:: #mythtv

Daily chat history

Current users (72):

aloril, andreaz, Anssi, blahdodo, brfransen, Captain_Murdoch, ChanServ, Chutt_, clever, coling, CyberJacob, dblain, dekarl, drewzhrodague, eharris, ElmerFudd, enyc, GhostOfRaven, ghoti, gregbert, GreyFoxx, Guest65292, Hydr0p0nX, ikevin, J-e-f-f-A, jab416171, jams, jheizer, jmcentee, jnylen, joki, jpabq, jpharvey, jst_, jya, kc, knightr, lautriv_, libsci, mad_enz, markspieth, MythBuild, MythLogBot, natanojl, nephyrin, peper03, poptix, pppingme, raven42, rich0, Roklobotomy, saaki, Seeker, seld, ShapeShifter499, sphery, SpootDev, sraue, stuarta, stuartm, suffice, superm1, tgm4883, toeb, tonsofpcs, tris, unforgiven512, vincent42, XDS2010, xris, _charly_, _iwc
Friday, February 19th, 2016, 00:03 UTC
[00:03:22] dmfrey (dmfrey! has joined #mythtv
[00:06:13] amessina (amessina!~amessina@unaffiliated/amessina) has joined #mythtv
[00:07:36] Warped (Warped!~Warped@unaffiliated/warped) has quit (Ping timeout: 276 seconds)
[00:08:22] dmfrey (dmfrey! has quit (Ping timeout: 255 seconds)
[00:19:21] raven42 (raven42! has joined #mythtv
[00:20:04] dmfrey (dmfrey! has joined #mythtv
[00:30:41] raven42 (raven42! has quit (Read error: Connection reset by peer)
[00:32:17] Warped (Warped!~Warped@unaffiliated/warped) has joined #mythtv
[00:32:29] raven42 (raven42! has joined #mythtv
[00:36:01] dmfrey (dmfrey! has quit (Ping timeout: 240 seconds)
[00:37:11] raven42 (raven42! has quit (Ping timeout: 250 seconds)
[00:43:06] raven42 (raven42! has joined #mythtv
[00:45:31] raven42 (raven42! has quit (Read error: Connection reset by peer)
[01:24:24] dmfrey (dmfrey! has joined #mythtv
[01:37:52] jams (jams! has quit (Ping timeout: 250 seconds)
[01:40:44] jams (jams! has joined #mythtv
[02:13:02] pppingme (pppingme!~pppingme@unaffiliated/pppingme) has joined #mythtv
[02:37:23] raven42 (raven42! has joined #mythtv
[03:29:33] dmfrey (dmfrey! has quit (Ping timeout: 240 seconds)
[03:33:55] dekarl (dekarl!~dekarl@mythtv/developer/dekarl) has quit (Read error: Connection reset by peer)
[03:35:36] dekarl (dekarl!~dekarl@mythtv/developer/dekarl) has joined #mythtv
[03:42:03] gregL (gregL! has quit (Ping timeout: 240 seconds)
[03:46:13] tgm4883 (tgm4883!uid23806@ubuntu/member/tgm4883) has quit (Quit: Connection closed for inactivity)
[03:48:27] tgm4883 (tgm4883!uid23806@ubuntu/member/tgm4883) has joined #mythtv
[03:52:38] gregL (gregL! has joined #mythtv
[04:26:40] SteveGoodey (SteveGoodey! has joined #mythtv
[04:50:39] SteveGoodey (SteveGoodey! has quit (Quit: Konversation terminated!)
[06:16:25] raven42 (raven42! has quit (Ping timeout: 240 seconds)
[06:35:54] raven42 (raven42! has joined #mythtv
[06:38:44] GhostOfRaven (GhostOfRaven! has joined #mythtv
[06:40:33] raven42 (raven42! has quit (Ping timeout: 240 seconds)
[06:53:27] Tobbe5178 (Tobbe5178! has joined #mythtv
[07:34:28] sraue_ (sraue_! has quit (Ping timeout: 248 seconds)
[07:52:56] Roklobotomy (Roklobotomy! has joined #mythtv
[08:11:57] lautriv_ (lautriv_!~lautriv@funtoo/user/lautriv) has joined #mythtv
[08:14:51] SteveGoodey (SteveGoodey! has joined #mythtv
[08:15:33] lautriv__ (lautriv__!~lautriv@funtoo/user/lautriv) has quit (Ping timeout: 244 seconds)
[08:46:37] Warped_ (Warped_!~Warped@ has joined #mythtv
[08:48:35] Warped (Warped!~Warped@unaffiliated/warped) has quit (Ping timeout: 240 seconds)
[08:55:08] stuarta: stuartm: any objection to me pushing the patch to fix the port hardcoding(s)
[08:58:34] Merlin83b (Merlin83b!~Daniel@2a00:1ee0:3:1337:8869:3e59:640:f79a) has joined #mythtv
[08:59:00] stuartm: no objections :)
[08:59:06] stuarta: ok
[08:59:51] joki (joki! has quit (Ping timeout: 265 seconds)
[09:00:11] stuarta: websocket port is +5 now, to leave the upnp ssl port at +10, since the websocket port will go away
[09:02:47] willcooke (willcooke! has joined #mythtv
[09:02:48] willcooke (willcooke! has quit (Changing host)
[09:02:48] willcooke (willcooke!~willcooke@ubuntu/member/willcooke) has joined #mythtv
[09:04:13] dekarl (dekarl!~dekarl@mythtv/developer/dekarl) has quit (Read error: No route to host)
[09:04:28] joki (joki! has joined #mythtv
[09:04:44] SteveGoodey (SteveGoodey! has quit (Quit: Konversation terminated!)
[09:05:01] dekarl (dekarl!~dekarl@mythtv/developer/dekarl) has joined #mythtv
[09:46:29] stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has quit (Ping timeout: 244 seconds)
[09:50:06] stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has joined #mythtv
[10:01:07] stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has quit (Ping timeout: 248 seconds)
[10:02:38] enyc: tgm4883: can i run an extra buildd to help with this, for the legitimate cases of it being useful to run 0.27 on newer box, migration not working, 0.28 QT5-bugginess incompatible with other OSes in use, etc.. ?
[10:03:04] stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has joined #mythtv
[10:03:10] enyc: tgm4883: completely accept makes sense to push new installes/users/focus/features into new build btw!
[10:06:50] GhostOfRaven (GhostOfRaven! has quit (Read error: Connection reset by peer)
[10:07:14] GhostOfRaven (GhostOfRaven! has joined #mythtv
[10:09:12] stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has quit (Ping timeout: 244 seconds)
[10:11:25] stuarta: enyc: we'll never get away from 0.27 if we didn't do it. people would start complaining about lack of hevc support which is already in master
[10:11:55] enyc: stuarta: nodsnods =)
[10:12:02] stuartm (stuartm!~stuartm@mythtv/developer/stuartm) has joined #mythtv
[10:12:17] enyc: stuarta: ok i didn't see that discussed here either, for some reason i saw a different previous-consensus
[10:12:54] enyc: stuarta: does 0.28 run well on qt 5.3 systems (debian stable etc.)?
[10:13:06] enyc: stuarta: i can see this being a cross-copmatibility snag
[10:13:38] stuarta: enyc: there was a different previous concensus, those previously against were those pushing for it to be released
[10:13:51] enyc: stuarta: interesting =)
[10:16:13] enyc: stuarta: i had this sort of fun, with retting up newservers at old-work and it did generalyl pay-off to work off the "to-become-" LTS-release etc... ahead
[10:20:53] enyc: #12634.
[10:20:53] ** MythLogBot **
[10:21:02] enyc: oops didn't mean to paste newline
[10:22:01] jmcentee: Not that my opinion matters, but I would really push to have 0.28 in the lts release. people wanting to run 0.27 need to accept an older distro, or the effort of self-build or other packages from somewhere else. New users or those that want an easy life should be on 0.28 as the default.
[10:22:13] enyc: ^^ I don't see any reason can't commit thase patches and +gdb line into fixes/0.27 for those who need a 0.27 during migrations in a heterogenuous base, nor to e.g. test/compare for bug in 0.28 vs 0.27 quite normal.
[10:25:00] jmcentee: stuarta: Did I sent my buildbot public key to the right e-mail address?
[10:25:54] enyc: a buildd// repo for debian/raspbian would not be a bad idea in my view... I may well be able to get some VM space to run these =)
[10:34:19] stuarta: jmcentee: yes i have it
[10:49:42] Roklobotomy (Roklobotomy! has quit (Quit: Leaving)
[11:01:41] Warped_ (Warped_!~Warped@ has quit (Quit: ChatZilla 0.9.92 [Firefox 44.0.2/20160210153822])
[11:02:54] Warped (Warped!~Warped@unaffiliated/warped) has joined #mythtv
[11:50:59] dekarl: enyc: what is the legitimate case for updating the OS to a new major version without touching the mythtv version? If it works just leave it alone.
[11:51:31] dekarl: 14.04 LTS is good for another 3 years
[11:51:58] enyc: dekarl: because of other packages on the new OS (ont mythtv itself), compatibility with newer hardware, ... you are simply installing/reinstalling new machines, and have other parts in the system that aren't yet compatible with running 0.28
[11:52:38] dekarl: well, then the new machines better be compatible with 14.04 :D
[11:52:44] enyc: dekarl: or indeed, you try to do this then discover a bug, and its' quite fair/reasonable to compare between 0.27 and 0.28 behaviour and/or run 0.27 while it gets sorted-out
[11:52:50] stuarta: enyc: are you aware there is a kernel update stream for 14.04 which means you can get the hardware support without leaving 14.04 ??
[11:53:18] enyc: stuarta: yes, but its' only kernel... not whole package set
[11:53:20] dekarl: but as stuarta noted, HEVC is being rolled out in pilot networks already
[11:53:29] enyc: dekarl: is that a USA thing?
[11:53:36] stuarta: europe mainly
[11:53:37] enyc: dekarl: european HDTV thing ?
[11:53:47] stuarta: yes europe hdtv
[11:53:48] dekarl: USA is just now switching slowly to AVC
[11:54:03] stuarta: well they are targetting 4k HD with it
[11:54:13] enyc: oooooo that makes more sense =)
[11:54:42] dekarl: stuarta, the "DVB-T2 HD" that our users started asking about will be 100% HEVC for 1080p (maybe also SD, I don't know)
[11:54:58] stuarta: ah right
[11:55:16] enyc: dekarl: oooooooh thats intreseting lots of deliberate-hardware-obsolescence sell-consumers-new-kit fun in EU? ;-)
[11:55:42] stuarta: ie. business as usual
[11:55:49] dekarl: enyc: no, lots of auctioning off more airspace to telcos, then wondering what to do about the other users ;)
[11:55:54] enyc: still, i think its' going to be interesting =)
[11:56:24] dekarl: [] fast youtube on your mobile / [] slow hardware cycle, chose one
[11:56:33] stuarta: use of DVB-T2/S2 combined with advanced codecs like h264, h265 (hevc) mean more crappy channels in the same bandwidth
[11:59:13] enyc: particular interop case I'm thinking of, running 0.28 across many boxes, we should either be reasonably sure 0.28 works with debian stable's qt5, 5.3.2 or that it can compile with a static/included varant of qt libs if thats' neccesary
[11:59:21] enyc: stuarta: yeus, more idiot box channels what fun =)
[12:00:09] stuarta: enyc: you should be fine on debian stable
[12:00:16] stuarta: ie. jessie
[12:00:27] stuarta: anything earlier doesn't work
[12:01:01] stuarta: it's only a specific version of qt 5.4.2 iirc that has an issue, and that's not used in 16.04
[12:02:50] enyc: (frontend on raspbian etc.)
[12:04:25] stuarta: you'll need master / 0.28 for that anyway, as it's only alpha right now
[12:05:57] dekarl: no new tricks for old dogs ;)
[12:07:29] ** stuarta flogs a dead horse **
[12:08:02] enyc: stuarta: anyway, so... i acn submit the pull-reqs relevant to 0.27 fixes builds, and usefully feedback my build-checks across 0.27 0.28 master on more architectures... i can also ask questions r.e. VM space or my own box so a buildd to a repo for debian could be created, usefully?
[12:08:42] stuarta: yes sure, see what we already have here
[12:22:16] stuarta: this might actually be time to migrate the buildmaster over to mizar while i'm at it
[12:24:43] enyc: stuarta: are we likely to get any answer from jya over the ubuntu-patches to 0.27 ? from my pov they work across all builds i've been able to make, i'd happily put them into the packaging diff's if thats the best approach
[12:25:11] jya: enyc: we are doing 0.28 for 16.04
[12:25:13] stuarta: enyc: in the short time that's the best option
[12:25:30] jya: so i don't see the point spending any time on 0.27
[12:25:31] enyc: jya: i know
[12:25:54] jya: so you can use your local patch if you want to, don't see the need to make them in our tree
[12:26:06] enyc: have any interessting cross-architecture-patches been discovered?
[12:26:07] enyc: err
[12:26:37] stuarta: jya: then can you close that ticket as wontfix and ask distros to carry the patches if they need them
[12:26:41] stuarta: ??
[12:27:07] enyc: cross-architceture bugs been discovered of trying to build on more and more arches =)
[12:27:30] jya: stuarta: will do
[12:27:33] enyc: i noticed qemu arm64 emulation actually works ;p
[12:27:43] enyc: on 32bit ,which doesn't for any other 64bit guest
[12:28:31] enyc: given what I know elsewhere, I suspect arm-insiders helping, even though, arm can't officially supporting anything like qemu due to licensing fun ;p
[12:41:28] enyc: stuarta: there are debian build-slaves but I haven't seen them creating packaging repositories per-se ...
[12:42:52] stuarta: enyc: no, they only build the code, they don't do packaging
[12:43:09] stuarta: historically the project does not build packages, we leave that to the packagers
[12:43:19] enyc: stuarta: aaaaaaaaaaah
[12:43:36] enyc: stuarta: yes as launchpad does the builds for PPA ?
[12:54:23] ** stuarta curses the damn firewall **
[13:03:02] enyc: lol not again??
[13:18:12] SteveGoodey (SteveGoodey! has joined #mythtv
[13:22:26] dmfrey (dmfrey!~dmfrey@ has joined #mythtv
[13:32:15] stuarta: enyc: i'm doing a planned migration on my personal system
[13:33:03] stuarta: there's an open bz that when you use the masquerade setting (like some of the zones do by default) your traffic from localhost gets nat'd to your external ip address as well, which is just plain wrong
[13:33:43] stuarta: and on my system i need masquerade, to nat the vm guests out, i have to work out how to bend firewalld to my will and stop masquerading localhost traffic
[13:35:48] jya: stuarta: do you know what libmythbase use for unzipping (as used by mythcoreutil.cpp)
[13:36:15] stuarta: jya: probably
[13:36:22] enyc: stuarta: argh!
[13:36:30] enyc: stuarta: i like my custom firewall scripts =)
[13:36:34] jya: no i mean what he uses to unzip
[13:36:46] jya: not the file it unzip!
[13:36:56] stuarta: not off the top of my head no
[13:36:59] jya:
[13:37:44] ** stuarta insults firewalld again **
[13:44:59] pppingme (pppingme!~pppingme@unaffiliated/pppingme) has quit (Ping timeout: 250 seconds)
[13:45:32] jya: right, i'll continue the resync with ffmpeg 3.0 tomorrow, for some reason i get linkage error in libmythbase related to zip
[13:46:08] stuarta: that's what the link line is like for me
[13:47:56] stuarta: maybe libz?
[13:48:08] stuarta: which is zlib
[13:56:27] peterbennett (peterbennett! has joined #mythtv
[13:58:02] peterbennett: Raspberry pi – there is a hang in mythfrontend running on Linux Mint with the PPA version of mythtv.
[13:58:09] peterbennett: It hangs on setuid
[13:58:35] peterbennett: Is the setuid necessary?
[13:59:52] peterbennett: Sorry Linux Mate not Linux Mint
[14:01:43] dekarl: jya, I think unzip.cpp is our local copy of some external pkzip decompressor using libz for the actual compression
[14:01:49] pppingme (pppingme! has joined #mythtv
[14:01:58] pppingme (pppingme! has quit (Changing host)
[14:01:58] pppingme (pppingme!~pppingme@unaffiliated/pppingme) has joined #mythtv
[14:02:17] dekarl: we have another copy of pkzip decompression in mythgame
[14:06:12] dekarl: peterbennett: on a raspberry?
[14:10:09] peterbennett: dekarl: Yes raspberry pi 2
[14:16:00] dekarl: peterbennett: . . . in.cpp#L1793
[14:16:19] amessina (amessina!~amessina@unaffiliated/amessina) has quit (Quit: Konversation terminated!)
[14:16:26] dekarl: its testing for QT5.0 or better and "egl" as part of some string, maybe that test fails on your install?
[14:18:16] peterbennett: dekarl: That test is not good enough evidently. egl is not in the string and yet it still fails
[14:18:40] peterbennett: dekarl: egl only applies when running standalone without xwindows
[14:19:07] peterbennett: dekarl: I am running under xwindows and the string is xcb
[14:20:25] peterbennett: dekarl: the log contains the message QT_QPA_PLATFORM=xcb just before the hang
[14:22:44] peterbennett: dekarl: I don't know why setuid should hang in the first place – LVR has the comment that it will hang with eglfs but why?
[14:30:13] stuarta: woot, testing package fixes firewalld issues
[14:41:10] dekarl: peterbennett: I don't know. Maybe he has a patch in his stash that fixes it? who knows
[14:42:29] peterbennett: dekarl: I will try to figure out a condition to bypass the setuid on all raspberry pi
[14:48:51] dekarl: that setuid(getuid()) stuff could use a cleanup... layers of hacks on layers of hacks... and duplicate stuff . . . 4f1bcc2784d3 . . . c31247b45523
[14:49:10] stuarta: burn it
[14:49:43] dekarl: what's the concept behind setuid(getuid()) in the first place? dropping SUID root privileges back to the running user after fiddling with thread priorities?
[14:49:56] dekarl: burn it all!
[14:51:58] peterbennett: dekarl: I cannot figure any use for it. The only thing I could think is in case a module is "setuid root" and you wanted to revert it.
[14:53:10] peterbennett: dekarl: OK I will try that.
[14:53:46] dekarl: hmm protect it with if (geteuid() != getuid()) { setuid(getuid()) }; ?
[14:54:00] stuarta: dekarl: realtime priority for the frontend
[14:54:05] dekarl: aka only try to setuid back to caller if the effective uid (due to suid root) is not equal to the caller?
[14:54:24] stuarta: tbh, i think it's useless these days
[14:55:10] dekarl: I agree, but ripping out "realtime prio" is something for fixes/0.29?
[14:55:59] dekarl: mythbuntu packaging doesn't install mythfrontend as suid root.
[14:56:26] stuarta: the call should just fail, and it should move on
[14:56:56] stuarta: it's unlikely the frontend / mythtv user has the capability set to allow the realtime prio set without the setuid
[14:57:06] peterbennett: It is mysterious why it hangs at all
[14:57:09] Seeker` is now known as Seeker
[14:57:27] dekarl: if someone wants funky stuff like priority fiddling, isn't that done via capabilities nowadays? => rip it out for good
[14:58:49] tgm4883: peterbennett: is this on the raspberry pi 2 on linux mate?
[14:59:07] peterbennett: Correct, Linux mate
[14:59:31] stuarta: dekarl: well capabilities give you the permission to set the prio without using setuid
[14:59:41] stuarta: dekarl: but yes we can rip all that crap out in 0.29
[15:00:37] tgm4883: peterbennett: hanging during startup? That's as far as I got before I had to work on other stuff
[15:00:52] peterbennett: So to get it working I can try #ifdef __arm__ conditional compile or else dekarl's suggestion if (geteuid() != getuid())
[15:01:11] dekarl: peterbennett: revoceRoot from . . . &t=16897 looks better
[15:01:57] dekarl: well, convert the funky "UID + EUID == 0" to "UID==0 && EUID==0"... the next developer has to understand what the intention was
[15:02:26] dekarl: peterbennett: if that fixes it for you I can push such a change
[15:03:32] peterbennett: OK I will try that code from revokeRoot.
[15:03:45] peterbennett: Thanks
[15:04:58] jpabq_ is now known as jpabq
[15:05:09] jpabq (jpabq! has quit (Changing host)
[15:05:09] jpabq (jpabq!~quassel@mythtv/developer/jpabq) has joined #mythtv
[15:15:00] enyc: stuarta: takling deb versions compatibility, it looks like actually 0.28 can build on wheezy-backports (oldstable), qt5 and libc and kernel are all backported, and my packaging changes include the old init needed too...
[15:16:07] enyc: but this is less useful for pi frontends etc. because they aren't being LTS anyway on non-i386/amd64 ;p
[15:20:04] peterbennett: enyc: What qt version? I think it needs 5.4 or newer
[15:20:25] dekarl: Ha, much more interesting, will Mythbuntu 16.04LTS configure ZFS for storage by default? . . . d-linux.html (<- may contain a fun idea that takes some time to engineer the solution for)
[15:20:48] tgm4883: no we will not
[15:21:46] tgm4883: well, I shouldn't say that so definite. No, we probably wont
[15:22:17] tgm4883: If Ubuntu does, we probably would though
[15:25:11] dekarl: tgm4883: just saw that pass by on the newsticker that ZFS will come preinstalled with Ubuntu 16.04. as in anyone can configure a ZFS if the feel like it without compiling modules themselves... But its not a "Ubuntu will come with root-on-zfs by default"
[15:25:50] tgm4883: dekarl: I've not read anything that says it will be preinstalled. Everything I've read says you still have to install zfs-utils
[15:26:11] tgm4883: which you can do via apt now, which is better than before
[15:26:26] enyc: peterbennett: 5.3 and -no-, it seems qt5.2 is minimum, but tehre is some known bug in a particular 5.4.1 version
[15:26:46] enyc: what about btrfs-as-root ?
[15:26:47] dekarl1 (dekarl1!~dekarl@mythtv/developer/dekarl) has joined #mythtv
[15:26:58] enyc: or btrfs for storage for that matter
[15:27:50] peterbennett: enyc: When I try to run on 5.3 it fails with OMX errors and downgrades the video player. I don't know for sure if that is because of 5.3 or something else
[15:28:17] dekarl1: enyc: I'm reluctant to that not-invented-here-syndrome-software ;)
[15:28:28] peterbennett: enyc: the 5.4.1 bug only affects backend as far as I know
[15:28:45] enyc: dekarl1: humm err what?? don't understand
[15:28:48] dekarl1: enyc: been using xfs and jfs for video storage earlier
[15:29:16] dekarl1: enyc: I consider btrfs to be a child of not-invented-here syndrome
[15:29:33] dekarl (dekarl!~dekarl@mythtv/developer/dekarl) has quit (Ping timeout: 240 seconds)
[15:29:54] enyc: dekarl1: errrr loking up errr... like other fs's did this already but we want to reinvent ?
[15:30:26] dekarl1: enyc: yes, yet another "lets rewrite it for linux in an incompatible way" software
[15:30:34] dekarl1 is now known as dekarl
[15:31:50] enyc: dekarl1: hrrrrrrrm
[15:32:17] enyc: dekarl: i wonder if this partly stemmed from licensing related headaches, who knows!
[15:33:26] dekarl: isn't that moot since oracle (btrfs) bought sun (zfs)?
[15:35:23] dekarl: its also moot as "zfs.ko" can have a different license as "kernel"
[15:35:41] dekarl: but I bet it was a cool exercise to build it ;)
[15:37:40] stefan_ (stefan_! has quit (Ping timeout: 248 seconds)
[15:40:41] jmcentee: if oracle wanted to be nice they could change the ZFS licens to be GPL like thier btrfs is. I still think it won't make it into the linux kernel as it duplicates the RAID stuff.
[15:41:28] stuarta: oracle have never been, and never will be nice
[15:41:33] jmcentee: I have used ZFS for about 6 years now and really like it
[15:43:33] jmcentee: oracle are great at capatialism, but I hate them more than microsoft.
[15:44:12] enyc: hah ;p
[15:44:55] enyc: my college managed computers now use btrfs for upgrades
[15:45:11] enyc: they snapshot and commit all dpkg updates with some wrapper around sync() to not actually sync hack ;p
[15:45:17] stuarta: i hope you backup your data
[15:45:28] enyc: stuarta: user data is separate from rootfs image
[15:45:58] stuarta: ah... that's the only use case most of the linux vendors are supporting. os updates
[15:46:16] stuarta: data? no no, don't put that shit anywhere near btrfs
[15:46:24] enyc: stuarta: then, when writable snapshot is all sync'ed to disk, this becomes next main snapshot and so on...
[15:47:36] jmcentee: opensolaris has/had a great upgrade, probably the same. it would clone the rootfs. you can then just chose a different grub boot item to go back to pre-update (as long yas you did not manually upgrade the zpool version)
[15:49:13] ** jmcentee very happy, just been able to order FTTC. **
[15:49:27] stuarta: \o/ i have that, it's lovely
[15:49:38] stuarta: 40/10 and i could get 80/20 if i needed it
[15:52:26] jmcentee: going for the extreme 80/20, I won't know what to do with it. currently have 4/1 and I always forget how so package downloads go compared to work.
[15:53:16] ** jmcentee remembers the time of mivng from a cable modem area out into a dail-up only village. **
[15:53:18] stuarta: you will find it faster to drive home, download stuff, and come back
[15:53:41] jheizer_: I tried to recored to btrfs one time. It was a total diseaster.
[15:53:57] jheizer_: Went back to ZFS.
[15:53:59] jmcentee: work have a 80/20 so no need to drive home.
[15:56:11] jmcentee: I am very unlikely to move away from ZFS, now with OpenZFS. The lonly problem I have is the VMware esx / NFS / ZFS -on-linux combimation is very slow, and I don't like the performance workaround I have implemented.
[15:57:10] jheizer_: I've used ZFS for a long time as well. Was starting a new array and tried brtfs so I could single drive expand the array over time. It couldn't handle more than one recording hitting it at once.
[15:57:43] jheizer_: ended up just buying 2 more drives and making a ZFS array that is waaaay bigger than I need now, but its done. Just means more snapshots. :)
[16:05:20] jmcentee: If ubuntu makes booting a zpool root easy I may give it a go, but currently I don't want the effort of grub and zfs myself.
[16:30:11] enyc: jheizer_: how long ago was the btrfs-mess ?
[16:31:43] jheizer_ is now known as jheizer
[16:31:52] jheizer: Looks like May 2014 mased on order confirmation
[16:31:56] jheizer: based
[16:32:56] jheizer: Not sure how much I tried to find work arounds. Looks liek 2 months later I ordered 2 more drives and decided that would be enough (and max out my sata ports)
[16:39:48] ** stuarta ponders upgrading prod system to 0.28 **
[16:57:29] ** drewzhrodague uses XFS – seems okay for 6T across two disks, raid 0. **
[16:57:42] ** drewzhrodague loves MythTV **
[17:09:52] tgm4883: 6TB seems like a lot of data to raid 0 IMO
[17:10:27] drewzhrodague: It is!  – all for MythTV.
[17:10:40] drewzhrodague: 500+ episodes of Doctor Who =_)
[17:11:43] tgm4883: drewzhrodague: why raid 0
[17:11:59] drewzhrodague: I don't need redundancy for my MythTV recordings.
[17:12:14] tgm4883: drewzhrodague: ok, but that still isn't a need for raid 0
[17:12:44] drewzhrodague: What would you suggest? I'm on CentOS/mdadm
[17:13:27] jheizer: Mythtv storage groups means you can have 2 individual disks so all is not lost if one fails.
[17:13:36] drewzhrodague: Oh, I see.
[17:13:37] tgm4883: drewzhrodague: well if it's only for mythtv recordings, then you could just do separate mount points and put both in the same storage group
[17:13:41] tgm4883: that's much better
[17:13:45] tgm4883: what jheizer said
[17:14:17] jheizer: And what he said, if it is just got mythtv. :)
[17:14:29] jheizer: jsut for mythtv
[17:14:39] drewzhrodague: Roger. I was going for high-speed throughput also – I have HDHR Prime, and HDHR ATSC + DVB dish. Plus I flag and delete commercials.
[17:14:51] drewzhrodague: Sometimes it records 6–8 things at once (!!)
[17:15:20] ** drewzhrodague is real glad we don't have to use analog tuners anymore **
[17:15:35] tgm4883: drewzhrodague: which it a little better, but mythtv would/could record/read from both disks still
[17:16:09] drewzhrodague: True. I didn't think of that when I put the fs together.
[17:19:08] drewzhrodague: I am still on .26-fixes – atrpms went down, and that was the source of MythTV binaries for EL6.
[17:28:38] stuarta: el6 isn't going to be of any use for 0.28
[17:29:21] peterbennett: Is it possible and not too hard to use dpkg-buildpackage to cross-compile?
[17:29:42] peterbennett: Build on raspberry pi takes a couple hours
[17:29:53] stuarta: theoretically possible, dunno if it works or not
[17:30:57] peterbennett: Also raspberry pi is a little short on memory
[17:34:03] peterbennett: I can cross compile using lvr's script but that will give different results, it does lots of different things
[17:36:52] tgm4883: peterbennett: you're testing out some patch?
[17:37:21] peterbennett: I want to test the fix for the setuid hang
[17:37:59] tgm4883: peterbennett: if you've got a patch for it, and are OK waiting 3–4 hours for the build I can push it to our patch testing PPA
[17:38:08] tgm4883: arm builds take a really long time :/
[17:38:28] tgm4883: the patch would need to apply cleanly to 0.28
[17:38:29] peterbennett: Since it happens on ubuntu mate with the package, I want to recreate the same type package
[17:39:26] peterbennett: OK it will apply cleanly, the build got past that point
[17:40:25] tgm4883: peterbennett: where are you getting the package for mate?
[17:40:44] peterbennett: From the 0.28 PPA
[17:41:19] tgm4883: peterbennett: ok then yea, give me the patch and It will be built the same way
[17:42:45] peterbennett: . . . waGmlya?dl=0
[17:43:34] peterbennett: there is the setuid patch. I would like my other two patches as well, they are currently sitting in tickets waiting
[17:45:27] peterbennett: tgm4883: Those patches are all for raspberry pi
[17:47:19] tgm4883: peterbennett: for which ubuntu version?
[17:47:44] peterbennett: wily
[17:48:05] tgm4883: ok
[17:48:05] Merlin83b (Merlin83b!~Daniel@2a00:1ee0:3:1337:8869:3e59:640:f79a) has quit (Quit: Leaving)
[17:48:16] tgm4883: I don't think we can do multiple patches at once, let me check
[17:49:11] tgm4883: wait, looks like it just wants them listed separately
[17:49:22] tgm4883: ok give me a minute
[17:50:09] peterbennett: i usually put them in the patches directory and add file names to series. They can go in any order
[17:50:22] warpme (warpme! has joined #mythtv
[17:50:25] tgm4883: right, but I was talking about our build scripts
[17:50:32] warpme: hi *
[17:50:44] tgm4883: I've added them, it's building the source package now
[17:51:02] peterbennett: OK I have to leave for lunch for a while
[17:51:07] peterbennett: Be back soon
[17:53:14] tgm4883: Uploading. The builds will show up here . . . es/+packages
[17:53:30] warpme: regarding my upgrade to Qt5.5.1 and screen flickering on UI operations: it looks like issue is only on some themes: MythMediaStream, MythAeon
[17:55:34] tgm4883: peterbennett: . . . uild/9038874
[17:55:38] warpme: markspieth: may You pls try on Yours 9400GS/BLOB thost themes? I want to be sure issue isn't on my side....
[17:55:59] warpme: s/thost/those/
[17:58:19] jpabq: Is git working for everyone else? I am suddenly getting "Received disconnect from 2: Too many authentication failures for git"
[18:05:09] enyc: peterbennett: hangon, do you really mean for 'armhf' (as included with ubuntu, not compatble with Pi A/B/B+) or actually for rasbpian 'pi' architecture ?
[18:08:20] peterbennett: enyc: raspberry pi
[18:10:38] enyc: peterbennett: aaaaaaah i can see rpi2 images for ubuntu
[18:10:43] enyc: peterbennett: that is armhf arch
[18:11:23] enyc: peterbennett: this is not to be muddled up with '-rpi' arch *labelled* armhf* from raspbian where they have all different compile flags and stuff to work on the older pi's
[18:11:46] enyc: peterbennett: I have various chroots of e.g. jessie-armhf which is NOT the same as jessie-rpi for this reason ...
[18:12:52] peterbennett: enyc: I think the arm builds in the PPA are designed for ubuntu Mate on raspberry pi
[18:13:07] enyc: peterbennett: and ubuntu mate, is only for pi2 ?
[18:13:26] peterbennett: enyc: I installed it yesterday and not looking bad at all, apart from this setuid issue
[18:13:54] peterbennett: enyc: I am using pi 2 I think anything less will be pushing it
[18:14:19] enyc: peterbennett: accourding to they ONLY support pi2
[18:15:23] peterbennett: Makes sense. the pi has not much memory
[18:15:27] enyc: peterbennett: that is because its on the armtf architecture. the ''rpi'' architecture is a made-up-thing by raspbian compile flags (NOT provided in ubuntu) in order raspbian to be compatible with older rpi.
[18:15:48] enyc: peterbennett: no, its' because it actually won't run, the original pi isn't compatible with debian/ubuntu standard armhf instruction-set
[18:19:10] willcooke (willcooke!~willcooke@ubuntu/member/willcooke) has quit (Quit: Do your hobbies)
[18:20:36] tgm4883: peterbennett: looks like something in those patches is causing builds to fail
[18:21:00] tgm4883: . . . LDING.txt.gz
[18:21:06] tgm4883: I don't have time to look at that now
[18:25:33] taylorr (taylorr!~taylorr@unaffiliated/elmojo) has quit (*.net *.split)
[18:25:33] tris (tris! has quit (*.net *.split)
[18:25:33] vincent42 (vincent42! has quit (*.net *.split)
[18:25:33] aloril (aloril! has quit (*.net *.split)
[18:25:33] gigem (gigem!~david@mythtv/developer/gigem) has quit (*.net *.split)
[18:25:40] vincent42 (vincent42! has joined #mythtv
[18:25:55] taylorr (taylorr!~taylorr@ has joined #mythtv
[18:25:55] taylorr (taylorr!~taylorr@unaffiliated/elmojo) has joined #mythtv
[18:25:55] taylorr (taylorr!~taylorr@ has quit (Changing host)
[18:26:02] gigem (gigem! has joined #mythtv
[18:26:02] gigem (gigem! has quit (Changing host)
[18:26:02] gigem (gigem!~david@mythtv/developer/gigem) has joined #mythtv
[18:26:05] tris (tris! has joined #mythtv
[18:26:27] aloril (aloril! has joined #mythtv
[18:27:21] rich0 (rich0!~quassel@gentoo/developer/rich0) has quit (*.net *.split)
[18:27:22] jpharvey (jpharvey!~jpharvey@ has quit (*.net *.split)
[18:27:36] rich0_ (rich0_!~quassel@gentoo/developer/rich0) has joined #mythtv
[18:29:50] jpharvey (jpharvey!~jpharvey@ has joined #mythtv
[18:32:45] peterbennett: tgm4883: looking at the output I cannot see where the error is
[18:35:08] peterbennett: tgm4883: Oh yes I see it if (revokeRoot()) != 0) parentheses messed up sorry
[18:42:39] peterbennett: tgm4883: I have updated the patch for that and the missing return statement.
[18:43:17] peterbennett: I guess I should build it myself before handing off like that :(
[18:46:43] jheizer: Building on a pi2 isn't too bad. First build sucks. After that it is usually around an hour.
[18:47:25] jheizer: Mine is currently longer, but probably a result of trying to have a local cache in front of the NFS share. I should try removing it.
[18:53:03] peterbennett: The trouble is, dpkg-buildpackage insists on doing a full build each time
[18:55:37] jheizer: ah, didn't read enough to see you wanted debs
[18:57:18] warpme (warpme! has quit (Ping timeout: 276 seconds)
[18:59:10] gregL (gregL! has quit (Read error: Connection reset by peer)
[19:03:26] andreaz (andreaz! has joined #mythtv
[19:09:13] enyc: =)
[19:12:59] ShapeShifter499 (ShapeShifter499!~ShapeShif@unaffiliated/shapeshifter499) has quit (Ping timeout: 240 seconds)
[19:21:53] natanojl (natanojl!~jonatan@mythtv/developer/natanojl) has joined #mythtv
[19:34:44] sraue (sraue!~stephan@kodi/staff/sraue) has joined #mythtv
[19:38:52] dekarl: enyc: peterbennett: ubuntu armhf is only for armv7+ with the rpi1 being armv6, but both include a fp unit...
[19:41:10] peterbennett: dekarl: enyc: So what do you want to call it in the wiki? There are build instructions that say "raspberry pi"
[19:43:33] dekarl: peterbennett: the rpi1/2 is the only (IIUIC) relevant device due to the combination of working HDMI-CEC, OpenMax, etc
[19:47:40] peterbennett: tgm4883: The build dependencies for armhf did not install libomxil-bellagio-dev, with the result that openmax would be disabled. I think that may be an omission?
[19:49:13] dekarl: peterbennett: you want the broadcom openmax packages, not the mesa/bellagio ones
[19:54:04] KungFuJesus (KungFuJesus! has quit (Ping timeout: 244 seconds)
[19:54:06] peterbennett: do you know which package that is?
[19:55:21] Chutt_ is now known as Chutt
[20:00:27] dekarl: libraspberrypi-dev I'm using image and PPA from here
[20:04:03] peterbennett: I have libraspberrypi-dev installed but the configure still disables openmax
[20:05:10] tgm4883: peterbennett: are you using our build scripts?
[20:05:53] peterbennett: tgm4883: I am now trying with ./configure and make
[20:06:16] tgm4883: peterbennett: here's one of our builds from the PPA, . . . LDING.txt.gz
[20:06:20] peterbennett: The trouble with dpkg was it would need to be cleaned out before every compile
[20:06:21] tgm4883: notice openmax is enabled
[20:07:03] GhostOfRaven (GhostOfRaven! has quit (Ping timeout: 240 seconds)
[20:07:22] peterbennett: Yes it is strange, when I run ./configure it disables openmax unless I install the bellagio
[20:08:49] ShapeShifter499 (ShapeShifter499!~ShapeShif@unaffiliated/shapeshifter499) has joined #mythtv
[20:08:55] tgm4883: probably looking in the wrong spot
[20:09:28] tgm4883: peterbennett: . . . an/rules#L63
[20:10:31] peterbennett: Isee ok we need some extra configure parameters
[20:10:32] dekarl: ahh peterbennet notice the cflags (hints only for configure) . . . y/logs/stdio
[20:10:48] dekarl: have I cursed at our build today?
[20:12:02] dekarl: that was ubuntu, heres the CFLAGS for Raspbian . . . y/logs/stdio
[20:13:33] peterbennett: ok it starts to make sense...
[20:40:03] dmfrey (dmfrey!~dmfrey@ has quit (Ping timeout: 240 seconds)
[20:47:58] enyc: peterbennett: I would call it something along lines of armhf (for rpi2 and others...) not sure how amny should be put in that list
[20:48:58] enyc: dekarl, peterbennett : in effect raspbian have created a different arctitecture, often seem that called 'rpi' which confusingly is still -armhf deb files in debian terms...
[20:49:11] dekarl: enyc: feel free to add relevant arm systems to
[20:49:44] dekarl: that's where I started to collect the relevant bits and pieces about the various popular / cheap arm boards for htpc use
[20:50:20] jheizer: We just need enough time to pass that everyone moves off rasbian to debian proper and/or ubuntu
[20:50:27] enyc: ... not sure ogf the rest of the dails for it, but 'hummingboard' be added =)
[20:50:47] jheizer: Including myself who was too lazy to redo a working sd card image when I had to get a new rpi 2
[20:50:59] enyc: jheizer: ;p
[20:51:18] dekarl: enyc: seeing its imx6, did they fix the HDMI-CEC wiring?
[20:51:29] enyc: dekarl: no clue....
[20:51:33] dmfrey (dmfrey!~dmfrey@ has joined #mythtv
[20:51:43] enyc: dekarl: can you give me some context? this is a story i don't know
[20:51:46] Roklobotomy (Roklobotomy! has joined #mythtv
[20:53:38] enyc: claim the hummingboard i2ex has "HDMI CEC 1080p"
[20:54:00] dekarl: enyc: the reference design had a bug in the HDMI-CEC connectiong that everyone copied into their boards until some users actually tried to used it and noticed the hardware issue
[20:54:03] enyc: interesting about the hummingboard is SATA and pci-e  ;p
[20:54:11] enyc: dekarl: oops ;p
[20:57:31] enyc: i'm really curious how well the hummingboard's pic-e works though
[20:57:57] enyc: notably, i understood from ARM engineers there is a big difeference with in-order // out-of-order bus/memory technologies
[20:58:07] enyc: apparentyl it must have some kind of buffering or similar
[20:58:59] dekarl: enyc: eg
[20:59:07] enyc: also, apparently, arm and intel behave differently in off-aligned memory-mapped-IO accesses ... intel flows over into next word, arm wraps around within one, which can cause chaos depending on the driver ;p
[20:59:59] jheizer: so as good as usb and nic on rpi
[21:00:00] jheizer: haha
[21:00:12] dekarl: sounds similiar funny as reverse-big-endian...
[21:01:47] enyc: dekarl: errrr what??
[21:02:35] dekarl: enyc: the ppc on the Wii (U?) is connected to its devices with a 32bit bus with bytes 1–4 in reverse
[21:03:53] dekarl: a hardware hack to avoid byte swapping for 32bit accesses, obviosly needs drivers that know that all 8/16bit aligned data may be 1,2 or 3 bytes off
[21:04:23] enyc: aorgh ;p
[21:04:36] warpme (warpme! has joined #mythtv
[21:12:50] dekarl:
[21:14:49] rich0_ (rich0_!~quassel@gentoo/developer/rich0) has quit (Remote host closed the connection)
[21:17:29] dmfrey (dmfrey!~dmfrey@ has quit (Ping timeout: 240 seconds)
[21:22:23] enyc: dekarl: but but but what happened to 'u8 b'
[21:22:31] enyc: dekarl: " a sequence of three registers {u8 a; u8 b; u16 c;} packed into a 32-bit word will look like {u16 c; u8 a; u8 a;}"
[21:22:55] dekarl: enyc: its a wiki, be bold and fix it :)
[21:23:37] dekarl: sounds like the messed up HDMI-CEC reference design... "To make things even more confusing, there are implementation bugs and the byte enable lines don't operate as they should."
[21:24:39] rich0_ (rich0_!~quassel@gentoo/developer/rich0) has joined #mythtv
[21:25:13] Warped (Warped!~Warped@unaffiliated/warped) has quit (Ping timeout: 240 seconds)
[21:32:37] enyc: the trouble with read-modify-write access on registers is simultanerous accesses surely ??
[21:38:00] dekarl: i don't think that's an issue one a multi-single-core system. each device will likely only be accessed by one of the cpus which both are single core
[21:40:19] dekarl: privileged stuff runs on the arm "kernel" cpu and the user stuff on the ppc
[21:40:53] stuarta: dekarl: that reversed little endian stuff is just crap
[21:41:10] stuarta: seriously, what were they smoking when they did that
[21:42:42] dekarl: more important, why did they make their next version binary compatible to that instead of moving to a highend mobile SoC?
[21:46:06] dekarl: back on topic. I like the idea of a PullRequestTemplate asking for the trac ticket number :) anything else to add?
[21:46:53] stuarta: ooo nice
[21:55:16] Chutt_ (Chutt_!~ijr@2605:a000:1225:7e:c457:ef1d:60fd:2177) has joined #mythtv
[21:56:29] Chutt (Chutt!~ijr@2605:a000:1225:7e:c457:ef1d:60fd:2177) has quit (Ping timeout: 240 seconds)
[22:07:04] sraue (sraue!~stephan@kodi/staff/sraue) has quit (Ping timeout: 252 seconds)
[22:07:37] sraue (sraue!~stephan@kodi/staff/sraue) has joined #mythtv
[22:08:17] warpme (warpme! has quit (Remote host closed the connection)
[22:08:54] peterbennett (peterbennett! has left #mythtv ()
[22:48:00] SteveGoodey (SteveGoodey! has quit (Quit: Konversation terminated!)
[23:40:39] raven42 (raven42! has joined #mythtv
[23:40:45] Tobbe5178 (Tobbe5178! has quit (Read error: Connection reset by peer)
[23:47:33] rich0_ is now known as rich0
[23:58:58] GhostOfRaven (GhostOfRaven! has joined #mythtv

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