/srv/irclogs.ubuntu.com/2015/02/14/#ubuntu-devel.txt

hallynslangasek: the point of the separate ns is so that /run/cgmanager/fs/* mounts get cleaned up when cgmanager exits00:28
slangasekah, ok00:28
hallynyeah i had almost convinced myself to just not do it, but it'll lead in redisual mounts and complicate restart00:29
ahoneybunRiddell, I found a wordpress theme making who has a company who makes a few free themes from non profit companies00:31
hallynsigh, i'm really quite bad at remembering to quilt add files to a patch00:32
keesxnox: say... can you fix mplayer in trusty? it's really really busted.01:31
xnoxkees: bug #?01:31
keeshttps://bugs.launchpad.net/ubuntu/+source/mplayer/+bug/135244701:31
ubottuLaunchpad bug 1352447 in mplayer (Ubuntu Trusty) "regression: mplayer detects video frame rate as -nan" [High,Confirmed]01:31
keesand now it likes to segfault on perfectly valid streams. had to downgrade _another_ machine today :(01:31
keesxnox: also, mencoder is needed.01:32
xnoxkees: =((((01:32
xnoxkees: i actually hate all of that audio/video stuff, but did do "binNMUs" and patches to get them "fixed" a lot....01:32
keesxnox: easy fix! revert to precise's mplayer ;)01:33
keesactually, I have no idea. I should see if the current debian mplayer works correctly.01:33
keesregardless, mencoder is seriously needed. it's the only thing that does deinterlacing correctly.01:34
keesI suppose I should go complain to Reinhard :)01:35
slangasekkees: --> siretart ? :)01:40
aeorildarkxst I have found out that some init code in vte.c (vte_terminal_init()) which initializes the terminal gtk3 widget uses default values of 80x24 for columns and rows the call to vte_terminal_set_size().  A widget method is set to vte_terminal_set_size() also, so I would think that as the gtk3 widget responds to resizing, it probably "automatically" calls that method I guess?  Somehow04:16
aeorilmaybe the timing is such between the initial setup of gnome-terminal, the resizing done via the --geometry command line argument and the loading of the vim child process within the gnome terminal sometimes the resize causes all the lines to be shown, sometimes only 24 (the default value hard coded into vte.c).  I could use a little help/direction on where to look next to figure out how04:16
aeorilall this stuff works together and where in the code I should be looking - I am thinking of downloading the gnome-terminal code to see how it uses libvte (vte3, the code I am looking at now)04:16
darkxstaeoril, but are you getting the default size? I though it was just off by a couple of lines or something?05:05
aeorildarkxst no, I am getting the default size - sorry if that was not clear - 24 lines are shown, the rest of vim down to the bottom of the enlarged gnome-terminal is blank lines, or invisible05:06
aeorilrows == 24, not sure about columns - I guess I need to test that05:07
darkxstif its a race perhaps running something like "sleep 5;vim" might identify it05:08
darkxstor you could try and hook into the resize event with gdb and see what is going on05:09
aeorildarkxst yahhhoooo!  "sleep 1;vi ..." fixes the problem05:13
darkxstaeoril, turn on debug messages that might help05:14
aeorildarkxst yes, I have been planning on that05:14
aeorildarkxst there do not seem to be command line switches to turn on debugging in gnome-terminal or vim - do I need to compile that in?05:17
darkxstaeoril, probably a env var to turn it on05:17
aeorildarkxst for gnome-terminal, I have to do both - compile debugging in and set an environment variable:  https://wiki.gnome.org/Apps/Terminal/Debugging  for vim, -g option on compile05:22
aeorildarkxst apparently, gnome-terminal has a client-server achitecture.  To debug, you have to start up the server, then hook to it with a client within 10 seconds05:23
aeorilusing gdb05:23
aeorilprobably make the timing issue impossible to reliably reproduce05:23
darkxstaeoril, vte has a VTE_DEBUG that should turn on debug message05:24
aeoriloh, didn't think about vte ... thanks05:24
aeorilActually, I thought about it, but found no manpage and forgot after that05:25
aeorilthanks anyway05:25
darkxstgrep -R getenv05:25
aeorildarkxst I was wondering how you found that ...05:26
darkxstaeoril, the code is often the best documentation ;)05:27
aeorildarkxst yes, I am finding that out ... :)05:28
aeorildarkxst I still need to define VTE_DEBUG when I compile, though05:34
aeoril(#ifdef)05:35
darkxstaeoril, ok05:37
darkxstI have never really worked with vte or gnome-terminal05:38
aeorildarkxst yes, but we are closer now than two days ago - thanks for the help again.  I'll tell you a secret - I have never worked with them before either ha ha!05:43
darkxstI pretty much assumed that05:43
aeorildarkxst that was a joke :)05:43
darkxstyes05:44
aeorilwell, afk.  Tomorrow is another day05:46
darkxstmaybe this was fixed as part of the re-wrap implementation? don't think that was in trusty?05:46
aeorilre-wrap?05:47
darkxstaeoril, http://blogs.gnome.org/mclasen/2013/12/09/a-terminal-surprise/05:48
aeorilCool, thanks.  Will look into it tomorrow :)05:49
aeorildarkxst I think I saw some stuff about wrapping in the latest code on vte maybe?05:50
darkxstaeoril, probably05:50
darkxstlikely in 14.10 also05:50
aeorildarkxst 14.10 exhibits this problem too, though much less often (about 5 % of the time as opposed to 60-70%)05:51
darkxstaeoril, races will always be tricky05:52
aeorildarkxst but races must be wone!05:52
aeorilwon*05:52
aeorildarkxst I wish I could get an audience with chpe!05:53
darkxstaeoril, many of the GNOME devs are on US Business Hoursy-ish05:54
darkxstsome are also EU05:54
darkxstvery few around on weekends05:54
aeorildarkxst good to know - I will post Tuesday earlier my same queries (Monday is a holiday)05:54
aeorilI should know the code better by then as well, and have looked at the gnome-terminal code some05:56
aeorildarkxst I am not wondering if it has something to do with the client-server architecture of gnome-terminal - server/client hookup timing or something like that05:57
aeorils/hookup/interactions/05:59
darkxstno idea05:59
aeorildarkxst I may be able to figure some of that out from code05:59
darkxsttbh  I don't even know what the client-server split involves (ie which bits do what)05:59
aeorildarkxst yes, I was just thinking that to myself, actually - I need to figure that out really06:00
darkxstaeoril, can you build trusty gnome-terminal with current vte?06:00
aeorildifficult I think - different so name/number/whatever06:01
aeorilnot really sure06:01
darkxstaeoril, thats not actually a problem in itself at build time, however often it means there are incompatible api calls06:01
aeorildarkxst wouldn't I have to screw with the build setup though if the library has a different name?06:02
darkxstaeoril, no its automatic06:02
darkxstthe build depend would be libvte3-dev06:02
aeorilahhh ... cool - I can give that a try then tomorrow06:02
aeorildarkxst so how do I get the vivid libvte3-dev on a trusty system to replace the old one?06:03
aeorilcan I do that with apt-get?06:03
darkxstno06:04
darkxstyou will need to rebuild it on trusty06:04
darkxstprobably in a ppa, so its easy to backout the changes with ppa-purge06:04
aeorilok, so build from a ppa source package?06:04
darkxstaeoril, grab the vivid vte, edit changelog and make it target trusty06:05
darkxstthen build under trusty (or upload to a ppa to build)06:05
aeorilok, I'll try that then06:05
aeorildarkxst thanks again!  I should have enough to go on for now, but like I said 20 minutes ago, need to head on ... :P06:07
darkxstbye06:07
aeorildarkxst thanks! :)  You have again been incredibly helpful!06:08
=== not_phunyguy is now known as phunyguy
=== Mr0rris0 is now known as nic778
=== ValicekB_ is now known as ValicekB
darkxstpitti, why on earth would I get asked for permission to unmount a partition when extracting a file via file-roller?11:42
hjdAnyone know or have a suggestion why https://tracker.debian.org/pkg/python-click doesn't seem to have been synced to Ubuntu? I can't find it in the sync blacklist.11:58
ogra_hjd, most likely because click is uploaded to ubuntu first ?11:59
Laneyit's a different 'click'12:00
Laneybut it was removed due to overlapping binary packages with the other one it seems: https://launchpad.net/ubuntu/+source/python-click/+publishinghistory12:00
ogra_oh12:00
ogra_heh12:00
darkxstpitti, bug 142193812:03
ubottubug 1421938 in systemd (Ubuntu) "asked for permission to unmount permission while extracting file with file-roller" [Undecided,New] https://launchpad.net/bugs/142193812:03
darkxstthat should of course be partition though!12:03
hjdogra_: Laney ah ok. I thought about (ubuntu-)click but didn't know that had packages with conflicting names.12:43
=== e^0 is now known as Guest62239
melodiehello15:26
melodiecould someone point me to the page where I can find the filesystem.manifest of Ubuntu 14.04.2 please?15:27
cjwatsonmelodie: Nowhere, because there's no such thing as Ubuntu 14.04.2 yet.  The current builds are at http://cdimage.ubuntu.com/trusty/daily-live/current/ and the .manifest files are alongside them.19:25
melodiehi cjwatson thanks, that's what I meant19:37
melodiecjwatson what about the other flavors? what would be the link like for them?19:38
sladenmelodie: http://cdimage.ubuntu.com/kubuntu/trusty/daily-live/current/  etc19:43
sladenmelodie: (follow up to the top level directory, and then back down again)19:43
melodiesladen thank you19:43
melodieyes, I get it19:43
Bl4ckD34tHwhy i am banned on ubuntu?22:16
xnoxpitti: the set of openened lazr/launchpad bugs are fixed. Your move. ;-)23:27
melodiegood night23:53

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!