoro:

np

[19:35:31]

stuartm:

I was hopeing to get everything ported for 0.25, but if we miss that deadline it won't be the end of the world

[19:35:42]

jya:

stuartm: it's a dumb warning about settings6

[19:35:51]

jya:

it can only ever be used

[19:35:52]

iamlindoro:

I'm eager to see it made consistent/more capable and it would be a fun project, so I think I'll get to it before .25

[19:36:04]

stuartm:

and when I say everything, I mean plugins and elements of the frontend which are still using the old stuff

[19:37:54]

stuartm:

jya: yeah looks that way, gcc bug then

[19:38:04]

stuartm:

still it's easy to silence it

[19:38:18]

jya:

it is

[19:38:29]

jya:

but it breaks my usual code style

[19:38:32]

jya:

declare ..

[19:38:37]

stuartm:

jya: assign null?

[19:38:41]

jya:

add it all at once

[19:38:51]

jya:

nah i'm going to move everything in the if (mpcm)

[19:39:09]

jya:

when I clicked on the buildbot

[19:39:14]

jya:

for the 136 warnings

[19:39:20]

jya:

i didn't see those two

[19:39:27]

jya:

where were they showing?

[19:39:39]

stuartm:

right that's what I thought you meant when you said "breaks my usual code style", figured assigning NULL would avoid reformatting

[19:40:33]

jya:

stuarm: it's 6:40AM here... i'm a tad ... slow

[19:40:50]

jya:

and i have to get up in 1h

[19:40:56]

stuartm:

jya: http://code.mythtv.org/buildbot/builders/mast . . . ings%20(136) – about halfway down the page

[19:41:12]

jya:

no such child resource

[19:41:38]

stuartm:

audiogeneralsettings warnings are just after the filter errors

[19:41:57]

stuartm:

try the text version – http://code.mythtv.org/buildbot/builders/mast . . . 20(136)/text

[19:43:00]

stuartm:

jya: get some sleep, I'll fix them

[19:43:10]

jya:

let me commit the two

[19:43:21]

jya:

damn, my youngest is up

[19:43:46]

jya:

if I got to bed now, my wife will have some doubts and believe i went to bed, just to avoid getting to him

[19:44:39]

jya:

done

[19:48:47]

stuartm:

warnings (131) << Thanks

[20:00:20]

stuartm:

heh, pretty much all of the mythplugin warnings are in mythmusic

[20:01:04]

stuartm:

does anyone know what replaces "int avcodec_decode_video(AVCodecContext*, AVFrame*, int*, const uint8_t*, int)"?

[20:01:55]

stuartm:

n/m – http://cekirdek.uludag.org.tr/~ismail/ffmpeg- . . . recated.html

[20:12:22]

jya:

stuartm

[20:12:35]

jya:

it's usually pretty simple to replace a deprecated call

[20:12:46]

jya:

become avcodec_decode_video3

[20:13:04]

jya:

though you have now to construct a "packet"

[20:13:32]

stuartm:

yup, it's covered in that link I posted, which I only discovered after asking :)

[20:13:55]

jya:

check in libavformat

[20:14:03]

jya:

it has some example on ow to build the packet

[20:14:08]

stuartm:

thanks

[20:14:22]

Beirdo:

jya: did you want me to add a test compile of the newaudio branch as well?

[20:14:41]

jya:

Beirdo: no, I've merged

[20:15:01]

Beirdo:

oh, it's in master now?

[20:15:06]

Beirdo:

I missed that one :)

[20:15:08]

jya:

yes

[20:15:13]

stuartm:

Beirdo: how?

[20:15:22]

jya:

didn't you see the 1000s of commit logs ?

[20:15:45]

Beirdo:

I saw a burst, but I didn't pay attention at the time, I need to catch up on my email

[20:17:04]

stuartm:

paul-h: before I start porting the deprecated ffmpeg stuff in mytharchive, is that something you are already working on?

[20:18:36]

Beirdo:

oh wow, that merge really changed the earlier mini-merges around

[20:19:48]

jya:

I can't wait to see you merging your system branch :)

[20:20:05]

jya:

but you haven't done much merge in between, so it should be okay

[20:21:47]

stuartm:

I still think git is retarded for duplicating commits in that way, it should know that those were merged from master so when merging back again it disregards them – at least in the logs etc

[20:22:12]

kth (kth!~kth@dyndsl-080-228-190-123.ewe-ip-backbone.de) has joined #mythtv

[20:22:25]

jya:

stuartm: you just aren't thinking the right way :)

[20:22:34]

Chutt:

you're not using it right!

[20:22:45]

Beirdo:

it's smart enough not to double-merge anyways

[20:22:45]

kth (kth!~kth@dyndsl-080-228-190-123.ewe-ip-backbone.de) has quit (Client Quit)

[20:22:47]

Chutt:

or something

[20:22:52]

jya:

haven't you got it yet ? git = right, you/us = wrong

[20:22:52]

Beirdo:

or something :)

[20:23:14]

jya:

i have to say, i was *very* impressed on how it merged

[20:23:20]

jya:

amazing how easy it went

[20:23:29]

jya:

I only had like 2 conflicts in the lot

[20:23:40]

jya:

which I knew before i started

[20:23:41]

Beirdo:

how did you merge and keep the ffmpeg stuff out?

[20:23:58]

jya:

:) that's where my great skill got involved

[20:24:03]

jya:

at first I merge in master

[20:24:05]

jya:

then rebase

[20:24:17]

jya:

but all the SHA got changed, including others than mine

[20:24:46]

jya:

so what I did is rebase newaudiosetup in a private branch, remove the ffmpeg one

[20:24:51]

jya:

then merge that branch instead

[20:24:59]

Beirdo:

ahhh

[20:25:11]

Beirdo:

that's likely why it ended up a wee bit messy :)

[20:25:15]

Beirdo:

but cool

[20:25:29]

jya:

the problem is that you don't see that until you push

[20:25:39]

jya:

my local logs were all clean

[20:25:45]

jya:

so I was very happy to find a way around

[20:26:06]

jya:

cause I'm sure I would have gotten even more grief from people if I had rebased and push

[20:26:11]

jya:

master

[20:26:22]

Beirdo:

I think the mythsystem one will be a lot cleaner of a merge :)

[20:26:41]

Beirdo:

if you rebased master, you'd have a LOT of people mad, for sure :)

[20:28:09]

Beirdo:

oooh, look at those warning counts go down. Nice.

[20:28:15]

jya:

well, not duplicate logs then

[20:28:29]

jya:

people are never satisfied for some reason

[20:28:34]

Beirdo:

hehe

[20:28:42]

Beirdo:

there is that

[20:29:34]

Kunalagon (Kunalagon!~Kunalagon@212.200.241.209) has joined #mythtv

[20:30:14]

Beirdo:

mind you, when mythsystem-rewrite is merged, there will be a burst of a lot of commits. That branch has a pile of commits on it

[20:30:24]

jya:

did you check the --mythroot patch ?

[20:30:31]

Beirdo:

anyways, it still needs testing, primarily in Windows

[20:30:53]

Beirdo:

nope, I haven't looked at any patches in the last couple days, really

[20:39:19]

jya:

it would have made your life easier for building

[20:39:22]

stuartm:

danielk22: do we care about warnings in replex? if not we can permanently disable warnings for that in the build and save the hassle of filtering them out

[20:39:34]

jya:

lets you build mythplugins without having to install mythtv first

[20:48:49]

Beirdo:

gigem: nice way to fix #9363. I can't believe the user wanted us to do that right in mythsystem... that would have been nasty :0

[20:49:24]

Beirdo:

oooh, I missed a spot

[20:50:02]

iamlindoro:

stuartm: I'll need to test tonight, but I should have all the non-MythUI stuff except for the progress dialogs done in MythGame now

[20:50:24]

stuartm:

nice

[20:50:57]

stuartm:

progress dialogs require threading if done properly, I can't remember if I started on that already

[20:51:34]

iamlindoro:

stuartm: I've found the progress bar widget to be a bit complicated in the past... the threaded stuff in it has confused me a bit

[20:51:47]

iamlindoro:

I'm sure I'll get it

[20:52:11]

Beirdo:

sorry about the missing warnings on core compile for 0.24

[20:52:16]

Beirdo:

tis fixed

[20:52:37]

gigem:

Beirdo: thanks. btw, unless i missed something, i don't think the user intended for us to changed myth_system. i think his intent was pretty much what i commited.

[20:52:38]

superm1:

xris, 2:0.25.0~master.20101214.a888720–0ubuntu0mythbuntu1 is a sample. so that's basically Epoch:upstream version ~ master.date.hash-0ubuntuversion

[20:52:53]

superm1:

can't really have more than one build a day though doing it that way, but it works well for autobuilds

[20:53:28]

Beirdo:

I guess I misread it :)

[20:53:53]

Beirdo:

yeah, tweaking around the call makes sense. Somehow I read it as he wanted it inside myth_system

[20:54:07]

stuartm:

iamlindoro: basically you want anything that might block the UI and therefore updates of the progress bar to go into the background, so a thread for scanning which posts ProgressUpdateEvents back to the UI thread to 'drive' the progress dialog

[20:54:14]

Beirdo:

which would royally mess things up. I apologize for misreading :)

[20:54:53]

iamlindoro:

stuartm: Yeah, in the case of mythgame I think that's going to take a rewrite of the scanner-- which is needed anyway

[20:54:59]

iamlindoro:

so might be a bit delayed on those

[20:55:38]

stuartm:

threading isn't strictly required with MythProgressDialog but it's very strongly encouraged, anything that calls for a progress or busy dialog will also tend to block the UI and therefore shouldn't be in the UI thread

[20:56:41]

stuartm:

repeating myself a little there :)

[20:57:03]

Beirdo:

stuartm: you may want to look over the mytharchive stuff in mythsystem-rewrite branch at some point. It is also using a progress bar, and I don't recall which way. I essentially tried to make it just work, but it may need prettifying.

[20:57:05]

iamlindoro:

I'd like the game and video scanners to behave more or less alike, ven if I think the video scanner is a bit of a pain in the ass

[20:58:41]

stuartm:

video scanner should be threaded, pretty sure I'd have done that during the port, but there are other examples around, music scan maybe ... can't remember :)

[20:58:53]

stuartm:

Beirdo: I've made a note

[20:58:56]

sphery:

xris: any thoughts on the idea of removing http://www.mythtv.org/download and, instead having it go to http://www.mythtv.org/wiki/Packages . It a) would encourage users to think of packages first, b) has a link to Source in the nav bar (which tells them to use -fixes, instead of outdated/broken tarballs), c) doesn't have outdated info/links for the MacOS and Windows packages

[20:59:31]

sphery:

Then, all we need to do is stop making tarballs for releases :) Asking because users don't like the "you will find git checkout instructions at https://github.com/MythTV" part – http://mythtv.org/pipermail/mythtv-users/2010 . . . /305505.html

[20:59:34]

Beirdo:

thanks :)

[21:00:03]

iamlindoro:

stuartm: Video scanner is threaded

[21:00:09]

iamlindoro:

game scanner, not so much

[21:00:21]

stuartm:

I'm going to be fighting this for a while, it's not difficult to write something from the beginning to be threaded but not everyone is going to consider what needs to be threaded before submitting patches/code

[21:00:41]

iamlindoro:

thus my desire to make it behave the same as video, at least externally-- create teh scanner object, run it, the end

[21:00:46]

iamlindoro:

(and maybe catch some events)

[21:01:59]

stuartm:

iamlindoro: ok, figured you might appreciate some examples, not sure the video scanner is the best one around, the stuff I did in mythmusic? for the rip/import? screens might be easier to follow

[21:02:22]

stuartm:

might be something in mythgallery too

[21:02:29]

iamlindoro:

stuartm: Ah, no, I wasn't clear-- I know how it works, I just find myself needing to relearn it each time and it hurts :)

[21:02:37]

stuartm:

ah, heh

[21:03:45]

stuartm:

I quite like it, but it's probably more natural to see this stuff as linear

[21:04:12]

Beirdo:

hehe. mythgallery needs to be SG-ized too

[21:04:51]

Beirdo:

my list is never-ending :)

[21:05:45]

iamlindoro:

stuartm: I do understand threading quite a lot better (though still not expertly) than I did last time, so I suspect I'll have an easier time this time

[21:05:45]

stuartm:

Beirdo: I'd owe you a few beers if you get mythgallery ported to use storage groups, I'd love to see all plugins made SG friendly for 0.25 as it would bring us much closer to a truely plug and play frontend

[21:06:22]

iamlindoro:

Having done the MythVideo Image SG conversion, I could probably throw together a patch in a night or two

[21:06:34]

iamlindoro:

Not a promise to do it tonight, but if you don't get to it, I can do it fairly easily I think

[21:07:01]

Beirdo:

Cool, I'd be happy to have a hand on that :)

[21:07:32]

iamlindoro:

yeah, it's fairly straightforward when you have the hang of it-- let me get these couple dialogs ported and then I'll try to have a look

[21:28:51]

Kunalagon (Kunalagon!~Kunalagon@212.200.241.209) has quit (Quit: Leaving.)

[21:33:47]

stuartm:

so how many people have had the automated job offer from the Google bots trawling github?

[21:35:29]

high-rez:

like actual google job offers?

[21:37:21]

stuartm:

like "please send us a C.V."

[21:37:40]

high-rez:

and the email to send them to is groogle.cn ?

[21:38:43]

stuartm:

as I understand it they solicit job applications from people found via data mining/html scraping sites looking for certain keywords/experience, seems github is one of those sites

[21:39:22]

high-rez:

I think I need a github account

[21:39:31]

Feico (Feico!~Feico@sd5116475.adsl.wanadoo.nl) has quit (Quit: Toedeledokie)

[21:44:29]

natanojl (natanojl!~jonatan@c83-252-237-63.bredband.comhem.se) has quit (Ping timeout: 264 seconds)

[22:00:48]

stuartm:

kormoc: a belated happy birthday ;)

[22:01:07]

kormoc:

Thanks! :)

[22:05:17]

kormoc:

stuartm, what timing! I just got notice of winning a google tv device too! You and google share the same schedule :)

[22:06:01]

stuartm:

heh

[22:10:05]

Steve_Goodey (Steve_Goodey!~steve@host86-145-237-137.range86-145.btcentralplus.com) has quit (Remote host closed the connection)

[22:19:41]

JEDIDIAH__ (JEDIDIAH__!~jedi@cpe-76-185-72-21.tx.res.rr.com) has quit (Quit: Ex-Chat)

[22:20:41]

stuartm:

Captain_Murdoch, xris, danielk22, kormoc, GreyFoxx, stuarta et al: LinkedIn is operating a beta to create company pages/groups as a seperate entity to normal groups, I've created a MythTV company page if you'd like to formally associate yourself as an 'employee'

[22:20:55]

kormoc:

Cool

[22:22:07]

Beirdo:

cool

[22:23:05]

stuartm:

I can make it so that anyone with a mythtv.org email address can admin it but for now it's locked down to named users, occurred to me that because the mailing lists are @mythtv.org it's possible for anyone to give themselves access :)

[22:23:12]

stuartm:

well possibly

[22:23:25]

kormoc:

heh, makes sense

[22:24:19]

stuartm:

might not be possible since the lists are subscription only, but we can re-visit the issue later

[22:26:38]

stuartm:

oh that's clever, since I didn't explicitly add myself to the admin list I'm now locked out :)

[22:27:32]

stuartm:

it's a good job I added people to that list otherwise we'd be pretty stuck

[22:28:56]

kormoc:

whoops!

[22:35:13]

misto (misto!~misto@64.17.253.218) has joined #mythtv

[22:37:19]

stuartm:

bad design, but then it is a beta

[22:38:08]

stuartm:

you kinda assume as the person creating the group/company that it will add you to the list of admins

[22:40:31]

kormoc:

yeah

[22:42:27]

paul-h:

Is there a correct way to reference a trac ticket from a commit message or is it still broken?

[22:43:41]

Beirdo:

it's still not parsing them correctly (on hte trac side)

[22:44:00]

Beirdo:

and if you put #1234, github will point you at a bogus ticket in their system

[22:44:39]

Beirdo:

Ticket 1234 is likely your best bet, and we'll try to unbork the github trac plugin soon

[22:45:16]

paul-h:

Beirdo: ok thanks, I'll do that and close the ticket manually

[22:51:06]

stuartm:

sphery: you are commiting with the email address mtdean@ but your alias is currently mdean@ so github isn't able to connect up your commits with your username

[22:51:52]

stuartm:

do you want change your email alias, or just fix git to use the existing address?

[22:52:16]

sphery:

I want my git to use mdean@

[22:52:52]

sphery:

I changed it there (and see it's correct in ~/.gitconfig), but don't know why it did that one wrong

[22:53:42]

sphery:

all my others, save the first, were correct...

[22:54:29]

stuartm:

hmm, do you have a .gitconfig in the repo somewhere which is overriding the global?

[22:54:53]

sphery:

did a grep, and can't find mtdean anywhere under the repo

[22:55:12]

JEDIDIAH__ (JEDIDIAH__!~jedi@cpe-76-185-72-21.tx.res.rr.com) has joined #mythtv

[22:55:38]

stuartm:

strange stuff

[22:56:30]

stuartm:

I'll let xris, jannau or Beirdo speculate on git behaviour :)

[22:56:40]

iamlindoro:

oh MythGallery

[22:56:42]

iamlindoro:

QString blah = dir;

[22:56:42]

iamlindoro:

QDir d(blah);

[22:57:18]

sphery:

git config user.email shows the right one, too

[22:57:44]

kormoc:

sphery, if you have a global one, you can't use per repo ones

[22:57:51]

Beirdo:

you committed ONCE wrong

[22:57:59]

Beirdo:

umm, yeah you can.

[22:58:00]

kormoc:

I had to unset my global one to allow the mythtv repo one to be used correctly

[22:58:18]

Beirdo:

repo takes precedence over global

[22:58:27]

kormoc:

Beirdo, not for me, git showed both in the output and it was using the global one

[22:58:29]

** kormoc shrugs **

[22:58:36]

Beirdo:

frigging odd

[22:59:09]

sphery:

ah, the problem is that I just backported the one that used the wrong address--i.e. that was the backport of my first commit

[22:59:55]

stuartm:

Beirdo: two instances are wrong – https://github.com/MythTV/mythtv/commit/c91d869 and https://github.com/MythTV/mythtv/commit/7a644c4d8

[23:00:02]

stuartm:

sphery: ah-hah

[23:00:08]

sphery:

sorry for confusing us all, again

[23:00:19]

sphery:

the mistake that keeps on giving

[23:00:22]

Beirdo:

yeah, he had it wrong... committed... then cherry-picked the incorrect :)

[23:00:35]

stuartm:

sphery: np, sorry for jumping on it :)

[23:00:36]

Beirdo:

tis fun

[23:00:49]

Beirdo:

good to catch though. Typos happen

[23:00:55]

sphery:

no, glad you mentioned it--I didn't notice that it happened, and now I know something new about git

[23:02:39]

Beirdo:

you could have amended it before pushing... had you noticed before pushing :)

[23:02:53]

stuartm:

heh, well jya seems to think I spend all my time just waiting to catch people making a mistake, so I've done nothing to dispel that notion :)

[23:03:02]

Beirdo:

hehe

[23:03:22]

Beirdo:

someone's gotta keep us honest

[23:04:45]

stuartm:

I can legitimately be accused of procrastinating today, watching the emails roll in and playing with github/buildbot/etc instead of doing some real work

[23:04:58]

sphery:

OK, so I have a question. I'm doing the patch to move the mythvideo schema into the TV/main schema. It's about 590 lines. I'll need access to a performActualUpdate() and UpdateDBVersionNumber() function, which are both static in libs/libmythtv/dbcheck.cpp (and not in the header). So, should I put the lines into libs/libmythtv/dbcheck.cpp or ???

[23:06:43]

j-rod is now known as j-rod|afk

[23:07:25]

sphery:

I could do another file for InitializeVideoSchema() and doUpgradeVideoDatabaseSchema(), but would have to open up performActualUpdate() and UpdateDBVersionNumber() to do it that way. However, that has the added benefit of allowing us to break dbcheck.cpp up into smaller chunks--i.e. one for InitializeMythSchema() which almost never changes--which could be good for compilation speed.

[23:07:32]

stuartm:

schema should all be in the one place, I'm not sure that place is dbcheck.cpp but that's not really a discussion that needs to be had now

[23:08:21]

stuartm:

dbcheck.cpp should be broken up in the longer term, I thought we had a plan for that at one point

[23:09:20]

sphery:

I remember the idea being mentioned, but don't remember any actual plans coming from it

[23:09:56]

elmojo:

are the @mythtv.org e-mail addresses supposed to work?

[23:09:58]

sphery:

I could do an update for InitializeMythSchema(), too, that does a post-UTF-8 initial schema. Would probably be a good thing

[23:10:33]

stuartm:

sphery: could be that no concrete plans came from it

[23:10:55]

stuartm:

elmojo: your alias hasn't been setup yet

[23:11:15]

stuartm:

we need a forwarding address and consent

[23:13:12]

stuartm:

sphery: it would be good to update the initial schema to match the current status, then change the starting schema version, there is no reason for the schema version for new users not to match the latest

[23:14:18]

stuartm:

i.e. we avoid the 'updating schema' step on a fresh install because there is no existing database to be updated

[23:14:35]

sphery:

stuartm: yeah, it's just that it's kind of a pain to update InitializeMythSchema() frequently, so it hasn't been done in a long while.

[23:14:59]

stuartm:

yup

[23:15:14]

sphery:

it basically requires creating a new schema, then dumping the schema create output from mythfilldatabase, so pretty much completely replaces the SQL in the function

[23:16:00]

sphery:

at least that's probably safer than having devs edit the initial schema SQL with each schema change--which would wreak havok if new schemas and updated schemas diverged

[23:16:52]

sphery:

I'll plan to update it soon. I think I'm going to add performActualUpdate() and UpdateDBVersionNumber() to dbcheck.h, and do a videodbcheck file that includes the historic mythvideo schema/update

[23:17:28]

sphery:

then maybe split out InitializeMythSchema() when I update it

[23:20:46]

stuartm:

whatever you do will be the right thing ;)

[23:21:26]

sphery:

heh, or can easily be turned into the right thing when someone else points me in that direction :)

[23:50:31]

stuartm:

does anyone know why we get strict aliasing errors from the quickdnr filter even though strict aliasing is disabled for that code?

[23:55:31]

stuartm:

because the flag is specified in the .pro but is not being added to the Makefile

[23:56:00]

the_alien (the_alien!~till@xbmc/staff/thealien) has joined #mythtv

MythLogBot@irc.freenode.net :: #mythtv

Current users (104):