:: #mythtv

Daily chat history

Current users (88):

aloril, Anssi, anykey_, Beirdo, brfransen, brtb, CaCtus491, cattelan_away, cesman, Chutt, clever, coling, Cougar, damaltor, danielk22, Dave123, davide, dekarl, dlblog, ElmerFudd, foobum, foxbuntu, ghoti, gregL, GreyFoxx, Guest76506, hank, highzeth, J-e-f-f-A, j-rod|afk, jams, jcarlos, joe_, joki, jpabq, jpabq-, jpabq_, jstenback, justinh, jwhite, jya, k-man, kenni, knightr, kurre2, kwmonroe, laga, lolcat`, mag0o, Malard, mrand, MythBuild, MythLogBot, mzanetti, natanojl, noahric, peitolm, pheld, poptix, purserj, rhpot1991, rsiebert, skd5aner, Slasher`, sphery, sraue, stuarta, stuartm, superm1, sutula, tgm4883, ThisNewGuy, tomimo, tris, Unhelpful, vallor, wagnerrp, wahrhaft, wseltzer, xavierh, xavierh_, XChatMav, XDS2010_, xris, ybot_, zCougar, zombor, _charly_
Saturday, March 3rd, 2012, 00:00 UTC
[00:00:58] wookey_: hello people, mythtv won't tune 'Dave' (UK, sandy-heath), even after a full rescan
[00:01:09] stuartm: UX is more than skin deep :) It's lots of little niggling issues, that everyone learns to live with but which still cause irritation e.g. upnp returning data in the wrong fields or failing auto-discovery, mythcommflag cutting too early or late on ad breaks
[00:01:09] wookey_: It found the channel, with correct ID
[00:01:36] wookey_: but it still doesn;t seem to work. Other channels are OK. Guess I should check other channels on same mux
[00:01:59] stuartm: wookey_: working for me on Waltham
[00:02:20] wookey_: OK. so it does still exist then at least so it must be me..
[00:03:13] wookey_: I seem to have cruft in my database that I can't delete (like dave-ja-vu which isn;t listed here anymore
[00:03:34] wookey_: and I have about 10 transports, not just the 6 that still exist. I can;t work out how to get rid of the dead ones
[00:03:45] stuartm: wookey_: probably best to discuss this in #mythtv-users
[00:03:56] wookey_: ah. OK. this the dev channel?
[00:04:00] stuartm: yes
[00:04:14] wookey_: sorry for the noise
[00:10:51] noahric (noahric!~noahric@ has joined #mythtv
[00:11:59] Beirdo: stuartm: yeah, gotcha. I'm not a UI person, but I can certainly be useful under the covers :)
[00:13:29] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has quit (Quit: jpabq_)
[00:16:01] zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[00:23:29] wookey_ (wookey_! has left #mythtv ()
[00:24:03] kormoc (kormoc!~kormoc@mythtv/developer/kormoc) has quit (Quit: kormoc)
[00:35:46] xavierh_: Is thee a way to know wich object emit the signal ?
[00:35:58] xavierh_: s/thee/there
[00:36:34] xavierh_: I believed I read this somewhere but can't remember it, neither find it in the doc
[00:37:17] wagnerrp: yeah, there is... give me a second
[00:38:00] xavierh_: would be great so I would not have to stupidely pass it as a parameter
[00:38:09] wagnerrp: sender()
[00:38:17] wagnerrp: its just a generic QObject
[00:38:30] wagnerrp: so youll need to know what its supposed to be ahead of time, and cast it as such
[00:38:46] xavierh_: ha, big thx
[00:40:07] wagnerrp:
[00:43:00] gigem (gigem!~gigem@mythtv/developer/gigem) has quit ()
[00:44:28] gigem (gigem! has joined #mythtv
[00:44:28] gigem (gigem!~gigem@mythtv/developer/gigem) has joined #mythtv
[00:44:28] gigem (gigem! has quit (Changing host)
[00:53:47] OldEnK (OldEnK! has left #mythtv ("Leaving")
[00:57:39] stichnot (stichnot!~chatzilla@mythtv/developer/stichnot) has quit (Ping timeout: 245 seconds)
[01:02:41] noahric (noahric!~noahric@ has quit (Quit: noahric)
[01:32:41] warped (warped!~piotro@ has quit (Quit: warped)
[01:58:04] noahric (noahric!~noahric@ has joined #mythtv
[01:58:11] jpabq: LiveTV with the hd-pvr in master really is quite broken. I will see what I can figure out this weekend.
[02:04:42] davide (davide! has quit (Remote host closed the connection)
[02:05:04] davide (davide! has joined #mythtv
[02:08:16] zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv
[02:12:31] zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[02:32:21] stichnot (stichnot!chatzilla@mythtv/developer/stichnot) has joined #mythtv
[02:45:09] danielk22: jpabq: I'll mostly be doing taxes and volunteer stuff this weekend, but if you want to fire questions my way I'll try to reply in a reasonable amount of time..
[02:47:14] jpabq: danielk22, Cool. It may not be as broken as I first thought. I have been having trouble with one of my HD-PVRs, and *it* really has trouble with LiveTV. They other one I tried actually seems to work okay — although it will eventually fall back to the menu on a channel change, so is not perfect either.
[02:53:30] danielk22: jpabq: LiveTV has regressed for all recorders when there is a problem with the tuning. The UI become completely unresponsive. I believe it has something to do with the MythUI conversion of the OSD where everything was placed in the UI thread.
[02:54:06] jpabq: That seems likely.
[02:54:31] danielk22: jpabq: I wasn't really planning on even looking at that for this release as I think it will require a bit rethink of playback to really get that to work properly...
[02:55:40] danielk22: but maybe there is something that can be done in this particular case, my assumption is that we're blocking somewhere waiting for the 'other' thread to do something, but there is no other thread so we block until playback times out completely.
[02:56:59] danielk22: If you find out where we are blocking then we can figure out what to do about it...
[02:57:04] jpabq: danielk22, Actually, with my "good" HD-PVR, LiveTV is working better than it did the last time I tried it a few months ago.
[02:57:42] danielk22: that's good to hear :)
[02:58:58] jpabq: I am going to try the latest firmware on my "bad" hdpvr, and see it if works any better. But that will have to wait until tomorrow.
[03:41:54] cattelan is now known as cattelan_away
[03:42:00] saintdev (saintdev!~saint@unaffiliated/saintdev) has quit (Read error: Connection reset by peer)
[03:43:41] stichnot: danielk22: I just posted #10401, one of the patches I've been using for a long time. You're probably best qualified to decide whether it's an actual bug.
[04:01:25] saintdev (saintdev!~saint@unaffiliated/saintdev) has joined #mythtv
[04:11:22] jya: danielk22: I have noticed that if there is a tuning issue… You'll have to wait over 40s before returning to the main menu...
[04:11:47] jya: I see in the log to series of waiting 1s for blah (20 times)
[04:12:11] jya: and then another count of waiting 1s for blah to exit (20 times)
[04:13:41] jya: I have a DVB-T adapter that is not as good as the other, and now with the new liveTV card order, it's the first one that is tried by default. Often, it times breaks in the middle, and myth fail… going back to the menu does take a frustratingly long time
[04:48:22] lolcat` (lolcat`! has quit (Ping timeout: 252 seconds)
[04:49:08] lolcat` (lolcat`! has joined #mythtv
[05:05:00] saintdev (saintdev!~saint@unaffiliated/saintdev) has quit (Read error: No route to host)
[05:06:14] Rasperin (Rasperin!~ray@ has left #mythtv ()
[05:06:18] saintdev (saintdev! has joined #mythtv
[05:06:18] saintdev (saintdev! has quit (Changing host)
[05:06:19] saintdev (saintdev!~saint@unaffiliated/saintdev) has joined #mythtv
[06:29:05] jya: mythplugins' configure script is fundamentally broken, and many of the plugins aren't included on the mac for that reason. To check if a feature is available, it only looks in a static list of folders. If using anything different to those folders, it will just fail… going to work on fixing that...
[06:50:49] vallor (vallor!~Ponzo@pdpc/supporter/monthlygold/vallor) has quit (Read error: Connection reset by peer)
[06:50:56] vallor (vallor!~Ponzo@pdpc/supporter/monthlygold/vallor) has joined #mythtv
[07:01:55] Goga777 (Goga777! has joined #mythtv
[07:14:46] xris: stuartm: you had to just go and make me figure out how to make my schedule plan readable.
[07:28:18] noahric (noahric!~noahric@ has quit (Quit: noahric)
[07:39:45] stoffel (stoffel! has joined #mythtv
[07:58:11] wagnerrp: Beirdo or xris: around?
[07:58:32] Beirdo: yup
[07:58:46] xris: yup
[07:58:59] Beirdo: still stress testing my fixes to mythtranscode
[07:59:26] wagnerrp: im using this git-subtree thing
[08:00:05] wagnerrp: i extracted a copy of the tmdbv3 stuff into
[08:00:25] wagnerrp: but when pulling the remote tree in, its duplicating all the old commits
[08:00:47] Beirdo: pulling what remote tree?
[08:00:59] wagnerrp: the above link
[08:01:09] Beirdo: you would have two, I would assume, to be able to create this
[08:01:29] wagnerrp: im guessing theres no way around that without fubaring the tree with a rebase?
[08:02:12] Beirdo: Hmm, I'm not quite sure what you're trying to do. After pushing the extracted subtree, did you reclone from that?
[08:02:17] xris: git-subtree is over my head
[08:03:00] wagnerrp: its not exactly a clone
[08:03:11] Beirdo: yah, I've never tried it myself.
[08:03:34] wagnerrp: its not using submodules, so there is no need to independently pull it when you clone a new copy of the repo
[08:04:11] wagnerrp: i guess since theyre independent, theyre going to create independent hashes, and duplicate commits
[08:04:15] wagnerrp: theres just no way around it
[08:05:10] Beirdo: dunno, doesn't sound right to me
[08:06:44] Beirdo: I think to pull a subtree, you need to do a git pull -s subtree
[08:07:38] Beirdo: I've never bothered with it, too many other things to keep my brain full and all
[08:07:41] wagnerrp: git subtree pull --prefix=<path>, rather
[08:07:58] wagnerrp: which is what i did, and generated about a dozen duplicates
[08:08:18] wagnerrp: im thinking it doesnt seem to have altered any previous hashes, so it hasnt done anything bad
[08:08:28] wagnerrp: just live with the handful of duplicates
[08:08:31] Beirdo: met the guy who wrote that support
[08:13:06] Beirdo: Unfortunately, using this option results in 'git log' showing an extra copy of every new commit that was created (the original, and the synthetic one).
[08:13:35] Beirdo: that's in the docs for --rejoin on the git subtree split
[08:14:06] wagnerrp: i didnt use rejoin specifically because it stated that
[08:14:38] wagnerrp: or should i have?
[08:15:03] Beirdo: Yeah,you might wanna email apenwarr and get some clarification
[08:15:44] Beirdo: not sure. It's not standard git to start with, not that all of git is crystal clear to understand on the surface anyways
[08:16:08] wagnerrp: how do i back out to current master:HEAD?
[08:17:07] Beirdo: generally? git reset does taht
[08:17:20] Beirdo: you mean to origin's master:HEAD?
[08:17:25] wagnerrp: yes
[08:17:35] wagnerrp: dump those 12 commits on the merge
[08:17:41] Beirdo: git reset origin/master I think
[08:18:06] wagnerrp: that did it, thanks
[08:18:13] Beirdo: no problemo :)
[08:18:33] Beirdo: heh, I love his description
[08:18:44] Beirdo: An experimental alternative to the git-submodule command.
[08:22:04] wagnerrp: lets try this a second time...
[08:42:33] Goga777 (Goga777! has quit (Remote host closed the connection)
[08:55:27] warped (warped!~piotro@ has joined #mythtv
[08:59:36] warped (warped!~piotro@ has quit (Client Quit)
[09:14:41] xris: there. pretty.
[09:19:37] warped (warped!~piotro@ has joined #mythtv
[09:52:18] stoffel (stoffel! has quit (Ping timeout: 272 seconds)
[10:24:36] natanojl (natanojl! has joined #mythtv
[10:32:36] Goga777 (Goga777! has joined #mythtv
[10:35:45] warped (warped!~piotro@ has quit (Quit: warped)
[11:05:02] mike|2 (mike|2! has quit (Remote host closed the connection)
[11:05:48] mike (mike! has joined #mythtv
[11:06:14] mike is now known as Guest76506
[11:45:40] jeff999 (jeff999! has joined #mythtv
[11:46:12] jeff999 (jeff999! has quit (Client Quit)
[12:10:19] Goga777 (Goga777! has quit (Remote host closed the connection)
[12:16:26] knightr (knightr!~knightr@mythtv/developer/knightr) has quit (Read error: Connection reset by peer)
[12:16:41] knightr (knightr!~knightr@mythtv/developer/knightr) has joined #mythtv
[12:17:32] stuartm: some sage advice here about the need to design simple UIs for TVs – . . . _their_sets/
[12:17:56] stuartm: of course he doesn't tell you how that's done ...
[12:55:02] Seeker` (Seeker`!~cjo20@unaffiliated/seeker) has quit (Remote host closed the connection)
[13:04:33] davide (davide! has quit (Remote host closed the connection)
[13:04:54] davide (davide! has joined #mythtv
[13:19:11] Seeker` (Seeker`!~cjo20@unaffiliated/seeker) has joined #mythtv
[13:29:16] knightr (knightr!~knightr@mythtv/developer/knightr) has quit (Read error: Connection reset by peer)
[13:29:39] knightr (knightr!~knightr@mythtv/developer/knightr) has joined #mythtv
[13:35:27] xavierh_: xris: are ..26 and 0.28 plan for this year too?
[13:35:46] xavierh_: xris: are 0.26 and 0.27 plan for this year too?
[13:36:45] stuartm: xavierh_: no, that's an example of one proposed workflow not an agreed schedule
[13:37:14] stuartm: xavierh_: 0.26 should be this year, we're still arguing over when in the year
[14:09:53] aloril (aloril! has quit (Ping timeout: 255 seconds)
[14:23:32] aloril (aloril!~aloril@ has joined #mythtv
[14:27:50] warped (warped!~piotro@ has joined #mythtv
[14:38:59] warped (warped!~piotro@ has quit (Quit: warped)
[15:09:40] jarle (jarle! has quit (Remote host closed the connection)
[15:37:13] danielk22: stichnot: Sounds like a bug. But I don't really know the DVD code at all. I don't use MythTV for DVD's except for a couple Potty DVD's that have now mostly outlived their usefulness (and I had to patch avlib for one of those to work.)
[15:39:45] xavierh_: stuartm: Is there a way to get a reference of the current dialog ?
[15:41:58] danielk22: jya: Yeah, I don't doubt there are bugs in the playback code. I get pauses during playback, the VDPAU and OpenGL video renders don't resize when they should, I don't always see the tuning OSD when I should.
[15:44:31] jya: I think I;m hitting some more annoying bug with llvm-g++.. I can see that it is started with -I../libmythbase, yet #include "util.h" fail, when #include "../libmythbase/util.h" works.. no error about the file missing, just that it gives me error about undeclared symbols
[15:44:47] skd5aner: what is a "potty" dvd?
[15:44:55] skd5aner: or do I even want to know
[15:45:12] jya: harry potter ?
[15:45:33] jya: though danielk22's kid is way too young… or way too advanced
[15:45:54] jarle (jarle! has joined #mythtv
[15:46:45] skd5aner: I take my dog out to potty, but I sure don't film it and put it on DVD
[15:54:16] jya: seems to be related to the option -isysroot ; when I have it, some of the -I../to/path are effectively disabled
[15:56:37] danielk22: skd5aner: Two year olds don't believe anything you tell them. You need to show them a video or a book that says the same thing and then they believe it. Hence the "Potty Video" industry and a book called "Everyone Poops"
[15:57:13] skd5aner: yea, I've heard of the book before... I've got a 9 month old, not quite there yet
[15:57:18] skd5aner: :)
[15:57:18] markk (markk! has joined #mythtv
[15:57:41] markk (markk! has left #mythtv ()
[16:00:09] stichnot: danielk22: yeah, I find that book really disturbing. thank goodness no one in the house knew of a DVD...
[16:04:14] stichnot: danielk22: I suspect the DVD navigation code is there because of a bug/limitation in an older version of dvdnav. If you have no objection, I'll commit it.
[16:09:16] zombor_ (zombor_!~zombor_@kohana/developer/zombor) has joined #mythtv
[16:11:42] rsiebert (rsiebert! has joined #mythtv
[16:11:43] rsiebert_ (rsiebert_! has quit (Read error: Connection reset by peer)
[16:19:14] danielk22: stichnot: no objection here
[16:21:47] stuartm: stichnot: if there's no clear reason for the code, it's clearly breaking normal functionality and no-one bothered to add in-line comments explaining it, then ripping it out would always be my first instinct
[16:23:29] stuartm: conversely if you're adding code which might need explanation then add comments explaining in detail why it was necessary so that it doesn't get reverted in the future :)
[16:23:31] jya: anyone? . . . header-files
[16:24:42] zombor_ (zombor_!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[16:26:20] danielk22: jya: In trunk myth_system is not defined in util.h it's defined in mythsystem.h
[16:26:35] jya: yeah… that's just an example :)
[16:26:47] jya: actually I have exactly the same issue with another header
[16:27:22] jya: so the issue for mythmediamonitor is "mythsystem.h" vs "../libmythbase/mythsystem.h"
[16:27:33] danielk22: jya: My guess would be the symbol visibility stuff.
[16:27:45] jya: the first one doesn't work, while the 2nd does, even though there is -I../libmythbase
[16:28:09] jya: if I remove the -isysroot then there's no problem
[16:28:26] jya: but in which case, what I compile on OS 10.7, only works on 10.7
[16:29:19] danielk22: maybe there is header file name aliasing? i.e. OSX has a header called "util.h" which is included by mythsystem.h and without the ../libmythbase it pulls in the wrong header?
[16:29:51] jya: I have another file that gives me an error about symbols found in util.h , and that's one of the enum of util.h
[16:30:12] jya: danielk22: ahhh.. that would make sense...
[16:30:25] jya: util.h -> mythutil.h ?
[16:30:50] jya: ah ah !:
[16:30:55] jya: /usr/include/util.h
[16:30:57] danielk22: jya: I've been meaning to do that rename for a while...
[16:31:19] jya: why didn't I think of this earlier ?
[16:31:55] danielk22: stuartm: stichnot: Well there is one place we should look.. the commit message. git blame should find out when and why that was committed...
[16:32:09] stichnot: danielk22: good point, I'll check
[16:32:20] jya: funny on how you can bang your head for a while, only to find that the solution is very obvious
[16:32:28] stuartm: danielk22: if the original commit hasn't been obliterated by a refactor
[16:32:55] jya: may I rename util.h into mythutil.h and update all the files calling it ?
[16:32:58] stuartm: which is why I'd still advocate in-line comments where they might be necessary
[16:37:05] danielk22: stuartm: It is sometimes hard to know if a comment will be necessary. I try to put them around hacks like this, but sometimes even code that isn't a hack gets "rebroken" after a fix and only when fixed again earns a comment.
[16:37:12] stichnot: interesting ... the conditional used to be commented out, then it was uncommented in . . . fa0e#L50L744
[16:37:20] stichnot: so I need to look back further
[16:39:40] danielk22: stichnot: It looks like this was broken in a bad 'resync'; go ahead and fix it.
[16:41:07] stichnot: ok, and I'll leave some comments to document the confusion
[16:50:49] jya: for i in `find . -name "*.cpp" -exec grep -q \"util.h\" {} \; -print`; do echo $i; sed -i -e 's/"util.h"/"mythutil.h"/g' $i ; done
[16:53:01] danielk22: jya: look at the plugins too.. make sure you delete any old util.h in /usr/{local/,}include.
[16:54:19] jya: I ran it on .h, .pro… manually adapting now
[16:54:39] jya: I'm guessing it would error where it hasn't been updated. I recompile from 0 on the mac
[16:55:11] jya: zmclient.cpp
[16:55:38] danielk22: jya: yeah, it will error if the old util.h is gone.. but it isn't automatically deleted. You might want to do the work on a Linux system for faster compiles with ccache and distcc (I'm assuming the Mac is slow, but if it isn't ignore this.)
[16:56:22] jya: actually, my mac is a good 3 times faster than my linux box (i7 quad 3.4Ghz vs i5 2.6
[16:56:29] jya: also use ccache
[16:57:02] jya: 1m58 for a clean build now (that's with the new build script)
[16:57:48] jya:
[17:00:47] jya: I get tons of those warning when compiling with -isysroot
[17:00:48] jya:
[17:01:18] jya: not on every file.. but close
[17:01:42] noahric (noahric!~noahric@ has joined #mythtv
[17:03:33] jya: ah! in mythfrontend/progdetails.cpp, it was even doing:
[17:03:35] jya: #include <util.h>
[17:04:43] ponyofdeath (ponyofdeath! has left #mythtv ()
[17:05:43] zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv
[17:06:03] jya: actually, about twenty are using the <..> notation instead
[17:08:34] Seeker`: Were there any forced setting changes recently? Just upgraded to the latest mythbuntu packages, and it set the required pin to blank (instead of 0000)
[17:09:16] jya: Re Captain_Murdoch email.. I feel like I'm back at work !
[17:09:40] stuartm: Seeker`: just the MaxDeletedAge or whatever it was called
[17:10:17] Seeker`: stuartm: I upgraded my packages, then it wouldn't connect to the backend, with no obvious indication that i'd have to go in to mythtv-setup and set this pin
[17:12:49] stuartm: Seeker`: no idea what might have caused that
[17:13:16] stuartm: although the frontend log should be indicating the reason for the failure
[17:18:23] stuartm: if it doesn't that needs fixing
[17:26:40] Seeker`: Either the server is down or the master server settings
[17:26:41] Seeker`: in mythtv-settings does not contain the proper IP address
[17:27:10] Seeker`: I looked at the IP setting, but managed to completely overlook the PIN setting
[17:31:33] frederickjh (frederickjh!~frederick@rivendell/member/frederickjh) has joined #mythtv
[17:35:34] frederickjh: I have the latest Mythbuntu installed and working. My multimedia keys on the keyboard and my remote( that looks like a keyboard to the computer) when the MythTV Frontend is NOT running. However when I start the MythTV Frontend the multimedia volume keys(Volume Up, Volume Down, Volume Mute) no longer work.
[17:35:54] frederickjh: Any one seen this before or know what to do to configure it?
[17:36:30] frederickjh: F9, F10, and F11 do work for these functions.
[17:40:46] zombor (zombor!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[17:47:05] Captain_Murdoch: jya, I think you are referring to xris. I'm Pinkham.
[17:48:17] stichnot: stuartm: re #9593, isn't it a little cruel to assign me something for 0.25 that even markk couldn't figure out? :)
[17:48:54] stichnot: (just kidding, in case it isn't clear...)
[17:52:13] jpabq: stuartm, have you tried the channel icon in the OSD on a remote frontend? On my remote frontend, the channel icon shows fine in the guide, but does not show in the OSD. If I run the frontend on the same machine as the backend, then the icon shows in the OSD.
[17:53:59] warped (warped!~piotro@ has joined #mythtv
[17:58:46] natanojl (natanojl! has quit (Ping timeout: 276 seconds)
[18:00:44] stuartm: jpabq: that's a really old bug, the frontend will only load icons hosted locally, it won't pull them in over the protocol
[18:01:49] warped (warped!~piotro@ has quit (Quit: warped)
[18:02:07] stuartm: I tried a fix a few years back that made it use the same icon-over-protocol code as the guide, but at the time that caused so much latency in video streaming so that using browse mode caused the video to stutter
[18:02:36] stuartm: it might be worth re-visiting and finally fixing that for 0.25
[18:05:57] davide (davide! has quit (Remote host closed the connection)
[18:06:21] davide (davide! has joined #mythtv
[18:10:23] natanojl (natanojl! has joined #mythtv
[18:22:14] stuartm: hmm, a clean fix prior to 0.25 might not be do-able, maybe best left as a goal for 0.26
[18:22:26] stuartm: we can create a channel icon storage group and leverage that common code
[18:26:27] stuartm: in fact ... creating a storage group for 0.25 might be the simplest and cleanest fix, MythUIImage can load direct from a storage group, channel icons are in a know location on existing installs, so a new channel icon storage group pointing at ~/.mythtv/channels/ and we'd have a working solution with minimal changes to existing code
[18:28:49] sphery: fwiw, the other benefit of doing that for 0.25 is that we've been saying we're going to do it since 0.22 and haven't yet done it :)
[18:29:09] sphery: (the channel icon storage group that is)
[18:31:41] stuartm: yeah, as I understand storage groups it's almost as simple as inserting a new row into the storagegroup table and ensuring that we prepend myth://channelicon/ or similar to the beginning of the icon url we get from the database (or a db update to add the prefix)
[18:33:13] stuartm: can we create an uneditable storage group? If we allow the location to be changed then the icon downloader and mythfilldatabase would need changing, they have the location hardcoded atm
[18:34:32] zombor_ (zombor_!~zombor_@kohana/developer/zombor) has joined #mythtv
[18:36:28] sphery: stuartm: the "uneditable" storage groups are the "built-in" storage groups
[18:37:17] stuartm: sphery: ok cool, I didn't know whether such a thing existed, let's just say that I haven't been paying enough attention ;)
[18:37:18] sphery: no db updates for those... And it seems we have a builtin ChannelIcons group... . . . roup.cpp#L68
[18:37:44] sphery: so maybe just converting the OSD code to use that?
[18:37:47] stuartm: sphery: even better
[18:38:01] stuartm: yeah, if I can figure out the url format
[18:39:10] sphery: stuartm: seems it came from
[18:39:12] stuartm: sphery: although that's pointing that the wrong directory for existing users, so either we change it; require users to redownload/move icons or do a one-off mv ourselves
[18:40:14] stuartm: Captain_Murdoch: any thoughts on that last bit?
[18:41:38] sphery: FWIW, if you break the HTML setup stuff, it won't affect 0.25 (since that's all disabled :)
[18:43:24] jarle (jarle! has quit (Remote host closed the connection)
[19:06:08] zombor_ (zombor_!~zombor_@kohana/developer/zombor) has quit (Remote host closed the connection)
[19:10:07] frederickjh (frederickjh!~frederick@rivendell/member/frederickjh) has left #mythtv ("Leaving.")
[19:27:50] noahric (noahric!~noahric@ has quit (Quit: noahric)
[19:28:19] RogerM (RogerM!~chatzilla@ has joined #mythtv
[19:28:28] taylorr (taylorr! has joined #mythtv
[19:28:28] taylorr (taylorr! has quit (Changing host)
[19:28:28] taylorr (taylorr!~taylorr@unaffiliated/elmojo) has joined #mythtv
[19:32:07] taylorr: jpabq, jpabq-: fyi, the hdpvr color control patch made it into 3.3rc5 and is now is the media_tree git repo..... so if you re-run for the media_build repo you should have my changes
[19:33:47] warped (warped!~piotro@ has joined #mythtv
[19:39:39] stuartm: ok, well it works in local tests, I don't have a remote frontend running master available right now, I'm a little concerned about performance when scrolling through a large number of channels in browse mode :/
[19:40:13] stuartm: Captain_Murdoch: we appear to create a new socket for every file transferred from a storage group, wouldn't it be better to share a socket?
[19:41:04] stuartm: my frontend logs look like this –
[19:46:50] sphery: stuartm: doesn't it get written to the local image cache? ... though I guess it could rotate out. Maybe we could have the caching code cache channel icons separately?
[19:47:30] stuartm: the disk cache has no expiry mechanism, once something is in there it stays in there
[19:47:57] stuartm: so yes, once it's cached for a theme it reduces the load
[19:48:22] stuartm: although we still open sockets to do last modified checks against the remote copy
[19:48:41] sphery: ah, I see.
[19:48:55] stuartm: that last bit might be too aggressive, I know it is for http images
[19:50:32] stuartm: fwiw there is separate channel icon caching, under {config_dir}/RemoteCache or similar, I added that some time back to speed up scrolling in the guide, but this change to storage groups makes that redundant
[19:51:02] stuartm: the image cache should in theory be sufficient, it works well enough for fanart etc
[19:55:43] stuartm: OT, anyone know how to suppress warnings from ld about incompatible libs? In trying to find some requirements it's looking at 32bit libs and complaining loudly
[20:13:51] warped (warped!~piotro@ has quit (Quit: warped)
[20:22:54] wagnerrp: stuartm: yes, one open file means one open socket
[20:23:22] ** stuartm takes a stand on 'preserve' and prepares to take a beating **
[20:23:40] stuartm: stuarta: avenge me when I'm gone!
[20:23:40] wagnerrp: are you suggesting some manner of reusing sockets, recently made available by the ability to change remote files, for use by HLS?
[20:23:58] wagnerrp: or some mechanism to multiplex multiple file streams into a single socket?
[20:25:20] wagnerrp:
[20:25:51] stuartm: wagnerrp: well I was thinking a pool of open sockets, sockets staying open and connected for a period after their use and remaining available for any subsequent transfers made within the timeout period
[20:26:24] stuartm: but whatever works I guess
[20:26:58] xavierh_ (xavierh_! has quit (Ping timeout: 276 seconds)
[20:27:05] wagnerrp: the above commit is maybe a quarter of the way there
[20:28:40] wagnerrp: i wonder how much different it actually makes though
[20:28:53] wagnerrp: how expensive is it actually to start up a new connection?
[20:29:14] stuartm: probably not as expensive as I think
[20:29:54] stuartm: which isn't that expensive to begin with, but for dozens of back to back transfers I figured that overhead would add up
[20:31:16] xavierh_ (xavierh_! has joined #mythtv
[20:31:18] wagnerrp: is that artwork in PBB?
[20:31:48] stuartm: channel icons in the guide and browse mode
[20:32:03] stuartm: but it would be artwork in mythvideo or pbb
[20:32:31] wagnerrp: odd that it would be opening two connections like that in rapid succession
[20:33:44] sphery: So something like the HTTP Keep-Alive
[20:36:19] stuartm: jpabq: can you test
[20:36:31] stuartm: sphery: yeah
[20:37:56] stuartm: wagnerrp: why odd? It's opening a connection for every remote file currently being displayed, that's up to four different bits of artwork and a channel icon for each channel/recording, then when you factor in scrolling through the list ...
[20:39:11] wagnerrp: stuartm: the channel info can now display more than just the channel icon?
[20:39:19] xavierh_ (xavierh_! has quit (Ping timeout: 276 seconds)
[20:39:43] aloril (aloril!~aloril@ has quit (Ping timeout: 246 seconds)
[20:40:14] wagnerrp: youve got two new connections opening within a few ms of each other, so thats not you scrolling through the list
[20:40:35] wagnerrp: the frontend uses the persistent connection to the frontend to check that the channel icon exists
[20:40:47] wagnerrp: and then opens a new connection to actually download it
[20:40:48] sphery: stuartm: I love your idea on removing preserve--great from a "principle of least surprise" standpoint (as a /lot/ of users think--and say on list--that disabling autoexpire will keep recordings from being deleted on Max Episodes rules). And, block deletion is a great idea--and useful all around.
[20:41:06] sphery: count me as all for it
[20:41:34] wagnerrp: stuartm: im just curious what would be causing the second new connection
[20:41:50] sphery: based on his commits, it sounds like he has posters and fan art in his osds
[20:42:15] wagnerrp: s/fan art/banners/?
[20:42:27] wagnerrp: 'fanart' is the full screen backdrops
[20:44:35] sphery: seems coverart and poster, specifically... couldn't remember exactly which ones he mentioned (it's all just images to me :)
[20:45:14] wagnerrp: coverart == poster
[20:45:44] stuartm: wagnerrp: the guide shows multiple icons at once
[20:45:56] stuartm: one for each visible channel
[20:46:41] stuartm: wagnerrp: sorry, I misunderstood the original question, the log I pasted was from the guide grid
[20:47:14] stuartm: I thought the question was a more generic 'When would we be requested lots of files in sequence?'
[20:47:17] wagnerrp: but youre still scrolling by one channel at a time
[20:47:34] wagnerrp: which would be the ~200–300ms gaps between blocks of transfers
[20:52:27] aloril (aloril!~aloril@ has joined #mythtv
[20:54:10] xavierh_ (xavierh_! has joined #mythtv
[21:03:36] Captain_Murdoch (Captain_Murdoch!~cpinkham@mythtv/developer/CaptainMurdoch) has quit (Ping timeout: 255 seconds)
[21:07:46] noahric (noahric! has joined #mythtv
[21:12:27] jpabq: stuartm, does that need built on both the frontend and backend?
[21:13:45] stuartm: jpabq: yes, since I changed the hardcoded storage group location to point to the actual icon directory
[21:14:12] stuartm: although if you're just testing you could copy over your icons to the ChannelIcons directory and avoid rebuilding the backend
[21:14:19] jpabq: taylorr, Back when the hdpvr driver was first written, I was able to build it outside of the kernel source tree. For the last year or so, that has not worked for me. I think I will just build a whole new kernel. Could you send me your patch, and I will apply it to the kernel source tree.
[21:15:21] kenni: stuartm: I've closed #9508 as an upstream bug (you can read the details in my comment)
[21:15:22] stuartm: natanojl: nice catch
[21:16:32] natanojl: stuarta: :)
[21:16:50] natanojl: oops, wrong Stuart :)
[21:16:59] jpabq: My system is recording, so I would rather not restart the backend. When you say ChannelIcons directory, is that defined as a storagegroup?
[21:17:04] Captain_Murdoch (Captain_Murdoch! has joined #mythtv
[21:17:04] Captain_Murdoch (Captain_Murdoch!~cpinkham@mythtv/developer/CaptainMurdoch) has joined #mythtv
[21:17:04] Captain_Murdoch (Captain_Murdoch! has quit (Changing host)
[21:17:27] stuartm: natanojl: probably my fault originally
[21:18:29] stuartm: jpabq: ~/.mythtv/ChannelIcons – it's been defined as a hardcoded storage group for some time evidentally, it's just not where channel icons are actually installed (~/.mythtv/channels) so the patch changes that location to match reality
[21:18:30] natanojl: stuartm: Yes, I think it was your commit in late January dealing with MythUI dialogs
[21:19:40] jpabq: stuartm, got it. Will try here in a minute.
[21:30:40] jpabq: Odd: 2012-03–03 14:29:20.783397 E [9048/9048] CoreContext mythuihelper.cpp:1378 (LoadScaleImage) – MythUIHelper: LoadScaleImage(myth://ChannelIcon@ failed to load image
[21:30:49] jpabq: What is odd about that, is the IP address is wrong.
[21:33:40] knightr (knightr!~knightr@mythtv/developer/knightr) has quit (Read error: Connection reset by peer)
[21:33:54] knightr (knightr!~knightr@mythtv/developer/knightr) has joined #mythtv
[21:54:51] skd5aner: minor suggestion... when outputing --version it reminds people to provide it as an attachement in bug reports, might be useful to add a line in there that says "... and pastebin when asking for assistance in irc at #mythtv-users on freenode" or something
[21:55:17] skd5aner: maybe not, but just had someone copy and paste the output in there, frequent enough occurance
[21:55:30] jpabq: stuartm, As long as it can find the backend via IP, it works very well. I guess that frontend has been using upnp for a while. When the frontend can't find the backend via IP, trying to browse channels is very painful.
[21:55:34] stuartm: jpabq: ok, I've fixed it not to try for the icon if the path is empty
[21:56:17] stuartm: as for the IP, there must be a better way to get the backend ip than using the setting
[21:58:33] wagnerrp: jpabq: is that the master backend address?
[21:58:37] jpabq: stuartm, anyway, on a properly configured system, the OSD does not pause any more than it did before. The first time I bring up the OSD, there is always a quick pause as it loads the statetype icons. After the first time, there is no delay.
[21:59:11] jpabq: wagnerrp, yes. My masterbackend has not lived at that address for a very long time. I never noticed before today that it was still trying to use it.
[21:59:17] wagnerrp: skd5aner: channel rules already say to use pastebin for anything >2 lines
[21:59:35] wagnerrp: if they cant read the channel rules, what likelihood is it theyll read that printout
[21:59:57] wagnerrp: they already paste the version directly into the ticket, even though it has said for two years to attach it as a file
[22:00:34] joki- (joki-! has joined #mythtv
[22:00:42] joki (joki! has quit (Ping timeout: 265 seconds)
[22:00:42] joki- is now known as joki
[22:01:21] stuartm: jpabq: what about browse mode in livetv?
[22:01:44] jpabq: Seems to be fine. Same behavior.
[22:03:32] jpabq: A bunch of my channels don't have icons, and it "browsed" over them find, as well. No noticeable pause in the playback.
[22:03:39] jpabq: s/find/fine/
[22:04:42] jpabq: I don't know if it matters, but I am using VDPAU with a 3.0GHz Core2 duo on the frontend.
[22:09:29] stuartm: ok, I've amended the patch a little to reduce some code duplication in the creation of the url and to prevent it trying to load an icon when one isn't defined, I'll commit after I've compile tested
[22:35:14] RogerM (RogerM!~chatzilla@ has left #mythtv ()
[22:36:09] jya: wagnerrp: I have rewritten the configure script for mythplugin.. Would you mind running it on FreeBSD and let me know if it works okay? (I tested it on mac and linux, but /bin/sh is actually bash and not the good old sh)
[22:36:40] jya: I made sure to use tests and string commands that are sh compatible
[22:36:45] wagnerrp: sure
[22:37:28] Beirdo: why?
[22:37:44] Beirdo: what was wrong with the previous implementation?
[22:37:55] wagnerrp: freebsd doesnt come with bash?
[22:38:10] jya: it does if you install it from ports.. not by default
[22:38:19] wagnerrp: yeah, but then its in /usr/local/bin/bash
[22:38:22] Beirdo: and please don't assume sh = bash
[22:38:23] jya: and as this configure starts with #!/bin/sh
[22:38:38] Beirdo: as that is not true for modern Linux distros
[22:39:04] wagnerrp: Beirdo: yes, thats the idea... to test it on normal 'sh' rather than 'bash'
[22:39:27] Beirdo: right, but why was it changed at all is my original question
[22:39:44] Beirdo: I'm sure there were deficiencies, but wondering what they are :)
[22:39:48] jya: Beirdo: the previous implementation was just plain broken and some plugins never compiled on mac because they weren't detected… The core issue is that to determine if a feature exists, it only looked at if the header or lib existed by searching for the files in a specifc set of locations
[22:39:48] wagnerrp: cant answer that one
[22:40:36] Beirdo: hmmm, not working on the mac != just plain broken, but I see what ya mean. Fair enough.
[22:40:37] jya: Beirdo: I *know* bash != sh, I just explained what I've done to make sure it would work, and asked for wagnerrp to test
[22:41:09] Beirdo: it can always use tweaking, was just curious what this latest tweak was
[22:41:11] Beirdo: :)
[22:41:27] jya: Beirdo: looking if a library exists in a particular location, is no guarantee that this feature would exist
[22:41:39] Beirdo: hopefully it will behave well as we don't have much time for ironing out wrinkles
[22:41:57] jya: in particular, on a mac, some libs could be 32 bits only, so if you compile on 64 bits, it would just break
[22:42:09] Beirdo: right, what a mess
[22:42:36] Beirdo: cool. Sounds like fun
[22:42:58] Beirdo: gotta fix things as you run across them, of course :)
[22:43:14] jya: But why I started to rework it, was simply because I never understood why some plugins weren't available on the mac: and the reason was simply because some libs weren't found, where they were actually there (mac uses Frameworks, not all libs are in /usr/lib
[22:44:07] jya: while at it, I made the error messages consistent. Like for MythWeather is would show "WARNING: blah"
[22:44:16] Beirdo: yeah, Apple really made a mess of things with that :)
[22:44:26] jya: for mythnetvision, another text etc… for some reasons, that annoyed me to no end
[22:44:35] ** wagnerrp wonders when hes going to get this new configure to test **
[22:44:37] wagnerrp: :)
[22:45:20] jya: Beirdo: I actually find the framework idea brilliant. everything, from programs down to libraries are all self contained. You don't need to copy your files in particular locations to work. Framework is to libs, what .app bundle are to applications
[22:45:39] Beirdo: yeah, non-unix :)
[22:45:40] Beirdo: hehe
[22:45:49] jya: wagnerrp: otherwise I'll just commit it and see what the buildbot states
[22:45:51] Beirdo: there is some sense to it, of course
[22:46:17] jya: I just love that I can delete a single "file" and it gets rid of the whole lot: header, libraries, application you name it
[22:46:18] joki (joki! has quit (Ping timeout: 245 seconds)
[22:47:06] jya: All that cleanup started because I started to use Qt libraries package from Qt web site.
[22:47:16] Beirdo: hmmm, why won't JW Player install?!
[22:47:49] wagnerrp: jya: either works, although it looks like the buildbot is currently missing some dependencies for mythmusic and mythweather
[22:47:51] jya: Qt is compiled as frameworks, so each Qt lib is a framework, containing, libs, developement lib and headers…
[22:47:54] stuartm: Beirdo: I had trouble with it when trying it out for the first time last week, fixed it manually in the end
[22:48:51] jya: wagnerrp: that's actually perfect that it's missing stuff… best way to test that it's properly detected. I'm fairly confident with the change.
[22:48:58] stuartm: and it didn't do anything, so I figured the actual working bit to transcode/stream hasn't been finished yet
[22:50:01] ** jya now looking why mythgallery now doesn't compile on the mac **
[22:50:38] Beirdo: stuartm: hmm, not sure how to manually fix this :)
[22:51:06] joki (joki! has joined #mythtv
[22:52:34] jya: Beirdo: when I do a git pull --rebase. Does git rename the commit number ?
[22:52:48] jya: the ones that is pushed back to the top of the list that is
[22:52:49] Beirdo: if it gets rebased, yes
[22:53:09] jya: ahhh, that would explain why github can't find some of my references in my log :(
[22:53:13] Beirdo: as the parent SHA1 is different, the commit's SHA1 has to be different
[22:54:08] jya: I usually have a comment like: following 2414141 , redo bah. The last push where I use git pull --rebase, github gives a 404 error when I click on the commit number
[22:54:49] stuartm: Beirdo: sorry, I really don't remember what I did to fix it :/
[22:55:13] jya: so really, that --rebase thing, while it makes the git log clearer, it breaks heaps of stuff. so github scripts do not work, and my commit's comment are now invalid
[22:55:54] stuartm: jya: I use it before every commit, it's never done that for me
[22:56:03] Beirdo: it doesn't update comments. it has nothing at all do do with github's scripts
[22:56:28] jya: Beirdo: I'm referring to what you wrote about the post-commit hook not updating trac
[22:56:44] Beirdo: found it.
[22:56:45] Beirdo:
[22:56:47] jya: stuartm: but you don't use in the commit log a reference to a previous commit you've done
[22:57:07] Beirdo: the post-commit hook doesn't always send us data
[22:57:08] stuartm: jya: no, that's true, I don't
[22:57:49] stuartm: sorry, I'm not paying enough attention
[22:57:56] jya: stuartm: That's just a habit I've taken… if I commit something that depends on another commit, I write a little word about it
[22:58:05] Beirdo: this is still not doing what I want. Hmmm.
[22:58:39] Beirdo: oooh, I have to unzip it on the backend machine
[22:58:46] Beirdo: hahah
[22:59:39] ** jya decided he liked perl very much **
[22:59:41] stuartm: jya: that would mean searching for the sha1 of that previous commit and that's just too much hassle, I would have done it back in the SVN days when the previous commit was a nice memorable 5 digit number
[23:00:39] stuartm: it's easy to underestimate just how lazy I can be ;)
[23:00:50] jya: stuartm: when I make a commit, I take not of the commit number and put it on my list. I like to make several logical commits for one main feature added. makes it more structures that way
[23:09:15] gregL (gregL! has quit (Read error: Connection reset by peer)
[23:26:07] gregL (gregL! has joined #mythtv
[23:26:22] zombor (zombor!~zombor_@kohana/developer/zombor) has joined #mythtv
[23:26:52] Beirdo: OK, I guess I have it partially working anyways
[23:27:09] stuartm: Beirdo: you have video?
[23:27:19] Beirdo: I'll leave it to Captain_Murdoch to test more heavily for HLS as I'm not sure I have everything setup right
[23:27:22] Beirdo: yeah
[23:27:32] stuartm: I couldn't get that to work and lost interest
[23:27:53] stuartm: is it still using mythffmpeg in the background?
[23:28:04] Beirdo: had to pull from their svn to get adaptiveProveider.swf
[23:28:22] Beirdo: no, it uses code right in mythtranscode that the Captain put in
[23:28:40] stuartm: hmm, ok, no idea why it didn't work for me
[23:30:40] Beirdo: it's not 100%, I'll give ya that :)
[23:36:34] skd5aner: Any anticipated date for RC1? Just want to have a heads up to bring release notes up to speed from the last few weeks
[23:37:06] skd5aner: general speculation is ok for now unless one is nailed donw
[23:37:07] skd5aner: down
[23:37:52] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has joined #mythtv
[23:40:45] Beirdo: Beta is tomorrow
[23:41:14] Beirdo: RC1 March 18
[23:41:22] Beirdo: Final release Apr 2
[23:41:24] stuartm: skd5aner: release schedule hasn't changed from the original announcement
[23:41:38] Beirdo: assuming nothing show-stopping happens in between
[23:42:08] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has quit (Ping timeout: 240 seconds)
[23:43:53] stuartm: more bug fixing would be good
[23:44:35] Beirdo: OK, I'm going to push this mythtranscode fix. It may yet need follow up work for non-fifo modes, but HLS seems to be operational at least (best I can tell)
[23:44:46] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has joined #mythtv
[23:45:05] Beirdo: I'll leave it for a few hours as I'm going out for a bit. Captain_Murdoch if you can give it a shot if you're home, that would be best :)
[23:45:52] Beirdo: unfortunately, to make it work again required a touch of overhaul, so I want it in as soon as possible to let any remaining bugs get squashed in it
[23:46:14] Beirdo: basically, fifo mode (and thus nuvexport) was completely and utterly broken
[23:48:36] Beirdo: soon enough (0.26, 0.27, whatever) I'd like to make nuvexport just a file selector and move all teh transcoding right into mythtranscode, no more need to run mythffmpeg externally
[23:48:45] Beirdo: then fifo mode can die
[23:48:49] Beirdo: but until then...
[23:52:15] Beirdo: be back in a few hours.
[23:53:08] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has quit (Ping timeout: 260 seconds)
[23:54:09] jpabq_ (jpabq_!~jpabq@mythtv/developer/jpabq) has joined #mythtv
[23:59:51] jya: is there a way to convert a commit and put it in stash ? I could do this with mercurial...

IRC Logs collected by BeirdoBot.
Please use the above link to report any bugs.